sha
stringlengths 40
40
| text
stringlengths 1
13.4M
| id
stringlengths 2
117
| tags
listlengths 1
7.91k
| created_at
stringlengths 25
25
| metadata
stringlengths 2
875k
| last_modified
stringlengths 25
25
| arxiv
listlengths 0
25
| languages
listlengths 0
7.91k
| tags_str
stringlengths 17
159k
| text_str
stringlengths 1
447k
| text_lists
listlengths 0
352
| processed_texts
listlengths 1
353
| tokens_length
listlengths 1
353
| input_texts
listlengths 1
40
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
61d0351be8caaad1702ab1f97b74d3f119529469
|
# Dataset Card for "data-standardized_cluster_15_alpaca"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_15_alpaca
|
[
"region:us"
] |
2023-10-23T01:49:15+00:00
|
{"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 6684630, "num_examples": 3122}], "download_size": 2841840, "dataset_size": 6684630}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:49:17+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_15_alpaca"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_15_alpaca\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_15_alpaca\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_15_alpaca\"\n\nMore Information needed"
] |
302ea5617250d423ba95469cb94935281db31b55
|
# Dataset Card for "data-standardized_cluster_15"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_15
|
[
"region:us"
] |
2023-10-23T01:49:17+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "embedding", "sequence": "float64"}, {"name": "cluster", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 32223194, "num_examples": 3123}], "download_size": 9263830, "dataset_size": 32223194}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:49:20+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_15"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_15\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_15\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_15\"\n\nMore Information needed"
] |
64b908d2f01f33ce999f0567d5648e0d25a24a8e
|
# Dataset Card for "data-standardized_cluster_16_std"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_16_std
|
[
"region:us"
] |
2023-10-23T01:50:09+00:00
|
{"dataset_info": {"features": [{"name": "message", "dtype": "string"}, {"name": "message_type", "dtype": "string"}, {"name": "message_id", "dtype": "int64"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "cluster", "dtype": "float64"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 9175738, "num_examples": 8716}], "download_size": 3946637, "dataset_size": 9175738}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:50:13+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_16_std"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_16_std\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_16_std\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_16_std\"\n\nMore Information needed"
] |
8759766d90c2d59e60848dadec9853dd7815a4cf
|
# Dataset Card for "data-standardized_cluster_16_alpaca"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_16_alpaca
|
[
"region:us"
] |
2023-10-23T01:50:14+00:00
|
{"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 8929857, "num_examples": 4357}], "download_size": 3739633, "dataset_size": 8929857}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:50:16+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_16_alpaca"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_16_alpaca\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_16_alpaca\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_16_alpaca\"\n\nMore Information needed"
] |
a1cb1c147f7127b38a31b7f59bfafe4fa63c8d38
|
# Dataset Card for "data-standardized_cluster_16"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_16
|
[
"region:us"
] |
2023-10-23T01:50:16+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "embedding", "sequence": "float64"}, {"name": "cluster", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 44567056, "num_examples": 4358}], "download_size": 12737777, "dataset_size": 44567056}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:50:19+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_16"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_16\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_16\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_16\"\n\nMore Information needed"
] |
966e0efc0721392a2e4b2360e4b1f04ce4648290
|
# Dataset Card for "data-standardized_cluster_17_std"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_17_std
|
[
"region:us"
] |
2023-10-23T01:51:09+00:00
|
{"dataset_info": {"features": [{"name": "message", "dtype": "string"}, {"name": "message_type", "dtype": "string"}, {"name": "message_id", "dtype": "int64"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "cluster", "dtype": "float64"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 4535547, "num_examples": 3716}], "download_size": 1978928, "dataset_size": 4535547}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:51:11+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_17_std"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_17_std\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_17_std\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_17_std\"\n\nMore Information needed"
] |
6337bf576bb43d37bc8e7df04b7829740c090e58
|
# Dataset Card for "data-standardized_cluster_17_alpaca"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_17_alpaca
|
[
"region:us"
] |
2023-10-23T01:51:12+00:00
|
{"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 4428976, "num_examples": 1857}], "download_size": 1900676, "dataset_size": 4428976}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:51:14+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_17_alpaca"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_17_alpaca\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_17_alpaca\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_17_alpaca\"\n\nMore Information needed"
] |
a435a2897c5ec35b06434b438a91a1468d92f40f
|
# Dataset Card for "data-standardized_cluster_17"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_17
|
[
"region:us"
] |
2023-10-23T01:51:14+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "embedding", "sequence": "float64"}, {"name": "cluster", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 19624365, "num_examples": 1858}], "download_size": 5710907, "dataset_size": 19624365}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:51:17+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_17"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_17\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_17\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_17\"\n\nMore Information needed"
] |
3e1798d2d32092da29c554e12dbf8ac1453385ed
|
# Dataset Card for "data-standardized_cluster_18_std"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_18_std
|
[
"region:us"
] |
2023-10-23T01:52:09+00:00
|
{"dataset_info": {"features": [{"name": "message", "dtype": "string"}, {"name": "message_type", "dtype": "string"}, {"name": "message_id", "dtype": "int64"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "cluster", "dtype": "float64"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 9051592, "num_examples": 8532}], "download_size": 3910771, "dataset_size": 9051592}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:52:11+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_18_std"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_18_std\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_18_std\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_18_std\"\n\nMore Information needed"
] |
94390d25603139bda8519d5b9c6824146d7bc34d
|
# Dataset Card for "data-standardized_cluster_18_alpaca"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_18_alpaca
|
[
"region:us"
] |
2023-10-23T01:52:12+00:00
|
{"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 8810239, "num_examples": 4265}], "download_size": 3718272, "dataset_size": 8810239}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:52:14+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_18_alpaca"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_18_alpaca\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_18_alpaca\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_18_alpaca\"\n\nMore Information needed"
] |
4ec436eedf89f3954e7e06621a7a954841c8a248
|
# Dataset Card for "data-standardized_cluster_18"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_18
|
[
"region:us"
] |
2023-10-23T01:52:14+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "embedding", "sequence": "float64"}, {"name": "cluster", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 43695778, "num_examples": 4266}], "download_size": 12523641, "dataset_size": 43695778}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:52:17+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_18"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_18\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_18\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_18\"\n\nMore Information needed"
] |
0d177cd1f7c5cc5d0ad150b8826d76f915e05b6b
|
# Dataset Card for "data-standardized_cluster_19_std"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_19_std
|
[
"region:us"
] |
2023-10-23T01:53:07+00:00
|
{"dataset_info": {"features": [{"name": "message", "dtype": "string"}, {"name": "message_type", "dtype": "string"}, {"name": "message_id", "dtype": "int64"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "cluster", "dtype": "float64"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 9165871, "num_examples": 8554}], "download_size": 4002670, "dataset_size": 9165871}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:53:09+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_19_std"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_19_std\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_19_std\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_19_std\"\n\nMore Information needed"
] |
c401dc00f9c8dd82acb8d78db42abfedd5715a2e
|
# Dataset Card for "data-standardized_cluster_19_alpaca"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_19_alpaca
|
[
"region:us"
] |
2023-10-23T01:53:10+00:00
|
{"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 8924497, "num_examples": 4276}], "download_size": 3801177, "dataset_size": 8924497}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:53:13+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_19_alpaca"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_19_alpaca\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_19_alpaca\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_19_alpaca\"\n\nMore Information needed"
] |
c94f4df606d6cbb6bb65ef8545cd8046e8fc34f8
|
# Dataset Card for "data-standardized_cluster_19"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_19
|
[
"region:us"
] |
2023-10-23T01:53:13+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "embedding", "sequence": "float64"}, {"name": "cluster", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 43899388, "num_examples": 4277}], "download_size": 12613003, "dataset_size": 43899388}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:53:16+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_19"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_19\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_19\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_19\"\n\nMore Information needed"
] |
4a2d38473fba29394eefa2330486fce15b67eb41
|
# Dataset Card for "data-standardized_cluster_20_std"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_20_std
|
[
"region:us"
] |
2023-10-23T01:54:07+00:00
|
{"dataset_info": {"features": [{"name": "message", "dtype": "string"}, {"name": "message_type", "dtype": "string"}, {"name": "message_id", "dtype": "int64"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "cluster", "dtype": "float64"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 19282668, "num_examples": 16764}], "download_size": 8304570, "dataset_size": 19282668}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:54:11+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_20_std"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_20_std\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_20_std\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_20_std\"\n\nMore Information needed"
] |
e35b50a342a94c7c4eadc9696155da3cc75f1e59
|
# Dataset Card for "data-standardized_cluster_20_alpaca"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_20_alpaca
|
[
"region:us"
] |
2023-10-23T01:54:12+00:00
|
{"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 18811510, "num_examples": 8381}], "download_size": 7922311, "dataset_size": 18811510}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:54:15+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_20_alpaca"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_20_alpaca\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_20_alpaca\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_20_alpaca\"\n\nMore Information needed"
] |
6994ad72389498c925e9a4e890d487ff77eddc52
|
# Dataset Card for "data-standardized_cluster_20"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_20
|
[
"region:us"
] |
2023-10-23T01:54:15+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "embedding", "sequence": "float64"}, {"name": "cluster", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 87352890, "num_examples": 8382}], "download_size": 25168511, "dataset_size": 87352890}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:54:19+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_20"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_20\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_20\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_20\"\n\nMore Information needed"
] |
c9cc30a4a585762a530af8a1d21432c963d8f38a
|
# Dataset Card for "data-standardized_cluster_21_std"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_21_std
|
[
"region:us"
] |
2023-10-23T01:55:10+00:00
|
{"dataset_info": {"features": [{"name": "message", "dtype": "string"}, {"name": "message_type", "dtype": "string"}, {"name": "message_id", "dtype": "int64"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "cluster", "dtype": "float64"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 9681643, "num_examples": 9462}], "download_size": 4191325, "dataset_size": 9681643}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:55:12+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_21_std"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_21_std\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_21_std\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_21_std\"\n\nMore Information needed"
] |
1b4c15efa1c18a04c117d51970694c743f46e117
|
# Dataset Card for "data-standardized_cluster_21_alpaca"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_21_alpaca
|
[
"region:us"
] |
2023-10-23T01:55:13+00:00
|
{"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 9415114, "num_examples": 4730}], "download_size": 3969543, "dataset_size": 9415114}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:55:16+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_21_alpaca"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_21_alpaca\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_21_alpaca\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_21_alpaca\"\n\nMore Information needed"
] |
0a7ab7bf23d8c92c97d03f909b4f3475c1b47d70
|
# Dataset Card for "data-standardized_cluster_21"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_21
|
[
"region:us"
] |
2023-10-23T01:55:16+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "embedding", "sequence": "float64"}, {"name": "cluster", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 48102094, "num_examples": 4731}], "download_size": 13694364, "dataset_size": 48102094}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:55:18+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_21"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_21\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_21\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_21\"\n\nMore Information needed"
] |
735cfeb1ef64c9d54bb9ddff460df85e3f1c4e7e
|
# Dataset Card for "data-standardized_cluster_22_std"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_22_std
|
[
"region:us"
] |
2023-10-23T01:56:10+00:00
|
{"dataset_info": {"features": [{"name": "message", "dtype": "string"}, {"name": "message_type", "dtype": "string"}, {"name": "message_id", "dtype": "int64"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "cluster", "dtype": "float64"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 27474399, "num_examples": 25474}], "download_size": 11969072, "dataset_size": 27474399}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:56:13+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_22_std"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_22_std\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_22_std\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_22_std\"\n\nMore Information needed"
] |
c574c16639a549381f4c2417e4b6ca66511586f1
|
# Dataset Card for "data-standardized_cluster_22_alpaca"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_22_alpaca
|
[
"region:us"
] |
2023-10-23T01:56:15+00:00
|
{"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 26759579, "num_examples": 12736}], "download_size": 11363431, "dataset_size": 26759579}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:56:17+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_22_alpaca"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_22_alpaca\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_22_alpaca\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_22_alpaca\"\n\nMore Information needed"
] |
e755e4ee4bbbb35bbf4909473e428f0bb4678efa
|
# Dataset Card for "data-standardized_cluster_22"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_22
|
[
"region:us"
] |
2023-10-23T01:56:18+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "embedding", "sequence": "float64"}, {"name": "cluster", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 130911576, "num_examples": 12737}], "download_size": 37520503, "dataset_size": 130911576}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:56:22+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_22"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_22\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_22\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_22\"\n\nMore Information needed"
] |
e2f8c6bab268cea523a7f6c2722689b015d9023d
|
# Dataset Card for "data-standardized_cluster_23_std"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_23_std
|
[
"region:us"
] |
2023-10-23T01:57:13+00:00
|
{"dataset_info": {"features": [{"name": "message", "dtype": "string"}, {"name": "message_type", "dtype": "string"}, {"name": "message_id", "dtype": "int64"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "cluster", "dtype": "float64"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 7951640, "num_examples": 6750}], "download_size": 3438336, "dataset_size": 7951640}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:57:15+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_23_std"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_23_std\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_23_std\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_23_std\"\n\nMore Information needed"
] |
e52aa1934daf862f634cdf476f92ae5eda2d57fc
|
# Dataset Card for "data-standardized_cluster_23_alpaca"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_23_alpaca
|
[
"region:us"
] |
2023-10-23T01:57:16+00:00
|
{"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 7761001, "num_examples": 3374}], "download_size": 3287663, "dataset_size": 7761001}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:57:18+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_23_alpaca"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_23_alpaca\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_23_alpaca\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_23_alpaca\"\n\nMore Information needed"
] |
5bb30afb5ac7ff958f61d7df6956e09705e791a8
|
# Dataset Card for "data-standardized_cluster_23"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_23
|
[
"region:us"
] |
2023-10-23T01:57:19+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "embedding", "sequence": "float64"}, {"name": "cluster", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 35360015, "num_examples": 3375}], "download_size": 10257244, "dataset_size": 35360015}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:57:22+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_23"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_23\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_23\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_23\"\n\nMore Information needed"
] |
9c37604fd29b89f1bb6b426a9cf4d6f83956e1ef
|
# Dataset Card for "data-standardized_cluster_24_std"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_24_std
|
[
"region:us"
] |
2023-10-23T01:58:11+00:00
|
{"dataset_info": {"features": [{"name": "message", "dtype": "string"}, {"name": "message_type", "dtype": "string"}, {"name": "message_id", "dtype": "int64"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "cluster", "dtype": "float64"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 18943226, "num_examples": 16868}], "download_size": 8113695, "dataset_size": 18943226}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:58:14+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_24_std"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_24_std\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_24_std\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_24_std\"\n\nMore Information needed"
] |
6283f231edeab30ac7a37bc6306138d500928cbc
|
# Dataset Card for "data-standardized_cluster_24_alpaca"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_24_alpaca
|
[
"region:us"
] |
2023-10-23T01:58:16+00:00
|
{"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 18469147, "num_examples": 8433}], "download_size": 7725454, "dataset_size": 18469147}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:58:18+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_24_alpaca"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_24_alpaca\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_24_alpaca\"\n\nMore Information needed"
] |
[
6,
22
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_24_alpaca\"\n\nMore Information needed"
] |
3a13d14f0b88174e4e14dadad9401a32a1cb03fe
|
# Dataset Card for "data-standardized_cluster_24"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
AdapterOcean/data-standardized_cluster_24
|
[
"region:us"
] |
2023-10-23T01:58:18+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "conversation_id", "dtype": "int64"}, {"name": "embedding", "sequence": "float64"}, {"name": "cluster", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 87435740, "num_examples": 8434}], "download_size": 25087384, "dataset_size": 87435740}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T01:58:21+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "data-standardized_cluster_24"
More Information needed
|
[
"# Dataset Card for \"data-standardized_cluster_24\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"data-standardized_cluster_24\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"data-standardized_cluster_24\"\n\nMore Information needed"
] |
5ee00f3fad17b76f19ec943c3b982ce9e0442f0b
|
Part of the OBELISC data set, including 32 Million samples, please refer to `dataset.py` to use this data
|
AILab-CVC/obelics_seed2_tokens
|
[
"license:cc-by-4.0",
"region:us"
] |
2023-10-23T02:08:19+00:00
|
{"license": "cc-by-4.0"}
|
2023-10-25T01:10:47+00:00
|
[] |
[] |
TAGS
#license-cc-by-4.0 #region-us
|
Part of the OBELISC data set, including 32 Million samples, please refer to 'URL' to use this data
|
[] |
[
"TAGS\n#license-cc-by-4.0 #region-us \n"
] |
[
15
] |
[
"passage: TAGS\n#license-cc-by-4.0 #region-us \n"
] |
d0364e9b4451761012e548b4269b7c6c02234c49
|
# Dataset Card for TD-MPC2
Official dataset release for the paper
[Scalable, Robust World Models for Continuous Control](https://nicklashansen.github.io/td-mpc2) by
[Nicklas Hansen](https://nicklashansen.github.io), [Hao Su](https://cseweb.ucsd.edu/~haosu)\*, [Xiaolong Wang](https://xiaolonw.github.io)\* (UC San Diego)
**Quick links:** [[Website]](https://nicklashansen.github.io/td-mpc2) [[Paper]](https://arxiv.org/abs/2310.16828) [[Models]](https://huggingface.co/nicklashansen/tdmpc2)
## Dataset Details
We open-source all data corresponding to the 80-task and 30-task datasets used in our multi-task experiments. The two datasets contain 545M and 345M transitions, respectively. The data is obtained from the replay buffers of 240 single-task TD-MPC2 agents, and thus contain a wide variety of behaviors ranging from random to expert policies. This section aims to provide further details about the released datasets.
### Dataset Description
- **Curated by:** [Nicklas Hansen](https://nicklashansen.github.io) (UC San Diego)
- **License:** MIT
### Dataset Sources
- **Repository:** [https://github.com/nicklashansen/tdmpc2](https://github.com/nicklashansen/tdmpc2)
- **Paper:** [https://arxiv.org/abs/2310.16828](https://arxiv.org/abs/2310.16828)
### Source Data
Our data is collected by 240 single-task TD-MPC2 agents trained on 104 continuous control tasks from DMControl, Meta-World, Maniskill2, and MyoSuite.
| Dataset | Embodiments | Max obs | Max action | Episodes | Transitions | Size |
|---------|:-----------:|:-------:|:----------:|:--------:|:-----------:|:----:|
| MT80 | 12 | 39 | 6 | 2.69M | 545M | 34 GB |
| MT30 | 11 | 24 | 6 | 690k | 345M | 20 GB |
See the [official website](https://nicklashansen.github.io/td-mpc2/dataset) for visualization of TD-MPC2 agents performing all of 80 tasks from the dataset.
## Citation
If you find our work useful, please consider citing the paper as follows:
**BibTeX:**
```
@misc{hansen2023tdmpc2,
title={TD-MPC2: Scalable, Robust World Models for Continuous Control},
author={Nicklas Hansen and Hao Su and Xiaolong Wang},
year={2023},
eprint={2310.16828},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
## Contact
Correspondence to: [Nicklas Hansen](https://nicklashansen.github.io)
|
nicklashansen/tdmpc2
|
[
"license:mit",
"arxiv:2310.16828",
"region:us"
] |
2023-10-23T03:26:38+00:00
|
{"license": "mit"}
|
2023-10-26T00:04:59+00:00
|
[
"2310.16828"
] |
[] |
TAGS
#license-mit #arxiv-2310.16828 #region-us
|
Dataset Card for TD-MPC2
========================
Official dataset release for the paper
Scalable, Robust World Models for Continuous Control by
Nicklas Hansen, Hao Su\*, Xiaolong Wang\* (UC San Diego)
Quick links: [[Website]](URL [[Paper]](URL [[Models]](URL
Dataset Details
---------------
We open-source all data corresponding to the 80-task and 30-task datasets used in our multi-task experiments. The two datasets contain 545M and 345M transitions, respectively. The data is obtained from the replay buffers of 240 single-task TD-MPC2 agents, and thus contain a wide variety of behaviors ranging from random to expert policies. This section aims to provide further details about the released datasets.
### Dataset Description
* Curated by: Nicklas Hansen (UC San Diego)
* License: MIT
### Dataset Sources
* Repository: URL
* Paper: URL
### Source Data
Our data is collected by 240 single-task TD-MPC2 agents trained on 104 continuous control tasks from DMControl, Meta-World, Maniskill2, and MyoSuite.
See the official website for visualization of TD-MPC2 agents performing all of 80 tasks from the dataset.
If you find our work useful, please consider citing the paper as follows:
BibTeX:
Contact
-------
Correspondence to: Nicklas Hansen
|
[
"### Dataset Description\n\n\n* Curated by: Nicklas Hansen (UC San Diego)\n* License: MIT",
"### Dataset Sources\n\n\n* Repository: URL\n* Paper: URL",
"### Source Data\n\n\nOur data is collected by 240 single-task TD-MPC2 agents trained on 104 continuous control tasks from DMControl, Meta-World, Maniskill2, and MyoSuite.\n\n\n\nSee the official website for visualization of TD-MPC2 agents performing all of 80 tasks from the dataset.\n\n\nIf you find our work useful, please consider citing the paper as follows:\n\n\nBibTeX:\n\n\nContact\n-------\n\n\nCorrespondence to: Nicklas Hansen"
] |
[
"TAGS\n#license-mit #arxiv-2310.16828 #region-us \n",
"### Dataset Description\n\n\n* Curated by: Nicklas Hansen (UC San Diego)\n* License: MIT",
"### Dataset Sources\n\n\n* Repository: URL\n* Paper: URL",
"### Source Data\n\n\nOur data is collected by 240 single-task TD-MPC2 agents trained on 104 continuous control tasks from DMControl, Meta-World, Maniskill2, and MyoSuite.\n\n\n\nSee the official website for visualization of TD-MPC2 agents performing all of 80 tasks from the dataset.\n\n\nIf you find our work useful, please consider citing the paper as follows:\n\n\nBibTeX:\n\n\nContact\n-------\n\n\nCorrespondence to: Nicklas Hansen"
] |
[
20,
22,
16,
109
] |
[
"passage: TAGS\n#license-mit #arxiv-2310.16828 #region-us \n### Dataset Description\n\n\n* Curated by: Nicklas Hansen (UC San Diego)\n* License: MIT### Dataset Sources\n\n\n* Repository: URL\n* Paper: URL### Source Data\n\n\nOur data is collected by 240 single-task TD-MPC2 agents trained on 104 continuous control tasks from DMControl, Meta-World, Maniskill2, and MyoSuite.\n\n\n\nSee the official website for visualization of TD-MPC2 agents performing all of 80 tasks from the dataset.\n\n\nIf you find our work useful, please consider citing the paper as follows:\n\n\nBibTeX:\n\n\nContact\n-------\n\n\nCorrespondence to: Nicklas Hansen"
] |
28814a2d621ae766860a961d414c02f77024aecf
|
<img src="https://huggingface.co/datasets/hkust-nlp/deita-images/resolve/main/logo-final.png" alt="Deita banner" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
# Dataset Card for Deita 6K V0
[GitHub](https://github.com/hkust-nlp/deita) | [Paper](https://arxiv.org/abs/2312.15685)
Deita is an open-sourced project designed to facilitate **Automatic Data Selection** for instruction tuning in Large Language Models (LLMs).
This dataset includes 6k of **lightweight, high-quality** alignment SFT data, mainly automatically selected from the following datasets:
- [ShareGPT](https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered) (Apache 2.0 listed, no official repo found): Use the 58 K ShareGPT dataset for selection.
- [UltraChat](https://huggingface.co/datasets/stingning/ultrachat) (MIT): Sample 105 K UltraChat dataset for selection.
- [WizardLM](https://huggingface.co/datasets/WizardLM/WizardLM_evol_instruct_V2_196k) : Use the evolved data of Alpaca and ShareGPT with 143 K mixture for selection.
**Model Family**: Other models and the dataset are found in the [Deita Collection](https://huggingface.co/collections/hkust-nlp/deita-6569c198c174808d94cf5bd4)
## Performance
| Model | Align | Data Size | MT-Bench | AlpacaEval(%) | OpenLLM (Avg.) |
|------------------------------------------------|-----------|------------|----------|---------------|----------------|
| **Proprietary Models** | | | | | |
| GPT-4-Turbo | ? | -- | 9.32 | 97.70 | -- |
| GPT-4 | SFT + PPO | -- | 8.99 | 95.03 | -- |
| Claude-2 | SFT + PPO | -- | 8.06 | 91.36 | -- |
| GPT-3.5-turbo | SFT + PPO | -- | 7.94 | 89.37 | -- |
| **Open-sourced Models based on LLaMA-1-13B** | | | | | |
| LIMA | SFT | 1K SFT | 4.29 | 41.98 | 59.82 |
| WizardLM-13B | SFT | 70K SFT | 6.35 | 75.31 | 58.96 |
| Vicuna-13B-v1.3 | SFT | 125K SFT | 6.39 | 82.11 | 60.01 |
| Random | SFT | 10K SFT | 6.03 | 71.52 | 60.14 |
| DEITA-LLaMA1-13B-v1.0-sft | SFT | 10K SFT | 6.60 | 78.01 | 64.27 |
| **Open-sourced Models based on LLaMA-2-13B** | | | | | |
| Tulu-2-13B | SFT | 326K SFT | 6.70 | 78.90 | -- |
| Tulu-2-13B+DPO | SFT + DPO | 326K SFT + 60K DPO | 7.00 | 89.50 | -- |
| LLaMA2-13B-Chat | SFT + PPO | -- | 6.65 | 81.09 | -- |
| WizardLM-13B-v1.2 | SFT | >70K SFT | 7.09 | 89.17 | -- |
| Vicuna-13B-v1.5 | SFT | 125K SFT | 6.57 | 78.80 | 61.63 |
| Random | SFT | 10K SFT | 5.78 | 65.19 | 61.32 |
| DEITA-LLaMA2-13B-v1.0-sft | SFT | 10K SFT | 6.79 | 81.09 | 62.71 |
| **Open-sourced Models based on Mistral-7B** | | | | | |
| Mistral-7B-Instruct-v0.1 | -- | -- | 6.84 | 69.65 | 60.45 |
| Zephyr-7B-sft | SFT | 200K SFT | 5.32 | 75.12 | 60.93 |
| $\text{Zephyr-7B-}\beta$ | SFT + DPO | 200K SFT + 60K DPO | 7.34 | 90.60 | 66.36 |
| OpenChat-3.5 | C-RLFT | >> 70K C-RLFT | 7.81 | 88.51 | -- |
| Starling-7B | C-RLFT + APA | >>70K C-RLFT + 183K APA | 8.09 | 91.99 | -- |
| Random | SFT | 10K SFT | 5.89 | 56.90 | 61.72 |
| DEITA-7B-v1.0-sft (6K) | SFT | 6K SFT | 7.22 | 80.78 | 64.94 |
| DEITA-7B-v1.0-sft (10K) | SFT | 10K SFT | 7.32 | 81.67 | 64.00 |
| DEITA-7B-v1.0 | SFT + DPO | 6K SFT + 10K DPO | 7.55 | 90.06 | 69.86 |
## Citation
If you find the content of this project helpful, please cite our paper as follows:
```
@misc{liu2023what,
title={What Makes Good Data for Alignment? A Comprehensive Study of Automatic Data Selection in Instruction Tuning},
author={Wei Liu and Weihao Zeng and Keqing He and Yong Jiang and Junxian He},
year={2023},
eprint={2312.15685},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
|
hkust-nlp/deita-6k-v0
|
[
"task_categories:conversational",
"size_categories:1K<n<10K",
"language:en",
"license:mit",
"arxiv:2312.15685",
"region:us"
] |
2023-10-23T03:40:08+00:00
|
{"language": ["en"], "license": "mit", "size_categories": ["1K<n<10K"], "task_categories": ["conversational"]}
|
2023-12-31T02:52:08+00:00
|
[
"2312.15685"
] |
[
"en"
] |
TAGS
#task_categories-conversational #size_categories-1K<n<10K #language-English #license-mit #arxiv-2312.15685 #region-us
|
<img src="URL alt="Deita banner" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
Dataset Card for Deita 6K V0
============================
GitHub | Paper
Deita is an open-sourced project designed to facilitate Automatic Data Selection for instruction tuning in Large Language Models (LLMs).
This dataset includes 6k of lightweight, high-quality alignment SFT data, mainly automatically selected from the following datasets:
* ShareGPT (Apache 2.0 listed, no official repo found): Use the 58 K ShareGPT dataset for selection.
* UltraChat (MIT): Sample 105 K UltraChat dataset for selection.
* WizardLM : Use the evolved data of Alpaca and ShareGPT with 143 K mixture for selection.
Model Family: Other models and the dataset are found in the Deita Collection
Performance
-----------
If you find the content of this project helpful, please cite our paper as follows:
|
[] |
[
"TAGS\n#task_categories-conversational #size_categories-1K<n<10K #language-English #license-mit #arxiv-2312.15685 #region-us \n"
] |
[
46
] |
[
"passage: TAGS\n#task_categories-conversational #size_categories-1K<n<10K #language-English #license-mit #arxiv-2312.15685 #region-us \n"
] |
7eb9eb5bbadb69b2aac2cd0402f83986657ee849
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj
## Dataset Description
- **Homepage:**
- **Repository:** https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj
- **Paper:**
- **Leaderboard:** https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
- **Point of Contact:** [email protected]
### Dataset Summary
Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).
To load the details from a run, you can for instance do the following:
```python
from datasets import load_dataset
data = load_dataset("open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w-q_k_v_o_proj",
"harness_winogrande_5",
split="train")
```
## Latest results
These are the [latest results from run 2023-10-23T04:40:08.123829](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w-q_k_v_o_proj/blob/main/results_2023-10-23T04-40-08.123829.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
```python
{
"all": {
"em": 0.35979446308724833,
"em_stderr": 0.0049150348561349125,
"f1": 0.39900062919463114,
"f1_stderr": 0.004825664226319409,
"acc": 0.46081554692422627,
"acc_stderr": 0.01059065324860096
},
"harness|drop|3": {
"em": 0.35979446308724833,
"em_stderr": 0.0049150348561349125,
"f1": 0.39900062919463114,
"f1_stderr": 0.004825664226319409
},
"harness|gsm8k|5": {
"acc": 0.14025777103866566,
"acc_stderr": 0.009565108281428673
},
"harness|winogrande|5": {
"acc": 0.7813733228097869,
"acc_stderr": 0.011616198215773246
}
}
```
### Supported Tasks and Leaderboards
[More Information Needed]
### Languages
[More Information Needed]
## Dataset Structure
### Data Instances
[More Information Needed]
### Data Fields
[More Information Needed]
### Data Splits
[More Information Needed]
## Dataset Creation
### Curation Rationale
[More Information Needed]
### Source Data
#### Initial Data Collection and Normalization
[More Information Needed]
#### Who are the source language producers?
[More Information Needed]
### Annotations
#### Annotation process
[More Information Needed]
#### Who are the annotators?
[More Information Needed]
### Personal and Sensitive Information
[More Information Needed]
## Considerations for Using the Data
### Social Impact of Dataset
[More Information Needed]
### Discussion of Biases
[More Information Needed]
### Other Known Limitations
[More Information Needed]
## Additional Information
### Dataset Curators
[More Information Needed]
### Licensing Information
[More Information Needed]
### Citation Information
[More Information Needed]
### Contributions
[More Information Needed]
|
open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w-q_k_v_o_proj
|
[
"region:us"
] |
2023-10-23T03:40:12+00:00
|
{"pretty_name": "Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj", "dataset_summary": "Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\nTo load the details from a run, you can for instance do the following:\n```python\nfrom datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w-q_k_v_o_proj\",\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\nThese are the [latest results from run 2023-10-23T04:40:08.123829](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w-q_k_v_o_proj/blob/main/results_2023-10-23T04-40-08.123829.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.35979446308724833,\n \"em_stderr\": 0.0049150348561349125,\n \"f1\": 0.39900062919463114,\n \"f1_stderr\": 0.004825664226319409,\n \"acc\": 0.46081554692422627,\n \"acc_stderr\": 0.01059065324860096\n },\n \"harness|drop|3\": {\n \"em\": 0.35979446308724833,\n \"em_stderr\": 0.0049150348561349125,\n \"f1\": 0.39900062919463114,\n \"f1_stderr\": 0.004825664226319409\n },\n \"harness|gsm8k|5\": {\n \"acc\": 0.14025777103866566,\n \"acc_stderr\": 0.009565108281428673\n },\n \"harness|winogrande|5\": {\n \"acc\": 0.7813733228097869,\n \"acc_stderr\": 0.011616198215773246\n }\n}\n```", "repo_url": "https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj", "leaderboard_url": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard", "point_of_contact": "[email protected]", "configs": [{"config_name": "harness_drop_3", "data_files": [{"split": "2023_10_23T04_40_08.123829", "path": ["**/details_harness|drop|3_2023-10-23T04-40-08.123829.parquet"]}, {"split": "latest", "path": ["**/details_harness|drop|3_2023-10-23T04-40-08.123829.parquet"]}]}, {"config_name": "harness_gsm8k_5", "data_files": [{"split": "2023_10_23T04_40_08.123829", "path": ["**/details_harness|gsm8k|5_2023-10-23T04-40-08.123829.parquet"]}, {"split": "latest", "path": ["**/details_harness|gsm8k|5_2023-10-23T04-40-08.123829.parquet"]}]}, {"config_name": "harness_winogrande_5", "data_files": [{"split": "2023_10_23T04_40_08.123829", "path": ["**/details_harness|winogrande|5_2023-10-23T04-40-08.123829.parquet"]}, {"split": "latest", "path": ["**/details_harness|winogrande|5_2023-10-23T04-40-08.123829.parquet"]}]}, {"config_name": "results", "data_files": [{"split": "2023_10_23T04_40_08.123829", "path": ["results_2023-10-23T04-40-08.123829.parquet"]}, {"split": "latest", "path": ["results_2023-10-23T04-40-08.123829.parquet"]}]}]}
|
2023-10-23T03:40:19+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj
## Dataset Description
- Homepage:
- Repository: URL
- Paper:
- Leaderboard: URL
- Point of Contact: clementine@URL
### Dataset Summary
Dataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj on the Open LLM Leaderboard.
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).
To load the details from a run, you can for instance do the following:
## Latest results
These are the latest results from run 2023-10-23T04:40:08.123829(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
### Supported Tasks and Leaderboards
### Languages
## Dataset Structure
### Data Instances
### Data Fields
### Data Splits
## Dataset Creation
### Curation Rationale
### Source Data
#### Initial Data Collection and Normalization
#### Who are the source language producers?
### Annotations
#### Annotation process
#### Who are the annotators?
### Personal and Sensitive Information
## Considerations for Using the Data
### Social Impact of Dataset
### Discussion of Biases
### Other Known Limitations
## Additional Information
### Dataset Curators
### Licensing Information
### Contributions
|
[
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T04:40:08.123829(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T04:40:08.123829(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
6,
39,
31,
187,
67,
10,
4,
6,
6,
5,
5,
5,
7,
4,
10,
10,
5,
5,
9,
8,
8,
7,
8,
7,
5,
6,
6,
5
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w-q_k_v_o_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:## Latest results\n\nThese are the latest results from run 2023-10-23T04:40:08.123829(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):### Supported Tasks and Leaderboards### Languages## Dataset Structure### Data Instances### Data Fields### Data Splits## Dataset Creation### Curation Rationale### Source Data#### Initial Data Collection and Normalization#### Who are the source language producers?### Annotations#### Annotation process#### Who are the annotators?### Personal and Sensitive Information## Considerations for Using the Data### Social Impact of Dataset### Discussion of Biases### Other Known Limitations## Additional Information### Dataset Curators### Licensing Information### Contributions"
] |
fdb88575d6d3b0e5b8fc9934778b62313c7837ba
|
<img src="https://huggingface.co/datasets/hkust-nlp/deita-images/resolve/main/logo-final.png" alt="Deita banner" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
# Dataset Card for Deita 10K V0
[GitHub](https://github.com/hkust-nlp/deita) | [Paper](https://arxiv.org/abs/2312.15685)
Deita is an open-sourced project designed to facilitate **Automatic Data Selection** for instruction tuning in Large Language Models (LLMs).
This dataset includes 10k of **lightweight, high-quality** alignment SFT data, mainly automatically selected from the following datasets:
- [ShareGPT](https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered) (Apache 2.0 listed, no official repo found): Use the 58 K ShareGPT dataset for selection.
- [UltraChat](https://huggingface.co/datasets/stingning/ultrachat) (MIT): Sample 105 K UltraChat dataset for selection.
- [WizardLM](https://huggingface.co/datasets/WizardLM/WizardLM_evol_instruct_V2_196k) : Use the evolved data of Alpaca and ShareGPT with 143 K mixture for selection.
**Model Family**: Other models and the dataset are found in the [Deita Collection](https://huggingface.co/collections/hkust-nlp/deita-6569c198c174808d94cf5bd4)
## Performance
| Model | Align | Data Size | MT-Bench | AlpacaEval(%) | OpenLLM (Avg.) |
|------------------------------------------------|-----------|------------|----------|---------------|----------------|
| **Proprietary Models** | | | | | |
| GPT-4-Turbo | ? | -- | 9.32 | 97.70 | -- |
| GPT-4 | SFT + PPO | -- | 8.99 | 95.03 | -- |
| Claude-2 | SFT + PPO | -- | 8.06 | 91.36 | -- |
| GPT-3.5-turbo | SFT + PPO | -- | 7.94 | 89.37 | -- |
| **Open-sourced Models based on LLaMA-1-13B** | | | | | |
| LIMA | SFT | 1K SFT | 4.29 | 41.98 | 59.82 |
| WizardLM-13B | SFT | 70K SFT | 6.35 | 75.31 | 58.96 |
| Vicuna-13B-v1.3 | SFT | 125K SFT | 6.39 | 82.11 | 60.01 |
| Random | SFT | 10K SFT | 6.03 | 71.52 | 60.14 |
| DEITA-LLaMA1-13B-v1.0-sft | SFT | 10K SFT | 6.60 | 78.01 | 64.27 |
| **Open-sourced Models based on LLaMA-2-13B** | | | | | |
| Tulu-2-13B | SFT | 326K SFT | 6.70 | 78.90 | -- |
| Tulu-2-13B+DPO | SFT + DPO | 326K SFT + 60K DPO | 7.00 | 89.50 | -- |
| LLaMA2-13B-Chat | SFT + PPO | -- | 6.65 | 81.09 | -- |
| WizardLM-13B-v1.2 | SFT | >70K SFT | 7.09 | 89.17 | -- |
| Vicuna-13B-v1.5 | SFT | 125K SFT | 6.57 | 78.80 | 61.63 |
| Random | SFT | 10K SFT | 5.78 | 65.19 | 61.32 |
| DEITA-LLaMA2-13B-v1.0-sft | SFT | 10K SFT | 6.79 | 81.09 | 62.71 |
| **Open-sourced Models based on Mistral-7B** | | | | | |
| Mistral-7B-Instruct-v0.1 | -- | -- | 6.84 | 69.65 | 60.45 |
| Zephyr-7B-sft | SFT | 200K SFT | 5.32 | 75.12 | 60.93 |
| $\text{Zephyr-7B-}\beta$ | SFT + DPO | 200K SFT + 60K DPO | 7.34 | 90.60 | 66.36 |
| OpenChat-3.5 | C-RLFT | >> 70K C-RLFT | 7.81 | 88.51 | -- |
| Starling-7B | C-RLFT + APA | >>70K C-RLFT + 183K APA | 8.09 | 91.99 | -- |
| Random | SFT | 10K SFT | 5.89 | 56.90 | 61.72 |
| DEITA-7B-v1.0-sft (6K) | SFT | 6K SFT | 7.22 | 80.78 | 64.94 |
| DEITA-7B-v1.0-sft (10K) | SFT | 10K SFT | 7.32 | 81.67 | 64.00 |
| DEITA-7B-v1.0 | SFT + DPO | 6K SFT + 10K DPO | 7.55 | 90.06 | 69.86 |
## Citation
If you find the content of this project helpful, please cite our paper as follows:
```
@misc{liu2023what,
title={What Makes Good Data for Alignment? A Comprehensive Study of Automatic Data Selection in Instruction Tuning},
author={Wei Liu and Weihao Zeng and Keqing He and Yong Jiang and Junxian He},
year={2023},
eprint={2312.15685},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
|
hkust-nlp/deita-10k-v0
|
[
"task_categories:conversational",
"size_categories:1K<n<10K",
"language:en",
"license:mit",
"arxiv:2312.15685",
"region:us"
] |
2023-10-23T03:49:14+00:00
|
{"language": ["en"], "license": "mit", "size_categories": ["1K<n<10K"], "task_categories": ["conversational"]}
|
2023-12-31T02:57:18+00:00
|
[
"2312.15685"
] |
[
"en"
] |
TAGS
#task_categories-conversational #size_categories-1K<n<10K #language-English #license-mit #arxiv-2312.15685 #region-us
|
<img src="URL alt="Deita banner" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
Dataset Card for Deita 10K V0
=============================
GitHub | Paper
Deita is an open-sourced project designed to facilitate Automatic Data Selection for instruction tuning in Large Language Models (LLMs).
This dataset includes 10k of lightweight, high-quality alignment SFT data, mainly automatically selected from the following datasets:
* ShareGPT (Apache 2.0 listed, no official repo found): Use the 58 K ShareGPT dataset for selection.
* UltraChat (MIT): Sample 105 K UltraChat dataset for selection.
* WizardLM : Use the evolved data of Alpaca and ShareGPT with 143 K mixture for selection.
Model Family: Other models and the dataset are found in the Deita Collection
Performance
-----------
If you find the content of this project helpful, please cite our paper as follows:
|
[] |
[
"TAGS\n#task_categories-conversational #size_categories-1K<n<10K #language-English #license-mit #arxiv-2312.15685 #region-us \n"
] |
[
46
] |
[
"passage: TAGS\n#task_categories-conversational #size_categories-1K<n<10K #language-English #license-mit #arxiv-2312.15685 #region-us \n"
] |
daafc844c6b7ba5d0d5b58cd01c5b524f4b7d4fb
|
# Dataset Card for "wiki_with_embedding1"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
arminmrm93/wiki_with_embedding1
|
[
"region:us"
] |
2023-10-23T03:51:22+00:00
|
{"dataset_info": {"features": [{"name": "id", "dtype": "string"}, {"name": "url", "dtype": "string"}, {"name": "title", "dtype": "string"}, {"name": "text", "dtype": "string"}, {"name": "embedding", "sequence": "float32"}], "splits": [{"name": "train", "num_bytes": 834134749, "num_examples": 64586}], "download_size": 597776213, "dataset_size": 834134749}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T03:51:54+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "wiki_with_embedding1"
More Information needed
|
[
"# Dataset Card for \"wiki_with_embedding1\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"wiki_with_embedding1\"\n\nMore Information needed"
] |
[
6,
17
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"wiki_with_embedding1\"\n\nMore Information needed"
] |
9a8ceda992c129fa6792ef8b0031308c53cdd3e4
|
# Dataset Card for "repobench_ablation"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
tianyang/repobench_ablation
|
[
"region:us"
] |
2023-10-23T04:03:19+00:00
|
{"configs": [{"config_name": "default", "data_files": [{"split": "cross_file_first", "path": "data/cross_file_first-*"}, {"split": "cross_file_random", "path": "data/cross_file_random-*"}, {"split": "in_file", "path": "data/in_file-*"}]}], "dataset_info": {"features": [{"name": "repo_name", "dtype": "string"}, {"name": "file_path", "dtype": "string"}, {"name": "context", "list": [{"name": "identifier", "dtype": "string"}, {"name": "path", "dtype": "string"}, {"name": "snippet", "dtype": "string"}]}, {"name": "import_statement", "dtype": "string"}, {"name": "token_num", "dtype": "int64"}, {"name": "cropped_code", "dtype": "string"}, {"name": "all_code", "dtype": "string"}, {"name": "next_line", "dtype": "string"}, {"name": "gold_snippet_index", "dtype": "int64"}], "splits": [{"name": "cross_file_first", "num_bytes": 144850826, "num_examples": 1695}, {"name": "cross_file_random", "num_bytes": 115858056, "num_examples": 1549}, {"name": "in_file", "num_bytes": 126244757, "num_examples": 1612}], "download_size": 116113239, "dataset_size": 386953639}}
|
2023-10-23T04:57:39+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "repobench_ablation"
More Information needed
|
[
"# Dataset Card for \"repobench_ablation\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"repobench_ablation\"\n\nMore Information needed"
] |
[
6,
17
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"repobench_ablation\"\n\nMore Information needed"
] |
dd11f8ae6bee4d81b548b39784af72cd3d182971
|
# Dataset Card for "arc_hella"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
Yehoon/arc_hella
|
[
"region:us"
] |
2023-10-23T04:06:44+00:00
|
{"dataset_info": {"features": [{"name": "question", "dtype": "string"}, {"name": "options", "sequence": "string"}, {"name": "answer", "dtype": "string"}, {"name": "label", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 8854506, "num_examples": 12418}], "download_size": 5407350, "dataset_size": 8854506}}
|
2023-10-23T04:06:48+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "arc_hella"
More Information needed
|
[
"# Dataset Card for \"arc_hella\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"arc_hella\"\n\nMore Information needed"
] |
[
6,
15
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"arc_hella\"\n\nMore Information needed"
] |
eb2d083ce866b5bceb6aa9f0925a13665e49f23d
|
# Dataset Card for "tulu_merge"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
Rocinante/tulu_merge
|
[
"region:us"
] |
2023-10-23T04:16:37+00:00
|
{"dataset_info": {"features": [{"name": "output", "dtype": "string"}, {"name": "instruction", "dtype": "string"}, {"name": "input", "dtype": "string"}, {"name": "data_source", "dtype": "string"}, {"name": "history", "sequence": {"sequence": "string"}}], "splits": [{"name": "train", "num_bytes": 306750727, "num_examples": 203886}], "download_size": 174953486, "dataset_size": 306750727}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T04:18:40+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "tulu_merge"
More Information needed
|
[
"# Dataset Card for \"tulu_merge\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"tulu_merge\"\n\nMore Information needed"
] |
[
6,
15
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"tulu_merge\"\n\nMore Information needed"
] |
ae8c2ffd31ed3e1de26ccb580c85642427972a56
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w
## Dataset Description
- **Homepage:**
- **Repository:** https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w
- **Paper:**
- **Leaderboard:** https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
- **Point of Contact:** [email protected]
### Dataset Summary
Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE1_17w](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).
To load the details from a run, you can for instance do the following:
```python
from datasets import load_dataset
data = load_dataset("open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w",
"harness_winogrande_5",
split="train")
```
## Latest results
These are the [latest results from run 2023-10-23T05:19:29.097153](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w/blob/main/results_2023-10-23T05-19-29.097153.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
```python
{
"all": {
"em": 0.36398909395973156,
"em_stderr": 0.00492738070289941,
"f1": 0.4025178271812084,
"f1_stderr": 0.004834808125344133,
"acc": 0.45268383345111485,
"acc_stderr": 0.010561270649266652
},
"harness|drop|3": {
"em": 0.36398909395973156,
"em_stderr": 0.00492738070289941,
"f1": 0.4025178271812084,
"f1_stderr": 0.004834808125344133
},
"harness|gsm8k|5": {
"acc": 0.1326762699014405,
"acc_stderr": 0.009343929131442216
},
"harness|winogrande|5": {
"acc": 0.7726913970007893,
"acc_stderr": 0.011778612167091087
}
}
```
### Supported Tasks and Leaderboards
[More Information Needed]
### Languages
[More Information Needed]
## Dataset Structure
### Data Instances
[More Information Needed]
### Data Fields
[More Information Needed]
### Data Splits
[More Information Needed]
## Dataset Creation
### Curation Rationale
[More Information Needed]
### Source Data
#### Initial Data Collection and Normalization
[More Information Needed]
#### Who are the source language producers?
[More Information Needed]
### Annotations
#### Annotation process
[More Information Needed]
#### Who are the annotators?
[More Information Needed]
### Personal and Sensitive Information
[More Information Needed]
## Considerations for Using the Data
### Social Impact of Dataset
[More Information Needed]
### Discussion of Biases
[More Information Needed]
### Other Known Limitations
[More Information Needed]
## Additional Information
### Dataset Curators
[More Information Needed]
### Licensing Information
[More Information Needed]
### Citation Information
[More Information Needed]
### Contributions
[More Information Needed]
|
open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w
|
[
"region:us"
] |
2023-10-23T04:19:33+00:00
|
{"pretty_name": "Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w", "dataset_summary": "Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE1_17w](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\nTo load the details from a run, you can for instance do the following:\n```python\nfrom datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w\",\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\nThese are the [latest results from run 2023-10-23T05:19:29.097153](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w/blob/main/results_2023-10-23T05-19-29.097153.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.36398909395973156,\n \"em_stderr\": 0.00492738070289941,\n \"f1\": 0.4025178271812084,\n \"f1_stderr\": 0.004834808125344133,\n \"acc\": 0.45268383345111485,\n \"acc_stderr\": 0.010561270649266652\n },\n \"harness|drop|3\": {\n \"em\": 0.36398909395973156,\n \"em_stderr\": 0.00492738070289941,\n \"f1\": 0.4025178271812084,\n \"f1_stderr\": 0.004834808125344133\n },\n \"harness|gsm8k|5\": {\n \"acc\": 0.1326762699014405,\n \"acc_stderr\": 0.009343929131442216\n },\n \"harness|winogrande|5\": {\n \"acc\": 0.7726913970007893,\n \"acc_stderr\": 0.011778612167091087\n }\n}\n```", "repo_url": "https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w", "leaderboard_url": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard", "point_of_contact": "[email protected]", "configs": [{"config_name": "harness_drop_3", "data_files": [{"split": "2023_10_23T05_19_29.097153", "path": ["**/details_harness|drop|3_2023-10-23T05-19-29.097153.parquet"]}, {"split": "latest", "path": ["**/details_harness|drop|3_2023-10-23T05-19-29.097153.parquet"]}]}, {"config_name": "harness_gsm8k_5", "data_files": [{"split": "2023_10_23T05_19_29.097153", "path": ["**/details_harness|gsm8k|5_2023-10-23T05-19-29.097153.parquet"]}, {"split": "latest", "path": ["**/details_harness|gsm8k|5_2023-10-23T05-19-29.097153.parquet"]}]}, {"config_name": "harness_winogrande_5", "data_files": [{"split": "2023_10_23T05_19_29.097153", "path": ["**/details_harness|winogrande|5_2023-10-23T05-19-29.097153.parquet"]}, {"split": "latest", "path": ["**/details_harness|winogrande|5_2023-10-23T05-19-29.097153.parquet"]}]}, {"config_name": "results", "data_files": [{"split": "2023_10_23T05_19_29.097153", "path": ["results_2023-10-23T05-19-29.097153.parquet"]}, {"split": "latest", "path": ["results_2023-10-23T05-19-29.097153.parquet"]}]}]}
|
2023-10-23T04:19:39+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w
## Dataset Description
- Homepage:
- Repository: URL
- Paper:
- Leaderboard: URL
- Point of Contact: clementine@URL
### Dataset Summary
Dataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w on the Open LLM Leaderboard.
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).
To load the details from a run, you can for instance do the following:
## Latest results
These are the latest results from run 2023-10-23T05:19:29.097153(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
### Supported Tasks and Leaderboards
### Languages
## Dataset Structure
### Data Instances
### Data Fields
### Data Splits
## Dataset Creation
### Curation Rationale
### Source Data
#### Initial Data Collection and Normalization
#### Who are the source language producers?
### Annotations
#### Annotation process
#### Who are the annotators?
### Personal and Sensitive Information
## Considerations for Using the Data
### Social Impact of Dataset
### Discussion of Biases
### Other Known Limitations
## Additional Information
### Dataset Curators
### Licensing Information
### Contributions
|
[
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T05:19:29.097153(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T05:19:29.097153(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
6,
28,
31,
176,
67,
10,
4,
6,
6,
5,
5,
5,
7,
4,
10,
10,
5,
5,
9,
8,
8,
7,
8,
7,
5,
6,
6,
5
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:## Latest results\n\nThese are the latest results from run 2023-10-23T05:19:29.097153(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):### Supported Tasks and Leaderboards### Languages## Dataset Structure### Data Instances### Data Fields### Data Splits## Dataset Creation### Curation Rationale### Source Data#### Initial Data Collection and Normalization#### Who are the source language producers?### Annotations#### Annotation process#### Who are the annotators?### Personal and Sensitive Information## Considerations for Using the Data### Social Impact of Dataset### Discussion of Biases### Other Known Limitations## Additional Information### Dataset Curators### Licensing Information### Contributions"
] |
74b9012801799472b5ca8a5619a18dc5153283b7
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj
## Dataset Description
- **Homepage:**
- **Repository:** https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj
- **Paper:**
- **Leaderboard:** https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
- **Point of Contact:** [email protected]
### Dataset Summary
Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).
To load the details from a run, you can for instance do the following:
```python
from datasets import load_dataset
data = load_dataset("open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w-q_k_v_o_proj",
"harness_winogrande_5",
split="train")
```
## Latest results
These are the [latest results from run 2023-10-23T05:27:14.109658](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w-q_k_v_o_proj/blob/main/results_2023-10-23T05-27-14.109658.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
```python
{
"all": {
"em": 0.3719588926174497,
"em_stderr": 0.004949726013193945,
"f1": 0.4084679110738261,
"f1_stderr": 0.004843145937750956,
"acc": 0.4480415851620389,
"acc_stderr": 0.010535274120903989
},
"harness|drop|3": {
"em": 0.3719588926174497,
"em_stderr": 0.004949726013193945,
"f1": 0.4084679110738261,
"f1_stderr": 0.004843145937750956
},
"harness|gsm8k|5": {
"acc": 0.1281273692191054,
"acc_stderr": 0.009206398549980031
},
"harness|winogrande|5": {
"acc": 0.7679558011049724,
"acc_stderr": 0.011864149691827948
}
}
```
### Supported Tasks and Leaderboards
[More Information Needed]
### Languages
[More Information Needed]
## Dataset Structure
### Data Instances
[More Information Needed]
### Data Fields
[More Information Needed]
### Data Splits
[More Information Needed]
## Dataset Creation
### Curation Rationale
[More Information Needed]
### Source Data
#### Initial Data Collection and Normalization
[More Information Needed]
#### Who are the source language producers?
[More Information Needed]
### Annotations
#### Annotation process
[More Information Needed]
#### Who are the annotators?
[More Information Needed]
### Personal and Sensitive Information
[More Information Needed]
## Considerations for Using the Data
### Social Impact of Dataset
[More Information Needed]
### Discussion of Biases
[More Information Needed]
### Other Known Limitations
[More Information Needed]
## Additional Information
### Dataset Curators
[More Information Needed]
### Licensing Information
[More Information Needed]
### Citation Information
[More Information Needed]
### Contributions
[More Information Needed]
|
open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w-q_k_v_o_proj
|
[
"region:us"
] |
2023-10-23T04:27:17+00:00
|
{"pretty_name": "Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj", "dataset_summary": "Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\nTo load the details from a run, you can for instance do the following:\n```python\nfrom datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w-q_k_v_o_proj\",\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\nThese are the [latest results from run 2023-10-23T05:27:14.109658](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w-q_k_v_o_proj/blob/main/results_2023-10-23T05-27-14.109658.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.3719588926174497,\n \"em_stderr\": 0.004949726013193945,\n \"f1\": 0.4084679110738261,\n \"f1_stderr\": 0.004843145937750956,\n \"acc\": 0.4480415851620389,\n \"acc_stderr\": 0.010535274120903989\n },\n \"harness|drop|3\": {\n \"em\": 0.3719588926174497,\n \"em_stderr\": 0.004949726013193945,\n \"f1\": 0.4084679110738261,\n \"f1_stderr\": 0.004843145937750956\n },\n \"harness|gsm8k|5\": {\n \"acc\": 0.1281273692191054,\n \"acc_stderr\": 0.009206398549980031\n },\n \"harness|winogrande|5\": {\n \"acc\": 0.7679558011049724,\n \"acc_stderr\": 0.011864149691827948\n }\n}\n```", "repo_url": "https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj", "leaderboard_url": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard", "point_of_contact": "[email protected]", "configs": [{"config_name": "harness_drop_3", "data_files": [{"split": "2023_10_23T05_27_14.109658", "path": ["**/details_harness|drop|3_2023-10-23T05-27-14.109658.parquet"]}, {"split": "latest", "path": ["**/details_harness|drop|3_2023-10-23T05-27-14.109658.parquet"]}]}, {"config_name": "harness_gsm8k_5", "data_files": [{"split": "2023_10_23T05_27_14.109658", "path": ["**/details_harness|gsm8k|5_2023-10-23T05-27-14.109658.parquet"]}, {"split": "latest", "path": ["**/details_harness|gsm8k|5_2023-10-23T05-27-14.109658.parquet"]}]}, {"config_name": "harness_winogrande_5", "data_files": [{"split": "2023_10_23T05_27_14.109658", "path": ["**/details_harness|winogrande|5_2023-10-23T05-27-14.109658.parquet"]}, {"split": "latest", "path": ["**/details_harness|winogrande|5_2023-10-23T05-27-14.109658.parquet"]}]}, {"config_name": "results", "data_files": [{"split": "2023_10_23T05_27_14.109658", "path": ["results_2023-10-23T05-27-14.109658.parquet"]}, {"split": "latest", "path": ["results_2023-10-23T05-27-14.109658.parquet"]}]}]}
|
2023-10-23T04:27:24+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj
## Dataset Description
- Homepage:
- Repository: URL
- Paper:
- Leaderboard: URL
- Point of Contact: clementine@URL
### Dataset Summary
Dataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj on the Open LLM Leaderboard.
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).
To load the details from a run, you can for instance do the following:
## Latest results
These are the latest results from run 2023-10-23T05:27:14.109658(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
### Supported Tasks and Leaderboards
### Languages
## Dataset Structure
### Data Instances
### Data Fields
### Data Splits
## Dataset Creation
### Curation Rationale
### Source Data
#### Initial Data Collection and Normalization
#### Who are the source language producers?
### Annotations
#### Annotation process
#### Who are the annotators?
### Personal and Sensitive Information
## Considerations for Using the Data
### Social Impact of Dataset
### Discussion of Biases
### Other Known Limitations
## Additional Information
### Dataset Curators
### Licensing Information
### Contributions
|
[
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T05:27:14.109658(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T05:27:14.109658(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
6,
39,
31,
187,
66,
10,
4,
6,
6,
5,
5,
5,
7,
4,
10,
10,
5,
5,
9,
8,
8,
7,
8,
7,
5,
6,
6,
5
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w-q_k_v_o_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:## Latest results\n\nThese are the latest results from run 2023-10-23T05:27:14.109658(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):### Supported Tasks and Leaderboards### Languages## Dataset Structure### Data Instances### Data Fields### Data Splits## Dataset Creation### Curation Rationale### Source Data#### Initial Data Collection and Normalization#### Who are the source language producers?### Annotations#### Annotation process#### Who are the annotators?### Personal and Sensitive Information## Considerations for Using the Data### Social Impact of Dataset### Discussion of Biases### Other Known Limitations## Additional Information### Dataset Curators### Licensing Information### Contributions"
] |
9fa9f7cc1d880c4883c7b71cfa496f302b68882f
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w
## Dataset Description
- **Homepage:**
- **Repository:** https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w
- **Paper:**
- **Leaderboard:** https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
- **Point of Contact:** [email protected]
### Dataset Summary
Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE2_3w](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).
To load the details from a run, you can for instance do the following:
```python
from datasets import load_dataset
data = load_dataset("open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w",
"harness_winogrande_5",
split="train")
```
## Latest results
These are the [latest results from run 2023-10-23T05:30:33.523762](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w/blob/main/results_2023-10-23T05-30-33.523762.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
```python
{
"all": {
"em": 0.19316275167785235,
"em_stderr": 0.004042912227684817,
"f1": 0.2405106963087243,
"f1_stderr": 0.004012764038516629,
"acc": 0.44387175953656505,
"acc_stderr": 0.010404181547690496
},
"harness|drop|3": {
"em": 0.19316275167785235,
"em_stderr": 0.004042912227684817,
"f1": 0.2405106963087243,
"f1_stderr": 0.004012764038516629
},
"harness|gsm8k|5": {
"acc": 0.1197877179681577,
"acc_stderr": 0.008944213403553045
},
"harness|winogrande|5": {
"acc": 0.7679558011049724,
"acc_stderr": 0.011864149691827948
}
}
```
### Supported Tasks and Leaderboards
[More Information Needed]
### Languages
[More Information Needed]
## Dataset Structure
### Data Instances
[More Information Needed]
### Data Fields
[More Information Needed]
### Data Splits
[More Information Needed]
## Dataset Creation
### Curation Rationale
[More Information Needed]
### Source Data
#### Initial Data Collection and Normalization
[More Information Needed]
#### Who are the source language producers?
[More Information Needed]
### Annotations
#### Annotation process
[More Information Needed]
#### Who are the annotators?
[More Information Needed]
### Personal and Sensitive Information
[More Information Needed]
## Considerations for Using the Data
### Social Impact of Dataset
[More Information Needed]
### Discussion of Biases
[More Information Needed]
### Other Known Limitations
[More Information Needed]
## Additional Information
### Dataset Curators
[More Information Needed]
### Licensing Information
[More Information Needed]
### Citation Information
[More Information Needed]
### Contributions
[More Information Needed]
|
open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w
|
[
"region:us"
] |
2023-10-23T04:30:37+00:00
|
{"pretty_name": "Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w", "dataset_summary": "Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE2_3w](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\nTo load the details from a run, you can for instance do the following:\n```python\nfrom datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w\",\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\nThese are the [latest results from run 2023-10-23T05:30:33.523762](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w/blob/main/results_2023-10-23T05-30-33.523762.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.19316275167785235,\n \"em_stderr\": 0.004042912227684817,\n \"f1\": 0.2405106963087243,\n \"f1_stderr\": 0.004012764038516629,\n \"acc\": 0.44387175953656505,\n \"acc_stderr\": 0.010404181547690496\n },\n \"harness|drop|3\": {\n \"em\": 0.19316275167785235,\n \"em_stderr\": 0.004042912227684817,\n \"f1\": 0.2405106963087243,\n \"f1_stderr\": 0.004012764038516629\n },\n \"harness|gsm8k|5\": {\n \"acc\": 0.1197877179681577,\n \"acc_stderr\": 0.008944213403553045\n },\n \"harness|winogrande|5\": {\n \"acc\": 0.7679558011049724,\n \"acc_stderr\": 0.011864149691827948\n }\n}\n```", "repo_url": "https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w", "leaderboard_url": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard", "point_of_contact": "[email protected]", "configs": [{"config_name": "harness_drop_3", "data_files": [{"split": "2023_10_23T05_30_33.523762", "path": ["**/details_harness|drop|3_2023-10-23T05-30-33.523762.parquet"]}, {"split": "latest", "path": ["**/details_harness|drop|3_2023-10-23T05-30-33.523762.parquet"]}]}, {"config_name": "harness_gsm8k_5", "data_files": [{"split": "2023_10_23T05_30_33.523762", "path": ["**/details_harness|gsm8k|5_2023-10-23T05-30-33.523762.parquet"]}, {"split": "latest", "path": ["**/details_harness|gsm8k|5_2023-10-23T05-30-33.523762.parquet"]}]}, {"config_name": "harness_winogrande_5", "data_files": [{"split": "2023_10_23T05_30_33.523762", "path": ["**/details_harness|winogrande|5_2023-10-23T05-30-33.523762.parquet"]}, {"split": "latest", "path": ["**/details_harness|winogrande|5_2023-10-23T05-30-33.523762.parquet"]}]}, {"config_name": "results", "data_files": [{"split": "2023_10_23T05_30_33.523762", "path": ["results_2023-10-23T05-30-33.523762.parquet"]}, {"split": "latest", "path": ["results_2023-10-23T05-30-33.523762.parquet"]}]}]}
|
2023-10-23T04:30:44+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w
## Dataset Description
- Homepage:
- Repository: URL
- Paper:
- Leaderboard: URL
- Point of Contact: clementine@URL
### Dataset Summary
Dataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w on the Open LLM Leaderboard.
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).
To load the details from a run, you can for instance do the following:
## Latest results
These are the latest results from run 2023-10-23T05:30:33.523762(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
### Supported Tasks and Leaderboards
### Languages
## Dataset Structure
### Data Instances
### Data Fields
### Data Splits
## Dataset Creation
### Curation Rationale
### Source Data
#### Initial Data Collection and Normalization
#### Who are the source language producers?
### Annotations
#### Annotation process
#### Who are the annotators?
### Personal and Sensitive Information
## Considerations for Using the Data
### Social Impact of Dataset
### Discussion of Biases
### Other Known Limitations
## Additional Information
### Dataset Curators
### Licensing Information
### Contributions
|
[
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T05:30:33.523762(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T05:30:33.523762(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
6,
28,
31,
176,
67,
10,
4,
6,
6,
5,
5,
5,
7,
4,
10,
10,
5,
5,
9,
8,
8,
7,
8,
7,
5,
6,
6,
5
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:## Latest results\n\nThese are the latest results from run 2023-10-23T05:30:33.523762(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):### Supported Tasks and Leaderboards### Languages## Dataset Structure### Data Instances### Data Fields### Data Splits## Dataset Creation### Curation Rationale### Source Data#### Initial Data Collection and Normalization#### Who are the source language producers?### Annotations#### Annotation process#### Who are the annotators?### Personal and Sensitive Information## Considerations for Using the Data### Social Impact of Dataset### Discussion of Biases### Other Known Limitations## Additional Information### Dataset Curators### Licensing Information### Contributions"
] |
45e37659738f66f867ece1ab3604456aaa3e4514
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_complex_qa_ground_onetime` is the data for training **grounding** module on **complex QA** task in **Lumos-Onetime (Lumos-O)** formulation.
The source of the training annotation training data is shown below:
| Datasets | Number |
|---|---|
|StrategyQA|1777|
|Musique|17632|
## Models Trained with the Data
`lumos_complex_qa_ground_onetime` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_complex_qa_ground_onetime`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_complex_qa_ground_onetime) |
|`lumos_unified_ground_onetime`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_ground_onetime) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_complex_qa_ground_onetime
|
[
"task_categories:conversational",
"task_categories:text-generation",
"task_categories:question-answering",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"language-agent",
"reasoning",
"question-answering",
"grounding",
"region:us"
] |
2023-10-23T04:31:31+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["conversational", "text-generation", "question-answering"], "tags": ["language-agent", "reasoning", "question-answering", "grounding"]}
|
2023-10-26T05:00:28+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #grounding #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_complex\_qa\_ground\_onetime' is the data for training grounding module on complex QA task in Lumos-Onetime (Lumos-O) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_complex\_qa\_ground\_onetime' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #grounding #region-us \n"
] |
[
80
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #grounding #region-us \n"
] |
4c66097a4b990a4ad218c5bef2ee1de6bdd4e804
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_complex_qa_ground_iterative` is the data for training **grounding** module on **complex QA** task in **Lumos-Iterative (Lumos-I)** formulation.
The source of the training annotation training data is shown below:
| Datasets | Number |
|---|---|
|StrategyQA|1777|
|Musique|17632|
## Models Trained with the Data
`lumos_complex_qa_ground_iterative` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_complex_qa_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_complex_qa_ground_iterative) |
|`lumos_unified_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_ground_iterative) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_complex_qa_ground_iterative
|
[
"task_categories:conversational",
"task_categories:text-generation",
"task_categories:question-answering",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"language-agent",
"reasoning",
"question-answering",
"grounding",
"region:us"
] |
2023-10-23T04:33:52+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["conversational", "text-generation", "question-answering"], "tags": ["language-agent", "reasoning", "question-answering", "grounding"]}
|
2023-10-26T04:59:30+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #grounding #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_complex\_qa\_ground\_iterative' is the data for training grounding module on complex QA task in Lumos-Iterative (Lumos-I) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_complex\_qa\_ground\_iterative' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #grounding #region-us \n"
] |
[
80
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #grounding #region-us \n"
] |
64dfb3c42090ef7ee7bf2d25cc033c1b0bfdfaeb
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_complex_qa_plan_iterative` is the data for training **planning** module on **complex QA** task in **Lumos-Iterative (Lumos-I)** formulation.
The source of the training annotation training data is shown below:
| Datasets | Number |
|---|---|
|StrategyQA|1777|
|Musique|17632|
## Models Trained with the Data
`lumos_complex_qa_plan_iterative` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_complex_qa_plan_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_complex_qa_plan_iterative) |
|`lumos_unified_plan_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_plan_iterative) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_complex_qa_plan_iterative
|
[
"task_categories:conversational",
"task_categories:text-generation",
"task_categories:question-answering",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"language-agent",
"reasoning",
"question-answering",
"planning",
"region:us"
] |
2023-10-23T04:35:41+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["conversational", "text-generation", "question-answering"], "tags": ["language-agent", "reasoning", "question-answering", "planning"]}
|
2023-10-23T21:35:57+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #planning #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_complex\_qa\_plan\_iterative' is the data for training planning module on complex QA task in Lumos-Iterative (Lumos-I) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_complex\_qa\_plan\_iterative' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #planning #region-us \n"
] |
[
80
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #planning #region-us \n"
] |
0910bf29760c18575dbe36358e3920d5c306a5e6
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_complex_qa_plan_onetime` is the data for training **planning** module on **complex QA** task in **Lumos-Onetime (Lumos-O)** formulation.
The source of the training annotation training data is shown below:
| Datasets | Number |
|---|---|
|StrategyQA|1777|
|Musique|17632|
## Models Trained with the Data
`lumos_complex_qa_plan_onetime` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_complex_qa_plan_onetime`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_complex_qa_plan_onetime) |
|`lumos_unified_plan_onetime`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_plan_onetime) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_complex_qa_plan_onetime
|
[
"task_categories:conversational",
"task_categories:text-generation",
"task_categories:question-answering",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"language-agent",
"reasoning",
"question-answering",
"planning",
"region:us"
] |
2023-10-23T04:36:48+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["conversational", "text-generation", "question-answering"], "tags": ["language-agent", "reasoning", "question-answering", "planning"]}
|
2023-10-23T21:34:12+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #planning #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_complex\_qa\_plan\_onetime' is the data for training planning module on complex QA task in Lumos-Onetime (Lumos-O) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_complex\_qa\_plan\_onetime' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #planning #region-us \n"
] |
[
80
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #reasoning #question-answering #planning #region-us \n"
] |
c66601379060872d5632471da4ffa1953b181eef
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_unified_plan_iterative` is the data for training **planning** module on **maths**, **complex QA** and **web agent** tasks in **Lumos-Iterative (Lumos-I)** formulation.
The source of the training annotation training data is shown below:
| Task | Number |
|---|---|
|PRM800K|10000|
|GSM8K|7473|
|ASDiv|2305|
|StrategyQA|1777|
|Musique|17632|
|Mind2Web|1009|
## Models Trained with the Data
`lumos_unified_plan_iterative` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_unified_plan_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_plan_iterative) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_unified_plan_iterative
|
[
"task_categories:conversational",
"task_categories:text-generation",
"task_categories:question-answering",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"language-agent",
"maths",
"reasoning",
"question-answering",
"web-agent",
"planning",
"region:us"
] |
2023-10-23T04:38:03+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["conversational", "text-generation", "question-answering"], "tags": ["language-agent", "maths", "reasoning", "question-answering", "web-agent", "planning"]}
|
2023-10-23T21:27:04+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #question-answering #web-agent #planning #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_unified\_plan\_iterative' is the data for training planning module on maths, complex QA and web agent tasks in Lumos-Iterative (Lumos-I) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_unified\_plan\_iterative' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #question-answering #web-agent #planning #region-us \n"
] |
[
87
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #question-answering #web-agent #planning #region-us \n"
] |
d8d8fbd3a2b076eb1ac3cc85c3f124a715b754ee
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_unified_ground_iterative` is the data for training **grounding** module on **maths**, **complex QA** and **web agent** tasks in **Lumos-Iterative (Lumos-I)** formulation.
The source of the training annotation training data is shown below:
| Task | Number |
|---|---|
|PRM800K|10000|
|GSM8K|7473|
|ASDiv|2305|
|StrategyQA|1777|
|Musique|17632|
|Mind2Web|1009|
## Models Trained with the Data
`lumos_unified_ground_iterative` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_unified_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_ground_iterative) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_unified_ground_iterative
|
[
"task_categories:conversational",
"task_categories:text-generation",
"task_categories:question-answering",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"language-agent",
"maths",
"reasoning",
"question-answering",
"web-agent",
"grounding",
"region:us"
] |
2023-10-23T04:39:02+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["conversational", "text-generation", "question-answering"], "tags": ["language-agent", "maths", "reasoning", "question-answering", "web-agent", "grounding"]}
|
2023-10-26T05:06:47+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #question-answering #web-agent #grounding #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_unified\_ground\_iterative' is the data for training grounding module on maths, complex QA and web agent tasks in Lumos-Iterative (Lumos-I) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_unified\_ground\_iterative' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #question-answering #web-agent #grounding #region-us \n"
] |
[
87
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #task_categories-question-answering #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #question-answering #web-agent #grounding #region-us \n"
] |
b8698ac6745692b7b8868380552a0175fb67b150
|
# Dataset Card for "sharegpt-binary"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
atmallen/sharegpt-binary
|
[
"region:us"
] |
2023-10-23T04:40:21+00:00
|
{"configs": [{"config_name": "default", "data_files": [{"split": "test", "path": "data/test-*"}]}], "dataset_info": {"features": [{"name": "statement", "dtype": "string"}, {"name": "label", "dtype": {"class_label": {"names": {"0": "false", "1": "true"}}}}, {"name": "model", "dtype": "string"}], "splits": [{"name": "test", "num_bytes": 1090167, "num_examples": 243}], "download_size": 188810, "dataset_size": 1090167}}
|
2023-10-23T20:50:35+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "sharegpt-binary"
More Information needed
|
[
"# Dataset Card for \"sharegpt-binary\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"sharegpt-binary\"\n\nMore Information needed"
] |
[
6,
16
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"sharegpt-binary\"\n\nMore Information needed"
] |
3b4c1b5034d21026946c52cb347981ebc5d521a1
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj
## Dataset Description
- **Homepage:**
- **Repository:** https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj
- **Paper:**
- **Leaderboard:** https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
- **Point of Contact:** [email protected]
### Dataset Summary
Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).
To load the details from a run, you can for instance do the following:
```python
from datasets import load_dataset
data = load_dataset("open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w-gate_up_down_proj",
"harness_winogrande_5",
split="train")
```
## Latest results
These are the [latest results from run 2023-10-23T05:41:52.177937](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w-gate_up_down_proj/blob/main/results_2023-10-23T05-41-52.177937.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
```python
{
"all": {
"em": 0.08137583892617449,
"em_stderr": 0.0027999889835206245,
"f1": 0.13315016778523447,
"f1_stderr": 0.0029419319985989354,
"acc": 0.4454347335673805,
"acc_stderr": 0.010395126943573653
},
"harness|drop|3": {
"em": 0.08137583892617449,
"em_stderr": 0.0027999889835206245,
"f1": 0.13315016778523447,
"f1_stderr": 0.0029419319985989354
},
"harness|gsm8k|5": {
"acc": 0.12054586808188021,
"acc_stderr": 0.008968608285309067
},
"harness|winogrande|5": {
"acc": 0.7703235990528808,
"acc_stderr": 0.011821645601838238
}
}
```
### Supported Tasks and Leaderboards
[More Information Needed]
### Languages
[More Information Needed]
## Dataset Structure
### Data Instances
[More Information Needed]
### Data Fields
[More Information Needed]
### Data Splits
[More Information Needed]
## Dataset Creation
### Curation Rationale
[More Information Needed]
### Source Data
#### Initial Data Collection and Normalization
[More Information Needed]
#### Who are the source language producers?
[More Information Needed]
### Annotations
#### Annotation process
[More Information Needed]
#### Who are the annotators?
[More Information Needed]
### Personal and Sensitive Information
[More Information Needed]
## Considerations for Using the Data
### Social Impact of Dataset
[More Information Needed]
### Discussion of Biases
[More Information Needed]
### Other Known Limitations
[More Information Needed]
## Additional Information
### Dataset Curators
[More Information Needed]
### Licensing Information
[More Information Needed]
### Citation Information
[More Information Needed]
### Contributions
[More Information Needed]
|
open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w-gate_up_down_proj
|
[
"region:us"
] |
2023-10-23T04:41:56+00:00
|
{"pretty_name": "Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj", "dataset_summary": "Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\nTo load the details from a run, you can for instance do the following:\n```python\nfrom datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w-gate_up_down_proj\",\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\nThese are the [latest results from run 2023-10-23T05:41:52.177937](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE2_3w-gate_up_down_proj/blob/main/results_2023-10-23T05-41-52.177937.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.08137583892617449,\n \"em_stderr\": 0.0027999889835206245,\n \"f1\": 0.13315016778523447,\n \"f1_stderr\": 0.0029419319985989354,\n \"acc\": 0.4454347335673805,\n \"acc_stderr\": 0.010395126943573653\n },\n \"harness|drop|3\": {\n \"em\": 0.08137583892617449,\n \"em_stderr\": 0.0027999889835206245,\n \"f1\": 0.13315016778523447,\n \"f1_stderr\": 0.0029419319985989354\n },\n \"harness|gsm8k|5\": {\n \"acc\": 0.12054586808188021,\n \"acc_stderr\": 0.008968608285309067\n },\n \"harness|winogrande|5\": {\n \"acc\": 0.7703235990528808,\n \"acc_stderr\": 0.011821645601838238\n }\n}\n```", "repo_url": "https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj", "leaderboard_url": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard", "point_of_contact": "[email protected]", "configs": [{"config_name": "harness_drop_3", "data_files": [{"split": "2023_10_23T05_41_52.177937", "path": ["**/details_harness|drop|3_2023-10-23T05-41-52.177937.parquet"]}, {"split": "latest", "path": ["**/details_harness|drop|3_2023-10-23T05-41-52.177937.parquet"]}]}, {"config_name": "harness_gsm8k_5", "data_files": [{"split": "2023_10_23T05_41_52.177937", "path": ["**/details_harness|gsm8k|5_2023-10-23T05-41-52.177937.parquet"]}, {"split": "latest", "path": ["**/details_harness|gsm8k|5_2023-10-23T05-41-52.177937.parquet"]}]}, {"config_name": "harness_winogrande_5", "data_files": [{"split": "2023_10_23T05_41_52.177937", "path": ["**/details_harness|winogrande|5_2023-10-23T05-41-52.177937.parquet"]}, {"split": "latest", "path": ["**/details_harness|winogrande|5_2023-10-23T05-41-52.177937.parquet"]}]}, {"config_name": "results", "data_files": [{"split": "2023_10_23T05_41_52.177937", "path": ["results_2023-10-23T05-41-52.177937.parquet"]}, {"split": "latest", "path": ["results_2023-10-23T05-41-52.177937.parquet"]}]}]}
|
2023-10-23T04:42:03+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj
## Dataset Description
- Homepage:
- Repository: URL
- Paper:
- Leaderboard: URL
- Point of Contact: clementine@URL
### Dataset Summary
Dataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj on the Open LLM Leaderboard.
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).
To load the details from a run, you can for instance do the following:
## Latest results
These are the latest results from run 2023-10-23T05:41:52.177937(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
### Supported Tasks and Leaderboards
### Languages
## Dataset Structure
### Data Instances
### Data Fields
### Data Splits
## Dataset Creation
### Curation Rationale
### Source Data
#### Initial Data Collection and Normalization
#### Who are the source language producers?
### Annotations
#### Annotation process
#### Who are the annotators?
### Personal and Sensitive Information
## Considerations for Using the Data
### Social Impact of Dataset
### Discussion of Biases
### Other Known Limitations
## Additional Information
### Dataset Curators
### Licensing Information
### Contributions
|
[
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T05:41:52.177937(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T05:41:52.177937(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
6,
37,
31,
185,
67,
10,
4,
6,
6,
5,
5,
5,
7,
4,
10,
10,
5,
5,
9,
8,
8,
7,
8,
7,
5,
6,
6,
5
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE2_3w-gate_up_down_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:## Latest results\n\nThese are the latest results from run 2023-10-23T05:41:52.177937(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):### Supported Tasks and Leaderboards### Languages## Dataset Structure### Data Instances### Data Fields### Data Splits## Dataset Creation### Curation Rationale### Source Data#### Initial Data Collection and Normalization#### Who are the source language producers?### Annotations#### Annotation process#### Who are the annotators?### Personal and Sensitive Information## Considerations for Using the Data### Social Impact of Dataset### Discussion of Biases### Other Known Limitations## Additional Information### Dataset Curators### Licensing Information### Contributions"
] |
14365aee2bb5acf3d9f250aebb5b220d0756ac96
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_maths_ground_iterative` is the data for training **grounding** module on **maths** task in **Lumos-Iterative (Lumos-I)** formulation.
The source of the training annotation training data is shown below:
| Task | Number |
|---|---|
|PRM800K|10000|
|GSM8K|7473|
|ASDiv|2305|
## Models Trained with the Data
`lumos_maths_ground_iterative` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_maths_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_maths_ground_iterative) |
|`lumos_unified_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_ground_iterative) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_maths_ground_iterative
|
[
"task_categories:conversational",
"task_categories:text-generation",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"language-agent",
"maths",
"reasoning",
"grounding",
"region:us"
] |
2023-10-23T04:42:36+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["conversational", "text-generation"], "tags": ["language-agent", "maths", "reasoning", "grounding"]}
|
2023-10-23T21:19:25+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #grounding #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_maths\_ground\_iterative' is the data for training grounding module on maths task in Lumos-Iterative (Lumos-I) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_maths\_ground\_iterative' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #grounding #region-us \n"
] |
[
65
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #grounding #region-us \n"
] |
4b5fa0e01ca7188f22433e70c1f159b2c8e25b48
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_maths_plan_iterative` is the data for training **planning** module on **maths** task in **Lumos-Iterative (Lumos-I)** formulation.
The source of the training annotation training data is shown below:
| Task | Number |
|---|---|
|PRM800K|10000|
|GSM8K|7473|
|ASDiv|2305|
## Models Trained with the Data
`lumos_maths_plan_iterative` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_maths_plan_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_maths_plan_iterative) |
|`lumos_unified_plan_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_plan_iterative) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_maths_plan_iterative
|
[
"task_categories:conversational",
"task_categories:text-generation",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"language-agent",
"maths",
"reasoning",
"planning",
"region:us"
] |
2023-10-23T04:43:59+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["conversational", "text-generation"], "tags": ["language-agent", "maths", "reasoning", "planning"]}
|
2023-10-23T21:20:52+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #planning #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_maths\_plan\_iterative' is the data for training planning module on maths task in Lumos-Iterative (Lumos-I) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_maths\_plan\_iterative' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #planning #region-us \n"
] |
[
65
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #planning #region-us \n"
] |
c2efa514789e2b9d990d77bec7711955a6b7174a
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_maths_ground_onetime` is the data for training **grounding** module on **maths** task in **Lumos-Onetime (Lumos-O)** formulation.
The source of the training annotation training data is shown below:
| Task | Number |
|---|---|
|PRM800K|10000|
|GSM8K|7473|
|ASDiv|2305|
## Models Trained with the Data
`lumos_maths_ground_onetime` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_maths_ground_onetime`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_maths_ground_onetime) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_maths_ground_onetime
|
[
"task_categories:conversational",
"task_categories:text-generation",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"language-agent",
"maths",
"reasoning",
"grounding",
"region:us"
] |
2023-10-23T04:45:30+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["conversational", "text-generation"], "tags": ["language-agent", "maths", "reasoning", "grounding"]}
|
2023-10-23T21:15:31+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #grounding #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_maths\_ground\_onetime' is the data for training grounding module on maths task in Lumos-Onetime (Lumos-O) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_maths\_ground\_onetime' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #grounding #region-us \n"
] |
[
65
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #grounding #region-us \n"
] |
b38e09eea87d01b275fdddacf164c24f103772ed
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_maths_plan_onetime` is the data for training **planning** module on **maths** task in **Lumos-Onetime (Lumos-O)** formulation.
The source of the training annotation training data is shown below:
| Task | Number |
|---|---|
|PRM800K|10000|
|GSM8K|7473|
|ASDiv|2305|
## Models Trained with the Data
`lumos_maths_plan_onetime` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_maths_plan_onetime`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_maths_plan_onetime) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_maths_plan_onetime
|
[
"task_categories:conversational",
"task_categories:text-generation",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"language-agent",
"maths",
"reasoning",
"region:us"
] |
2023-10-23T04:46:41+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["conversational", "text-generation"], "tags": ["language-agent", "maths", "reasoning"]}
|
2023-10-23T21:13:54+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_maths\_plan\_onetime' is the data for training planning module on maths task in Lumos-Onetime (Lumos-O) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_maths\_plan\_onetime' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #region-us \n"
] |
[
62
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #language-agent #maths #reasoning #region-us \n"
] |
bcd7fa3b452a14289ef532df6187f0cfe55592ce
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_web_agent_plan_iterative` is the data for training **planning** module on **web agent** task in **Lumos-Iterative (Lumos-I)** formulation.
The source of the training annotation training data is shown below:
| Task | Number |
|---|---|
|Mind2Web|1009|
## Models Trained with the Data
`lumos_web_agent_plan_iterative` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_web_agent_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_web_agent_plan_iterative) |
|`lumos_unified_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_plan_iterative) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_web_agent_plan_iterative
|
[
"task_categories:conversational",
"task_categories:text-generation",
"size_categories:1K<n<10K",
"language:en",
"license:apache-2.0",
"language-agent",
"web-agent",
"reasoning",
"planning",
"region:us"
] |
2023-10-23T04:48:14+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["1K<n<10K"], "task_categories": ["conversational", "text-generation"], "tags": ["language-agent", "web-agent", "reasoning", "planning"]}
|
2023-10-23T21:37:14+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #size_categories-1K<n<10K #language-English #license-apache-2.0 #language-agent #web-agent #reasoning #planning #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_web\_agent\_plan\_iterative' is the data for training planning module on web agent task in Lumos-Iterative (Lumos-I) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_web\_agent\_plan\_iterative' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-1K<n<10K #language-English #license-apache-2.0 #language-agent #web-agent #reasoning #planning #region-us \n"
] |
[
66
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-1K<n<10K #language-English #license-apache-2.0 #language-agent #web-agent #reasoning #planning #region-us \n"
] |
185041ca57422eb76ea8e9facba4c9e73cedcf71
|
# 🪄 Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
<p align="center">
🌐<a href="https://allenai.github.io/lumos">[Website]</a>
📝<a href="">[Paper]</a>
🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a>
🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a>
</p>
We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
**Lumos** has following features:
* 🧩 **Modular Architecture**:
- **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* 🌍 **Diverse Training Data**:
- **Lumos** is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* 🚀 **Competitive Performance**:
- 🚀 **Lumos** outperforms **GPT-4/3.5-based** agents on complex QA and web agent tasks, and **larger open agents** on maths tasks.
- 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **unmodularized** training.
- 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
## Data Overview
`lumos_web_agent_ground_iterative` is the data for training **grounding** module on **web agent** task in **Lumos-Iterative (Lumos-I)** formulation.
The source of the training annotation training data is shown below:
| Task | Number |
|---|---|
|Mind2Web|1009|
## Models Trained with the Data
`lumos_web_agent_ground_iterative` is used to train the following models.
|Model|Huggingface Repo|
|---|---|
|`lumos_web_agent_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_web_agent_ground_iterative) |
|`lumos_unified_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_ground_iterative) |
## Citation
If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
title={Lumos: Towards Language Agents that are Unified, Modular, and Open Source},
author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
year={2023}
}
```
|
ai2lumos/lumos_web_agent_ground_iterative
|
[
"task_categories:conversational",
"task_categories:text-generation",
"size_categories:1K<n<10K",
"language:en",
"license:apache-2.0",
"language-agent",
"web-agent",
"reasoning",
"grounding",
"region:us"
] |
2023-10-23T04:48:48+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["1K<n<10K"], "task_categories": ["conversational", "text-generation"], "tags": ["language-agent", "web-agent", "reasoning", "grounding"]}
|
2023-10-23T21:37:54+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-conversational #task_categories-text-generation #size_categories-1K<n<10K #language-English #license-apache-2.0 #language-agent #web-agent #reasoning #grounding #region-us
|
Lumos: Language Agents with Unified Formats, Modular Design, and Open-Source LLMs
=================================================================================
[[Paper]](URL
<a href=)
[We introduce Lumos, Language Agents with Unified Formats, Modular Design, and Open-Source LLMs. Lumos unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents.
Lumos has following features:
* Modular Architecture:
* Lumos consists of planning, grounding, and execution modules built based on LLAMA-2-7B.
* Diverse Training Data:
* Lumos is trained with ~40K high-quality annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
* Competitive Performance:
* Lumos outperforms GPT-4/3.5-based agents on complex QA and web agent tasks, and larger open agents on maths tasks.
* Lumos performs better than open agent baseline formulations including chain-of-thoughts and unmodularized training.
* Lumos surpasses larger open LLM agents and domain-specific agents on an unseen task, WebShop.
Data Overview
-------------
'lumos\_web\_agent\_ground\_iterative' is the data for training grounding module on web agent task in Lumos-Iterative (Lumos-I) formulation.
The source of the training annotation training data is shown below:
Models Trained with the Data
----------------------------
'lumos\_web\_agent\_ground\_iterative' is used to train the following models.
If you find this work is relevant with your research, please feel free to cite our work!](URL
<a href=)
|
[] |
[
"TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-1K<n<10K #language-English #license-apache-2.0 #language-agent #web-agent #reasoning #grounding #region-us \n"
] |
[
66
] |
[
"passage: TAGS\n#task_categories-conversational #task_categories-text-generation #size_categories-1K<n<10K #language-English #license-apache-2.0 #language-agent #web-agent #reasoning #grounding #region-us \n"
] |
9fa37f0f0f0caa621f67f98c1a8b73a6de0d4102
|
# Dataset Card for "OccuQuest"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
mingfengxue/OccuQuest
|
[
"region:us"
] |
2023-10-23T04:56:00+00:00
|
{"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "validation", "path": "data/validation-*"}, {"split": "test", "path": "data/test-*"}, {"split": "estate", "path": "data/estate-*"}, {"split": "quora", "path": "data/quora-*"}]}], "dataset_info": {"features": [{"name": "category", "dtype": "string"}, {"name": "occupation", "dtype": "string"}, {"name": "topic", "dtype": "string"}, {"name": "messages", "list": [{"name": "role", "dtype": "string"}, {"name": "content", "dtype": "string"}]}], "splits": [{"name": "train", "num_bytes": 330314955, "num_examples": 114090}, {"name": "validation", "num_bytes": 7314741, "num_examples": 2500}, {"name": "test", "num_bytes": 718046, "num_examples": 250}, {"name": "estate", "num_bytes": 703613, "num_examples": 250}, {"name": "quora", "num_bytes": 45540, "num_examples": 250}], "download_size": 139074820, "dataset_size": 339096895}}
|
2023-10-23T05:17:07+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "OccuQuest"
More Information needed
|
[
"# Dataset Card for \"OccuQuest\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"OccuQuest\"\n\nMore Information needed"
] |
[
6,
14
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"OccuQuest\"\n\nMore Information needed"
] |
f30f4464c908aeeaf727c9c75ac3a517567a2f80
|
Are you on a quest for effective weight loss? Look no further than the **BodyRock Weight Loss Program**. This comprehensive article will provide insights into the program, its workings, benefits, and more.
[Click Here To Get It From Official Website](https://snoppymart.com/get-bodyrock-weight-loss-program)
What is the BodyRock Weight Loss Program?
-----------------------------------------
The BodyRock Weight Loss Program is a holistic approach to shedding unwanted pounds and improving your overall fitness. It combines structured workout plans with nutrition guidance and a supportive community to help you achieve your weight loss goals.
[.png)](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg7jsS74t__JEGqdZnwC7XrRBc7-de0S526lEFpxhOw1Sap1uEUSOq9rAP9mL97KQlTg6XnR_6gBDz8FEzQF5CPL2vaQBNLVemF4ArL-dnqNHiHPvQedO48HGgN0RSKtfQjQlgrVoC0cp8Klc1-rjCu9sN-WX5lg1Flv8fHW3dZfuKxTYXclKC2awrYXnM/s1671/Screenshot%20(1435).png)
How Does the Program Work?
--------------------------
The program works by offering a combination of high-intensity interval training (HIIT) workouts and a carefully curated meal plan. These workouts are designed to boost your metabolism, burn calories, and build lean muscle, while the meal plan focuses on balanced and nutritious eating.
[.png)](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhmRXGW0GCZp-tNiDOEJ46KVF-f-S8H6zlTJsMn7UInzCHTwf48S4QfwpsMmi18JjaVCI940qvn7AxYHcbtAEY8p80Kx3hNkyh01LGUxAQxYT-jBbJ0s7XlnCe2674gJaQuBUPQy6OZzSpTHirQiPLU3AWZXKlpp5M8FV4mGxeIUuo3xs-bFG3nkJWQMYw/s1681/Screenshot%20(1436).png)
Benefits of BodyRock Weight Loss Program
----------------------------------------
* Rapid fat loss
* Increased energy levels
* Enhanced endurance and strength
* Improved overall fitness
* Personalized approach to weight loss
Success Stories
---------------
Many individuals have achieved remarkable results through the BodyRock Weight Loss Program. Their inspiring stories attest to the program's effectiveness and its ability to transform lives.
[.png)](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhVRNGSktDlCwaIRZz_OSJJB5dN_O5iM4WW4ml7uyCaVfHkLeqEwUE3kZiQ9zd0Kl-Mj-FzyZtmlp13muvuYX_YdyIaQm5su9tdWUjsAgrnsSlxbqaZpCZ1byo5wNH3pumkLvyzPbjuUzE4RAgiCyBP7us2pZJJFXHXnhlIMC_3EBrDQ_nuDwYRE0aQmZw/s1751/Screenshot%20(1438).png)
⏩ Tips for Success
To succeed with the program, it's important to stay consistent, stay hydrated, and maintain a positive mindset. A support system can make a significant difference in your journey to a healthier you.
⏩ Customization Options
The program recognizes that one size does not fit all, and offers customization options to cater to individual needs and goals. This personalization ensures that you get the best results.
⏩ Workout Plans
BodyRock provides a wide variety of workout plans that can be tailored to your fitness level and preferences. From beginner to advanced, there's something for everyone.
⏩ Nutrition and Meal Planning
Eating right is a critical component of weight loss, and BodyRock's meal plans are designed to make this process easier. You'll have access to nutritious recipes and guidance on portion control.
⏩ Tracking Progress
Tracking your progress is essential for staying motivated. The program offers tools to monitor your weight loss journey, helping you celebrate your successes.
⏩ Community Support
The BodyRock community is a supportive and encouraging group of like-minded individuals who can offer advice, motivation, and inspiration throughout your journey.
⏩ Expert Guidance
The program is led by experienced trainers and nutritionists who provide expert guidance and answer questions, ensuring you have all the support you need.
[.png)](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgEIQ3QhxhvOsN0CT0YUTHZcZBolCATbt3TYOvmKrTbkpzSlkmrJHTFqWVKsHA5MowHK3g7HrBaqLWFSnCWgsduDFWm8B1Vj6TIdtG0vCk0QF2NzYqZlEgVm3J45XENzIiuWaq038CJjVntx6SGr6PFoDV0Iob76oTrghFYnkFxdPgLXpUatSMd-E8biq4/s1683/Screenshot%20(1437).pnghttps://snoppymart.com/get-bodyrock-weight-loss-program)
Common Questions
----------------
1. Is the program suitable for beginners?
2. How quickly can I expect to see results?
3. Can I do the workouts at home?
4. Is the meal plan customizable?
5. What kind of equipment do I need for the workouts?
Conclusion
----------
The BodyRock Weight Loss Program is a comprehensive solution for those looking to shed pounds and boost their overall fitness. With personalized plans, expert guidance, and a supportive community, it's a proven way to achieve your weight loss goals.
[Click Here To Get It From Official Website](https://snoppymart.com/get-bodyrock-weight-loss-program)
FAQs
----
Is the BodyRock program suitable for people of all fitness levels?
➣ Yes, the program offers workout plans for beginners to advanced levels.
How quickly can I expect to see results?
➣ Results may vary, but many participants notice positive changes within a few weeks.
Can I do the workouts at home?
➣ Absolutely, most of the workouts can be done in the comfort of your home.
Is the meal plan customizable?
➣ Yes, you can customize your meal plan based on your dietary preferences and goals.
In conclusion, the BodyRock Weight Loss Program offers a holistic approach to achieving your weight loss goals. With personalized workout and meal plans, expert guidance, and a supportive community, it's a proven path to a healthier and fitter you. [Start your journey today](https://snoppymart.com/get-bodyrock-weight-loss-program)!
|
bodyrockweightlossprogram/bodyrockweightlossprogram
|
[
"region:us"
] |
2023-10-23T05:01:52+00:00
|
{}
|
2023-10-23T05:06:38+00:00
|
[] |
[] |
TAGS
#region-us
|
Are you on a quest for effective weight loss? Look no further than the BodyRock Weight Loss Program. This comprehensive article will provide insights into the program, its workings, benefits, and more.
Click Here To Get It From Official Website
What is the BodyRock Weight Loss Program?
-----------------------------------------
The BodyRock Weight Loss Program is a holistic approach to shedding unwanted pounds and improving your overall fitness. It combines structured workout plans with nutrition guidance and a supportive community to help you achieve your weight loss goals.
 workouts and a carefully curated meal plan. These workouts are designed to boost your metabolism, burn calories, and build lean muscle, while the meal plan focuses on balanced and nutritious eating.
](URL
⏩ Tips for Success
To succeed with the program, it's important to stay consistent, stay hydrated, and maintain a positive mindset. A support system can make a significant difference in your journey to a healthier you.
⏩ Customization Options
The program recognizes that one size does not fit all, and offers customization options to cater to individual needs and goals. This personalization ensures that you get the best results.
⏩ Workout Plans
BodyRock provides a wide variety of workout plans that can be tailored to your fitness level and preferences. From beginner to advanced, there's something for everyone.
⏩ Nutrition and Meal Planning
Eating right is a critical component of weight loss, and BodyRock's meal plans are designed to make this process easier. You'll have access to nutritious recipes and guidance on portion control.
⏩ Tracking Progress
Tracking your progress is essential for staying motivated. The program offers tools to monitor your weight loss journey, helping you celebrate your successes.
⏩ Community Support
The BodyRock community is a supportive and encouraging group of like-minded individuals who can offer advice, motivation, and inspiration throughout your journey.
⏩ Expert Guidance
The program is led by experienced trainers and nutritionists who provide expert guidance and answer questions, ensuring you have all the support you need.
![.png)](URL/URL
Common Questions
----------------
1. Is the program suitable for beginners?
2. How quickly can I expect to see results?
3. Can I do the workouts at home?
4. Is the meal plan customizable?
5. What kind of equipment do I need for the workouts?
Conclusion
----------
The BodyRock Weight Loss Program is a comprehensive solution for those looking to shed pounds and boost their overall fitness. With personalized plans, expert guidance, and a supportive community, it's a proven way to achieve your weight loss goals.
Click Here To Get It From Official Website
FAQs
----
Is the BodyRock program suitable for people of all fitness levels?
Yes, the program offers workout plans for beginners to advanced levels.
How quickly can I expect to see results?
Results may vary, but many participants notice positive changes within a few weeks.
Can I do the workouts at home?
Absolutely, most of the workouts can be done in the comfort of your home.
Is the meal plan customizable?
Yes, you can customize your meal plan based on your dietary preferences and goals.
In conclusion, the BodyRock Weight Loss Program offers a holistic approach to achieving your weight loss goals. With personalized workout and meal plans, expert guidance, and a supportive community, it's a proven path to a healthier and fitter you. Start your journey today!
|
[] |
[
"TAGS\n#region-us \n"
] |
[
6
] |
[
"passage: TAGS\n#region-us \n"
] |
f6ac55df631a26981fef650616b8e8d684c84d0d
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj
## Dataset Description
- **Homepage:**
- **Repository:** https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj
- **Paper:**
- **Leaderboard:** https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
- **Point of Contact:** [email protected]
### Dataset Summary
Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).
To load the details from a run, you can for instance do the following:
```python
from datasets import load_dataset
data = load_dataset("open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w-gate_up_down_proj",
"harness_winogrande_5",
split="train")
```
## Latest results
These are the [latest results from run 2023-10-23T06:02:22.448569](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w-gate_up_down_proj/blob/main/results_2023-10-23T06-02-22.448569.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
```python
{
"all": {
"em": 0.10098573825503356,
"em_stderr": 0.0030856947694457384,
"f1": 0.15267407718120746,
"f1_stderr": 0.0031959753495490175,
"acc": 0.4472367612449459,
"acc_stderr": 0.010567855433819127
},
"harness|drop|3": {
"em": 0.10098573825503356,
"em_stderr": 0.0030856947694457384,
"f1": 0.15267407718120746,
"f1_stderr": 0.0031959753495490175
},
"harness|gsm8k|5": {
"acc": 0.1288855193328279,
"acc_stderr": 0.009229580761400269
},
"harness|winogrande|5": {
"acc": 0.7655880031570639,
"acc_stderr": 0.011906130106237985
}
}
```
### Supported Tasks and Leaderboards
[More Information Needed]
### Languages
[More Information Needed]
## Dataset Structure
### Data Instances
[More Information Needed]
### Data Fields
[More Information Needed]
### Data Splits
[More Information Needed]
## Dataset Creation
### Curation Rationale
[More Information Needed]
### Source Data
#### Initial Data Collection and Normalization
[More Information Needed]
#### Who are the source language producers?
[More Information Needed]
### Annotations
#### Annotation process
[More Information Needed]
#### Who are the annotators?
[More Information Needed]
### Personal and Sensitive Information
[More Information Needed]
## Considerations for Using the Data
### Social Impact of Dataset
[More Information Needed]
### Discussion of Biases
[More Information Needed]
### Other Known Limitations
[More Information Needed]
## Additional Information
### Dataset Curators
[More Information Needed]
### Licensing Information
[More Information Needed]
### Citation Information
[More Information Needed]
### Contributions
[More Information Needed]
|
open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w-gate_up_down_proj
|
[
"region:us"
] |
2023-10-23T05:02:26+00:00
|
{"pretty_name": "Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj", "dataset_summary": "Dataset automatically created during the evaluation run of model [CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj](https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\nTo load the details from a run, you can for instance do the following:\n```python\nfrom datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w-gate_up_down_proj\",\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\nThese are the [latest results from run 2023-10-23T06:02:22.448569](https://huggingface.co/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-FINETUNE1_17w-gate_up_down_proj/blob/main/results_2023-10-23T06-02-22.448569.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.10098573825503356,\n \"em_stderr\": 0.0030856947694457384,\n \"f1\": 0.15267407718120746,\n \"f1_stderr\": 0.0031959753495490175,\n \"acc\": 0.4472367612449459,\n \"acc_stderr\": 0.010567855433819127\n },\n \"harness|drop|3\": {\n \"em\": 0.10098573825503356,\n \"em_stderr\": 0.0030856947694457384,\n \"f1\": 0.15267407718120746,\n \"f1_stderr\": 0.0031959753495490175\n },\n \"harness|gsm8k|5\": {\n \"acc\": 0.1288855193328279,\n \"acc_stderr\": 0.009229580761400269\n },\n \"harness|winogrande|5\": {\n \"acc\": 0.7655880031570639,\n \"acc_stderr\": 0.011906130106237985\n }\n}\n```", "repo_url": "https://huggingface.co/CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj", "leaderboard_url": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard", "point_of_contact": "[email protected]", "configs": [{"config_name": "harness_drop_3", "data_files": [{"split": "2023_10_23T06_02_22.448569", "path": ["**/details_harness|drop|3_2023-10-23T06-02-22.448569.parquet"]}, {"split": "latest", "path": ["**/details_harness|drop|3_2023-10-23T06-02-22.448569.parquet"]}]}, {"config_name": "harness_gsm8k_5", "data_files": [{"split": "2023_10_23T06_02_22.448569", "path": ["**/details_harness|gsm8k|5_2023-10-23T06-02-22.448569.parquet"]}, {"split": "latest", "path": ["**/details_harness|gsm8k|5_2023-10-23T06-02-22.448569.parquet"]}]}, {"config_name": "harness_winogrande_5", "data_files": [{"split": "2023_10_23T06_02_22.448569", "path": ["**/details_harness|winogrande|5_2023-10-23T06-02-22.448569.parquet"]}, {"split": "latest", "path": ["**/details_harness|winogrande|5_2023-10-23T06-02-22.448569.parquet"]}]}, {"config_name": "results", "data_files": [{"split": "2023_10_23T06_02_22.448569", "path": ["results_2023-10-23T06-02-22.448569.parquet"]}, {"split": "latest", "path": ["results_2023-10-23T06-02-22.448569.parquet"]}]}]}
|
2023-10-23T05:02:33+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj
## Dataset Description
- Homepage:
- Repository: URL
- Paper:
- Leaderboard: URL
- Point of Contact: clementine@URL
### Dataset Summary
Dataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj on the Open LLM Leaderboard.
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).
To load the details from a run, you can for instance do the following:
## Latest results
These are the latest results from run 2023-10-23T06:02:22.448569(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
### Supported Tasks and Leaderboards
### Languages
## Dataset Structure
### Data Instances
### Data Fields
### Data Splits
## Dataset Creation
### Curation Rationale
### Source Data
#### Initial Data Collection and Normalization
#### Who are the source language producers?
### Annotations
#### Annotation process
#### Who are the annotators?
### Personal and Sensitive Information
## Considerations for Using the Data
### Social Impact of Dataset
### Discussion of Biases
### Other Known Limitations
## Additional Information
### Dataset Curators
### Licensing Information
### Contributions
|
[
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T06:02:22.448569(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T06:02:22.448569(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
6,
37,
31,
185,
66,
10,
4,
6,
6,
5,
5,
5,
7,
4,
10,
10,
5,
5,
9,
8,
8,
7,
8,
7,
5,
6,
6,
5
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for Evaluation run of CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL### Dataset Summary\n\nDataset automatically created during the evaluation run of model CHIH-HUNG/llama-2-13b-FINETUNE1_17w-gate_up_down_proj on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:## Latest results\n\nThese are the latest results from run 2023-10-23T06:02:22.448569(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):### Supported Tasks and Leaderboards### Languages## Dataset Structure### Data Instances### Data Fields### Data Splits## Dataset Creation### Curation Rationale### Source Data#### Initial Data Collection and Normalization#### Who are the source language producers?### Annotations#### Annotation process#### Who are the annotators?### Personal and Sensitive Information## Considerations for Using the Data### Social Impact of Dataset### Discussion of Biases### Other Known Limitations## Additional Information### Dataset Curators### Licensing Information### Contributions"
] |
06fe7ba739c7a30606f015e576c5db7c3a974789
|
# Dataset Card for "squad_v2"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
aai520-group6/squad_v2
|
[
"region:us"
] |
2023-10-23T05:05:06+00:00
|
{"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "validation", "path": "data/validation-*"}]}], "dataset_info": {"features": [{"name": "id", "dtype": "string"}, {"name": "title", "dtype": "string"}, {"name": "context", "dtype": "string"}, {"name": "question", "dtype": "string"}, {"name": "answers", "sequence": [{"name": "text", "dtype": "string"}, {"name": "answer_start", "dtype": "int32"}]}], "splits": [{"name": "train", "num_bytes": 116732025, "num_examples": 130319}, {"name": "validation", "num_bytes": 11661091, "num_examples": 11873}], "download_size": 0, "dataset_size": 128393116}}
|
2023-10-25T10:53:11+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "squad_v2"
More Information needed
|
[
"# Dataset Card for \"squad_v2\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"squad_v2\"\n\nMore Information needed"
] |
[
6,
15
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"squad_v2\"\n\nMore Information needed"
] |
b90a2b80e02641d6cd664c852f4c9de0107ebe94
|
# SVGen Vector Images Dataset Instruct Version
## Overview
SVGen is a comprehensive dataset containing 300,000 SVG vector codes from a diverse set of sources including SVG-Repo, Noto Emoji, and InstructSVG. The dataset aims to provide a wide range of SVG files suitable for various applications including web development, design, and machine learning research.
## Data Fields
```json
{
"text": "<s>[INST] Icon of Look Up here are the inputs Look Up [/INST] \\n <?xml version=\"1.0\" encoding=\"utf-8\"?>\r\n<!-- Uploaded to: SVG Repo, www.svgrepo.com, Generator: SVG Repo Mixer Tools -->\r\n<svg width=\"800px\" height=\"800px\" viewBox=\"0 0 1024 1024\" class=\"icon\" version=\"1.1\" xmlns=\"http://www.w3.org/2000/svg\"><path d=\"M460 459.3m-245.2 0a245.2 245.2 0 1 0 490.4 0 245.2 245.2 0 1 0-490.4 0Z\" fill=\"#E1F0FF\" /><path d=\"M460 719c-69.4 0-134.6-27-183.6-76.1s-76.1-114.3-76.1-183.6c0-69.4 27-134.6 76.1-183.6s114.3-76.1 183.6-76.1c69.4 0 134.6 27 183.6 76.1s76.1 114.3 76.1 183.6c0 69.4-27 134.6-76.1 183.6S529.4 719 460 719z m0-490.4c-61.6 0-119.6 24-163.1 67.6-43.6 43.6-67.6 101.5-67.6 163.1s24 119.6 67.6 163.1C340.4 666 398.4 690 460 690s119.6-24 163.1-67.6c43.6-43.6 67.6-101.5 67.6-163.1s-24-119.6-67.6-163.1c-43.5-43.6-101.5-67.6-163.1-67.6z\" fill=\"#446EB1\" /><path d=\"M640.6 630.6c8.7-8.7 22.8-8.7 31.5 0L802 760.5c8.6 8.9 8.3 23-0.5 31.5-8.6 8.3-22.3 8.3-31 0L640.6 662.1c-8.4-8.6-8.1-22.6 0-31.5z\" fill=\"#446EB1\" /></svg> </s>",
"description": "Icon of Look Up",
"input": "Look Up",
"output": "<?xml version=\"1.0\" encoding=\"utf-8\"?>\r\n<!-- Uploaded to: SVG Repo, www.svgrepo.com, Generator: SVG Repo Mixer Tools -->\r\n<svg width=\"800px\" height=\"800px\" viewBox=\"0 0 1024 1024\" class=\"icon\" version=\"1.1\" xmlns=\"http://www.w3.org/2000/svg\"><path d=\"M460 459.3m-245.2 0a245.2 245.2 0 1 0 490.4 0 245.2 245.2 0 1 0-490.4 0Z\" fill=\"#E1F0FF\" /><path d=\"M460 719c-69.4 0-134.6-27-183.6-76.1s-76.1-114.3-76.1-183.6c0-69.4 27-134.6 76.1-183.6s114.3-76.1 183.6-76.1c69.4 0 134.6 27 183.6 76.1s76.1 114.3 76.1 183.6c0 69.4-27 134.6-76.1 183.6S529.4 719 460 719z m0-490.4c-61.6 0-119.6 24-163.1 67.6-43.6 43.6-67.6 101.5-67.6 163.1s24 119.6 67.6 163.1C340.4 666 398.4 690 460 690s119.6-24 163.1-67.6c43.6-43.6 67.6-101.5 67.6-163.1s-24-119.6-67.6-163.1c-43.5-43.6-101.5-67.6-163.1-67.6z\" fill=\"#446EB1\" /><path d=\"M640.6 630.6c8.7-8.7 22.8-8.7 31.5 0L802 760.5c8.6 8.9 8.3 23-0.5 31.5-8.6 8.3-22.3 8.3-31 0L640.6 662.1c-8.4-8.6-8.1-22.6 0-31.5z\" fill=\"#446EB1\" /></svg>"
}
```
## Data Sources
- [SVG-Repo](https://www.svgrepo.com/)
- [Noto Emoji](https://huggingface.co/datasets/darknoon/noto-emoji-vector-512-svg)
- [InstructSVG](https://huggingface.co/datasets/uwunion/instruct_svg)
## Usage
The dataset is particularly useful for tasks such as icon classification, style transfer, image-to-vector translation, and much more. It serves as a rich resource for machine learning models that require high-quality SVG data.
## Help Wanted
I wanted to use BILP to generate `description`'s for each SVG, but It's not working well. If you have any ideas, please let me know. Here is the [Github](https://github.com/umuthopeyildirim/SVGenDataset) and it also contains Colab notebook links.
## License
The dataset incorporates SVG files with varying licenses. Users are advised to consult the `license` field of each record for specific usage rights.
## Contribution Guidelines
Contributions are welcome! If you find any issues or would like to add more SVGs to the dataset, please submit a pull request or open an issue in the repository.
## Acknowledgements
A huge thanks to SVGRepo, Noto Emoji, and InstructSVG for providing the SVG files that make up this dataset.
For more details and to download the dataset, visit the project repository.
|
umuthopeyildirim/svgen-500k-instruct
|
[
"task_categories:text-generation",
"size_categories:100K<n<1M",
"language:en",
"license:cc",
"SVG",
"vector",
"region:us"
] |
2023-10-23T06:19:29+00:00
|
{"language": ["en"], "license": "cc", "size_categories": ["100K<n<1M"], "task_categories": ["text-generation"], "pretty_name": "SVGen Instruct Dataset", "tags": ["SVG", "vector"]}
|
2023-10-23T06:38:15+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-cc #SVG #vector #region-us
|
# SVGen Vector Images Dataset Instruct Version
## Overview
SVGen is a comprehensive dataset containing 300,000 SVG vector codes from a diverse set of sources including SVG-Repo, Noto Emoji, and InstructSVG. The dataset aims to provide a wide range of SVG files suitable for various applications including web development, design, and machine learning research.
## Data Fields
## Data Sources
- SVG-Repo
- Noto Emoji
- InstructSVG
## Usage
The dataset is particularly useful for tasks such as icon classification, style transfer, image-to-vector translation, and much more. It serves as a rich resource for machine learning models that require high-quality SVG data.
## Help Wanted
I wanted to use BILP to generate 'description''s for each SVG, but It's not working well. If you have any ideas, please let me know. Here is the Github and it also contains Colab notebook links.
## License
The dataset incorporates SVG files with varying licenses. Users are advised to consult the 'license' field of each record for specific usage rights.
## Contribution Guidelines
Contributions are welcome! If you find any issues or would like to add more SVGs to the dataset, please submit a pull request or open an issue in the repository.
## Acknowledgements
A huge thanks to SVGRepo, Noto Emoji, and InstructSVG for providing the SVG files that make up this dataset.
For more details and to download the dataset, visit the project repository.
|
[
"# SVGen Vector Images Dataset Instruct Version",
"## Overview\n\nSVGen is a comprehensive dataset containing 300,000 SVG vector codes from a diverse set of sources including SVG-Repo, Noto Emoji, and InstructSVG. The dataset aims to provide a wide range of SVG files suitable for various applications including web development, design, and machine learning research.",
"## Data Fields",
"## Data Sources\n\n- SVG-Repo\n- Noto Emoji\n- InstructSVG",
"## Usage\n\nThe dataset is particularly useful for tasks such as icon classification, style transfer, image-to-vector translation, and much more. It serves as a rich resource for machine learning models that require high-quality SVG data.",
"## Help Wanted\n\nI wanted to use BILP to generate 'description''s for each SVG, but It's not working well. If you have any ideas, please let me know. Here is the Github and it also contains Colab notebook links.",
"## License\n\nThe dataset incorporates SVG files with varying licenses. Users are advised to consult the 'license' field of each record for specific usage rights.",
"## Contribution Guidelines\n\nContributions are welcome! If you find any issues or would like to add more SVGs to the dataset, please submit a pull request or open an issue in the repository.",
"## Acknowledgements\n\nA huge thanks to SVGRepo, Noto Emoji, and InstructSVG for providing the SVG files that make up this dataset.\n\nFor more details and to download the dataset, visit the project repository."
] |
[
"TAGS\n#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-cc #SVG #vector #region-us \n",
"# SVGen Vector Images Dataset Instruct Version",
"## Overview\n\nSVGen is a comprehensive dataset containing 300,000 SVG vector codes from a diverse set of sources including SVG-Repo, Noto Emoji, and InstructSVG. The dataset aims to provide a wide range of SVG files suitable for various applications including web development, design, and machine learning research.",
"## Data Fields",
"## Data Sources\n\n- SVG-Repo\n- Noto Emoji\n- InstructSVG",
"## Usage\n\nThe dataset is particularly useful for tasks such as icon classification, style transfer, image-to-vector translation, and much more. It serves as a rich resource for machine learning models that require high-quality SVG data.",
"## Help Wanted\n\nI wanted to use BILP to generate 'description''s for each SVG, but It's not working well. If you have any ideas, please let me know. Here is the Github and it also contains Colab notebook links.",
"## License\n\nThe dataset incorporates SVG files with varying licenses. Users are advised to consult the 'license' field of each record for specific usage rights.",
"## Contribution Guidelines\n\nContributions are welcome! If you find any issues or would like to add more SVGs to the dataset, please submit a pull request or open an issue in the repository.",
"## Acknowledgements\n\nA huge thanks to SVGRepo, Noto Emoji, and InstructSVG for providing the SVG files that make up this dataset.\n\nFor more details and to download the dataset, visit the project repository."
] |
[
44,
11,
72,
4,
20,
53,
57,
37,
45,
54
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-cc #SVG #vector #region-us \n# SVGen Vector Images Dataset Instruct Version## Overview\n\nSVGen is a comprehensive dataset containing 300,000 SVG vector codes from a diverse set of sources including SVG-Repo, Noto Emoji, and InstructSVG. The dataset aims to provide a wide range of SVG files suitable for various applications including web development, design, and machine learning research.## Data Fields## Data Sources\n\n- SVG-Repo\n- Noto Emoji\n- InstructSVG## Usage\n\nThe dataset is particularly useful for tasks such as icon classification, style transfer, image-to-vector translation, and much more. It serves as a rich resource for machine learning models that require high-quality SVG data.## Help Wanted\n\nI wanted to use BILP to generate 'description''s for each SVG, but It's not working well. If you have any ideas, please let me know. Here is the Github and it also contains Colab notebook links.## License\n\nThe dataset incorporates SVG files with varying licenses. Users are advised to consult the 'license' field of each record for specific usage rights.## Contribution Guidelines\n\nContributions are welcome! If you find any issues or would like to add more SVGs to the dataset, please submit a pull request or open an issue in the repository.## Acknowledgements\n\nA huge thanks to SVGRepo, Noto Emoji, and InstructSVG for providing the SVG files that make up this dataset.\n\nFor more details and to download the dataset, visit the project repository."
] |
3a85ad80c2c47649f1a2728d3f31bbdccb03d91c
|
# Dataset Card for "generated-ai-sample"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
rdiazconcha/generated-ai-sample
|
[
"region:us"
] |
2023-10-23T06:23:44+00:00
|
{"dataset_info": {"features": [{"name": "item", "dtype": "string"}, {"name": "description", "dtype": "string"}, {"name": "ad", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 6849, "num_examples": 5}], "download_size": 12525, "dataset_size": 6849}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T06:23:45+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "generated-ai-sample"
More Information needed
|
[
"# Dataset Card for \"generated-ai-sample\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"generated-ai-sample\"\n\nMore Information needed"
] |
[
6,
17
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"generated-ai-sample\"\n\nMore Information needed"
] |
c70a1543633720730d522df89926f0e07c38e6d5
|
# Visually Grounded embeddings for Fast-text and GloVe
This repository contains multiple visually grounded word embedding models.
All of these embeddings have been effectively infused with visual information from images.<br>
They have been proven to show stronger correlations (compared to textual embeddings)
to human judgments on various word similarities and relatedness benchmarks.
# Usage
All of the models are encoded in [gensim](https://pypi.org/project/gensim/) format.
Loading the model:
```python
import gensim
model_g = gensim.models.KeyedVectors.load_word2vec_format('path_to_embeddings' , binary=True)
#retrieve the most similar words
print(model_g.most_similar('together',topn=10))
[('togther', 0.6425853967666626), ('togehter', 0.6374243497848511), ('togeather', 0.6196791529655457),
('togather', 0.5998020172119141), ('togheter', 0.5819681882858276),('toghether', 0.5738174319267273),
('2gether', 0.5187329053878784), ('togethor', 0.501663088798523), ('gether', 0.49128714203834534),
('toegther', 0.48457157611846924)]
print(model_g.most_similar('sad',topn=10))
[('saddening', 0.6763913631439209), ('depressing', 0.6676110029220581), ('saddened', 0.6352651715278625),
('sorrowful', 0.6336953043937683), ('heartbreaking', 0.6180269122123718), ('heartbroken', 0.6099187135696411),
('tragic', 0.6039361953735352), ('pathetic', 0.5848405361175537), ('Sad', 0.5826965570449829),
('mournful', 0.5742306709289551)]
#find the outlier word
print(model_g.doesnt_match(['fire', 'water', 'land', 'sea', 'air', 'car']))
car
```
where 'path_to_embeddings' is the path to the embeddings you intend to use.
# Which embeddings to use
Under the **Files and Versions** tab, you can see the list of 4 available embeddings.
The following embedding files are from the paper [Learning Zero-Shot Multifaceted Visually Grounded Word Embeddings via Multi-Task Training](https://aclanthology.org/2021.conll-1.12/):
- v_glove_1024d_1.0
- v_fasttext_1024d_1.0
The following embedding files are from the paper [Language with Vision: a Study on Grounded Word and Sentence Embeddings](https://arxiv.org/pdf/2206.08823.pdf):
- v_glove_1024d_2.0
- v_glove_300_d_2.0
All of them come with 1024-dimensional word vectors except v_glove_300_d_2.0 which contains 300-dimensional word vectors.
|
fittar/visually_grounded_embeddings
|
[
"language:en",
"license:mit",
"arxiv:2206.08823",
"region:us"
] |
2023-10-23T06:40:00+00:00
|
{"language": ["en"], "license": "mit"}
|
2023-10-23T13:50:10+00:00
|
[
"2206.08823"
] |
[
"en"
] |
TAGS
#language-English #license-mit #arxiv-2206.08823 #region-us
|
# Visually Grounded embeddings for Fast-text and GloVe
This repository contains multiple visually grounded word embedding models.
All of these embeddings have been effectively infused with visual information from images.<br>
They have been proven to show stronger correlations (compared to textual embeddings)
to human judgments on various word similarities and relatedness benchmarks.
# Usage
All of the models are encoded in gensim format.
Loading the model:
where 'path_to_embeddings' is the path to the embeddings you intend to use.
# Which embeddings to use
Under the Files and Versions tab, you can see the list of 4 available embeddings.
The following embedding files are from the paper Learning Zero-Shot Multifaceted Visually Grounded Word Embeddings via Multi-Task Training:
- v_glove_1024d_1.0
- v_fasttext_1024d_1.0
The following embedding files are from the paper Language with Vision: a Study on Grounded Word and Sentence Embeddings:
- v_glove_1024d_2.0
- v_glove_300_d_2.0
All of them come with 1024-dimensional word vectors except v_glove_300_d_2.0 which contains 300-dimensional word vectors.
|
[
"# Visually Grounded embeddings for Fast-text and GloVe\n\nThis repository contains multiple visually grounded word embedding models.\nAll of these embeddings have been effectively infused with visual information from images.<br>\nThey have been proven to show stronger correlations (compared to textual embeddings) \nto human judgments on various word similarities and relatedness benchmarks.",
"# Usage\n\nAll of the models are encoded in gensim format.\nLoading the model:\n\n\nwhere 'path_to_embeddings' is the path to the embeddings you intend to use.",
"# Which embeddings to use\nUnder the Files and Versions tab, you can see the list of 4 available embeddings.\n\nThe following embedding files are from the paper Learning Zero-Shot Multifaceted Visually Grounded Word Embeddings via Multi-Task Training:\n- v_glove_1024d_1.0\n- v_fasttext_1024d_1.0\n\nThe following embedding files are from the paper Language with Vision: a Study on Grounded Word and Sentence Embeddings:\n\n- v_glove_1024d_2.0\n- v_glove_300_d_2.0\n\nAll of them come with 1024-dimensional word vectors except v_glove_300_d_2.0 which contains 300-dimensional word vectors."
] |
[
"TAGS\n#language-English #license-mit #arxiv-2206.08823 #region-us \n",
"# Visually Grounded embeddings for Fast-text and GloVe\n\nThis repository contains multiple visually grounded word embedding models.\nAll of these embeddings have been effectively infused with visual information from images.<br>\nThey have been proven to show stronger correlations (compared to textual embeddings) \nto human judgments on various word similarities and relatedness benchmarks.",
"# Usage\n\nAll of the models are encoded in gensim format.\nLoading the model:\n\n\nwhere 'path_to_embeddings' is the path to the embeddings you intend to use.",
"# Which embeddings to use\nUnder the Files and Versions tab, you can see the list of 4 available embeddings.\n\nThe following embedding files are from the paper Learning Zero-Shot Multifaceted Visually Grounded Word Embeddings via Multi-Task Training:\n- v_glove_1024d_1.0\n- v_fasttext_1024d_1.0\n\nThe following embedding files are from the paper Language with Vision: a Study on Grounded Word and Sentence Embeddings:\n\n- v_glove_1024d_2.0\n- v_glove_300_d_2.0\n\nAll of them come with 1024-dimensional word vectors except v_glove_300_d_2.0 which contains 300-dimensional word vectors."
] |
[
24,
91,
44,
168
] |
[
"passage: TAGS\n#language-English #license-mit #arxiv-2206.08823 #region-us \n# Visually Grounded embeddings for Fast-text and GloVe\n\nThis repository contains multiple visually grounded word embedding models.\nAll of these embeddings have been effectively infused with visual information from images.<br>\nThey have been proven to show stronger correlations (compared to textual embeddings) \nto human judgments on various word similarities and relatedness benchmarks.# Usage\n\nAll of the models are encoded in gensim format.\nLoading the model:\n\n\nwhere 'path_to_embeddings' is the path to the embeddings you intend to use.# Which embeddings to use\nUnder the Files and Versions tab, you can see the list of 4 available embeddings.\n\nThe following embedding files are from the paper Learning Zero-Shot Multifaceted Visually Grounded Word Embeddings via Multi-Task Training:\n- v_glove_1024d_1.0\n- v_fasttext_1024d_1.0\n\nThe following embedding files are from the paper Language with Vision: a Study on Grounded Word and Sentence Embeddings:\n\n- v_glove_1024d_2.0\n- v_glove_300_d_2.0\n\nAll of them come with 1024-dimensional word vectors except v_glove_300_d_2.0 which contains 300-dimensional word vectors."
] |
ca3ab8eabdd89259fdc39b78bada5090107477b8
|
**[BodyRock Weight Loss Program](https://snoppymart.com/get-bodyrock-weight-loss-program)** Health Club is a Fitness and Wellness outlet that offers a wide range of services to help you reach your fitness goals. It provides a variety of classes such as yoga, Pilates, Zumba, and more. It also offers personal training, nutrition counseling, and massage therapy. The club has a fully equipped gym with the latest equipment and a wide range of free weights. The club also has a sauna, steam room, and a swimming pool. Body Rock Health Club is dedicated to helping you achieve your fitness goals and live a healthier lifestyle.
[.png)](https://snoppymart.com/get-bodyrock-weight-loss-program)
In a world saturated with weight loss programs promising quick fixes and instant results, it can be challenging to navigate through the noise and find a program that not only helps you shed pounds but also promotes a sustainable and healthy lifestyle. One such program that has gained attention in recent years is the Bodyrock Weight Loss Program. In this article, we'll take a closer look at what Bodyrock is all about, and why it may be the right choice for those seeking a long-term solution to their weight loss journey.
### **[Official website Here >> BodyRock Weight Loss Program <<](https://snoppymart.com/get-bodyrock-weight-loss-program)**
### [BodyRock](https://www.facebook.com/people/BodyRock-Weight-Loss-Program/61552741376544/): A Fitness Pioneer
BodyRock is no stranger to the fitness world. Established as a fitness company in 2008, BodyRock has been at the forefront of creating effective and engaging workout programs. With a strong online presence and a legion of dedicated followers, the company is known for its unique approach to fitness.
### The Burn & Boost Bootcamp Concept
The Burn & Boost Bootcamp is a comprehensive weight loss program designed to help participants shed unwanted pounds, boost their fitness levels, and improve overall health. It is a carefully curated plan that combines cutting-edge workout routines, nutritional guidance, and a supportive community to create a holistic approach to weight loss.
### Key Features of the Burn & Boost Bootcamp
**Diverse Workouts:** One of the program’s most exciting aspects is its workout variety. The Burn & Boost Bootcamp offers a range of exercises, from high-intensity interval training (HIIT) to strength training, ensuring that participants never get bored and consistently challenge their bodies.
**Online Accessibility:** **[BodyRock](https://snoppymart.com/get-bodyrock-weight-loss-program)** recognizes the convenience of home workouts, and the Burn & Boost Bootcamp can be accessed from the comfort of your own space. This flexibility makes it an ideal choice for those with busy schedules.
**Nutritional Guidance:** Effective weight loss involves more than just exercise. The program includes nutritional guidance to help participants make healthier food choices, understand portion control, and maintain a balanced diet.
**Supportive Community:** The **[BodyRock](https://snoppymart.com/get-bodyrock-weight-loss-program)** community is a pivotal element of the program. Participants can interact with like-minded individuals, share their experiences, and receive encouragement and support throughout their weight loss journey.
**Results-Driven:** The Burn & Boost Bootcamp is designed to deliver results. Participants often report improved fitness levels, fat loss, muscle gain, and an overall sense of well-being.
**[Official website Here >> BodyRock Weight Loss Program <<](https://snoppymart.com/get-bodyrock-weight-loss-program)**
### Benefits of the Burn & Boost Bootcamp
**Sustainable Weight Loss:** The program focuses on long-term success, encouraging participants to adopt sustainable fitness and dietary habits.
**Convenience:** With no need for a gym membership, participants can work out at their own pace and on their own schedule.
**Motivation and Accountability:** The supportive community and regularly updated workout routines keep participants motivated and accountable.
**Health Improvement:** Beyond weight loss, the program can lead to improved cardiovascular health, increased energy levels, and reduced stress.
**Accessible to All:** The Burn & Boost Bootcamp is designed for individuals of all fitness levels, from beginners to experienced athletes.
[.png)](https://snoppymart.com/get-bodyrock-weight-loss-program)
### The [Bodyrock](https://snoppymart.com/get-bodyrock-weight-loss-program) Approach
The **[Bodyrock Weight Loss Program](https://snoppymart.com/get-bodyrock-weight-loss-program)** is not just about losing weight; it's about transforming your entire lifestyle. The program emphasizes the importance of a balanced approach, combining exercise, nutrition, and mindset to achieve holistic health and fitness.
**Exercise:** The cornerstone of **[Bodyrock](https://snoppymart.com/get-bodyrock-weight-loss-program)** is its workout regimen. These workouts, which can be accessed online, vary in intensity and duration, making them suitable for people of all fitness levels. Whether you're a beginner or an experienced athlete, Bodyrock has a wide range of workout routines that can be tailored to your needs. The workouts often blend elements of high-intensity interval training (HIIT) and strength training to maximize calorie burn and muscle development.
**Nutrition:** Alongside exercise, the **[Bodyrock program](https://snoppymart.com/get-bodyrock-weight-loss-program)** places a strong emphasis on a balanced diet. Their website offers a variety of meal plans, recipes, and nutritional guidance that help participants make healthier food choices. The program promotes a sustainable and sensible approach to eating, avoiding fad diets and instead focusing on real, whole foods.
**Mindset:** Weight loss is not just about physical transformation; it's about mental well-being as well. Bodyrock understands this and offers resources for developing a positive mindset. Their content includes motivational articles, meditation sessions, and advice on setting realistic goals. This holistic approach sets **[Bodyrock](https://snoppymart.com/get-bodyrock-weight-loss-program)** apart from many other weight loss programs.
### Sustainability
The **[Bodyrock Weight Loss Program](https://snoppymart.com/get-bodyrock-weight-loss-program)**'s focus on sustainability is one of its defining features. Unlike crash diets or excessive exercise programs, Bodyrock aims to help participants create lasting changes in their lives. This includes adopting healthier habits that will remain with them long after they've reached their weight loss goals.
### Realistic Expectations
**[BodyRock](https://snoppymart.com/get-bodyrock-weight-loss-program)** is transparent about what participants can expect. The program doesn't promise overnight results or drastic transformations within a few weeks. Instead, it encourages participants to set achievable goals and work steadily towards them. This approach is a refreshing departure from many programs that make unrealistic claims.
**[Official website Here >> BodyRock Weight Loss Program <<](https://snoppymart.com/get-bodyrock-weight-loss-program)**
### Is [BodyRock](https://snoppymart.com/get-bodyrock-weight-loss-program) Right for You?
While the **[BodyRock Weight Loss Program](https://snoppymart.com/get-bodyrock-weight-loss-program)** offers numerous benefits and has garnered a devoted following, it may not be the perfect fit for everyone. Before embarking on this program, it's essential to consider the following factors:
**Time Commitment:** The workouts can be intense, and you'll need to set aside time in your schedule to complete them.
**Budget:** Some participants may find the associated costs of workout equipment or premium access to the program's resources a bit high.
**Self-Motivation:** Like any fitness program, your success ultimately depends on your dedication and self-motivation.
[](https://snoppymart.com/get-bodyrock-weight-loss-program)
### Conclusion
The **[BodyRock Weight Loss Program](https://snoppymart.com/get-bodyrock-weight-loss-program)** is a comprehensive and holistic approach to weight loss and overall health. By combining exercise, nutrition, and mindset, it addresses the critical elements needed for sustainable change. The program's community support and emphasis on realistic expectations make it an appealing choice for those who are committed to transforming their lives in a healthy and balanced way. Before starting any weight loss program, it's advisable to consult with a healthcare professional to ensure it's a safe and suitable choice for your individual needs and circumstances.
|
bodyrockweightloss/bodyrockweightloss
|
[
"region:us"
] |
2023-10-23T06:51:50+00:00
|
{}
|
2023-10-23T06:52:37+00:00
|
[] |
[] |
TAGS
#region-us
|
BodyRock Weight Loss Program Health Club is a Fitness and Wellness outlet that offers a wide range of services to help you reach your fitness goals. It provides a variety of classes such as yoga, Pilates, Zumba, and more. It also offers personal training, nutrition counseling, and massage therapy. The club has a fully equipped gym with the latest equipment and a wide range of free weights. The club also has a sauna, steam room, and a swimming pool. Body Rock Health Club is dedicated to helping you achieve your fitness goals and live a healthier lifestyle.
 to strength training, ensuring that participants never get bored and consistently challenge their bodies.
Online Accessibility: BodyRock recognizes the convenience of home workouts, and the Burn & Boost Bootcamp can be accessed from the comfort of your own space. This flexibility makes it an ideal choice for those with busy schedules.
Nutritional Guidance: Effective weight loss involves more than just exercise. The program includes nutritional guidance to help participants make healthier food choices, understand portion control, and maintain a balanced diet.
Supportive Community: The BodyRock community is a pivotal element of the program. Participants can interact with like-minded individuals, share their experiences, and receive encouragement and support throughout their weight loss journey.
Results-Driven: The Burn & Boost Bootcamp is designed to deliver results. Participants often report improved fitness levels, fat loss, muscle gain, and an overall sense of well-being.
Official website Here >> BodyRock Weight Loss Program <<
### Benefits of the Burn & Boost Bootcamp
Sustainable Weight Loss: The program focuses on long-term success, encouraging participants to adopt sustainable fitness and dietary habits.
Convenience: With no need for a gym membership, participants can work out at their own pace and on their own schedule.
Motivation and Accountability: The supportive community and regularly updated workout routines keep participants motivated and accountable.
Health Improvement: Beyond weight loss, the program can lead to improved cardiovascular health, increased energy levels, and reduced stress.
Accessible to All: The Burn & Boost Bootcamp is designed for individuals of all fitness levels, from beginners to experienced athletes.
 and strength training to maximize calorie burn and muscle development.
Nutrition: Alongside exercise, the Bodyrock program places a strong emphasis on a balanced diet. Their website offers a variety of meal plans, recipes, and nutritional guidance that help participants make healthier food choices. The program promotes a sustainable and sensible approach to eating, avoiding fad diets and instead focusing on real, whole foods.
Mindset: Weight loss is not just about physical transformation; it's about mental well-being as well. Bodyrock understands this and offers resources for developing a positive mindset. Their content includes motivational articles, meditation sessions, and advice on setting realistic goals. This holistic approach sets Bodyrock apart from many other weight loss programs.
### Sustainability
The Bodyrock Weight Loss Program's focus on sustainability is one of its defining features. Unlike crash diets or excessive exercise programs, Bodyrock aims to help participants create lasting changes in their lives. This includes adopting healthier habits that will remain with them long after they've reached their weight loss goals.
### Realistic Expectations
BodyRock is transparent about what participants can expect. The program doesn't promise overnight results or drastic transformations within a few weeks. Instead, it encourages participants to set achievable goals and work steadily towards them. This approach is a refreshing departure from many programs that make unrealistic claims.
Official website Here >> BodyRock Weight Loss Program <<
### Is BodyRock Right for You?
While the BodyRock Weight Loss Program offers numerous benefits and has garnered a devoted following, it may not be the perfect fit for everyone. Before embarking on this program, it's essential to consider the following factors:
Time Commitment: The workouts can be intense, and you'll need to set aside time in your schedule to complete them.
Budget: Some participants may find the associated costs of workout equipment or premium access to the program's resources a bit high.
Self-Motivation: Like any fitness program, your success ultimately depends on your dedication and self-motivation.
![](URL
### Conclusion
The BodyRock Weight Loss Program is a comprehensive and holistic approach to weight loss and overall health. By combining exercise, nutrition, and mindset, it addresses the critical elements needed for sustainable change. The program's community support and emphasis on realistic expectations make it an appealing choice for those who are committed to transforming their lives in a healthy and balanced way. Before starting any weight loss program, it's advisable to consult with a healthcare professional to ensure it's a safe and suitable choice for your individual needs and circumstances.
|
[
"### Official website Here >> BodyRock Weight Loss Program <<",
"### BodyRock: A Fitness Pioneer\n\nBodyRock is no stranger to the fitness world. Established as a fitness company in 2008, BodyRock has been at the forefront of creating effective and engaging workout programs. With a strong online presence and a legion of dedicated followers, the company is known for its unique approach to fitness.",
"### The Burn & Boost Bootcamp Concept\n\nThe Burn & Boost Bootcamp is a comprehensive weight loss program designed to help participants shed unwanted pounds, boost their fitness levels, and improve overall health. It is a carefully curated plan that combines cutting-edge workout routines, nutritional guidance, and a supportive community to create a holistic approach to weight loss.",
"### Key Features of the Burn & Boost Bootcamp\n\nDiverse Workouts: One of the program’s most exciting aspects is its workout variety. The Burn & Boost Bootcamp offers a range of exercises, from high-intensity interval training (HIIT) to strength training, ensuring that participants never get bored and consistently challenge their bodies.\n\nOnline Accessibility: BodyRock recognizes the convenience of home workouts, and the Burn & Boost Bootcamp can be accessed from the comfort of your own space. This flexibility makes it an ideal choice for those with busy schedules.\n\nNutritional Guidance: Effective weight loss involves more than just exercise. The program includes nutritional guidance to help participants make healthier food choices, understand portion control, and maintain a balanced diet.\n\nSupportive Community: The BodyRock community is a pivotal element of the program. Participants can interact with like-minded individuals, share their experiences, and receive encouragement and support throughout their weight loss journey.\n\nResults-Driven: The Burn & Boost Bootcamp is designed to deliver results. Participants often report improved fitness levels, fat loss, muscle gain, and an overall sense of well-being.\n\nOfficial website Here >> BodyRock Weight Loss Program <<",
"### Benefits of the Burn & Boost Bootcamp\n\nSustainable Weight Loss: The program focuses on long-term success, encouraging participants to adopt sustainable fitness and dietary habits.\n\nConvenience: With no need for a gym membership, participants can work out at their own pace and on their own schedule.\n\nMotivation and Accountability: The supportive community and regularly updated workout routines keep participants motivated and accountable.\n\nHealth Improvement: Beyond weight loss, the program can lead to improved cardiovascular health, increased energy levels, and reduced stress.\n\nAccessible to All: The Burn & Boost Bootcamp is designed for individuals of all fitness levels, from beginners to experienced athletes.\n\n![.png)](URL",
"### The Bodyrock Approach\n\nThe Bodyrock Weight Loss Program is not just about losing weight; it's about transforming your entire lifestyle. The program emphasizes the importance of a balanced approach, combining exercise, nutrition, and mindset to achieve holistic health and fitness.\n\nExercise: The cornerstone of Bodyrock is its workout regimen. These workouts, which can be accessed online, vary in intensity and duration, making them suitable for people of all fitness levels. Whether you're a beginner or an experienced athlete, Bodyrock has a wide range of workout routines that can be tailored to your needs. The workouts often blend elements of high-intensity interval training (HIIT) and strength training to maximize calorie burn and muscle development.\n\nNutrition: Alongside exercise, the Bodyrock program places a strong emphasis on a balanced diet. Their website offers a variety of meal plans, recipes, and nutritional guidance that help participants make healthier food choices. The program promotes a sustainable and sensible approach to eating, avoiding fad diets and instead focusing on real, whole foods.\n\nMindset: Weight loss is not just about physical transformation; it's about mental well-being as well. Bodyrock understands this and offers resources for developing a positive mindset. Their content includes motivational articles, meditation sessions, and advice on setting realistic goals. This holistic approach sets Bodyrock apart from many other weight loss programs.",
"### Sustainability\n\nThe Bodyrock Weight Loss Program's focus on sustainability is one of its defining features. Unlike crash diets or excessive exercise programs, Bodyrock aims to help participants create lasting changes in their lives. This includes adopting healthier habits that will remain with them long after they've reached their weight loss goals.",
"### Realistic Expectations\n\nBodyRock is transparent about what participants can expect. The program doesn't promise overnight results or drastic transformations within a few weeks. Instead, it encourages participants to set achievable goals and work steadily towards them. This approach is a refreshing departure from many programs that make unrealistic claims.\n\nOfficial website Here >> BodyRock Weight Loss Program <<",
"### Is BodyRock Right for You?\n\nWhile the BodyRock Weight Loss Program offers numerous benefits and has garnered a devoted following, it may not be the perfect fit for everyone. Before embarking on this program, it's essential to consider the following factors:\n\nTime Commitment: The workouts can be intense, and you'll need to set aside time in your schedule to complete them.\n\nBudget: Some participants may find the associated costs of workout equipment or premium access to the program's resources a bit high.\n\nSelf-Motivation: Like any fitness program, your success ultimately depends on your dedication and self-motivation.\n\n![](URL",
"### Conclusion\n\nThe BodyRock Weight Loss Program is a comprehensive and holistic approach to weight loss and overall health. By combining exercise, nutrition, and mindset, it addresses the critical elements needed for sustainable change. The program's community support and emphasis on realistic expectations make it an appealing choice for those who are committed to transforming their lives in a healthy and balanced way. Before starting any weight loss program, it's advisable to consult with a healthcare professional to ensure it's a safe and suitable choice for your individual needs and circumstances."
] |
[
"TAGS\n#region-us \n",
"### Official website Here >> BodyRock Weight Loss Program <<",
"### BodyRock: A Fitness Pioneer\n\nBodyRock is no stranger to the fitness world. Established as a fitness company in 2008, BodyRock has been at the forefront of creating effective and engaging workout programs. With a strong online presence and a legion of dedicated followers, the company is known for its unique approach to fitness.",
"### The Burn & Boost Bootcamp Concept\n\nThe Burn & Boost Bootcamp is a comprehensive weight loss program designed to help participants shed unwanted pounds, boost their fitness levels, and improve overall health. It is a carefully curated plan that combines cutting-edge workout routines, nutritional guidance, and a supportive community to create a holistic approach to weight loss.",
"### Key Features of the Burn & Boost Bootcamp\n\nDiverse Workouts: One of the program’s most exciting aspects is its workout variety. The Burn & Boost Bootcamp offers a range of exercises, from high-intensity interval training (HIIT) to strength training, ensuring that participants never get bored and consistently challenge their bodies.\n\nOnline Accessibility: BodyRock recognizes the convenience of home workouts, and the Burn & Boost Bootcamp can be accessed from the comfort of your own space. This flexibility makes it an ideal choice for those with busy schedules.\n\nNutritional Guidance: Effective weight loss involves more than just exercise. The program includes nutritional guidance to help participants make healthier food choices, understand portion control, and maintain a balanced diet.\n\nSupportive Community: The BodyRock community is a pivotal element of the program. Participants can interact with like-minded individuals, share their experiences, and receive encouragement and support throughout their weight loss journey.\n\nResults-Driven: The Burn & Boost Bootcamp is designed to deliver results. Participants often report improved fitness levels, fat loss, muscle gain, and an overall sense of well-being.\n\nOfficial website Here >> BodyRock Weight Loss Program <<",
"### Benefits of the Burn & Boost Bootcamp\n\nSustainable Weight Loss: The program focuses on long-term success, encouraging participants to adopt sustainable fitness and dietary habits.\n\nConvenience: With no need for a gym membership, participants can work out at their own pace and on their own schedule.\n\nMotivation and Accountability: The supportive community and regularly updated workout routines keep participants motivated and accountable.\n\nHealth Improvement: Beyond weight loss, the program can lead to improved cardiovascular health, increased energy levels, and reduced stress.\n\nAccessible to All: The Burn & Boost Bootcamp is designed for individuals of all fitness levels, from beginners to experienced athletes.\n\n![.png)](URL",
"### The Bodyrock Approach\n\nThe Bodyrock Weight Loss Program is not just about losing weight; it's about transforming your entire lifestyle. The program emphasizes the importance of a balanced approach, combining exercise, nutrition, and mindset to achieve holistic health and fitness.\n\nExercise: The cornerstone of Bodyrock is its workout regimen. These workouts, which can be accessed online, vary in intensity and duration, making them suitable for people of all fitness levels. Whether you're a beginner or an experienced athlete, Bodyrock has a wide range of workout routines that can be tailored to your needs. The workouts often blend elements of high-intensity interval training (HIIT) and strength training to maximize calorie burn and muscle development.\n\nNutrition: Alongside exercise, the Bodyrock program places a strong emphasis on a balanced diet. Their website offers a variety of meal plans, recipes, and nutritional guidance that help participants make healthier food choices. The program promotes a sustainable and sensible approach to eating, avoiding fad diets and instead focusing on real, whole foods.\n\nMindset: Weight loss is not just about physical transformation; it's about mental well-being as well. Bodyrock understands this and offers resources for developing a positive mindset. Their content includes motivational articles, meditation sessions, and advice on setting realistic goals. This holistic approach sets Bodyrock apart from many other weight loss programs.",
"### Sustainability\n\nThe Bodyrock Weight Loss Program's focus on sustainability is one of its defining features. Unlike crash diets or excessive exercise programs, Bodyrock aims to help participants create lasting changes in their lives. This includes adopting healthier habits that will remain with them long after they've reached their weight loss goals.",
"### Realistic Expectations\n\nBodyRock is transparent about what participants can expect. The program doesn't promise overnight results or drastic transformations within a few weeks. Instead, it encourages participants to set achievable goals and work steadily towards them. This approach is a refreshing departure from many programs that make unrealistic claims.\n\nOfficial website Here >> BodyRock Weight Loss Program <<",
"### Is BodyRock Right for You?\n\nWhile the BodyRock Weight Loss Program offers numerous benefits and has garnered a devoted following, it may not be the perfect fit for everyone. Before embarking on this program, it's essential to consider the following factors:\n\nTime Commitment: The workouts can be intense, and you'll need to set aside time in your schedule to complete them.\n\nBudget: Some participants may find the associated costs of workout equipment or premium access to the program's resources a bit high.\n\nSelf-Motivation: Like any fitness program, your success ultimately depends on your dedication and self-motivation.\n\n![](URL",
"### Conclusion\n\nThe BodyRock Weight Loss Program is a comprehensive and holistic approach to weight loss and overall health. By combining exercise, nutrition, and mindset, it addresses the critical elements needed for sustainable change. The program's community support and emphasis on realistic expectations make it an appealing choice for those who are committed to transforming their lives in a healthy and balanced way. Before starting any weight loss program, it's advisable to consult with a healthcare professional to ensure it's a safe and suitable choice for your individual needs and circumstances."
] |
[
6,
13,
69,
80,
270,
156,
324,
74,
86,
144,
121
] |
[
"passage: TAGS\n#region-us \n### Official website Here >> BodyRock Weight Loss Program <<### BodyRock: A Fitness Pioneer\n\nBodyRock is no stranger to the fitness world. Established as a fitness company in 2008, BodyRock has been at the forefront of creating effective and engaging workout programs. With a strong online presence and a legion of dedicated followers, the company is known for its unique approach to fitness.### The Burn & Boost Bootcamp Concept\n\nThe Burn & Boost Bootcamp is a comprehensive weight loss program designed to help participants shed unwanted pounds, boost their fitness levels, and improve overall health. It is a carefully curated plan that combines cutting-edge workout routines, nutritional guidance, and a supportive community to create a holistic approach to weight loss.### Key Features of the Burn & Boost Bootcamp\n\nDiverse Workouts: One of the program’s most exciting aspects is its workout variety. The Burn & Boost Bootcamp offers a range of exercises, from high-intensity interval training (HIIT) to strength training, ensuring that participants never get bored and consistently challenge their bodies.\n\nOnline Accessibility: BodyRock recognizes the convenience of home workouts, and the Burn & Boost Bootcamp can be accessed from the comfort of your own space. This flexibility makes it an ideal choice for those with busy schedules.\n\nNutritional Guidance: Effective weight loss involves more than just exercise. The program includes nutritional guidance to help participants make healthier food choices, understand portion control, and maintain a balanced diet.\n\nSupportive Community: The BodyRock community is a pivotal element of the program. Participants can interact with like-minded individuals, share their experiences, and receive encouragement and support throughout their weight loss journey.\n\nResults-Driven: The Burn & Boost Bootcamp is designed to deliver results. Participants often report improved fitness levels, fat loss, muscle gain, and an overall sense of well-being.\n\nOfficial website Here >> BodyRock Weight Loss Program <<",
"passage: ### Benefits of the Burn & Boost Bootcamp\n\nSustainable Weight Loss: The program focuses on long-term success, encouraging participants to adopt sustainable fitness and dietary habits.\n\nConvenience: With no need for a gym membership, participants can work out at their own pace and on their own schedule.\n\nMotivation and Accountability: The supportive community and regularly updated workout routines keep participants motivated and accountable.\n\nHealth Improvement: Beyond weight loss, the program can lead to improved cardiovascular health, increased energy levels, and reduced stress.\n\nAccessible to All: The Burn & Boost Bootcamp is designed for individuals of all fitness levels, from beginners to experienced athletes.\n\n and strength training to maximize calorie burn and muscle development.\n\nNutrition: Alongside exercise, the Bodyrock program places a strong emphasis on a balanced diet. Their website offers a variety of meal plans, recipes, and nutritional guidance that help participants make healthier food choices. The program promotes a sustainable and sensible approach to eating, avoiding fad diets and instead focusing on real, whole foods.\n\nMindset: Weight loss is not just about physical transformation; it's about mental well-being as well. Bodyrock understands this and offers resources for developing a positive mindset. Their content includes motivational articles, meditation sessions, and advice on setting realistic goals. This holistic approach sets Bodyrock apart from many other weight loss programs.### Sustainability\n\nThe Bodyrock Weight Loss Program's focus on sustainability is one of its defining features. Unlike crash diets or excessive exercise programs, Bodyrock aims to help participants create lasting changes in their lives. This includes adopting healthier habits that will remain with them long after they've reached their weight loss goals.### Realistic Expectations\n\nBodyRock is transparent about what participants can expect. The program doesn't promise overnight results or drastic transformations within a few weeks. Instead, it encourages participants to set achievable goals and work steadily towards them. This approach is a refreshing departure from many programs that make unrealistic claims.\n\nOfficial website Here >> BodyRock Weight Loss Program <<"
] |
5e9a5c66676e5199ed4d3fd865cb1dc9d2047104
|
# Dataset Card for "cot"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
quyanh/cot
|
[
"region:us"
] |
2023-10-23T07:04:26+00:00
|
{"dataset_info": {"features": [{"name": "inputs", "dtype": "string"}, {"name": "response", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 3530400.0, "num_examples": 9000}], "download_size": 2120620, "dataset_size": 3530400.0}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-25T08:02:23+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "cot"
More Information needed
|
[
"# Dataset Card for \"cot\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"cot\"\n\nMore Information needed"
] |
[
6,
11
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"cot\"\n\nMore Information needed"
] |
79f4cc56cf5facfb2af96ca7844b57e25fc3952f
|
# Dataset Card for "layouts_donut"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
sankettgorey/layouts_donut
|
[
"region:us"
] |
2023-10-23T07:04:53+00:00
|
{"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "test", "path": "data/test-*"}, {"split": "validation", "path": "data/validation-*"}]}], "dataset_info": {"features": [{"name": "image", "dtype": "image"}, {"name": "ground_truth", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 1461607176.2173023, "num_examples": 5362}, {"name": "test", "num_bytes": 182370076.8402208, "num_examples": 671}, {"name": "validation", "num_bytes": 181812032.0684768, "num_examples": 670}], "download_size": 1524050233, "dataset_size": 1825789285.126}}
|
2023-10-23T07:06:03+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "layouts_donut"
More Information needed
|
[
"# Dataset Card for \"layouts_donut\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"layouts_donut\"\n\nMore Information needed"
] |
[
6,
15
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"layouts_donut\"\n\nMore Information needed"
] |
135fcac2cf82d7e70d52ec14b0d47c9f20889b1a
|
# The Pile -- PhilPaper (refined by Data-Juicer)
A refined version of PhilPaper dataset in The Pile by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/the-pile-philpaper-refine-result.jsonl) (About 1.7GB).
## Dataset Information
- Number of samples: 29,117 (Keep ~88.82% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'our-recipes-Philpaper'
dataset_path: '/path/to/the/original/dataset/' # path to your dataset directory or file
export_path: 'Philpaper-refine-result.jsonl' # path to your dataset result file
np: 50 # number of subprocess to process your dataset
ds_cache_dir: /cache # path to your dataset cache file
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.7 # <3sigma (0.72)
- average_line_length_filter:
max_len: 5e5 # >3sigma (406006)
- character_repetition_filter:
rep_len: 10
max_ratio: 0.2 # >3sigma (0.145)
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.0007 # 3sigma
- language_id_score_filter:
min_score: 0.6
- maximum_line_length_filter:
max_len: 1e6 # 3sigma
- perplexity_filter:
lang: en
max_ppl: 5000
- special_characters_filter:
max_ratio: 0.4 # > 3sigma (0.302)
- words_num_filter:
lang: en
tokenization: true
min_num: 1000
max_num: 2e5 # 3sigma
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.3 # > 3sigma (0.249)
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/the-pile-philpaper-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:16:41+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:33:30+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# The Pile -- PhilPaper (refined by Data-Juicer)
A refined version of PhilPaper dataset in The Pile by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 1.7GB).
## Dataset Information
- Number of samples: 29,117 (Keep ~88.82% from the original dataset)
## Refining Recipe
|
[
"# The Pile -- PhilPaper (refined by Data-Juicer)\n\nA refined version of PhilPaper dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 1.7GB).",
"## Dataset Information\n\n- Number of samples: 29,117 (Keep ~88.82% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# The Pile -- PhilPaper (refined by Data-Juicer)\n\nA refined version of PhilPaper dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 1.7GB).",
"## Dataset Information\n\n- Number of samples: 29,117 (Keep ~88.82% from the original dataset)",
"## Refining Recipe"
] |
[
49,
101,
25,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# The Pile -- PhilPaper (refined by Data-Juicer)\n\nA refined version of PhilPaper dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 1.7GB).## Dataset Information\n\n- Number of samples: 29,117 (Keep ~88.82% from the original dataset)## Refining Recipe"
] |
cf95360da4ea5bc220495a77723b51acef7f1d60
|
# Bangumi Image Base of Kimi To Boku No Saigo No Senjou Arui Wa Sekai Ga Hajimaru Seisen
This is the image base of bangumi Kimi to Boku no Saigo no Senjou Arui wa Sekai ga Hajimaru Seisen, we detected 20 characters, 1108 images in total. The full dataset is [here](all.zip).
**Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual.** If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| # | Images | Download | Preview 1 | Preview 2 | Preview 3 | Preview 4 | Preview 5 | Preview 6 | Preview 7 | Preview 8 |
|:------|---------:|:---------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|
| 0 | 203 | [Download](0/dataset.zip) |  |  |  |  |  |  |  |  |
| 1 | 11 | [Download](1/dataset.zip) |  |  |  |  |  |  |  |  |
| 2 | 14 | [Download](2/dataset.zip) |  |  |  |  |  |  |  |  |
| 3 | 36 | [Download](3/dataset.zip) |  |  |  |  |  |  |  |  |
| 4 | 72 | [Download](4/dataset.zip) |  |  |  |  |  |  |  |  |
| 5 | 11 | [Download](5/dataset.zip) |  |  |  |  |  |  |  |  |
| 6 | 14 | [Download](6/dataset.zip) |  |  |  |  |  |  |  |  |
| 7 | 97 | [Download](7/dataset.zip) |  |  |  |  |  |  |  |  |
| 8 | 229 | [Download](8/dataset.zip) |  |  |  |  |  |  |  |  |
| 9 | 46 | [Download](9/dataset.zip) |  |  |  |  |  |  |  |  |
| 10 | 14 | [Download](10/dataset.zip) |  |  |  |  |  |  |  |  |
| 11 | 25 | [Download](11/dataset.zip) |  |  |  |  |  |  |  |  |
| 12 | 17 | [Download](12/dataset.zip) |  |  |  |  |  |  |  |  |
| 13 | 99 | [Download](13/dataset.zip) |  |  |  |  |  |  |  |  |
| 14 | 29 | [Download](14/dataset.zip) |  |  |  |  |  |  |  |  |
| 15 | 13 | [Download](15/dataset.zip) |  |  |  |  |  |  |  |  |
| 16 | 8 | [Download](16/dataset.zip) |  |  |  |  |  |  |  |  |
| 17 | 66 | [Download](17/dataset.zip) |  |  |  |  |  |  |  |  |
| 18 | 7 | [Download](18/dataset.zip) |  |  |  |  |  |  |  | N/A |
| noise | 97 | [Download](-1/dataset.zip) |  |  |  |  |  |  |  |  |
|
BangumiBase/kimitobokunosaigonosenjouaruiwasekaigahajimaruseisen
|
[
"size_categories:1K<n<10K",
"license:mit",
"art",
"region:us"
] |
2023-10-23T07:26:25+00:00
|
{"license": "mit", "size_categories": ["1K<n<10K"], "tags": ["art"]}
|
2023-10-23T08:27:08+00:00
|
[] |
[] |
TAGS
#size_categories-1K<n<10K #license-mit #art #region-us
|
Bangumi Image Base of Kimi To Boku No Saigo No Senjou Arui Wa Sekai Ga Hajimaru Seisen
======================================================================================
This is the image base of bangumi Kimi to Boku no Saigo no Senjou Arui wa Sekai ga Hajimaru Seisen, we detected 20 characters, 1108 images in total. The full dataset is here.
Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual. If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
|
[] |
[
"TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] |
[
25
] |
[
"passage: TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] |
13d2db3860adaf5ef6c53015659a52d028f9c04a
|
# RedPajama -- ArXiv (refined by Data-Juicer)
A refined version of ArXiv dataset in RedPajama by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-arxiv-refine-result.jsonl) (About 85GB).
## Dataset Information
- Number of samples: 1,655,259 (Keep ~95.99% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-arxivrecipes-arxiv'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.516 # 3sigma
max_ratio: 0.915 # 3sigma
- average_line_length_filter: # for code
max_len: 682 # 3sigma
- character_repetition_filter:
rep_len: 10
max_ratio: 0.3
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.00076 # 3sigma
#- language_id_score_filter: # remove language filter
- maximum_line_length_filter: # for code
max_len: 4000
- perplexity_filter:
lang: en
max_ppl: 8000
- special_characters_filter:
max_ratio: 0.6
- text_length_filter:
max_len: 350000
- words_num_filter:
lang: en
tokenization: true
min_num: 20
max_num: 100000
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.574 # 3sigma
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/redpajama-arxiv-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:1M<n<10M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:35:29+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["1M<n<10M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:37:41+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama -- ArXiv (refined by Data-Juicer)
A refined version of ArXiv dataset in RedPajama by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 85GB).
## Dataset Information
- Number of samples: 1,655,259 (Keep ~95.99% from the original dataset)
## Refining Recipe
|
[
"# RedPajama -- ArXiv (refined by Data-Juicer)\n\nA refined version of ArXiv dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 85GB).",
"## Dataset Information\n\n- Number of samples: 1,655,259 (Keep ~95.99% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama -- ArXiv (refined by Data-Juicer)\n\nA refined version of ArXiv dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 85GB).",
"## Dataset Information\n\n- Number of samples: 1,655,259 (Keep ~95.99% from the original dataset)",
"## Refining Recipe"
] |
[
49,
103,
28,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama -- ArXiv (refined by Data-Juicer)\n\nA refined version of ArXiv dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 85GB).## Dataset Information\n\n- Number of samples: 1,655,259 (Keep ~95.99% from the original dataset)## Refining Recipe"
] |
025ca60be97b6713cf28807e2401922df1dfb4fe
|
# Alpaca-CoT -- ZH (refined by Data-Juicer)
A refined Chinese version of Alpaca-CoT dataset by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to fine-tune a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/CFT/alpaca-cot-zh-refine_result.jsonl) (About 18.7GB).
## Dataset Information
- Number of samples: 9,873,214 (Keep ~46.58% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-alpaca-cot-zh'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- document_deduplicator:
lowercase: true
ignore_non_character: true
- alphanumeric_filter:
tokenization: false
min_ratio: 0.10
- character_repetition_filter:
rep_len: 10
max_ratio: 0.6
- flagged_words_filter:
lang: zh
tokenization: true
use_words_aug: true
max_ratio: 0.017
- text_length_filter:
min_len: 10
- document_simhash_deduplicator:
tokenization: character
window_size: 4
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 10
hamming_distance: 8
```
|
datajuicer/alpaca-cot-zh-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:1M<n<10M",
"language:zh",
"license:apache-2.0",
"data-juicer",
"Fine-tuning",
"region:us"
] |
2023-10-23T07:35:58+00:00
|
{"language": ["zh"], "license": "apache-2.0", "size_categories": ["1M<n<10M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "Fine-tuning"]}
|
2023-11-10T13:33:53+00:00
|
[] |
[
"zh"
] |
TAGS
#task_categories-text-generation #size_categories-1M<n<10M #language-Chinese #license-apache-2.0 #data-juicer #Fine-tuning #region-us
|
# Alpaca-CoT -- ZH (refined by Data-Juicer)
A refined Chinese version of Alpaca-CoT dataset by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to fine-tune a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 18.7GB).
## Dataset Information
- Number of samples: 9,873,214 (Keep ~46.58% from the original dataset)
## Refining Recipe
|
[
"# Alpaca-CoT -- ZH (refined by Data-Juicer)\n\nA refined Chinese version of Alpaca-CoT dataset by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to fine-tune a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 18.7GB).",
"## Dataset Information\n\n- Number of samples: 9,873,214 (Keep ~46.58% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-1M<n<10M #language-Chinese #license-apache-2.0 #data-juicer #Fine-tuning #region-us \n",
"# Alpaca-CoT -- ZH (refined by Data-Juicer)\n\nA refined Chinese version of Alpaca-CoT dataset by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to fine-tune a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 18.7GB).",
"## Dataset Information\n\n- Number of samples: 9,873,214 (Keep ~46.58% from the original dataset)",
"## Refining Recipe"
] |
[
53,
105,
29,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-1M<n<10M #language-Chinese #license-apache-2.0 #data-juicer #Fine-tuning #region-us \n# Alpaca-CoT -- ZH (refined by Data-Juicer)\n\nA refined Chinese version of Alpaca-CoT dataset by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to fine-tune a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 18.7GB).## Dataset Information\n\n- Number of samples: 9,873,214 (Keep ~46.58% from the original dataset)## Refining Recipe"
] |
7080c45c7a95296f080d4e5703abec8ef5918ea2
|
# RedPajama -- CommonCrawl-2023-06 (refined by Data-Juicer)
A refined version of CommonCrawl-2023-06 dataset in RedPajama by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-cc-refine-results/redpajama-cc-2023-06-refine-result.jsonl) (About 310GB).
## Dataset Information
- Number of samples: 50,643,699 (Keep ~45.46% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-cc-2013-06'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.7508 # 3sigma
max_ratio: 0.8591 # 3sigma -- 1036821
- average_line_length_filter: # for code
max_len: 1500 # < 3sigma -- 395868
- character_repetition_filter:
rep_len: 10
max_ratio: 0.3 # > 3sigma -- 195026
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.0015 # 3sigma -- 287896
- language_id_score_filter: # remove language filter
min_score: 0.793 # 3sigma -- 2173246
- maximum_line_length_filter: # for code
max_len: 5000 # < 3sigma -- 797111
- perplexity_filter:
lang: en
max_ppl: 5000 # 3sigma -- 942162
- special_characters_filter:
min_ratio: 0.15 # > 3sigma
max_ratio: 0.35 # > 3sigma -- 1155090
- text_length_filter:
max_len: 58187 # 3sigma -- 1165902
- words_num_filter:
lang: en
tokenization: true
min_num: 20
max_num: 11529 # 3sigma -- 1185363
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.2962 # 3sigma -- 2407282
```
|
datajuicer/redpajama-cc-2023-06-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10M<n<100M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:40:01+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10M<n<100M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:41:32+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama -- CommonCrawl-2023-06 (refined by Data-Juicer)
A refined version of CommonCrawl-2023-06 dataset in RedPajama by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 310GB).
## Dataset Information
- Number of samples: 50,643,699 (Keep ~45.46% from the original dataset)
## Refining Recipe
|
[
"# RedPajama -- CommonCrawl-2023-06 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2023-06 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 310GB).",
"## Dataset Information\n\n- Number of samples: 50,643,699 (Keep ~45.46% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama -- CommonCrawl-2023-06 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2023-06 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 310GB).",
"## Dataset Information\n\n- Number of samples: 50,643,699 (Keep ~45.46% from the original dataset)",
"## Refining Recipe"
] |
[
49,
111,
31,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama -- CommonCrawl-2023-06 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2023-06 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 310GB).## Dataset Information\n\n- Number of samples: 50,643,699 (Keep ~45.46% from the original dataset)## Refining Recipe"
] |
8efbb94352946eba6640f1dab12bde42208f0e15
|
# RedPajama -- CommonCrawl-2022-05 (refined by Data-Juicer)
A refined version of CommonCrawl-2022-05 dataset in RedPajama by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-cc-refine-results/redpajama-cc-2022-05-refine-result.jsonl) (About 265GB).
## Dataset Information
- Number of samples: 42,648,496 (Keep ~45.34% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-cc-2022-05'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.7514 # 3sigma
max_ratio: 0.8577 # 3sigmai -- 888003
- average_line_length_filter: # for code
max_len: 1500 # < 3sigma -- 447069
- character_repetition_filter:
rep_len: 10
max_ratio: 0.3 # > 3sigma -- 145890 samples
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.0012 # 3sigma -- 319395
- language_id_score_filter: # remove language filter
min_score: 0.791 # 3sigma -- 1823528
- maximum_line_length_filter: # for code
max_len: 5000 # < 3sigma -- 791612
- perplexity_filter:
lang: en
max_ppl: 5000 # < 3sigma -- 654459
- special_characters_filter:
min_ratio: 0.15 # > 3sigma
max_ratio: 0.35 # > 3sigma
- text_length_filter:
max_len: 59265 # 3sigma -- 1046590
- words_num_filter:
lang: en
tokenization: true
min_num: 20 # > 3sigma
max_num: 11860 # 3sigma -- 1036780
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.3117 # 3sigma -- 2089703
```
|
datajuicer/redpajama-cc-2022-05-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10M<n<100M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:42:57+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10M<n<100M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:44:21+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama -- CommonCrawl-2022-05 (refined by Data-Juicer)
A refined version of CommonCrawl-2022-05 dataset in RedPajama by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 265GB).
## Dataset Information
- Number of samples: 42,648,496 (Keep ~45.34% from the original dataset)
## Refining Recipe
|
[
"# RedPajama -- CommonCrawl-2022-05 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2022-05 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 265GB).",
"## Dataset Information\n\n- Number of samples: 42,648,496 (Keep ~45.34% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama -- CommonCrawl-2022-05 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2022-05 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 265GB).",
"## Dataset Information\n\n- Number of samples: 42,648,496 (Keep ~45.34% from the original dataset)",
"## Refining Recipe"
] |
[
49,
111,
29,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama -- CommonCrawl-2022-05 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2022-05 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 265GB).## Dataset Information\n\n- Number of samples: 42,648,496 (Keep ~45.34% from the original dataset)## Refining Recipe"
] |
43a7fb5fbc48a51af0f4f3fac6ad27ba747d7c27
|
# RedPajama & TheStack -- Github Code (refined by Data-Juicer)
A refined version of Github Code dataset in RedPajama & TheStack by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-stack-code-refine-result.jsonl) (About 232GB).
## Dataset Information
- Number of samples: 49,279,344 (Keep ~52.09% from the original dataset)
## Refining Recipe
### RedPajama code refinement
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-code-rp'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- clean_copyright_mapper:
- alphanumeric_filter:
tokenization: False
min_ratio: 0.4
max_ratio: 0.8
- alphanumeric_filter:
tokenization: True
min_ratio: 1.5
max_ratio: 3
- average_line_length_filter:
min_len: 15
max_len: 100
- character_repetition_filter:
rep_len: 10
min_ratio: 0.05
max_ratio: 0.3
- maximum_line_length_filter:
min_len: 50
max_len: 500
- text_length_filter:
min_len: 300
- words_num_filter:
lang: en
tokenization: False
min_num: 30
max_num: 5000
- word_repetition_filter:
lang: en
tokenization: False
rep_len: 10
max_ratio: 0.1
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
### TheStack code refinement (only max_stars_count >= 20)
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-the-stack'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
text_key: 'content'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- clean_copyright_mapper:
- alphanumeric_filter: # 18766
tokenization: false
min_ratio: 0.2 # < 3sigma (0.3791)
max_ratio: 0.9163 # 3sigma
- alphanumeric_filter: # 146432
tokenization: true
min_ratio: 0.546 # 3sigma
max_ratio: 3.65 # 3sigma
- average_line_length_filter: # for code
min_len: 10 # > 3sigma (0) -- 48790
max_len: 150 # < 3sigma (15603) -- 233275
- character_repetition_filter:
max_ratio: 0.36 # 3sigma -- 346875
- maximum_line_length_filter: # for code
max_len: 1000 # remove 256670 samples
- text_length_filter:
max_len: 96714 # 3sigma -- 190006
- words_num_filter:
min_num: 20 # remove 1504958 samples
max_num: 6640 # 3sigma -- remove 179847 samples
- word_repetition_filter:
rep_len: 10
max_ratio: 0.357 # 3sigma -- 598462
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
### Merge and Remove Duplicate Samples
```yaml
project_name: 'Data-Juicer-recipes-code'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl' # path to your dataset result file
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/redpajama-stack-code-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10M<n<100M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:45:44+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10M<n<100M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:47:08+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama & TheStack -- Github Code (refined by Data-Juicer)
A refined version of Github Code dataset in RedPajama & TheStack by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 232GB).
## Dataset Information
- Number of samples: 49,279,344 (Keep ~52.09% from the original dataset)
## Refining Recipe
### RedPajama code refinement
### TheStack code refinement (only max_stars_count >= 20)
### Merge and Remove Duplicate Samples
|
[
"# RedPajama & TheStack -- Github Code (refined by Data-Juicer)\n\nA refined version of Github Code dataset in RedPajama & TheStack by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 232GB).",
"## Dataset Information\n\n- Number of samples: 49,279,344 (Keep ~52.09% from the original dataset)",
"## Refining Recipe",
"### RedPajama code refinement",
"### TheStack code refinement (only max_stars_count >= 20)",
"### Merge and Remove Duplicate Samples"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama & TheStack -- Github Code (refined by Data-Juicer)\n\nA refined version of Github Code dataset in RedPajama & TheStack by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 232GB).",
"## Dataset Information\n\n- Number of samples: 49,279,344 (Keep ~52.09% from the original dataset)",
"## Refining Recipe",
"### RedPajama code refinement",
"### TheStack code refinement (only max_stars_count >= 20)",
"### Merge and Remove Duplicate Samples"
] |
[
49,
114,
28,
4,
9,
22,
11
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama & TheStack -- Github Code (refined by Data-Juicer)\n\nA refined version of Github Code dataset in RedPajama & TheStack by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 232GB).## Dataset Information\n\n- Number of samples: 49,279,344 (Keep ~52.09% from the original dataset)## Refining Recipe### RedPajama code refinement### TheStack code refinement (only max_stars_count >= 20)### Merge and Remove Duplicate Samples"
] |
6d93989056b3ae7e996e5a78a886358ef3dbf887
|
# The Pile -- EuroParl (refined by Data-Juicer)
A refined version of EuroParl dataset in The Pile by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/the-pile-europarl-refine-result.jsonl) (About 2.2GB).
## Dataset Information
- Number of samples: 61,601 (Keep ~88.23% from the original dataset)
## Refining Recipe
```yaml
# global parameters
# global parameters
project_name: 'Data-Juicer-recipes-EuroParl'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.75 # <3sigma (0.779)
max_ratio: 0.90 # >3sigma(0.878)
- average_line_length_filter: # for code
max_len: 588 # 3sigma
- character_repetition_filter:
rep_len: 10
max_ratio: 0.16 # >3sigma (0.114)
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.0007 # 3sigma
- language_id_score_filter:
min_score: 0.7
- maximum_line_length_filter: # for code
max_len: 4000 # >3sigma (3104)
- perplexity_filter:
lang: en
max_ppl: 7596 #(3sigma)
- special_characters_filter:
max_ratio: 0.3 # > 3sigma (0.243)
- text_length_filter:
max_len: 2e5
- words_num_filter:
tokenization: true
min_num: 20
max_num: 1e5 # 3sigma
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.2 # > 3sigma (0.185)
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/the-pile-europarl-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10K<n<100K",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:47:08+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T08:21:31+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# The Pile -- EuroParl (refined by Data-Juicer)
A refined version of EuroParl dataset in The Pile by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 2.2GB).
## Dataset Information
- Number of samples: 61,601 (Keep ~88.23% from the original dataset)
## Refining Recipe
|
[
"# The Pile -- EuroParl (refined by Data-Juicer)\n\nA refined version of EuroParl dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 2.2GB).",
"## Dataset Information\n\n- Number of samples: 61,601 (Keep ~88.23% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# The Pile -- EuroParl (refined by Data-Juicer)\n\nA refined version of EuroParl dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 2.2GB).",
"## Dataset Information\n\n- Number of samples: 61,601 (Keep ~88.23% from the original dataset)",
"## Refining Recipe"
] |
[
49,
103,
26,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10K<n<100K #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# The Pile -- EuroParl (refined by Data-Juicer)\n\nA refined version of EuroParl dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 2.2GB).## Dataset Information\n\n- Number of samples: 61,601 (Keep ~88.23% from the original dataset)## Refining Recipe"
] |
46a7a39c505d2f9c46d652d86aeed63b1342d55b
|
# RedPajama & The Pile -- StackExchange (refined by Data-Juicer)
A refined version of StackExchange dataset in RedPajama & The Pile by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original merged dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-pile-stackexchange-refine-result.jsonl) (About 71GB).
## Dataset Information
- Number of samples: 26,309,203 (Keep ~57.89% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-stack-exchange'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.35 # <3sigma
max_ratio: 0.943 # 3sigma
- average_line_length_filter: # for code
min_len: 20 # >3sigma
max_len: 400 # >3sigma
- character_repetition_filter:
rep_len: 10
max_ratio: 0.4 # >3sigma (0.12)
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.01 # >3sigma
- language_id_score_filter: # remove language filter
min_score: 0.1 # <3sigma
- maximum_line_length_filter: # for code
min_len: 80
- perplexity_filter:
lang: en
max_ppl: 10000 # >3sigma
- special_characters_filter:
min_ratio: 0.232 # 3sigma
max_ratio: 0.7 # >3sigma
- text_length_filter:
min_len: 200
- words_num_filter:
lang: en
tokenization: true
min_num: 100
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.8 # >3sigma
- document_simhash_deduplicator: #26309203 left
tokenization: space
window_size: 3
lowercase: true
ignore_pattern: '\n\n'
num_blocks: 9
hamming_distance: 7
```
|
datajuicer/redpajama-pile-stackexchange-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10M<n<100M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:48:37+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10M<n<100M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:50:23+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama & The Pile -- StackExchange (refined by Data-Juicer)
A refined version of StackExchange dataset in RedPajama & The Pile by Data-Juicer. Removing some "bad" samples from the original merged dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 71GB).
## Dataset Information
- Number of samples: 26,309,203 (Keep ~57.89% from the original dataset)
## Refining Recipe
|
[
"# RedPajama & The Pile -- StackExchange (refined by Data-Juicer)\n\nA refined version of StackExchange dataset in RedPajama & The Pile by Data-Juicer. Removing some \"bad\" samples from the original merged dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 71GB).",
"## Dataset Information\n\n- Number of samples: 26,309,203\t (Keep ~57.89% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama & The Pile -- StackExchange (refined by Data-Juicer)\n\nA refined version of StackExchange dataset in RedPajama & The Pile by Data-Juicer. Removing some \"bad\" samples from the original merged dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 71GB).",
"## Dataset Information\n\n- Number of samples: 26,309,203\t (Keep ~57.89% from the original dataset)",
"## Refining Recipe"
] |
[
49,
115,
28,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama & The Pile -- StackExchange (refined by Data-Juicer)\n\nA refined version of StackExchange dataset in RedPajama & The Pile by Data-Juicer. Removing some \"bad\" samples from the original merged dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 71GB).## Dataset Information\n\n- Number of samples: 26,309,203\t (Keep ~57.89% from the original dataset)## Refining Recipe"
] |
f55cae0e57555c2b9550ec62d7e1a9137d986987
|
# RedPajama -- Wikipedia (refined by Data-Juicer)
A refined version of Wikipedia dataset in RedPajama by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-wiki-refine-result.jsonl) (About 68GB).
## Dataset Information
- Number of samples: 26,990,659 (Keep ~90.47% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-wiki'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.6 # <3sigma (0.735)
max_ratio: 0.884 # 3sigma
- average_line_length_filter: # for code
max_len: 192 # 3sigma
- character_repetition_filter:
rep_len: 10
max_ratio: 0.4 # >3sigma (0.197)
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.0019 # 3sigma
- language_id_score_filter:
min_score: 0.689 # 3sigma
- maximum_line_length_filter: # for code
max_len: 1630 # 3sigma tbd
- perplexity_filter:
lang: en
max_ppl: 6887 # 3sigma
- special_characters_filter:
max_ratio: 0.5 # >3sigma (0.34)
- text_length_filter:
max_len: 18221 # 3sigma
- words_num_filter:
lang: en
tokenization: true
min_num: 20
max_num: 6086 # 3sigma
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.3 # 3sigma (0.194)
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/redpajama-wiki-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10M<n<100M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:50:55+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10M<n<100M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:52:08+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama -- Wikipedia (refined by Data-Juicer)
A refined version of Wikipedia dataset in RedPajama by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 68GB).
## Dataset Information
- Number of samples: 26,990,659 (Keep ~90.47% from the original dataset)
## Refining Recipe
|
[
"# RedPajama -- Wikipedia (refined by Data-Juicer)\n\nA refined version of Wikipedia dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 68GB).",
"## Dataset Information\n\n- Number of samples: 26,990,659 (Keep ~90.47% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama -- Wikipedia (refined by Data-Juicer)\n\nA refined version of Wikipedia dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 68GB).",
"## Dataset Information\n\n- Number of samples: 26,990,659 (Keep ~90.47% from the original dataset)",
"## Refining Recipe"
] |
[
49,
99,
27,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama -- Wikipedia (refined by Data-Juicer)\n\nA refined version of Wikipedia dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 68GB).## Dataset Information\n\n- Number of samples: 26,990,659 (Keep ~90.47% from the original dataset)## Refining Recipe"
] |
49c7ec9b67ff0a9c807583aa24b0208f53d5435a
|
This dataset represents some data that Ines annotated. I am adding this info manually.
|
koaning/fashion-test
|
[
"task_categories:token-classification",
"task_ids:named-entity-recognition",
"annotations_creators:expert-generated",
"language_creators:found",
"multilinguality:monolingual",
"size_categories:1K<n<10K",
"source_datasets:original",
"language:en",
"license:mit",
"reddit",
"fashion",
"region:us"
] |
2023-10-23T07:51:30+00:00
|
{"annotations_creators": ["expert-generated"], "language_creators": ["found"], "language": ["en"], "license": ["mit"], "multilinguality": ["monolingual"], "size_categories": ["1K<n<10K"], "source_datasets": ["original"], "task_categories": ["token-classification"], "task_ids": ["named-entity-recognition"], "pretty_name": "Detecting fashion substrings in text.", "dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "meta", "struct": [{"name": "section", "dtype": "string"}]}, {"name": "_input_hash", "dtype": "int64"}, {"name": "_task_hash", "dtype": "int64"}, {"name": "tokens", "list": [{"name": "end", "dtype": "int64"}, {"name": "id", "dtype": "int64"}, {"name": "start", "dtype": "int64"}, {"name": "text", "dtype": "string"}]}, {"name": "spans", "list": [{"name": "end", "dtype": "int64"}, {"name": "input_hash", "dtype": "int64"}, {"name": "label", "dtype": "string"}, {"name": "source", "dtype": "string"}, {"name": "start", "dtype": "int64"}, {"name": "text", "dtype": "string"}, {"name": "token_end", "dtype": "int64"}, {"name": "token_start", "dtype": "int64"}]}, {"name": "_session_id", "dtype": "null"}, {"name": "_view_id", "dtype": "string"}, {"name": "answer", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 3120984, "num_examples": 1735}], "download_size": 817069, "dataset_size": 3120984}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}], "tags": ["reddit", "fashion"]}
|
2023-10-23T09:57:03+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-token-classification #task_ids-named-entity-recognition #annotations_creators-expert-generated #language_creators-found #multilinguality-monolingual #size_categories-1K<n<10K #source_datasets-original #language-English #license-mit #reddit #fashion #region-us
|
This dataset represents some data that Ines annotated. I am adding this info manually.
|
[] |
[
"TAGS\n#task_categories-token-classification #task_ids-named-entity-recognition #annotations_creators-expert-generated #language_creators-found #multilinguality-monolingual #size_categories-1K<n<10K #source_datasets-original #language-English #license-mit #reddit #fashion #region-us \n"
] |
[
96
] |
[
"passage: TAGS\n#task_categories-token-classification #task_ids-named-entity-recognition #annotations_creators-expert-generated #language_creators-found #multilinguality-monolingual #size_categories-1K<n<10K #source_datasets-original #language-English #license-mit #reddit #fashion #region-us \n"
] |
e68c95f881d65f1dfba1ffaf976f9fe1c90bfb14
|
# RedPajama -- CommonCrawl-2021-04 (refined by Data-Juicer)
A refined version of CommonCrawl-2021-04 dataset in RedPajama by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-cc-refine-results/redpajama-cc-2021-04-refine-result.jsonl) (About 284GB).
## Dataset Information
- Number of samples: 44,724,752 (Keep ~45.23% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-cc-2021-04'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.7494 # 3sigma
max_ratio: 0.8595 # 3sigma -- 1001790
- average_line_length_filter: # for code
max_len: 1500 # < 3sigma (2817) -- 541131
- character_repetition_filter:
rep_len: 10
max_ratio: 0.3 # > 3sigma (0.1463) -- 159152
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.0019 # 3sigma -- 184714
- language_id_score_filter: # remove language filter
min_score: 0.786 # 3sigma -- 1995115
- maximum_line_length_filter: # for code
max_len: 5000 # < 3sigma -- 1076085
- perplexity_filter:
lang: en
max_ppl: 5000 # < 3sigma -- 906649
- special_characters_filter:
min_ratio: 0.15 # > 3sigma
max_ratio: 0.35 # > 3sigma -- 1046590
- text_length_filter:
max_len: 61592 # 3sigma -- 1114727
- words_num_filter:
lang: en
tokenization: true
min_num: 20 # > 3sigma
max_num: 12241 # 3sigma -- 1120334
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.3105 # 3sigma -- 2234933
```
|
datajuicer/redpajama-cc-2021-04-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10M<n<100M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:52:52+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10M<n<100M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:53:56+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama -- CommonCrawl-2021-04 (refined by Data-Juicer)
A refined version of CommonCrawl-2021-04 dataset in RedPajama by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 284GB).
## Dataset Information
- Number of samples: 44,724,752 (Keep ~45.23% from the original dataset)
## Refining Recipe
|
[
"# RedPajama -- CommonCrawl-2021-04 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2021-04 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 284GB).",
"## Dataset Information\n\n- Number of samples: 44,724,752 (Keep ~45.23% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama -- CommonCrawl-2021-04 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2021-04 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 284GB).",
"## Dataset Information\n\n- Number of samples: 44,724,752 (Keep ~45.23% from the original dataset)",
"## Refining Recipe"
] |
[
49,
114,
29,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama -- CommonCrawl-2021-04 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2021-04 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 284GB).## Dataset Information\n\n- Number of samples: 44,724,752 (Keep ~45.23% from the original dataset)## Refining Recipe"
] |
0e520b7bad4078f4bfd5f592a22842c24faed4bb
|
# RedPajama -- CommonCrawl-2020-05 (refined by Data-Juicer)
A refined version of CommonCrawl-2020-05 dataset in RedPajama by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-cc-refine-results/redpajama-cc-2020-05-refine-result.jsonl) (About 297GB).
## Dataset Information
- Number of samples: 42,612,596 (Keep ~46.90% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-cc-2020-05'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.7469 # 3sigma
max_ratio: 0.8609 # 3sigma
- average_line_length_filter: # for code
max_len: 1500 # < 3sigma -- 332621
- character_repetition_filter:
rep_len: 10
max_ratio: 0.3 # > 3sigma -- 170501
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.002 # 3sigma -- 167167
- language_id_score_filter: # remove language filter
min_score: 0.774 # 3sigma -- 1943513
- maximum_line_length_filter: # for code
max_len: 5000 # < 3sigma -- 845490
- perplexity_filter:
lang: en
max_ppl: 5000 # < 3sigma -- 909218
- special_characters_filter:
min_ratio: 0.15 # > 3sigma
max_ratio: 0.35 # > 3sigma -- 1134347
- text_length_filter:
max_len: 68161 # 3sigma -- 1145902
- words_num_filter:
lang: en
tokenization: true
min_num: 20 # remove 7913 samples
max_num: 13644 # 3sigma -- 1148810
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.328 # 3sigma -- 2125070
```
|
datajuicer/redpajama-cc-2020-05-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10M<n<100M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:54:30+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10M<n<100M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:55:35+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama -- CommonCrawl-2020-05 (refined by Data-Juicer)
A refined version of CommonCrawl-2020-05 dataset in RedPajama by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 297GB).
## Dataset Information
- Number of samples: 42,612,596 (Keep ~46.90% from the original dataset)
## Refining Recipe
|
[
"# RedPajama -- CommonCrawl-2020-05 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2020-05 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 297GB).",
"## Dataset Information\n\n- Number of samples: 42,612,596 (Keep ~46.90% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama -- CommonCrawl-2020-05 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2020-05 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 297GB).",
"## Dataset Information\n\n- Number of samples: 42,612,596 (Keep ~46.90% from the original dataset)",
"## Refining Recipe"
] |
[
49,
110,
28,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama -- CommonCrawl-2020-05 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2020-05 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 297GB).## Dataset Information\n\n- Number of samples: 42,612,596 (Keep ~46.90% from the original dataset)## Refining Recipe"
] |
cafb431872997c3926a99411d172d8e481056f0e
|
# Dataset Card for "celebrity_prompts"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
Falah/celebrity_prompts
|
[
"region:us"
] |
2023-10-23T07:55:32+00:00
|
{"dataset_info": {"features": [{"name": "prompts", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 296727, "num_examples": 1000}], "download_size": 3413, "dataset_size": 296727}}
|
2023-10-23T07:55:33+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "celebrity_prompts"
More Information needed
|
[
"# Dataset Card for \"celebrity_prompts\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"celebrity_prompts\"\n\nMore Information needed"
] |
[
6,
17
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"celebrity_prompts\"\n\nMore Information needed"
] |
686af020736016d1b8dfbb9cab833c864a7b1f9b
|
# RedPajama -- CommonCrawl-2019-30 (refined by Data-Juicer)
A refined version of CommonCrawl-2019-30 dataset in RedPajama by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-cc-refine-results/redpajama-cc-2019-30-refine-result.jsonl) (About 240GB).
## Dataset Information
- Number of samples: 36,557,283 (Keep ~45.08% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-cc-2019-30'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter: # 770218
tokenization: false
min_ratio: 0.7489 # 3sigma
max_ratio: 0.8585 # 3sigma
- average_line_length_filter: # for code
max_len: 1500 # < 3sigma (2689) -- 177520
- character_repetition_filter:
rep_len: 10
max_ratio: 0.3 # > 3sigma (0.1491) -- 151703
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.0025 # 3sigma -- 101540
- language_id_score_filter: # remove language filter
min_score: 0.788 # 3sigma -- 1622574
- maximum_line_length_filter: # for code
max_len: 5000 # < 3sigma (8775) -- 485806
- perplexity_filter:
lang: en
max_ppl: 5000 # < 3sigma (6723) -- 676914
- special_characters_filter:
min_ratio: 0.15 # > 3sigma (0.104)
max_ratio: 0.35 # > 3sigma (0.322) -- 859797
- text_length_filter:
max_len: 65589 # 3sigma -- 975142
- words_num_filter:
lang: en
tokenization: true
min_num: 20 # > 3sigma -- 196
max_num: 13030 # 3sigma -- 989078
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.279 # 3sigma -- 1716308
```
|
datajuicer/redpajama-cc-2019-30-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10M<n<100M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:56:10+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10M<n<100M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:57:10+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama -- CommonCrawl-2019-30 (refined by Data-Juicer)
A refined version of CommonCrawl-2019-30 dataset in RedPajama by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 240GB).
## Dataset Information
- Number of samples: 36,557,283 (Keep ~45.08% from the original dataset)
## Refining Recipe
|
[
"# RedPajama -- CommonCrawl-2019-30 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2019-30 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 240GB).",
"## Dataset Information\n\n- Number of samples: 36,557,283 (Keep ~45.08% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama -- CommonCrawl-2019-30 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2019-30 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 240GB).",
"## Dataset Information\n\n- Number of samples: 36,557,283 (Keep ~45.08% from the original dataset)",
"## Refining Recipe"
] |
[
49,
109,
29,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama -- CommonCrawl-2019-30 (refined by Data-Juicer)\n\nA refined version of CommonCrawl-2019-30 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 240GB).## Dataset Information\n\n- Number of samples: 36,557,283 (Keep ~45.08% from the original dataset)## Refining Recipe"
] |
b670526128b47cc815d5aae922c5cbcb3d0aa938
|
# Alpaca-CoT -- EN (refined by Data-Juicer)
A refined English version of Alpaca-CoT dataset by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to fine-tune a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/CFT/alpaca-cot-en-refine_result.jsonl) (About 226GB).
## Dataset Information
- Number of samples: 72,855,345 (Keep ~54.48% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-alpaca-cot-en'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- document_deduplicator:
lowercase: true
ignore_non_character: true
- alphanumeric_filter:
tokenization: false
min_ratio: 0.1
- character_repetition_filter:
rep_len: 10
max_ratio: 0.6
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.017
- maximum_line_length_filter:
min_len: 20
- text_length_filter:
min_len: 30
- document_simhash_deduplicator:
tokenization: space
window_size: 3
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 9
hamming_distance: 7
```
|
datajuicer/alpaca-cot-en-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:10M<n<100M",
"language:en",
"license:apache-2.0",
"data-juicer",
"fine-tuning",
"region:us"
] |
2023-10-23T07:56:37+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["10M<n<100M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "fine-tuning"]}
|
2023-11-10T13:34:15+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #fine-tuning #region-us
|
# Alpaca-CoT -- EN (refined by Data-Juicer)
A refined English version of Alpaca-CoT dataset by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to fine-tune a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 226GB).
## Dataset Information
- Number of samples: 72,855,345 (Keep ~54.48% from the original dataset)
## Refining Recipe
|
[
"# Alpaca-CoT -- EN (refined by Data-Juicer)\n\nA refined English version of Alpaca-CoT dataset by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to fine-tune a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 226GB).",
"## Dataset Information\n\n- Number of samples: 72,855,345 (Keep ~54.48% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #fine-tuning #region-us \n",
"# Alpaca-CoT -- EN (refined by Data-Juicer)\n\nA refined English version of Alpaca-CoT dataset by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to fine-tune a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 226GB).",
"## Dataset Information\n\n- Number of samples: 72,855,345 (Keep ~54.48% from the original dataset)",
"## Refining Recipe"
] |
[
51,
103,
28,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-10M<n<100M #language-English #license-apache-2.0 #data-juicer #fine-tuning #region-us \n# Alpaca-CoT -- EN (refined by Data-Juicer)\n\nA refined English version of Alpaca-CoT dataset by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to fine-tune a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 226GB).## Dataset Information\n\n- Number of samples: 72,855,345 (Keep ~54.48% from the original dataset)## Refining Recipe"
] |
edb68cf0b897ebdf43d40341c63c9181f7c8949d
|
# RedPajama -- Book (refined by Data-Juicer)
A refined version of Book dataset in RedPajama by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-book-refine-result.jsonl) (About 91GB).
## Dataset Information
- Number of samples: 195,983 (Keep ~95.51% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-book'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.55 # <3sigma (0.697)
max_ratio: 0.854 # 3sigma
- average_line_length_filter: # for code
max_len: 500 # >3sigma (364)
- character_repetition_filter:
rep_len: 10
max_ratio: 0.2 # >3sigma (0.12)
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.00047 # 3sigma
- language_id_score_filter: # remove language filter
min_score: 0.2
- maximum_line_length_filter: # for code
max_len: 13381 # 3sigma
- perplexity_filter:
lang: en
max_ppl: 6000 # <3sigma (16516)
- special_characters_filter:
max_ratio: 0.5 # >3sigma (0.32)
- words_num_filter:
lang: en
tokenization: true
min_num: 1000
max_num: 539754 # 3sigma
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.194 # 3sigma
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/redpajama-book-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:100K<n<1M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:57:47+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["100K<n<1M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T07:59:08+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama -- Book (refined by Data-Juicer)
A refined version of Book dataset in RedPajama by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 91GB).
## Dataset Information
- Number of samples: 195,983 (Keep ~95.51% from the original dataset)
## Refining Recipe
|
[
"# RedPajama -- Book (refined by Data-Juicer)\n\nA refined version of Book dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 91GB).",
"## Dataset Information\n\n- Number of samples: 195,983 (Keep ~95.51% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama -- Book (refined by Data-Juicer)\n\nA refined version of Book dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 91GB).",
"## Dataset Information\n\n- Number of samples: 195,983 (Keep ~95.51% from the original dataset)",
"## Refining Recipe"
] |
[
49,
99,
26,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama -- Book (refined by Data-Juicer)\n\nA refined version of Book dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 91GB).## Dataset Information\n\n- Number of samples: 195,983 (Keep ~95.51% from the original dataset)## Refining Recipe"
] |
43d4957fa6f10452a8735d5211d5a876efffcba3
|
# Dataset Card for "dolphin"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
chrisgru/dolphin
|
[
"region:us"
] |
2023-10-23T07:58:42+00:00
|
{"dataset_info": {"features": [{"name": "system_prompt", "dtype": "string"}, {"name": "question", "dtype": "string"}, {"name": "response", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 15782.991668743547, "num_examples": 9}], "download_size": 10003, "dataset_size": 15782.991668743547}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T08:17:46+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "dolphin"
More Information needed
|
[
"# Dataset Card for \"dolphin\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"dolphin\"\n\nMore Information needed"
] |
[
6,
13
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"dolphin\"\n\nMore Information needed"
] |
d6480a724e564bef3ddf157d648ce13d89e18b33
|
# Dataset Card for "SECOND_SUMMARY_RAW"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
jjonhwa/SECOND_SUMMARY_RAW
|
[
"region:us"
] |
2023-10-23T07:59:24+00:00
|
{"dataset_info": {"features": [{"name": "\ubb38\uc7a5", "dtype": "string"}, {"name": "\uc694\uc57d", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 104066240, "num_examples": 30979}], "download_size": 62857147, "dataset_size": 104066240}}
|
2023-10-23T07:59:39+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "SECOND_SUMMARY_RAW"
More Information needed
|
[
"# Dataset Card for \"SECOND_SUMMARY_RAW\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"SECOND_SUMMARY_RAW\"\n\nMore Information needed"
] |
[
6,
19
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"SECOND_SUMMARY_RAW\"\n\nMore Information needed"
] |
62e312865179dfa357c406edbdc81d34ec75da94
|
# RedPajama -- C4 (refined by Data-Juicer)
A refined version of C4 dataset in RedPajama by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/redpajama-c4-refine-result.jsonl) (About 832GB).
## Dataset Information
- Number of samples: 344,491,171 (Keep ~94.42% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-c4'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl' # path to your dataset result file
np: 50 # number of subprocess to process your dataset
open_tracer: True
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.65 # <3sigma (0.740)
max_ratio: 0.9 # >3sigma (0.867)
- average_line_length_filter: # for code
max_len: 3000 # >3sigma (1277)
- character_repetition_filter:
rep_len: 10
max_ratio: 0.3 # >3sigma (0.168)
- language_id_score_filter:
min_score: 0.6
- maximum_line_length_filter: # for code
max_len: 4000 # >3sigma (2017)
- perplexity_filter:
lang: en
max_ppl: 6000 #(>3sigma 4543)
- special_characters_filter:
max_ratio: 0.4 # > 3sigma (0.303)
- words_num_filter:
tokenization: true
min_num: 20
max_num: 10000
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.231 # 3sigma
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/redpajama-c4-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:100M<n<1B",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T07:59:29+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["100M<n<1B"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T08:00:40+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-100M<n<1B #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# RedPajama -- C4 (refined by Data-Juicer)
A refined version of C4 dataset in RedPajama by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 832GB).
## Dataset Information
- Number of samples: 344,491,171 (Keep ~94.42% from the original dataset)
## Refining Recipe
|
[
"# RedPajama -- C4 (refined by Data-Juicer)\n\nA refined version of C4 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 832GB).",
"## Dataset Information\n\n- Number of samples: 344,491,171 (Keep ~94.42% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-100M<n<1B #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# RedPajama -- C4 (refined by Data-Juicer)\n\nA refined version of C4 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 832GB).",
"## Dataset Information\n\n- Number of samples: 344,491,171 (Keep ~94.42% from the original dataset)",
"## Refining Recipe"
] |
[
49,
102,
29,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-100M<n<1B #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# RedPajama -- C4 (refined by Data-Juicer)\n\nA refined version of C4 dataset in RedPajama by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 832GB).## Dataset Information\n\n- Number of samples: 344,491,171 (Keep ~94.42% from the original dataset)## Refining Recipe"
] |
17c5c263701e953189157f91c8029da91f190faf
|
# Dataset Card for "llama-2-nuv-intent-noE-xl"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
Luciya/llama-2-nuv-intent-noE-xl
|
[
"region:us"
] |
2023-10-23T08:00:04+00:00
|
{"dataset_info": {"features": [{"name": "text", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 1420374, "num_examples": 3224}], "download_size": 224581, "dataset_size": 1420374}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]}
|
2023-10-23T08:00:06+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "llama-2-nuv-intent-noE-xl"
More Information needed
|
[
"# Dataset Card for \"llama-2-nuv-intent-noE-xl\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"llama-2-nuv-intent-noE-xl\"\n\nMore Information needed"
] |
[
6,
24
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"llama-2-nuv-intent-noE-xl\"\n\nMore Information needed"
] |
9e9101119a9ddfb4ddb7e28797d7447a2e88caba
|
# The Pile -- HackerNews (refined by Data-Juicer)
A refined version of HackerNews dataset in The Pile by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/the-pile-hackernews-refine-result.jsonl) (About 1.8G).
## Dataset Information
- Number of samples: 371,331 (Keep ~99.55% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-HackerNews'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 48 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
#- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.2 #<3sigma
- average_line_length_filter:
min_len: 15 # >3sigma
- character_repetition_filter:
rep_len: 10
max_ratio: 0.3 # >3sigma
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.05 # >3sigma
- language_id_score_filter:
min_score: 0.2 # <3sigma
- maximum_line_length_filter:
min_len: 20 # >3sigma
- perplexity_filter:
lang: en
max_ppl: 10000 # >3sigma
- special_characters_filter:
max_ratio: 0.7 # >3sigma
- text_length_filter:
min_len: 100 # > 3sigma
- words_num_filter:
lang: en
tokenization: true
min_num: 30 # > 3sigma
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.8 # > 3sigma
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/the-pile-hackernews-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:100K<n<1M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T08:02:21+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["100K<n<1M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T08:11:09+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# The Pile -- HackerNews (refined by Data-Juicer)
A refined version of HackerNews dataset in The Pile by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 1.8G).
## Dataset Information
- Number of samples: 371,331 (Keep ~99.55% from the original dataset)
## Refining Recipe
|
[
"# The Pile -- HackerNews (refined by Data-Juicer)\n\nA refined version of HackerNews dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 1.8G).",
"## Dataset Information\n\n- Number of samples: 371,331 (Keep ~99.55% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# The Pile -- HackerNews (refined by Data-Juicer)\n\nA refined version of HackerNews dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 1.8G).",
"## Dataset Information\n\n- Number of samples: 371,331 (Keep ~99.55% from the original dataset)",
"## Refining Recipe"
] |
[
49,
103,
28,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# The Pile -- HackerNews (refined by Data-Juicer)\n\nA refined version of HackerNews dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 1.8G).## Dataset Information\n\n- Number of samples: 371,331 (Keep ~99.55% from the original dataset)## Refining Recipe"
] |
ed70b0d0e0c243d04f98739d4f6a61ce1cf3cf88
|
# Recognize Anything Plus Tag Descriptions Card
## Dataset details
**Dataset type:**
These tag descriptions files come from the RAM++ by calling GPT api.
**Dataset date:**
Recognize Anything Plus Tag Descriptions was collected in October 2023
**Paper or resources for more information:**
https://github.com/xinyu1205/recognize-anything
**Where to send questions or comments about the model:**
https://github.com/xinyu1205/recognize-anything/issues
## Intended use
**Primary intended uses:**
The primary use of Recognize Anything Plus Model is research on fundamental image recognition models.
**Primary intended users:**
The primary intended users of the model are researchers and hobbyists in computer vision, natural language processing, machine learning, and artificial intelligence.
|
xinyu1205/recognize-anything-plus-model-tag-descriptions
|
[
"task_categories:zero-shot-classification",
"language:en",
"license:apache-2.0",
"image recognition",
"region:us"
] |
2023-10-23T08:05:21+00:00
|
{"language": ["en"], "license": "apache-2.0", "task_categories": ["zero-shot-classification"], "tags": ["image recognition"]}
|
2023-10-23T11:12:12+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-zero-shot-classification #language-English #license-apache-2.0 #image recognition #region-us
|
# Recognize Anything Plus Tag Descriptions Card
## Dataset details
Dataset type:
These tag descriptions files come from the RAM++ by calling GPT api.
Dataset date:
Recognize Anything Plus Tag Descriptions was collected in October 2023
Paper or resources for more information:
URL
Where to send questions or comments about the model:
URL
## Intended use
Primary intended uses:
The primary use of Recognize Anything Plus Model is research on fundamental image recognition models.
Primary intended users:
The primary intended users of the model are researchers and hobbyists in computer vision, natural language processing, machine learning, and artificial intelligence.
|
[
"# Recognize Anything Plus Tag Descriptions Card",
"## Dataset details\n\nDataset type:\nThese tag descriptions files come from the RAM++ by calling GPT api.\n\nDataset date:\nRecognize Anything Plus Tag Descriptions was collected in October 2023\n\nPaper or resources for more information:\nURL\n\nWhere to send questions or comments about the model:\nURL",
"## Intended use\nPrimary intended uses:\nThe primary use of Recognize Anything Plus Model is research on fundamental image recognition models.\n\nPrimary intended users:\nThe primary intended users of the model are researchers and hobbyists in computer vision, natural language processing, machine learning, and artificial intelligence."
] |
[
"TAGS\n#task_categories-zero-shot-classification #language-English #license-apache-2.0 #image recognition #region-us \n",
"# Recognize Anything Plus Tag Descriptions Card",
"## Dataset details\n\nDataset type:\nThese tag descriptions files come from the RAM++ by calling GPT api.\n\nDataset date:\nRecognize Anything Plus Tag Descriptions was collected in October 2023\n\nPaper or resources for more information:\nURL\n\nWhere to send questions or comments about the model:\nURL",
"## Intended use\nPrimary intended uses:\nThe primary use of Recognize Anything Plus Model is research on fundamental image recognition models.\n\nPrimary intended users:\nThe primary intended users of the model are researchers and hobbyists in computer vision, natural language processing, machine learning, and artificial intelligence."
] |
[
34,
11,
62,
65
] |
[
"passage: TAGS\n#task_categories-zero-shot-classification #language-English #license-apache-2.0 #image recognition #region-us \n# Recognize Anything Plus Tag Descriptions Card## Dataset details\n\nDataset type:\nThese tag descriptions files come from the RAM++ by calling GPT api.\n\nDataset date:\nRecognize Anything Plus Tag Descriptions was collected in October 2023\n\nPaper or resources for more information:\nURL\n\nWhere to send questions or comments about the model:\nURL## Intended use\nPrimary intended uses:\nThe primary use of Recognize Anything Plus Model is research on fundamental image recognition models.\n\nPrimary intended users:\nThe primary intended users of the model are researchers and hobbyists in computer vision, natural language processing, machine learning, and artificial intelligence."
] |
b097c8dd1de473f89b59464331b8d6bfd1f4019a
|
# Dataset Card for Evaluation run of golaxy/gogpt2-13b
## Dataset Description
- **Homepage:**
- **Repository:** https://huggingface.co/golaxy/gogpt2-13b
- **Paper:**
- **Leaderboard:** https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
- **Point of Contact:** [email protected]
### Dataset Summary
Dataset automatically created during the evaluation run of model [golaxy/gogpt2-13b](https://huggingface.co/golaxy/gogpt2-13b) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).
To load the details from a run, you can for instance do the following:
```python
from datasets import load_dataset
data = load_dataset("open-llm-leaderboard/details_golaxy__gogpt2-13b",
"harness_winogrande_5",
split="train")
```
## Latest results
These are the [latest results from run 2023-10-23T09:08:59.644904](https://huggingface.co/datasets/open-llm-leaderboard/details_golaxy__gogpt2-13b/blob/main/results_2023-10-23T09-08-59.644904.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
```python
{
"all": {
"em": 0.25041946308724833,
"em_stderr": 0.004436932707454965,
"f1": 0.319959102348994,
"f1_stderr": 0.004400567822301105,
"acc": 0.3496193990687978,
"acc_stderr": 0.00851200635523702
},
"harness|drop|3": {
"em": 0.25041946308724833,
"em_stderr": 0.004436932707454965,
"f1": 0.319959102348994,
"f1_stderr": 0.004400567822301105
},
"harness|gsm8k|5": {
"acc": 0.02047005307050796,
"acc_stderr": 0.003900413385915719
},
"harness|winogrande|5": {
"acc": 0.6787687450670876,
"acc_stderr": 0.013123599324558321
}
}
```
### Supported Tasks and Leaderboards
[More Information Needed]
### Languages
[More Information Needed]
## Dataset Structure
### Data Instances
[More Information Needed]
### Data Fields
[More Information Needed]
### Data Splits
[More Information Needed]
## Dataset Creation
### Curation Rationale
[More Information Needed]
### Source Data
#### Initial Data Collection and Normalization
[More Information Needed]
#### Who are the source language producers?
[More Information Needed]
### Annotations
#### Annotation process
[More Information Needed]
#### Who are the annotators?
[More Information Needed]
### Personal and Sensitive Information
[More Information Needed]
## Considerations for Using the Data
### Social Impact of Dataset
[More Information Needed]
### Discussion of Biases
[More Information Needed]
### Other Known Limitations
[More Information Needed]
## Additional Information
### Dataset Curators
[More Information Needed]
### Licensing Information
[More Information Needed]
### Citation Information
[More Information Needed]
### Contributions
[More Information Needed]
|
open-llm-leaderboard/details_golaxy__gogpt2-13b
|
[
"region:us"
] |
2023-10-23T08:09:03+00:00
|
{"pretty_name": "Evaluation run of golaxy/gogpt2-13b", "dataset_summary": "Dataset automatically created during the evaluation run of model [golaxy/gogpt2-13b](https://huggingface.co/golaxy/gogpt2-13b) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\nTo load the details from a run, you can for instance do the following:\n```python\nfrom datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_golaxy__gogpt2-13b\",\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\nThese are the [latest results from run 2023-10-23T09:08:59.644904](https://huggingface.co/datasets/open-llm-leaderboard/details_golaxy__gogpt2-13b/blob/main/results_2023-10-23T09-08-59.644904.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.25041946308724833,\n \"em_stderr\": 0.004436932707454965,\n \"f1\": 0.319959102348994,\n \"f1_stderr\": 0.004400567822301105,\n \"acc\": 0.3496193990687978,\n \"acc_stderr\": 0.00851200635523702\n },\n \"harness|drop|3\": {\n \"em\": 0.25041946308724833,\n \"em_stderr\": 0.004436932707454965,\n \"f1\": 0.319959102348994,\n \"f1_stderr\": 0.004400567822301105\n },\n \"harness|gsm8k|5\": {\n \"acc\": 0.02047005307050796,\n \"acc_stderr\": 0.003900413385915719\n },\n \"harness|winogrande|5\": {\n \"acc\": 0.6787687450670876,\n \"acc_stderr\": 0.013123599324558321\n }\n}\n```", "repo_url": "https://huggingface.co/golaxy/gogpt2-13b", "leaderboard_url": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard", "point_of_contact": "[email protected]", "configs": [{"config_name": "harness_drop_3", "data_files": [{"split": "2023_10_23T09_08_59.644904", "path": ["**/details_harness|drop|3_2023-10-23T09-08-59.644904.parquet"]}, {"split": "latest", "path": ["**/details_harness|drop|3_2023-10-23T09-08-59.644904.parquet"]}]}, {"config_name": "harness_gsm8k_5", "data_files": [{"split": "2023_10_23T09_08_59.644904", "path": ["**/details_harness|gsm8k|5_2023-10-23T09-08-59.644904.parquet"]}, {"split": "latest", "path": ["**/details_harness|gsm8k|5_2023-10-23T09-08-59.644904.parquet"]}]}, {"config_name": "harness_winogrande_5", "data_files": [{"split": "2023_10_23T09_08_59.644904", "path": ["**/details_harness|winogrande|5_2023-10-23T09-08-59.644904.parquet"]}, {"split": "latest", "path": ["**/details_harness|winogrande|5_2023-10-23T09-08-59.644904.parquet"]}]}, {"config_name": "results", "data_files": [{"split": "2023_10_23T09_08_59.644904", "path": ["results_2023-10-23T09-08-59.644904.parquet"]}, {"split": "latest", "path": ["results_2023-10-23T09-08-59.644904.parquet"]}]}]}
|
2023-10-23T08:09:11+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for Evaluation run of golaxy/gogpt2-13b
## Dataset Description
- Homepage:
- Repository: URL
- Paper:
- Leaderboard: URL
- Point of Contact: clementine@URL
### Dataset Summary
Dataset automatically created during the evaluation run of model golaxy/gogpt2-13b on the Open LLM Leaderboard.
The dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).
To load the details from a run, you can for instance do the following:
## Latest results
These are the latest results from run 2023-10-23T09:08:59.644904(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval):
### Supported Tasks and Leaderboards
### Languages
## Dataset Structure
### Data Instances
### Data Fields
### Data Splits
## Dataset Creation
### Curation Rationale
### Source Data
#### Initial Data Collection and Normalization
#### Who are the source language producers?
### Annotations
#### Annotation process
#### Who are the annotators?
### Personal and Sensitive Information
## Considerations for Using the Data
### Social Impact of Dataset
### Discussion of Biases
### Other Known Limitations
## Additional Information
### Dataset Curators
### Licensing Information
### Contributions
|
[
"# Dataset Card for Evaluation run of golaxy/gogpt2-13b",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model golaxy/gogpt2-13b on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T09:08:59.644904(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for Evaluation run of golaxy/gogpt2-13b",
"## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL",
"### Dataset Summary\n\nDataset automatically created during the evaluation run of model golaxy/gogpt2-13b on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:",
"## Latest results\n\nThese are the latest results from run 2023-10-23T09:08:59.644904(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):",
"### Supported Tasks and Leaderboards",
"### Languages",
"## Dataset Structure",
"### Data Instances",
"### Data Fields",
"### Data Splits",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Initial Data Collection and Normalization",
"#### Who are the source language producers?",
"### Annotations",
"#### Annotation process",
"#### Who are the annotators?",
"### Personal and Sensitive Information",
"## Considerations for Using the Data",
"### Social Impact of Dataset",
"### Discussion of Biases",
"### Other Known Limitations",
"## Additional Information",
"### Dataset Curators",
"### Licensing Information",
"### Contributions"
] |
[
6,
18,
31,
166,
67,
10,
4,
6,
6,
5,
5,
5,
7,
4,
10,
10,
5,
5,
9,
8,
8,
7,
8,
7,
5,
6,
6,
5
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for Evaluation run of golaxy/gogpt2-13b## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: \n- Leaderboard: URL\n- Point of Contact: clementine@URL### Dataset Summary\n\nDataset automatically created during the evaluation run of model golaxy/gogpt2-13b on the Open LLM Leaderboard.\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the Open LLM Leaderboard).\n\nTo load the details from a run, you can for instance do the following:## Latest results\n\nThese are the latest results from run 2023-10-23T09:08:59.644904(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):### Supported Tasks and Leaderboards### Languages## Dataset Structure### Data Instances### Data Fields### Data Splits## Dataset Creation### Curation Rationale### Source Data#### Initial Data Collection and Normalization#### Who are the source language producers?### Annotations#### Annotation process#### Who are the annotators?### Personal and Sensitive Information## Considerations for Using the Data### Social Impact of Dataset### Discussion of Biases### Other Known Limitations## Additional Information### Dataset Curators### Licensing Information### Contributions"
] |
e66e47840d8b4af53097f219dc76103137157765
|
# The Pile -- FreeLaw (refined by Data-Juicer)
A refined version of FreeLaw dataset in The Pile by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/the-pile-freelaw-refine-result.jsonl) (About 45GB).
## Dataset Information
- Number of samples: 2,942,612 (Keep ~82.61% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-freelaw'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.3 # <3sigma (0.436)
- average_line_length_filter: # for code
max_len: 697 # 3sigma TBD
- character_repetition_filter:
rep_len: 10
max_ratio: 0.4 # >3sigma (0.350)
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.0053 # 3sigma
- language_id_score_filter:
min_score: 0.5 # < 3sigma (0.583)
- maximum_line_length_filter: # for code
max_len: 4229 # 3sigma
- perplexity_filter:
lang: en
max_ppl: 5322 # 3sigma
- special_characters_filter:
max_ratio: 0.7 # > 3sigma (0.626)
- stopwords_filter: # not use
lang: en
tokenization: true
min_ratio: 0.1 # > 3sigma (0.07)
- text_length_filter:
max_len: 84026 # 3sigma
- words_num_filter:
lang: en
tokenization: true
min_num: 100
max_num: 15208 # 3sigma
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.155 # 3sigma
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/the-pile-freelaw-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:1M<n<10M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T08:10:03+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["1M<n<10M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T08:12:32+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# The Pile -- FreeLaw (refined by Data-Juicer)
A refined version of FreeLaw dataset in The Pile by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 45GB).
## Dataset Information
- Number of samples: 2,942,612 (Keep ~82.61% from the original dataset)
## Refining Recipe
|
[
"# The Pile -- FreeLaw (refined by Data-Juicer)\n\nA refined version of FreeLaw dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 45GB).",
"## Dataset Information\n\n- Number of samples: 2,942,612 (Keep ~82.61% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# The Pile -- FreeLaw (refined by Data-Juicer)\n\nA refined version of FreeLaw dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 45GB).",
"## Dataset Information\n\n- Number of samples: 2,942,612 (Keep ~82.61% from the original dataset)",
"## Refining Recipe"
] |
[
49,
103,
28,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# The Pile -- FreeLaw (refined by Data-Juicer)\n\nA refined version of FreeLaw dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 45GB).## Dataset Information\n\n- Number of samples: 2,942,612 (Keep ~82.61% from the original dataset)## Refining Recipe"
] |
f6d7b6e7ef850266304d15167e58eb60fbea09ac
|
# Dataset Card for "sft_test_custom_dataset_RLHF"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
sayan1101/sft_test_custom_dataset_RLHF
|
[
"region:us"
] |
2023-10-23T08:11:29+00:00
|
{"dataset_info": {"features": [{"name": "label", "dtype": "string"}, {"name": "prompt", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 34685, "num_examples": 51}, {"name": "test", "num_bytes": 34685, "num_examples": 51}, {"name": "valid", "num_bytes": 34685, "num_examples": 51}], "download_size": 86937, "dataset_size": 104055}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "test", "path": "data/test-*"}, {"split": "valid", "path": "data/valid-*"}]}]}
|
2023-10-24T04:59:21+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "sft_test_custom_dataset_RLHF"
More Information needed
|
[
"# Dataset Card for \"sft_test_custom_dataset_RLHF\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"sft_test_custom_dataset_RLHF\"\n\nMore Information needed"
] |
[
6,
23
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"sft_test_custom_dataset_RLHF\"\n\nMore Information needed"
] |
a676a64785a5fe69230a9fd7594544f4d7fdc61f
|
# The Pile -- USPTO (refined by Data-Juicer)
A refined version of USPTO dataset in The Pile by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/the-pile-uspto-refine-result.jsonl) (About 18G).
## Dataset Information
- Number of samples: 4,516,283 (Keep ~46.77% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-uspto'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl' # path to your dataset result file
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter:
tokenization: false
min_ratio: 0.7 # <3sigma (0.758)
- average_line_length_filter: # for code
max_len: 2000 # >3sigma (1307)
- character_repetition_filter:
rep_len: 10
max_ratio: 0.2 # >3sigma (0.189)
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.0016 # 3sigma
- language_id_score_filter:
min_score: 0.6
- maximum_line_length_filter: # for code
max_len: 3061 # 3sigma
- perplexity_filter:
lang: en
max_ppl: 4000 # 3sigma
- special_characters_filter:
max_ratio: 0.3 # > 3sigma (0.274)
- text_length_filter:
max_len: 21556 # 3sigma
- words_num_filter:
lang: en
tokenization: true
min_num: 100
max_num: 6000 # 3sigma
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.169 # 3sigma
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/the-pile-uspto-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:1M<n<10M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T08:15:24+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["1M<n<10M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T08:18:18+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# The Pile -- USPTO (refined by Data-Juicer)
A refined version of USPTO dataset in The Pile by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 18G).
## Dataset Information
- Number of samples: 4,516,283 (Keep ~46.77% from the original dataset)
## Refining Recipe
|
[
"# The Pile -- USPTO (refined by Data-Juicer)\n\nA refined version of USPTO dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 18G).",
"## Dataset Information\n\n- Number of samples: 4,516,283 (Keep ~46.77% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# The Pile -- USPTO (refined by Data-Juicer)\n\nA refined version of USPTO dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 18G).",
"## Dataset Information\n\n- Number of samples: 4,516,283 (Keep ~46.77% from the original dataset)",
"## Refining Recipe"
] |
[
49,
103,
28,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# The Pile -- USPTO (refined by Data-Juicer)\n\nA refined version of USPTO dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 18G).## Dataset Information\n\n- Number of samples: 4,516,283 (Keep ~46.77% from the original dataset)## Refining Recipe"
] |
a0ee12fefb14434d17b9803d6bd3570511ed225a
|
# The Pile -- PubMed Central (refined by Data-Juicer)
A refined version of PubMed Central dataset in The Pile by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/the-pile-pubmed-central-refine-result.jsonl) (About 83G).
## Dataset Information
- Number of samples: 2,694,860 (Keep ~86.96% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-pubmed-central'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter: # 89217
tokenization: false
min_ratio: 0.2787 # 3sigma
- average_line_length_filter: # for code
max_len: 1200 # < 3sigma (1478) -- 7410
- character_repetition_filter:
rep_len: 10
max_ratio: 0.3741 # 3sigma -- 65849
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.00195 # 3sigma -- 8305
- language_id_score_filter: # remove language filter
min_score: 0.5 # 272359
- maximum_line_length_filter: # for code
max_len: 7328 # remove 23808 samples
- perplexity_filter:
lang: en
max_ppl: 8000 # remove 173883 samples
- special_characters_filter:
max_ratio: 0.842 # remove 87661 samples
- text_length_filter:
max_len: 136028 # 3sigma -- 15118
- words_num_filter:
lang: en
tokenization: true
min_num: 20 # remove 176537 samples
max_num: 23305 # remove 15016 samples
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.5981 # 3sigma -- 93843
- document_simhash_deduplicator:
tokenization: space
window_size: 6
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 6
hamming_distance: 4
```
|
datajuicer/the-pile-pubmed-central-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:1M<n<10M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T08:26:13+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["1M<n<10M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T08:30:08+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# The Pile -- PubMed Central (refined by Data-Juicer)
A refined version of PubMed Central dataset in The Pile by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 83G).
## Dataset Information
- Number of samples: 2,694,860 (Keep ~86.96% from the original dataset)
## Refining Recipe
|
[
"# The Pile -- PubMed Central (refined by Data-Juicer)\n\nA refined version of PubMed Central dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 83G).",
"## Dataset Information\n\n- Number of samples: 2,694,860 (Keep ~86.96% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# The Pile -- PubMed Central (refined by Data-Juicer)\n\nA refined version of PubMed Central dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 83G).",
"## Dataset Information\n\n- Number of samples: 2,694,860 (Keep ~86.96% from the original dataset)",
"## Refining Recipe"
] |
[
49,
103,
28,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# The Pile -- PubMed Central (refined by Data-Juicer)\n\nA refined version of PubMed Central dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 83G).## Dataset Information\n\n- Number of samples: 2,694,860 (Keep ~86.96% from the original dataset)## Refining Recipe"
] |
0ceb2056b1e8a9b686a758062ebc87bc5e83da5c
|
# Dataset Card for "ner_tokens"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
getawayfrommeXD/ner_tokens
|
[
"region:us"
] |
2023-10-23T08:28:27+00:00
|
{"dataset_info": {"features": [{"name": "word", "dtype": "string"}, {"name": "label", "dtype": "string"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 4492364, "num_examples": 203621}, {"name": "validation", "num_bytes": 1133031, "num_examples": 51362}, {"name": "test", "num_bytes": 1022873, "num_examples": 46435}], "download_size": 3296837, "dataset_size": 6648268}}
|
2023-10-23T08:28:38+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "ner_tokens"
More Information needed
|
[
"# Dataset Card for \"ner_tokens\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"ner_tokens\"\n\nMore Information needed"
] |
[
6,
14
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"ner_tokens\"\n\nMore Information needed"
] |
0393f4885199e3bdb88be752575485f08b11949c
|
# Dataset Card for "ner_sentences"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
getawayfrommeXD/ner_sentences
|
[
"region:us"
] |
2023-10-23T08:28:38+00:00
|
{"dataset_info": {"features": [{"name": "sentence", "sequence": "string"}, {"name": "labels", "sequence": "string"}], "splits": [{"name": "train", "num_bytes": 2975724, "num_examples": 14041}, {"name": "validation", "num_bytes": 748135, "num_examples": 3250}, {"name": "test", "num_bytes": 679017, "num_examples": 3453}], "download_size": 1244231, "dataset_size": 4402876}}
|
2023-10-23T08:28:47+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "ner_sentences"
More Information needed
|
[
"# Dataset Card for \"ner_sentences\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"ner_sentences\"\n\nMore Information needed"
] |
[
6,
14
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"ner_sentences\"\n\nMore Information needed"
] |
7886ed5b2c983beaae706c474c7743aa0b21d669
|
# Dataset Card for Dataset Name
<!-- Provide a quick summary of the dataset. -->
This dataset card aims to be a base template for new datasets. It has been generated using [this raw template](https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/datasetcard_template.md?plain=1).
## Dataset Details
### Dataset Description
<!-- Provide a longer summary of what this dataset is. -->
- **Curated by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
### Dataset Sources [optional]
<!-- Provide the basic links for the dataset. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the dataset is intended to be used. -->
### Direct Use
<!-- This section describes suitable use cases for the dataset. -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
[More Information Needed]
## Dataset Structure
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
[More Information Needed]
## Dataset Creation
### Curation Rationale
<!-- Motivation for the creation of this dataset. -->
[More Information Needed]
### Source Data
<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
#### Data Collection and Processing
<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
[More Information Needed]
#### Who are the source data producers?
<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
[More Information Needed]
### Annotations [optional]
<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
#### Annotation process
<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
[More Information Needed]
#### Who are the annotators?
<!-- This section describes the people or systems who created the annotations. -->
[More Information Needed]
#### Personal and Sensitive Information
<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
## Citation [optional]
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Dataset Card Authors [optional]
[More Information Needed]
## Dataset Card Contact
[More Information Needed]
|
qwertysingh/data3
|
[
"region:us"
] |
2023-10-23T08:30:40+00:00
|
{}
|
2023-10-23T08:32:06+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for Dataset Name
This dataset card aims to be a base template for new datasets. It has been generated using this raw template.
## Dataset Details
### Dataset Description
- Curated by:
- Funded by [optional]:
- Shared by [optional]:
- Language(s) (NLP):
- License:
### Dataset Sources [optional]
- Repository:
- Paper [optional]:
- Demo [optional]:
## Uses
### Direct Use
### Out-of-Scope Use
## Dataset Structure
## Dataset Creation
### Curation Rationale
### Source Data
#### Data Collection and Processing
#### Who are the source data producers?
### Annotations [optional]
#### Annotation process
#### Who are the annotators?
#### Personal and Sensitive Information
## Bias, Risks, and Limitations
### Recommendations
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
[optional]
BibTeX:
APA:
## Glossary [optional]
## More Information [optional]
## Dataset Card Authors [optional]
## Dataset Card Contact
|
[
"# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.",
"## Dataset Details",
"### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.",
"## Dataset Details",
"### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] |
[
6,
34,
4,
40,
29,
3,
4,
9,
6,
5,
7,
4,
7,
10,
9,
5,
9,
8,
10,
46,
8,
7,
10,
5
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.## Dataset Details### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:## Uses### Direct Use### Out-of-Scope Use## Dataset Structure## Dataset Creation### Curation Rationale### Source Data#### Data Collection and Processing#### Who are the source data producers?### Annotations [optional]#### Annotation process#### Who are the annotators?#### Personal and Sensitive Information## Bias, Risks, and Limitations### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:## Glossary [optional]## More Information [optional]## Dataset Card Authors [optional]## Dataset Card Contact"
] |
b21ba152b29f0b4eec87c2ffa85aca3c79d93df6
|
# Dataset Card for Dataset Name
<!-- Provide a quick summary of the dataset. -->
## Dataset Details
### Dataset Description
<!-- Provide a longer summary of what this dataset is. -->
- **Curated by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
### Dataset Sources [optional]
<!-- Provide the basic links for the dataset. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the dataset is intended to be used. -->
### Direct Use
<!-- This section describes suitable use cases for the dataset. -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
[More Information Needed]
## Dataset Structure
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
[More Information Needed]
## Dataset Creation
### Curation Rationale
<!-- Motivation for the creation of this dataset. -->
[More Information Needed]
### Source Data
<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
#### Data Collection and Processing
<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
[More Information Needed]
#### Who are the source data producers?
<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
[More Information Needed]
### Annotations [optional]
<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
#### Annotation process
<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
[More Information Needed]
#### Who are the annotators?
<!-- This section describes the people or systems who created the annotations. -->
[More Information Needed]
#### Personal and Sensitive Information
<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
## Citation [optional]
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Dataset Card Authors [optional]
[More Information Needed]
## Dataset Card Contact
[More Information Needed]
|
Wauplin/test_empty_data_card
|
[
"prodigy",
"region:us"
] |
2023-10-23T08:31:18+00:00
|
{"tags": ["prodigy"]}
|
2023-10-23T08:31:20+00:00
|
[] |
[] |
TAGS
#prodigy #region-us
|
# Dataset Card for Dataset Name
## Dataset Details
### Dataset Description
- Curated by:
- Funded by [optional]:
- Shared by [optional]:
- Language(s) (NLP):
- License:
### Dataset Sources [optional]
- Repository:
- Paper [optional]:
- Demo [optional]:
## Uses
### Direct Use
### Out-of-Scope Use
## Dataset Structure
## Dataset Creation
### Curation Rationale
### Source Data
#### Data Collection and Processing
#### Who are the source data producers?
### Annotations [optional]
#### Annotation process
#### Who are the annotators?
#### Personal and Sensitive Information
## Bias, Risks, and Limitations
### Recommendations
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
[optional]
BibTeX:
APA:
## Glossary [optional]
## More Information [optional]
## Dataset Card Authors [optional]
## Dataset Card Contact
|
[
"# Dataset Card for Dataset Name",
"## Dataset Details",
"### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] |
[
"TAGS\n#prodigy #region-us \n",
"# Dataset Card for Dataset Name",
"## Dataset Details",
"### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] |
[
10,
8,
4,
40,
29,
3,
4,
9,
6,
5,
7,
4,
7,
10,
9,
5,
9,
8,
10,
46,
8,
7,
10,
5
] |
[
"passage: TAGS\n#prodigy #region-us \n# Dataset Card for Dataset Name## Dataset Details### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:## Uses### Direct Use### Out-of-Scope Use## Dataset Structure## Dataset Creation### Curation Rationale### Source Data#### Data Collection and Processing#### Who are the source data producers?### Annotations [optional]#### Annotation process#### Who are the annotators?#### Personal and Sensitive Information## Bias, Risks, and Limitations### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:## Glossary [optional]## More Information [optional]## Dataset Card Authors [optional]## Dataset Card Contact"
] |
da7ed95cdfb134bb4df8e8232adda5b168058411
|
# The Pile -- PubMed Abstracts (refined by Data-Juicer)
A refined version of PubMed Abstracts dataset in The Pile by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/the-pile-pubmed-abstract-refine-result.jsonl) (About 24G).
## Dataset Information
- Number of samples: 371,331 (Keep ~99.55% from the original dataset)
## Refining Recipe
```yaml
# global parameters
project_name: 'Data-Juicer-recipes-pubmed-abstract'
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
export_path: '/path/to/your/dataset.jsonl'
np: 50 # number of subprocess to process your dataset
open_tracer: true
# process schedule
# a list of several process operators with their arguments
process:
- clean_email_mapper:
- clean_links_mapper:
- fix_unicode_mapper:
- punctuation_normalization_mapper:
- whitespace_normalization_mapper:
- alphanumeric_filter: # 4068
tokenization: false
min_ratio: 0.7 # < 3sigma (0.773)
max_ratio: 0.881 # 3sigma
- average_line_length_filter: # for code
max_len: 2100 # > 3sigma (1471) -- 7410
- character_repetition_filter:
rep_len: 10
max_ratio: 0.2 # > 3sigma (0.1458) -- 6060
- flagged_words_filter:
lang: en
tokenization: true
max_ratio: 0.00232 # 3sigma
- language_id_score_filter: # remove language filter
min_score: 0.5
- maximum_line_length_filter: # for code
max_len: 4000 # remove 8202 samples
- perplexity_filter:
lang: en
max_ppl: 4000 # remove 10284 samples
- special_characters_filter:
max_ratio: 0.38 # remove 5532 samples
- text_length_filter:
max_len: 4000 # > 3sigma -- 10873
- words_num_filter:
lang: en
tokenization: true
min_num: 20 # remove 10790 samples
max_num: 700 # remove 22709 samples
- word_repetition_filter:
lang: en
tokenization: true
rep_len: 10
max_ratio: 0.0887 # 3sigma
- document_simhash_deduplicator:
tokenization: space
window_size: 3 # small window size for short texts
lowercase: true
ignore_pattern: '\p{P}'
num_blocks: 10
hamming_distance: 8 # larger hamming distance threshold for short texts
```
|
datajuicer/the-pile-pubmed-abstracts-refined-by-data-juicer
|
[
"task_categories:text-generation",
"size_categories:100K<n<1M",
"language:en",
"license:apache-2.0",
"data-juicer",
"pretraining",
"region:us"
] |
2023-10-23T08:31:40+00:00
|
{"language": ["en"], "license": "apache-2.0", "size_categories": ["100K<n<1M"], "task_categories": ["text-generation"], "tags": ["data-juicer", "pretraining"]}
|
2023-10-23T08:34:32+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us
|
# The Pile -- PubMed Abstracts (refined by Data-Juicer)
A refined version of PubMed Abstracts dataset in The Pile by Data-Juicer. Removing some "bad" samples from the original dataset to make it higher-quality.
This dataset is usually used to pretrain a Large Language Model.
Notice: Here is a small subset for previewing. The whole dataset is available here (About 24G).
## Dataset Information
- Number of samples: 371,331 (Keep ~99.55% from the original dataset)
## Refining Recipe
|
[
"# The Pile -- PubMed Abstracts (refined by Data-Juicer)\n\nA refined version of PubMed Abstracts dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 24G).",
"## Dataset Information\n\n- Number of samples: 371,331 (Keep ~99.55% from the original dataset)",
"## Refining Recipe"
] |
[
"TAGS\n#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n",
"# The Pile -- PubMed Abstracts (refined by Data-Juicer)\n\nA refined version of PubMed Abstracts dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 24G).",
"## Dataset Information\n\n- Number of samples: 371,331 (Keep ~99.55% from the original dataset)",
"## Refining Recipe"
] |
[
49,
105,
28,
4
] |
[
"passage: TAGS\n#task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #data-juicer #pretraining #region-us \n# The Pile -- PubMed Abstracts (refined by Data-Juicer)\n\nA refined version of PubMed Abstracts dataset in The Pile by Data-Juicer. Removing some \"bad\" samples from the original dataset to make it higher-quality.\n\nThis dataset is usually used to pretrain a Large Language Model.\n\nNotice: Here is a small subset for previewing. The whole dataset is available here (About 24G).## Dataset Information\n\n- Number of samples: 371,331 (Keep ~99.55% from the original dataset)## Refining Recipe"
] |
68e8f920cae792a5f1467837a89853f827e980bb
|
# LLaVA Instruct Mix
Added OCR and Chart QA dataset into this for more text extraction questions
|
theblackcat102/llava-instruct-mix
|
[
"task_categories:visual-question-answering",
"size_categories:100K<n<1M",
"language:en",
"license:cc-by-nc-4.0",
"multimodal",
"vision",
"region:us"
] |
2023-10-23T08:34:45+00:00
|
{"language": ["en"], "license": "cc-by-nc-4.0", "size_categories": ["100K<n<1M"], "task_categories": ["visual-question-answering"], "dataset_info": {"features": [{"name": "image", "dtype": "image"}, {"name": "conversations", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 46019106088.205, "num_examples": 272795}], "download_size": 20289135489, "dataset_size": 46019106088.205}, "tags": ["multimodal", "vision"]}
|
2023-10-23T09:14:27+00:00
|
[] |
[
"en"
] |
TAGS
#task_categories-visual-question-answering #size_categories-100K<n<1M #language-English #license-cc-by-nc-4.0 #multimodal #vision #region-us
|
# LLaVA Instruct Mix
Added OCR and Chart QA dataset into this for more text extraction questions
|
[
"# LLaVA Instruct Mix\n\nAdded OCR and Chart QA dataset into this for more text extraction questions"
] |
[
"TAGS\n#task_categories-visual-question-answering #size_categories-100K<n<1M #language-English #license-cc-by-nc-4.0 #multimodal #vision #region-us \n",
"# LLaVA Instruct Mix\n\nAdded OCR and Chart QA dataset into this for more text extraction questions"
] |
[
54,
25
] |
[
"passage: TAGS\n#task_categories-visual-question-answering #size_categories-100K<n<1M #language-English #license-cc-by-nc-4.0 #multimodal #vision #region-us \n# LLaVA Instruct Mix\n\nAdded OCR and Chart QA dataset into this for more text extraction questions"
] |
6d04bc0565cb88aff707f8334ca6390b614e5896
|
# Dataset Card for "christmas-edits_dataset2"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
StivenLancheros/christmas-edits_dataset2
|
[
"region:us"
] |
2023-10-23T08:35:24+00:00
|
{"dataset_info": {"features": [{"name": "input_image", "dtype": "image"}, {"name": "edit_prompt", "dtype": "string"}, {"name": "edited_image", "dtype": "image"}], "splits": [{"name": "train", "num_bytes": 33235736614.672, "num_examples": 84776}], "download_size": 32967759861, "dataset_size": 33235736614.672}}
|
2023-10-23T09:35:18+00:00
|
[] |
[] |
TAGS
#region-us
|
# Dataset Card for "christmas-edits_dataset2"
More Information needed
|
[
"# Dataset Card for \"christmas-edits_dataset2\"\n\nMore Information needed"
] |
[
"TAGS\n#region-us \n",
"# Dataset Card for \"christmas-edits_dataset2\"\n\nMore Information needed"
] |
[
6,
20
] |
[
"passage: TAGS\n#region-us \n# Dataset Card for \"christmas-edits_dataset2\"\n\nMore Information needed"
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.