Hugging Face's logo

sampras343
/
wav2vec2-base-ft-keyword-spotting

Audio Classification
Transformers
PyTorch
TensorBoard
Generated from Trainer
Model card Files
xet
Metrics Community
sampras343 commited on
Commit
07f519d
·
1 Parent(s): 80d2803

Initial Commit

Browse files
Files changed (15) hide show
  1. README.md +70 -0
  2. all_results.json +12 -0
  3. config.json +117 -0
  4. emissions.csv +2 -0
  5. eval_results.json +8 -0
  6. preprocessor_config.json +9 -0
  7. pytorch_model.bin +3 -0
  8. runs/Oct27_22-01-18_anton-gpu-v100/1635372097.669254/events.out.tfevents.1635372097.anton-gpu-v100.276919.1 +3 -0
  9. runs/Oct27_22-01-18_anton-gpu-v100/events.out.tfevents.1635372097.anton-gpu-v100.276919.0 +3 -0
  10. runs/Oct27_22-01-18_anton-gpu-v100/events.out.tfevents.1635372990.anton-gpu-v100.276919.2 +3 -0
  11. tokenizer_config.json +1 -0
  12. train_results.json +7 -0
  13. trainer_state.json +1264 -0
  14. training_args.bin +3 -0
  15. vocab.json +1 -0
README.md ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - audio-classification
5
+ - generated_from_trainer
6
+ datasets:
7
+ - superb
8
+ metrics:
9
+ - accuracy
10
+ model-index:
11
+ - name: wav2vec2-base-ft-keyword-spotting
12
+ results: []
13
+ ---
14
+
15
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
16
+ should probably proofread and complete it, then remove this comment. -->
17
+
18
+ # wav2vec2-base-ft-keyword-spotting
19
+
20
+ This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the superb dataset.
21
+ It achieves the following results on the evaluation set:
22
+ - Loss: 0.0824
23
+ - Accuracy: 0.9826
24
+
25
+ ## Model description
26
+
27
+ More information needed
28
+
29
+ ## Intended uses & limitations
30
+
31
+ More information needed
32
+
33
+ ## Training and evaluation data
34
+
35
+ More information needed
36
+
37
+ ## Training procedure
38
+
39
+ ### Training hyperparameters
40
+
41
+ The following hyperparameters were used during training:
42
+ - learning_rate: 3e-05
43
+ - train_batch_size: 32
44
+ - eval_batch_size: 32
45
+ - seed: 0
46
+ - gradient_accumulation_steps: 4
47
+ - total_train_batch_size: 128
48
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
49
+ - lr_scheduler_type: linear
50
+ - lr_scheduler_warmup_ratio: 0.1
51
+ - num_epochs: 5.0
52
+ - mixed_precision_training: Native AMP
53
+
54
+ ### Training results
55
+
56
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
57
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
58
+ | 0.8972 | 1.0 | 399 | 0.7023 | 0.8174 |
59
+ | 0.3274 | 2.0 | 798 | 0.1634 | 0.9773 |
60
+ | 0.1993 | 3.0 | 1197 | 0.1048 | 0.9788 |
61
+ | 0.1777 | 4.0 | 1596 | 0.0824 | 0.9826 |
62
+ | 0.1527 | 5.0 | 1995 | 0.0812 | 0.9810 |
63
+
64
+
65
+ ### Framework versions
66
+
67
+ - Transformers 4.12.0.dev0
68
+ - Pytorch 1.9.1+cu111
69
+ - Datasets 1.14.0
70
+ - Tokenizers 0.10.3
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_accuracy": 0.9826419535157399,
4
+ "eval_loss": 0.08240818232297897,
5
+ "eval_runtime": 14.1019,
6
+ "eval_samples_per_second": 482.063,
7
+ "eval_steps_per_second": 15.104,
8
+ "train_loss": 0.5217096392672164,
9
+ "train_runtime": 876.8675,
10
+ "train_samples_per_second": 291.344,
11
+ "train_steps_per_second": 2.275
12
+ }
config.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-base",
3
+ "activation_dropout": 0.0,
4
+ "apply_spec_augment": true,
5
+ "architectures": [
6
+ "Wav2Vec2ForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "classifier_proj_size": 256,
11
+ "codevector_dim": 256,
12
+ "contrastive_logits_temperature": 0.1,
13
+ "conv_bias": false,
14
+ "conv_dim": [
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512
22
+ ],
23
+ "conv_kernel": [
24
+ 10,
25
+ 3,
26
+ 3,
27
+ 3,
28
+ 3,
29
+ 2,
30
+ 2
31
+ ],
32
+ "conv_stride": [
33
+ 5,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2,
38
+ 2,
39
+ 2
40
+ ],
41
+ "ctc_loss_reduction": "sum",
42
+ "ctc_zero_infinity": false,
43
+ "diversity_loss_weight": 0.1,
44
+ "do_stable_layer_norm": false,
45
+ "eos_token_id": 2,
46
+ "feat_extract_activation": "gelu",
47
+ "feat_extract_norm": "group",
48
+ "feat_proj_dropout": 0.1,
49
+ "feat_quantizer_dropout": 0.0,
50
+ "final_dropout": 0.0,
51
+ "finetuning_task": "audio-classification",
52
+ "freeze_feat_extract_train": true,
53
+ "hidden_act": "gelu",
54
+ "hidden_dropout": 0.1,
55
+ "hidden_size": 768,
56
+ "id2label": {
57
+ "0": "yes",
58
+ "1": "no",
59
+ "10": "_silence_",
60
+ "11": "_unknown_",
61
+ "2": "up",
62
+ "3": "down",
63
+ "4": "left",
64
+ "5": "right",
65
+ "6": "on",
66
+ "7": "off",
67
+ "8": "stop",
68
+ "9": "go"
69
+ },
70
+ "initializer_range": 0.02,
71
+ "intermediate_size": 3072,
72
+ "label2id": {
73
+ "_silence_": "10",
74
+ "_unknown_": "11",
75
+ "down": "3",
76
+ "go": "9",
77
+ "left": "4",
78
+ "no": "1",
79
+ "off": "7",
80
+ "on": "6",
81
+ "right": "5",
82
+ "stop": "8",
83
+ "up": "2",
84
+ "yes": "0"
85
+ },
86
+ "layer_norm_eps": 1e-05,
87
+ "layerdrop": 0.05,
88
+ "mask_channel_length": 10,
89
+ "mask_channel_min_space": 1,
90
+ "mask_channel_other": 0.0,
91
+ "mask_channel_prob": 0.0,
92
+ "mask_channel_selection": "static",
93
+ "mask_feature_length": 10,
94
+ "mask_feature_prob": 0.0,
95
+ "mask_time_length": 10,
96
+ "mask_time_min_space": 1,
97
+ "mask_time_other": 0.0,
98
+ "mask_time_prob": 0.05,
99
+ "mask_time_selection": "static",
100
+ "model_type": "wav2vec2",
101
+ "no_mask_channel_overlap": false,
102
+ "no_mask_time_overlap": false,
103
+ "num_attention_heads": 12,
104
+ "num_codevector_groups": 2,
105
+ "num_codevectors_per_group": 320,
106
+ "num_conv_pos_embedding_groups": 16,
107
+ "num_conv_pos_embeddings": 128,
108
+ "num_feat_extract_layers": 7,
109
+ "num_hidden_layers": 12,
110
+ "num_negatives": 100,
111
+ "pad_token_id": 0,
112
+ "proj_codevector_dim": 256,
113
+ "torch_dtype": "float32",
114
+ "transformers_version": "4.12.0.dev0",
115
+ "use_weighted_layer_sum": false,
116
+ "vocab_size": 32
117
+ }
emissions.csv ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ timestamp,experiment_id,project_name,duration,emissions,energy_consumed,country_name,country_iso_code,region,on_cloud,cloud_provider,cloud_region
2
+ 2021-10-27T22:16:14,8843ae00-7e3b-4076-a01c-76400931e2dc,codecarbon,876.8788502216339,0.05179114541761414,0.09145531594139882,USA,USA,Iowa,Y,gcp,us-central1
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_accuracy": 0.9826419535157399,
4
+ "eval_loss": 0.08240818232297897,
5
+ "eval_runtime": 14.1019,
6
+ "eval_samples_per_second": 482.063,
7
+ "eval_steps_per_second": 15.104
8
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0.0,
7
+ "return_attention_mask": false,
8
+ "sampling_rate": 16000
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d24be12297c42c2247f6054bc1bdba9d2e0476340a3ecc2b0ec08d925d344ec
3
+ size 378371987
runs/Oct27_22-01-18_anton-gpu-v100/1635372097.669254/events.out.tfevents.1635372097.anton-gpu-v100.276919.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93bffd933f01ad579c67046dfa7088da35322f22dada83b3650b87dde344e509
3
+ size 4584
runs/Oct27_22-01-18_anton-gpu-v100/events.out.tfevents.1635372097.anton-gpu-v100.276919.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11a8e1c3a351751ef46aea187efa3611845e2c78f0bf4a5bf10dddaa43044764
3
+ size 38172
runs/Oct27_22-01-18_anton-gpu-v100/events.out.tfevents.1635372990.anton-gpu-v100.276919.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a2bc717a11cae449e453015d00f41f1bba886850c7a8d8278d7d3418d7c0ce89
3
+ size 363
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "<pad>", "do_lower_case": false, "return_attention_mask": false, "do_normalize": true}
train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "train_loss": 0.5217096392672164,
4
+ "train_runtime": 876.8675,
5
+ "train_samples_per_second": 291.344,
6
+ "train_steps_per_second": 2.275
7
+ }
trainer_state.json ADDED
@@ -0,0 +1,1264 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9826419535157399,
3
+ "best_model_checkpoint": "wav2vec2-base-ft-keyword-spotting/checkpoint-1596",
4
+ "epoch": 4.999373825923607,
5
+ "global_step": 1995,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.03,
12
+ "learning_rate": 1.5e-06,
13
+ "loss": 2.4949,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.05,
18
+ "learning_rate": 3e-06,
19
+ "loss": 2.4769,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.08,
24
+ "learning_rate": 4.5e-06,
25
+ "loss": 2.4248,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.1,
30
+ "learning_rate": 6e-06,
31
+ "loss": 2.3294,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.13,
36
+ "learning_rate": 7.5e-06,
37
+ "loss": 2.2127,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.15,
42
+ "learning_rate": 9e-06,
43
+ "loss": 2.0328,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.18,
48
+ "learning_rate": 1.05e-05,
49
+ "loss": 1.9213,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.2,
54
+ "learning_rate": 1.2e-05,
55
+ "loss": 1.851,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.23,
60
+ "learning_rate": 1.3500000000000001e-05,
61
+ "loss": 1.7601,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.25,
66
+ "learning_rate": 1.5e-05,
67
+ "loss": 1.6933,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.28,
72
+ "learning_rate": 1.65e-05,
73
+ "loss": 1.5549,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.3,
78
+ "learning_rate": 1.8e-05,
79
+ "loss": 1.5171,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.33,
84
+ "learning_rate": 1.95e-05,
85
+ "loss": 1.5173,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.35,
90
+ "learning_rate": 2.1e-05,
91
+ "loss": 1.5287,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.38,
96
+ "learning_rate": 2.25e-05,
97
+ "loss": 1.5134,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.4,
102
+ "learning_rate": 2.4e-05,
103
+ "loss": 1.5117,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.43,
108
+ "learning_rate": 2.55e-05,
109
+ "loss": 1.4358,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.45,
114
+ "learning_rate": 2.7000000000000002e-05,
115
+ "loss": 1.5335,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.48,
120
+ "learning_rate": 2.8499999999999998e-05,
121
+ "loss": 1.5543,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.5,
126
+ "learning_rate": 3e-05,
127
+ "loss": 1.5198,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.53,
132
+ "learning_rate": 2.9832869080779945e-05,
133
+ "loss": 1.5746,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.55,
138
+ "learning_rate": 2.9665738161559886e-05,
139
+ "loss": 1.5224,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.58,
144
+ "learning_rate": 2.9498607242339834e-05,
145
+ "loss": 1.4122,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.6,
150
+ "learning_rate": 2.933147632311978e-05,
151
+ "loss": 1.4064,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.63,
156
+ "learning_rate": 2.916434540389972e-05,
157
+ "loss": 1.3188,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.65,
162
+ "learning_rate": 2.8997214484679665e-05,
163
+ "loss": 1.2805,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.68,
168
+ "learning_rate": 2.8830083565459613e-05,
169
+ "loss": 1.2531,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.7,
174
+ "learning_rate": 2.8662952646239554e-05,
175
+ "loss": 1.1727,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.73,
180
+ "learning_rate": 2.84958217270195e-05,
181
+ "loss": 1.203,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.75,
186
+ "learning_rate": 2.8328690807799443e-05,
187
+ "loss": 1.0836,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.78,
192
+ "learning_rate": 2.8161559888579388e-05,
193
+ "loss": 1.1541,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.8,
198
+ "learning_rate": 2.7994428969359332e-05,
199
+ "loss": 1.0898,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.83,
204
+ "learning_rate": 2.7827298050139277e-05,
205
+ "loss": 1.1117,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.85,
210
+ "learning_rate": 2.7660167130919218e-05,
211
+ "loss": 1.0768,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.88,
216
+ "learning_rate": 2.7493036211699166e-05,
217
+ "loss": 1.0354,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.9,
222
+ "learning_rate": 2.732590529247911e-05,
223
+ "loss": 0.9769,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.93,
228
+ "learning_rate": 2.7158774373259055e-05,
229
+ "loss": 0.9714,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.95,
234
+ "learning_rate": 2.6991643454038996e-05,
235
+ "loss": 0.917,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.98,
240
+ "learning_rate": 2.6824512534818944e-05,
241
+ "loss": 0.8972,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 1.0,
246
+ "eval_accuracy": 0.8174463077375699,
247
+ "eval_loss": 0.7023297548294067,
248
+ "eval_runtime": 14.9373,
249
+ "eval_samples_per_second": 455.102,
250
+ "eval_steps_per_second": 14.26,
251
+ "step": 399
252
+ },
253
+ {
254
+ "epoch": 1.0,
255
+ "learning_rate": 2.665738161559889e-05,
256
+ "loss": 0.9539,
257
+ "step": 400
258
+ },
259
+ {
260
+ "epoch": 1.03,
261
+ "learning_rate": 2.649025069637883e-05,
262
+ "loss": 0.7725,
263
+ "step": 410
264
+ },
265
+ {
266
+ "epoch": 1.05,
267
+ "learning_rate": 2.6323119777158774e-05,
268
+ "loss": 0.8338,
269
+ "step": 420
270
+ },
271
+ {
272
+ "epoch": 1.08,
273
+ "learning_rate": 2.6155988857938722e-05,
274
+ "loss": 0.751,
275
+ "step": 430
276
+ },
277
+ {
278
+ "epoch": 1.1,
279
+ "learning_rate": 2.5988857938718663e-05,
280
+ "loss": 0.7343,
281
+ "step": 440
282
+ },
283
+ {
284
+ "epoch": 1.13,
285
+ "learning_rate": 2.5821727019498608e-05,
286
+ "loss": 0.7739,
287
+ "step": 450
288
+ },
289
+ {
290
+ "epoch": 1.15,
291
+ "learning_rate": 2.5654596100278553e-05,
292
+ "loss": 0.7262,
293
+ "step": 460
294
+ },
295
+ {
296
+ "epoch": 1.18,
297
+ "learning_rate": 2.5487465181058497e-05,
298
+ "loss": 0.6699,
299
+ "step": 470
300
+ },
301
+ {
302
+ "epoch": 1.2,
303
+ "learning_rate": 2.532033426183844e-05,
304
+ "loss": 0.7,
305
+ "step": 480
306
+ },
307
+ {
308
+ "epoch": 1.23,
309
+ "learning_rate": 2.5153203342618386e-05,
310
+ "loss": 0.6335,
311
+ "step": 490
312
+ },
313
+ {
314
+ "epoch": 1.25,
315
+ "learning_rate": 2.4986072423398327e-05,
316
+ "loss": 0.6279,
317
+ "step": 500
318
+ },
319
+ {
320
+ "epoch": 1.28,
321
+ "learning_rate": 2.4818941504178275e-05,
322
+ "loss": 0.6569,
323
+ "step": 510
324
+ },
325
+ {
326
+ "epoch": 1.3,
327
+ "learning_rate": 2.465181058495822e-05,
328
+ "loss": 0.6112,
329
+ "step": 520
330
+ },
331
+ {
332
+ "epoch": 1.33,
333
+ "learning_rate": 2.448467966573816e-05,
334
+ "loss": 0.6111,
335
+ "step": 530
336
+ },
337
+ {
338
+ "epoch": 1.35,
339
+ "learning_rate": 2.4317548746518106e-05,
340
+ "loss": 0.5783,
341
+ "step": 540
342
+ },
343
+ {
344
+ "epoch": 1.38,
345
+ "learning_rate": 2.415041782729805e-05,
346
+ "loss": 0.4803,
347
+ "step": 550
348
+ },
349
+ {
350
+ "epoch": 1.4,
351
+ "learning_rate": 2.3983286908077995e-05,
352
+ "loss": 0.4735,
353
+ "step": 560
354
+ },
355
+ {
356
+ "epoch": 1.43,
357
+ "learning_rate": 2.381615598885794e-05,
358
+ "loss": 0.5297,
359
+ "step": 570
360
+ },
361
+ {
362
+ "epoch": 1.45,
363
+ "learning_rate": 2.3649025069637884e-05,
364
+ "loss": 0.5026,
365
+ "step": 580
366
+ },
367
+ {
368
+ "epoch": 1.48,
369
+ "learning_rate": 2.3481894150417825e-05,
370
+ "loss": 0.4613,
371
+ "step": 590
372
+ },
373
+ {
374
+ "epoch": 1.5,
375
+ "learning_rate": 2.3314763231197773e-05,
376
+ "loss": 0.4743,
377
+ "step": 600
378
+ },
379
+ {
380
+ "epoch": 1.53,
381
+ "learning_rate": 2.3147632311977718e-05,
382
+ "loss": 0.4586,
383
+ "step": 610
384
+ },
385
+ {
386
+ "epoch": 1.55,
387
+ "learning_rate": 2.298050139275766e-05,
388
+ "loss": 0.4131,
389
+ "step": 620
390
+ },
391
+ {
392
+ "epoch": 1.58,
393
+ "learning_rate": 2.2813370473537603e-05,
394
+ "loss": 0.4347,
395
+ "step": 630
396
+ },
397
+ {
398
+ "epoch": 1.6,
399
+ "learning_rate": 2.264623955431755e-05,
400
+ "loss": 0.509,
401
+ "step": 640
402
+ },
403
+ {
404
+ "epoch": 1.63,
405
+ "learning_rate": 2.2479108635097492e-05,
406
+ "loss": 0.4038,
407
+ "step": 650
408
+ },
409
+ {
410
+ "epoch": 1.65,
411
+ "learning_rate": 2.2311977715877437e-05,
412
+ "loss": 0.3699,
413
+ "step": 660
414
+ },
415
+ {
416
+ "epoch": 1.68,
417
+ "learning_rate": 2.214484679665738e-05,
418
+ "loss": 0.3891,
419
+ "step": 670
420
+ },
421
+ {
422
+ "epoch": 1.7,
423
+ "learning_rate": 2.1977715877437326e-05,
424
+ "loss": 0.4092,
425
+ "step": 680
426
+ },
427
+ {
428
+ "epoch": 1.73,
429
+ "learning_rate": 2.181058495821727e-05,
430
+ "loss": 0.3817,
431
+ "step": 690
432
+ },
433
+ {
434
+ "epoch": 1.75,
435
+ "learning_rate": 2.1643454038997215e-05,
436
+ "loss": 0.3415,
437
+ "step": 700
438
+ },
439
+ {
440
+ "epoch": 1.78,
441
+ "learning_rate": 2.147632311977716e-05,
442
+ "loss": 0.3615,
443
+ "step": 710
444
+ },
445
+ {
446
+ "epoch": 1.8,
447
+ "learning_rate": 2.1309192200557104e-05,
448
+ "loss": 0.3131,
449
+ "step": 720
450
+ },
451
+ {
452
+ "epoch": 1.83,
453
+ "learning_rate": 2.114206128133705e-05,
454
+ "loss": 0.3565,
455
+ "step": 730
456
+ },
457
+ {
458
+ "epoch": 1.85,
459
+ "learning_rate": 2.0974930362116993e-05,
460
+ "loss": 0.3024,
461
+ "step": 740
462
+ },
463
+ {
464
+ "epoch": 1.88,
465
+ "learning_rate": 2.0807799442896935e-05,
466
+ "loss": 0.3239,
467
+ "step": 750
468
+ },
469
+ {
470
+ "epoch": 1.9,
471
+ "learning_rate": 2.0640668523676883e-05,
472
+ "loss": 0.3449,
473
+ "step": 760
474
+ },
475
+ {
476
+ "epoch": 1.93,
477
+ "learning_rate": 2.0473537604456827e-05,
478
+ "loss": 0.2942,
479
+ "step": 770
480
+ },
481
+ {
482
+ "epoch": 1.95,
483
+ "learning_rate": 2.0306406685236768e-05,
484
+ "loss": 0.3161,
485
+ "step": 780
486
+ },
487
+ {
488
+ "epoch": 1.98,
489
+ "learning_rate": 2.0139275766016713e-05,
490
+ "loss": 0.3274,
491
+ "step": 790
492
+ },
493
+ {
494
+ "epoch": 2.0,
495
+ "eval_accuracy": 0.9773462783171522,
496
+ "eval_loss": 0.16337834298610687,
497
+ "eval_runtime": 13.9496,
498
+ "eval_samples_per_second": 487.324,
499
+ "eval_steps_per_second": 15.269,
500
+ "step": 798
501
+ },
502
+ {
503
+ "epoch": 2.01,
504
+ "learning_rate": 1.997214484679666e-05,
505
+ "loss": 0.3454,
506
+ "step": 800
507
+ },
508
+ {
509
+ "epoch": 2.03,
510
+ "learning_rate": 1.9805013927576602e-05,
511
+ "loss": 0.3337,
512
+ "step": 810
513
+ },
514
+ {
515
+ "epoch": 2.06,
516
+ "learning_rate": 1.9637883008356546e-05,
517
+ "loss": 0.2997,
518
+ "step": 820
519
+ },
520
+ {
521
+ "epoch": 2.08,
522
+ "learning_rate": 1.947075208913649e-05,
523
+ "loss": 0.3173,
524
+ "step": 830
525
+ },
526
+ {
527
+ "epoch": 2.11,
528
+ "learning_rate": 1.9303621169916436e-05,
529
+ "loss": 0.3355,
530
+ "step": 840
531
+ },
532
+ {
533
+ "epoch": 2.13,
534
+ "learning_rate": 1.913649025069638e-05,
535
+ "loss": 0.2756,
536
+ "step": 850
537
+ },
538
+ {
539
+ "epoch": 2.16,
540
+ "learning_rate": 1.8969359331476325e-05,
541
+ "loss": 0.2612,
542
+ "step": 860
543
+ },
544
+ {
545
+ "epoch": 2.18,
546
+ "learning_rate": 1.8802228412256266e-05,
547
+ "loss": 0.2838,
548
+ "step": 870
549
+ },
550
+ {
551
+ "epoch": 2.21,
552
+ "learning_rate": 1.863509749303621e-05,
553
+ "loss": 0.329,
554
+ "step": 880
555
+ },
556
+ {
557
+ "epoch": 2.23,
558
+ "learning_rate": 1.846796657381616e-05,
559
+ "loss": 0.2528,
560
+ "step": 890
561
+ },
562
+ {
563
+ "epoch": 2.26,
564
+ "learning_rate": 1.83008356545961e-05,
565
+ "loss": 0.2994,
566
+ "step": 900
567
+ },
568
+ {
569
+ "epoch": 2.28,
570
+ "learning_rate": 1.8133704735376044e-05,
571
+ "loss": 0.2609,
572
+ "step": 910
573
+ },
574
+ {
575
+ "epoch": 2.31,
576
+ "learning_rate": 1.796657381615599e-05,
577
+ "loss": 0.2295,
578
+ "step": 920
579
+ },
580
+ {
581
+ "epoch": 2.33,
582
+ "learning_rate": 1.7799442896935933e-05,
583
+ "loss": 0.2758,
584
+ "step": 930
585
+ },
586
+ {
587
+ "epoch": 2.36,
588
+ "learning_rate": 1.7632311977715878e-05,
589
+ "loss": 0.2202,
590
+ "step": 940
591
+ },
592
+ {
593
+ "epoch": 2.38,
594
+ "learning_rate": 1.7465181058495822e-05,
595
+ "loss": 0.2715,
596
+ "step": 950
597
+ },
598
+ {
599
+ "epoch": 2.41,
600
+ "learning_rate": 1.7298050139275764e-05,
601
+ "loss": 0.2479,
602
+ "step": 960
603
+ },
604
+ {
605
+ "epoch": 2.43,
606
+ "learning_rate": 1.713091922005571e-05,
607
+ "loss": 0.2663,
608
+ "step": 970
609
+ },
610
+ {
611
+ "epoch": 2.46,
612
+ "learning_rate": 1.6963788300835656e-05,
613
+ "loss": 0.2533,
614
+ "step": 980
615
+ },
616
+ {
617
+ "epoch": 2.48,
618
+ "learning_rate": 1.6796657381615597e-05,
619
+ "loss": 0.248,
620
+ "step": 990
621
+ },
622
+ {
623
+ "epoch": 2.51,
624
+ "learning_rate": 1.6629526462395542e-05,
625
+ "loss": 0.23,
626
+ "step": 1000
627
+ },
628
+ {
629
+ "epoch": 2.53,
630
+ "learning_rate": 1.646239554317549e-05,
631
+ "loss": 0.2349,
632
+ "step": 1010
633
+ },
634
+ {
635
+ "epoch": 2.56,
636
+ "learning_rate": 1.6295264623955434e-05,
637
+ "loss": 0.2123,
638
+ "step": 1020
639
+ },
640
+ {
641
+ "epoch": 2.58,
642
+ "learning_rate": 1.6128133704735375e-05,
643
+ "loss": 0.2496,
644
+ "step": 1030
645
+ },
646
+ {
647
+ "epoch": 2.61,
648
+ "learning_rate": 1.596100278551532e-05,
649
+ "loss": 0.2394,
650
+ "step": 1040
651
+ },
652
+ {
653
+ "epoch": 2.63,
654
+ "learning_rate": 1.5793871866295268e-05,
655
+ "loss": 0.2138,
656
+ "step": 1050
657
+ },
658
+ {
659
+ "epoch": 2.66,
660
+ "learning_rate": 1.562674094707521e-05,
661
+ "loss": 0.2262,
662
+ "step": 1060
663
+ },
664
+ {
665
+ "epoch": 2.68,
666
+ "learning_rate": 1.5459610027855154e-05,
667
+ "loss": 0.228,
668
+ "step": 1070
669
+ },
670
+ {
671
+ "epoch": 2.71,
672
+ "learning_rate": 1.5292479108635098e-05,
673
+ "loss": 0.2126,
674
+ "step": 1080
675
+ },
676
+ {
677
+ "epoch": 2.73,
678
+ "learning_rate": 1.5125348189415043e-05,
679
+ "loss": 0.2439,
680
+ "step": 1090
681
+ },
682
+ {
683
+ "epoch": 2.76,
684
+ "learning_rate": 1.4958217270194987e-05,
685
+ "loss": 0.2913,
686
+ "step": 1100
687
+ },
688
+ {
689
+ "epoch": 2.78,
690
+ "learning_rate": 1.479108635097493e-05,
691
+ "loss": 0.2643,
692
+ "step": 1110
693
+ },
694
+ {
695
+ "epoch": 2.81,
696
+ "learning_rate": 1.4623955431754876e-05,
697
+ "loss": 0.1985,
698
+ "step": 1120
699
+ },
700
+ {
701
+ "epoch": 2.83,
702
+ "learning_rate": 1.445682451253482e-05,
703
+ "loss": 0.1995,
704
+ "step": 1130
705
+ },
706
+ {
707
+ "epoch": 2.86,
708
+ "learning_rate": 1.4289693593314764e-05,
709
+ "loss": 0.2218,
710
+ "step": 1140
711
+ },
712
+ {
713
+ "epoch": 2.88,
714
+ "learning_rate": 1.4122562674094708e-05,
715
+ "loss": 0.225,
716
+ "step": 1150
717
+ },
718
+ {
719
+ "epoch": 2.91,
720
+ "learning_rate": 1.3955431754874653e-05,
721
+ "loss": 0.2027,
722
+ "step": 1160
723
+ },
724
+ {
725
+ "epoch": 2.93,
726
+ "learning_rate": 1.3788300835654596e-05,
727
+ "loss": 0.2403,
728
+ "step": 1170
729
+ },
730
+ {
731
+ "epoch": 2.96,
732
+ "learning_rate": 1.362116991643454e-05,
733
+ "loss": 0.2154,
734
+ "step": 1180
735
+ },
736
+ {
737
+ "epoch": 2.98,
738
+ "learning_rate": 1.3454038997214485e-05,
739
+ "loss": 0.1993,
740
+ "step": 1190
741
+ },
742
+ {
743
+ "epoch": 3.0,
744
+ "eval_accuracy": 0.9788172992056487,
745
+ "eval_loss": 0.10475532710552216,
746
+ "eval_runtime": 14.0991,
747
+ "eval_samples_per_second": 482.159,
748
+ "eval_steps_per_second": 15.107,
749
+ "step": 1197
750
+ },
751
+ {
752
+ "epoch": 3.01,
753
+ "learning_rate": 1.3286908077994428e-05,
754
+ "loss": 0.1856,
755
+ "step": 1200
756
+ },
757
+ {
758
+ "epoch": 3.03,
759
+ "learning_rate": 1.3119777158774374e-05,
760
+ "loss": 0.2406,
761
+ "step": 1210
762
+ },
763
+ {
764
+ "epoch": 3.06,
765
+ "learning_rate": 1.2952646239554317e-05,
766
+ "loss": 0.1763,
767
+ "step": 1220
768
+ },
769
+ {
770
+ "epoch": 3.08,
771
+ "learning_rate": 1.2785515320334262e-05,
772
+ "loss": 0.1754,
773
+ "step": 1230
774
+ },
775
+ {
776
+ "epoch": 3.11,
777
+ "learning_rate": 1.2618384401114206e-05,
778
+ "loss": 0.1781,
779
+ "step": 1240
780
+ },
781
+ {
782
+ "epoch": 3.13,
783
+ "learning_rate": 1.245125348189415e-05,
784
+ "loss": 0.2419,
785
+ "step": 1250
786
+ },
787
+ {
788
+ "epoch": 3.16,
789
+ "learning_rate": 1.2284122562674095e-05,
790
+ "loss": 0.215,
791
+ "step": 1260
792
+ },
793
+ {
794
+ "epoch": 3.18,
795
+ "learning_rate": 1.211699164345404e-05,
796
+ "loss": 0.2068,
797
+ "step": 1270
798
+ },
799
+ {
800
+ "epoch": 3.21,
801
+ "learning_rate": 1.1949860724233983e-05,
802
+ "loss": 0.1869,
803
+ "step": 1280
804
+ },
805
+ {
806
+ "epoch": 3.23,
807
+ "learning_rate": 1.1782729805013929e-05,
808
+ "loss": 0.2321,
809
+ "step": 1290
810
+ },
811
+ {
812
+ "epoch": 3.26,
813
+ "learning_rate": 1.1615598885793872e-05,
814
+ "loss": 0.1984,
815
+ "step": 1300
816
+ },
817
+ {
818
+ "epoch": 3.28,
819
+ "learning_rate": 1.1448467966573816e-05,
820
+ "loss": 0.1766,
821
+ "step": 1310
822
+ },
823
+ {
824
+ "epoch": 3.31,
825
+ "learning_rate": 1.1281337047353761e-05,
826
+ "loss": 0.2103,
827
+ "step": 1320
828
+ },
829
+ {
830
+ "epoch": 3.33,
831
+ "learning_rate": 1.1114206128133705e-05,
832
+ "loss": 0.209,
833
+ "step": 1330
834
+ },
835
+ {
836
+ "epoch": 3.36,
837
+ "learning_rate": 1.0947075208913648e-05,
838
+ "loss": 0.183,
839
+ "step": 1340
840
+ },
841
+ {
842
+ "epoch": 3.38,
843
+ "learning_rate": 1.0779944289693595e-05,
844
+ "loss": 0.2289,
845
+ "step": 1350
846
+ },
847
+ {
848
+ "epoch": 3.41,
849
+ "learning_rate": 1.0612813370473537e-05,
850
+ "loss": 0.1645,
851
+ "step": 1360
852
+ },
853
+ {
854
+ "epoch": 3.43,
855
+ "learning_rate": 1.0445682451253482e-05,
856
+ "loss": 0.1816,
857
+ "step": 1370
858
+ },
859
+ {
860
+ "epoch": 3.46,
861
+ "learning_rate": 1.0278551532033427e-05,
862
+ "loss": 0.1885,
863
+ "step": 1380
864
+ },
865
+ {
866
+ "epoch": 3.48,
867
+ "learning_rate": 1.0111420612813371e-05,
868
+ "loss": 0.2352,
869
+ "step": 1390
870
+ },
871
+ {
872
+ "epoch": 3.51,
873
+ "learning_rate": 9.944289693593314e-06,
874
+ "loss": 0.1751,
875
+ "step": 1400
876
+ },
877
+ {
878
+ "epoch": 3.53,
879
+ "learning_rate": 9.77715877437326e-06,
880
+ "loss": 0.2142,
881
+ "step": 1410
882
+ },
883
+ {
884
+ "epoch": 3.56,
885
+ "learning_rate": 9.610027855153203e-06,
886
+ "loss": 0.1668,
887
+ "step": 1420
888
+ },
889
+ {
890
+ "epoch": 3.58,
891
+ "learning_rate": 9.44289693593315e-06,
892
+ "loss": 0.1587,
893
+ "step": 1430
894
+ },
895
+ {
896
+ "epoch": 3.61,
897
+ "learning_rate": 9.275766016713092e-06,
898
+ "loss": 0.1708,
899
+ "step": 1440
900
+ },
901
+ {
902
+ "epoch": 3.63,
903
+ "learning_rate": 9.108635097493037e-06,
904
+ "loss": 0.2169,
905
+ "step": 1450
906
+ },
907
+ {
908
+ "epoch": 3.66,
909
+ "learning_rate": 8.941504178272981e-06,
910
+ "loss": 0.159,
911
+ "step": 1460
912
+ },
913
+ {
914
+ "epoch": 3.68,
915
+ "learning_rate": 8.774373259052926e-06,
916
+ "loss": 0.1713,
917
+ "step": 1470
918
+ },
919
+ {
920
+ "epoch": 3.71,
921
+ "learning_rate": 8.607242339832869e-06,
922
+ "loss": 0.1802,
923
+ "step": 1480
924
+ },
925
+ {
926
+ "epoch": 3.73,
927
+ "learning_rate": 8.440111420612815e-06,
928
+ "loss": 0.1586,
929
+ "step": 1490
930
+ },
931
+ {
932
+ "epoch": 3.76,
933
+ "learning_rate": 8.272980501392758e-06,
934
+ "loss": 0.2148,
935
+ "step": 1500
936
+ },
937
+ {
938
+ "epoch": 3.78,
939
+ "learning_rate": 8.1058495821727e-06,
940
+ "loss": 0.2177,
941
+ "step": 1510
942
+ },
943
+ {
944
+ "epoch": 3.81,
945
+ "learning_rate": 7.938718662952647e-06,
946
+ "loss": 0.1665,
947
+ "step": 1520
948
+ },
949
+ {
950
+ "epoch": 3.83,
951
+ "learning_rate": 7.77158774373259e-06,
952
+ "loss": 0.1761,
953
+ "step": 1530
954
+ },
955
+ {
956
+ "epoch": 3.86,
957
+ "learning_rate": 7.604456824512535e-06,
958
+ "loss": 0.2171,
959
+ "step": 1540
960
+ },
961
+ {
962
+ "epoch": 3.88,
963
+ "learning_rate": 7.43732590529248e-06,
964
+ "loss": 0.1509,
965
+ "step": 1550
966
+ },
967
+ {
968
+ "epoch": 3.91,
969
+ "learning_rate": 7.2701949860724235e-06,
970
+ "loss": 0.1756,
971
+ "step": 1560
972
+ },
973
+ {
974
+ "epoch": 3.93,
975
+ "learning_rate": 7.103064066852368e-06,
976
+ "loss": 0.1705,
977
+ "step": 1570
978
+ },
979
+ {
980
+ "epoch": 3.96,
981
+ "learning_rate": 6.935933147632313e-06,
982
+ "loss": 0.1679,
983
+ "step": 1580
984
+ },
985
+ {
986
+ "epoch": 3.98,
987
+ "learning_rate": 6.768802228412256e-06,
988
+ "loss": 0.1777,
989
+ "step": 1590
990
+ },
991
+ {
992
+ "epoch": 4.0,
993
+ "eval_accuracy": 0.9826419535157399,
994
+ "eval_loss": 0.08240818232297897,
995
+ "eval_runtime": 13.7079,
996
+ "eval_samples_per_second": 495.918,
997
+ "eval_steps_per_second": 15.538,
998
+ "step": 1596
999
+ },
1000
+ {
1001
+ "epoch": 4.01,
1002
+ "learning_rate": 6.601671309192201e-06,
1003
+ "loss": 0.184,
1004
+ "step": 1600
1005
+ },
1006
+ {
1007
+ "epoch": 4.04,
1008
+ "learning_rate": 6.4345403899721455e-06,
1009
+ "loss": 0.1912,
1010
+ "step": 1610
1011
+ },
1012
+ {
1013
+ "epoch": 4.06,
1014
+ "learning_rate": 6.267409470752089e-06,
1015
+ "loss": 0.1981,
1016
+ "step": 1620
1017
+ },
1018
+ {
1019
+ "epoch": 4.09,
1020
+ "learning_rate": 6.100278551532034e-06,
1021
+ "loss": 0.1253,
1022
+ "step": 1630
1023
+ },
1024
+ {
1025
+ "epoch": 4.11,
1026
+ "learning_rate": 5.933147632311978e-06,
1027
+ "loss": 0.1602,
1028
+ "step": 1640
1029
+ },
1030
+ {
1031
+ "epoch": 4.14,
1032
+ "learning_rate": 5.766016713091923e-06,
1033
+ "loss": 0.1941,
1034
+ "step": 1650
1035
+ },
1036
+ {
1037
+ "epoch": 4.16,
1038
+ "learning_rate": 5.598885793871867e-06,
1039
+ "loss": 0.1622,
1040
+ "step": 1660
1041
+ },
1042
+ {
1043
+ "epoch": 4.19,
1044
+ "learning_rate": 5.43175487465181e-06,
1045
+ "loss": 0.1529,
1046
+ "step": 1670
1047
+ },
1048
+ {
1049
+ "epoch": 4.21,
1050
+ "learning_rate": 5.264623955431755e-06,
1051
+ "loss": 0.1969,
1052
+ "step": 1680
1053
+ },
1054
+ {
1055
+ "epoch": 4.24,
1056
+ "learning_rate": 5.0974930362116986e-06,
1057
+ "loss": 0.1511,
1058
+ "step": 1690
1059
+ },
1060
+ {
1061
+ "epoch": 4.26,
1062
+ "learning_rate": 4.930362116991643e-06,
1063
+ "loss": 0.2085,
1064
+ "step": 1700
1065
+ },
1066
+ {
1067
+ "epoch": 4.29,
1068
+ "learning_rate": 4.763231197771588e-06,
1069
+ "loss": 0.1925,
1070
+ "step": 1710
1071
+ },
1072
+ {
1073
+ "epoch": 4.31,
1074
+ "learning_rate": 4.596100278551532e-06,
1075
+ "loss": 0.1979,
1076
+ "step": 1720
1077
+ },
1078
+ {
1079
+ "epoch": 4.34,
1080
+ "learning_rate": 4.428969359331476e-06,
1081
+ "loss": 0.186,
1082
+ "step": 1730
1083
+ },
1084
+ {
1085
+ "epoch": 4.36,
1086
+ "learning_rate": 4.2618384401114205e-06,
1087
+ "loss": 0.1738,
1088
+ "step": 1740
1089
+ },
1090
+ {
1091
+ "epoch": 4.39,
1092
+ "learning_rate": 4.094707520891365e-06,
1093
+ "loss": 0.1755,
1094
+ "step": 1750
1095
+ },
1096
+ {
1097
+ "epoch": 4.41,
1098
+ "learning_rate": 3.927576601671309e-06,
1099
+ "loss": 0.1693,
1100
+ "step": 1760
1101
+ },
1102
+ {
1103
+ "epoch": 4.44,
1104
+ "learning_rate": 3.7604456824512533e-06,
1105
+ "loss": 0.17,
1106
+ "step": 1770
1107
+ },
1108
+ {
1109
+ "epoch": 4.46,
1110
+ "learning_rate": 3.593314763231198e-06,
1111
+ "loss": 0.1865,
1112
+ "step": 1780
1113
+ },
1114
+ {
1115
+ "epoch": 4.49,
1116
+ "learning_rate": 3.426183844011142e-06,
1117
+ "loss": 0.1612,
1118
+ "step": 1790
1119
+ },
1120
+ {
1121
+ "epoch": 4.51,
1122
+ "learning_rate": 3.259052924791086e-06,
1123
+ "loss": 0.1919,
1124
+ "step": 1800
1125
+ },
1126
+ {
1127
+ "epoch": 4.54,
1128
+ "learning_rate": 3.0919220055710307e-06,
1129
+ "loss": 0.1526,
1130
+ "step": 1810
1131
+ },
1132
+ {
1133
+ "epoch": 4.56,
1134
+ "learning_rate": 2.924791086350975e-06,
1135
+ "loss": 0.1758,
1136
+ "step": 1820
1137
+ },
1138
+ {
1139
+ "epoch": 4.59,
1140
+ "learning_rate": 2.7576601671309194e-06,
1141
+ "loss": 0.19,
1142
+ "step": 1830
1143
+ },
1144
+ {
1145
+ "epoch": 4.61,
1146
+ "learning_rate": 2.5905292479108636e-06,
1147
+ "loss": 0.1274,
1148
+ "step": 1840
1149
+ },
1150
+ {
1151
+ "epoch": 4.64,
1152
+ "learning_rate": 2.4233983286908077e-06,
1153
+ "loss": 0.1246,
1154
+ "step": 1850
1155
+ },
1156
+ {
1157
+ "epoch": 4.66,
1158
+ "learning_rate": 2.2562674094707523e-06,
1159
+ "loss": 0.1699,
1160
+ "step": 1860
1161
+ },
1162
+ {
1163
+ "epoch": 4.69,
1164
+ "learning_rate": 2.0891364902506964e-06,
1165
+ "loss": 0.1549,
1166
+ "step": 1870
1167
+ },
1168
+ {
1169
+ "epoch": 4.71,
1170
+ "learning_rate": 1.922005571030641e-06,
1171
+ "loss": 0.1858,
1172
+ "step": 1880
1173
+ },
1174
+ {
1175
+ "epoch": 4.74,
1176
+ "learning_rate": 1.7548746518105849e-06,
1177
+ "loss": 0.1561,
1178
+ "step": 1890
1179
+ },
1180
+ {
1181
+ "epoch": 4.76,
1182
+ "learning_rate": 1.5877437325905292e-06,
1183
+ "loss": 0.162,
1184
+ "step": 1900
1185
+ },
1186
+ {
1187
+ "epoch": 4.79,
1188
+ "learning_rate": 1.4206128133704736e-06,
1189
+ "loss": 0.2132,
1190
+ "step": 1910
1191
+ },
1192
+ {
1193
+ "epoch": 4.81,
1194
+ "learning_rate": 1.253481894150418e-06,
1195
+ "loss": 0.1326,
1196
+ "step": 1920
1197
+ },
1198
+ {
1199
+ "epoch": 4.84,
1200
+ "learning_rate": 1.0863509749303623e-06,
1201
+ "loss": 0.1339,
1202
+ "step": 1930
1203
+ },
1204
+ {
1205
+ "epoch": 4.86,
1206
+ "learning_rate": 9.192200557103064e-07,
1207
+ "loss": 0.1589,
1208
+ "step": 1940
1209
+ },
1210
+ {
1211
+ "epoch": 4.89,
1212
+ "learning_rate": 7.520891364902508e-07,
1213
+ "loss": 0.1711,
1214
+ "step": 1950
1215
+ },
1216
+ {
1217
+ "epoch": 4.91,
1218
+ "learning_rate": 5.84958217270195e-07,
1219
+ "loss": 0.1458,
1220
+ "step": 1960
1221
+ },
1222
+ {
1223
+ "epoch": 4.94,
1224
+ "learning_rate": 4.178272980501393e-07,
1225
+ "loss": 0.1643,
1226
+ "step": 1970
1227
+ },
1228
+ {
1229
+ "epoch": 4.96,
1230
+ "learning_rate": 2.506963788300836e-07,
1231
+ "loss": 0.1598,
1232
+ "step": 1980
1233
+ },
1234
+ {
1235
+ "epoch": 4.99,
1236
+ "learning_rate": 8.356545961002785e-08,
1237
+ "loss": 0.1527,
1238
+ "step": 1990
1239
+ },
1240
+ {
1241
+ "epoch": 5.0,
1242
+ "eval_accuracy": 0.9810238305383936,
1243
+ "eval_loss": 0.08117648959159851,
1244
+ "eval_runtime": 13.9006,
1245
+ "eval_samples_per_second": 489.042,
1246
+ "eval_steps_per_second": 15.323,
1247
+ "step": 1995
1248
+ },
1249
+ {
1250
+ "epoch": 5.0,
1251
+ "step": 1995,
1252
+ "total_flos": 2.31918157475328e+18,
1253
+ "train_loss": 0.5217096392672164,
1254
+ "train_runtime": 876.8675,
1255
+ "train_samples_per_second": 291.344,
1256
+ "train_steps_per_second": 2.275
1257
+ }
1258
+ ],
1259
+ "max_steps": 1995,
1260
+ "num_train_epochs": 5,
1261
+ "total_flos": 2.31918157475328e+18,
1262
+ "trial_name": null,
1263
+ "trial_params": null
1264
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:728c2915cf73e2daccf93b46fbf327a040618ea03977c5e322f1630df1218152
3
+ size 2927
vocab.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"<pad>": 0, "<s>": 1, "</s>": 2, "<unk>": 3, "|": 4, "E": 5, "T": 6, "A": 7, "O": 8, "N": 9, "I": 10, "H": 11, "S": 12, "R": 13, "D": 14, "L": 15, "U": 16, "M": 17, "W": 18, "C": 19, "F": 20, "G": 21, "Y": 22, "P": 23, "B": 24, "V": 25, "K": 26, "'": 27, "X": 28, "J": 29, "Q": 30, "Z": 31}