Hugging Face's logo

QtGroup
/
CodeLlama-7B-QML

Safetensors
GGUF
QML
Code-Completion
Model card Files
xet
Community
1
qt-spyro-hf commited on
Commit
2690367
·
verified ·
1 Parent(s): 05f6824

Upload 14 files (#1)

Browse files

- Upload 14 files (7f8f19cbedb82d5682a41d9c0e7380a52ef0871e)

Files changed (15) hide show
  1. .gitattributes +1 -0
  2. Modelfile +23 -0
  3. README.md +93 -3
  4. adapter_config.json +38 -0
  5. adapter_model.safetensors +3 -0
  6. codellama-7b-qml.gguf +3 -0
  7. optimizer.pt +3 -0
  8. rest.safetensors +3 -0
  9. rng_state.pth +3 -0
  10. scheduler.pt +3 -0
  11. special_tokens_map.json +30 -0
  12. tokenizer.json +0 -0
  13. tokenizer_config.json +84 -0
  14. trainer_state.json +3114 -0
  15. training_args.bin +3 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ codellama-7b-qml.gguf filter=lfs diff=lfs merge=lfs -text
Modelfile ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ FROM codellama:7b-code
2
+
3
+ # Base model name and adapter
4
+ ADAPTER ./codellama-7b-qml.gguf
5
+
6
+ # Parameters optimized for code generation
7
+ PARAMETER temperature 0.2
8
+ PARAMETER num_predict 500
9
+ PARAMETER top_p 0.9
10
+ PARAMETER stop "<SUF>"
11
+ PARAMETER stop "<PRE>"
12
+ PARAMETER stop "</PRE>"
13
+ PARAMETER stop "</SUF>"
14
+ PARAMETER stop "< EOT >"
15
+ PARAMETER stop "\\end"
16
+ PARAMETER stop "<MID>"
17
+ PARAMETER stop "</MID>"
18
+ PARAMETER stop "##"
19
+
20
+
21
+ # Template format for code interactions
22
+ TEMPLATE """{{if .Prompt}}{{ .Prompt }}{{end}}{{if .Response}}{{ .Response }}{{end}}"""
23
+
README.md CHANGED
@@ -1,3 +1,93 @@
1
- ---
2
- license: llama3.2
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: llama2
3
+ base_model:
4
+ - meta-llama/CodeLlama-7b-hf
5
+ base_model_relation: adapter
6
+ tags:
7
+ - QML
8
+ - Code-Completion
9
+ ---
10
+ # Model Overview
11
+
12
+ ## Description:
13
+ CodeLlama-7B-QML is a large language model customized by the Qt Company for Fill-In-The-Middle code completion tasks in the QML programming language, especially for Qt Quick Controls compliant with Qt 6 releases. The CodeLlama-7B-QML model is designed for companies and individuals that want to self-host their LLM for HMI (Human Machine Interface) software development instead of relying on third-party hosted LLMs.
14
+
15
+ This model reaches a score of 79% on the QML100 Fill-In-the-Middle code completion benchmark for Qt 6-compliant code. In comparison, the model scored:
16
+ - CodeLlama-13B-QML (finetuned model from Qt): 79%,
17
+ - Claude 3.7 Sonnet: 76%,
18
+ - Claude 3.5 Sonnet: 68%,
19
+ - CodeLlama 13B: 66%,
20
+ - GPT-4o: 62%,
21
+ - CodeLlama 7B: 61%.
22
+
23
+ This model was fine-tuned based on raw data from over 5000 human-created QML code snippets using the LoRa fine-tuning method. CodeLlama-7B-QML is not optimised for the creation of Qt5-release compliant, C++, or Python code.
24
+
25
+ ## Terms of use:
26
+ By accessing this model, you are agreeing to the Llama 2 terms and conditions of the [license](https://github.com/meta-llama/llama/blob/main/LICENSE), [acceptable use policy](https://github.com/meta-llama/llama/blob/main/USE_POLICY.md) and [Meta’s privacy policy](https://www.facebook.com/privacy/policy/). By using this model, you are furthermore agreeing to the [Qt AI Model terms & conditions](https://www.qt.io/terms-conditions/ai-services/model-use).
27
+
28
+ ## Usage:
29
+
30
+ CodeLlama-7B-QML is a medium-sized Language Model that requires significant computing resources to perform with inference (response) times suitable for automatic code completion. Therefore, it should be used with a GPU accelerator, either in the cloud environment such as AWS, Google Cloud, Microsoft Azure, or locally.
31
+
32
+ Large Language Models, including CodeLlama-7B-QML, are not designed to be deployed in isolation but instead should be deployed as part of an overall AI system with additional safety guardrails as required. Developers are expected to deploy system safeguards when building AI systems.
33
+
34
+ The repository contains multiple files with adapters.
35
+
36
+ ## How to run CodeLlama-7B-QML in cloud deployment:
37
+
38
+ The configuration depends on the chosen cloud technology.
39
+
40
+ Running a CodeLlama-7b-QML in the cloud requires working with Docker and vLLM for optimal performance. Make sure all required dependencies are installed (transformers, accelerate and peft modules). Use bfloat16 precision. The setup leverages the base model from Hugging Face (requiring an access token) combined with adapter weights from the repository. Using vLLM enables efficient inference with an OpenAI-compatible API endpoint, making integration straightforward. vLLM serves as a highly optimized backend that implements request batching and queuing mechanisms, providing excellent serving optimization. The docker container should be run on an instance with GPU accelerator. The configuration has been thoroughly tested on Ubuntu 22.04 LTS running NVIDIA driver with A100 80GB GPUs, demonstrating stable and efficient performance.
41
+
42
+ ## How to run CodeLlama-7B-QML in ollama:
43
+
44
+ The model can be downloaded either from Hugging Face or Ollama. If the choice is Hugging Face, follow all the instruction steps. In case of Ollama, execute steps 1 and 5.
45
+
46
+ #### 1. Install ollama
47
+ https://ollama.com/download
48
+
49
+ #### 2. Clone the model repository
50
+
51
+ #### 3. Open the terminal and go to the repository
52
+
53
+ #### 4. Build the model in ollama
54
+ ```
55
+ ollama create theqtcompany/codellama-7b-code-qml -f Modelfile
56
+ ```
57
+ The model's name must be exactly as above if one wants to use the model in the Qt Creator
58
+
59
+ #### 5. Run the model
60
+ ```
61
+ ollama run theqtcompany/codellama-7b-qml
62
+ ```
63
+ You can start writing prompts in the terminal or send curl requests now.
64
+
65
+ Here is a curl request example:
66
+ ```
67
+ curl -X POST http://localhost:11434/api/generate -d '{
68
+ "model": "theqtcompany/codellama-7b-qml",
69
+ "Prompt": "<SUF>\n title: qsTr(\"Hello World\")\n}<PRE>import QtQuick\n\nWindow {\n width: 640\n height: 480\n visible: true\n<MID>",
70
+ "stream": false,
71
+ "temperature": 0.2,
72
+ "top_p": 0.9,
73
+ "num_predict": 500,
74
+ "stop": ["<SUF>", "<PRE>", "</PRE>", "</SUF>", "< EOT >", "\\end", "<MID>", "</MID>", "##"]
75
+ }'
76
+ ```
77
+
78
+ The prompt format:
79
+ ```
80
+ "<SUF>{suffix}<PRE>{prefix}<MID>"
81
+ ```
82
+
83
+ If there is no suffix, please use:
84
+ ```
85
+ "<PRE>{prefix}<MID>"
86
+ ```
87
+
88
+
89
+ ## Model Version:
90
+ v1.0
91
+
92
+ ## Attribution:
93
+ CodeLlama-7B is a model of the Llama 2 family. Llama 2 is licensed under the LLAMA 2 Community License, Copyright (c) Meta Platforms, Inc. All Rights Reserved.
adapter_config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "codellama/CodeLlama-7b-hf",
5
+ "bias": "none",
6
+ "eva_config": null,
7
+ "exclude_modules": null,
8
+ "fan_in_fan_out": false,
9
+ "inference_mode": true,
10
+ "init_lora_weights": true,
11
+ "layer_replication": null,
12
+ "layers_pattern": null,
13
+ "layers_to_transform": null,
14
+ "loftq_config": {},
15
+ "lora_alpha": 64,
16
+ "lora_bias": false,
17
+ "lora_dropout": 0.1,
18
+ "megatron_config": null,
19
+ "megatron_core": "megatron.core",
20
+ "modules_to_save": null,
21
+ "peft_type": "LORA",
22
+ "r": 64,
23
+ "rank_pattern": {},
24
+ "revision": null,
25
+ "target_modules": [
26
+ "v_proj",
27
+ "o_proj",
28
+ "gate_proj",
29
+ "lm_head",
30
+ "down_proj",
31
+ "q_proj",
32
+ "up_proj",
33
+ "k_proj"
34
+ ],
35
+ "task_type": "CAUSAL_LM",
36
+ "use_dora": false,
37
+ "use_rslora": false
38
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec4924dfac7c38ed7e78d23d166c1bac46cc7cde6ef2ec3d70c741a4d5b11ea7
3
+ size 648936760
codellama-7b-qml.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:87a033f1b70147091d4a52d8d7a3938ca182b5efa701418adaddcaf7331dce4e
3
+ size 648907072
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8da06080bb75f0fe64cb3bb6897ff13b9b2f17a551222d1a4883a09ca52cf17e
3
+ size 325614754
rest.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:554787be54f537266241b356a8bbc672062afe4d218d66ca5d2ece3b41d590b7
3
+ size 524550264
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:054177611f2412e52a8cd0a4d0895e2cb9e4fef94374378f0bf6c4059d506903
3
+ size 14244
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef1fe48b040bf7678711805e0ccda18391a9ad1db687676a750066fbac4eecfb
3
+ size 1064
special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "▁<PRE>",
4
+ "▁<MID>",
5
+ "▁<SUF>",
6
+ "▁<EOT>"
7
+ ],
8
+ "bos_token": {
9
+ "content": "<s>",
10
+ "lstrip": false,
11
+ "normalized": false,
12
+ "rstrip": false,
13
+ "single_word": false
14
+ },
15
+ "eos_token": {
16
+ "content": "</s>",
17
+ "lstrip": false,
18
+ "normalized": false,
19
+ "rstrip": false,
20
+ "single_word": false
21
+ },
22
+ "pad_token": "</s>",
23
+ "unk_token": {
24
+ "content": "<unk>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ }
30
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,84 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "32007": {
30
+ "content": "▁<PRE>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "32008": {
38
+ "content": "▁<SUF>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "32009": {
46
+ "content": "▁<MID>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "32010": {
54
+ "content": "▁<EOT>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ }
61
+ },
62
+ "additional_special_tokens": [
63
+ "▁<PRE>",
64
+ "▁<MID>",
65
+ "▁<SUF>",
66
+ "▁<EOT>"
67
+ ],
68
+ "bos_token": "<s>",
69
+ "clean_up_tokenization_spaces": false,
70
+ "eos_token": "</s>",
71
+ "eot_token": "▁<EOT>",
72
+ "extra_special_tokens": {},
73
+ "fill_token": "<FILL_ME>",
74
+ "legacy": null,
75
+ "middle_token": "▁<MID>",
76
+ "model_max_length": 1000000000000000019884624838656,
77
+ "pad_token": "</s>",
78
+ "prefix_token": "▁<PRE>",
79
+ "sp_model_kwargs": {},
80
+ "suffix_token": "▁<SUF>",
81
+ "tokenizer_class": "CodeLlamaTokenizer",
82
+ "unk_token": "<unk>",
83
+ "use_default_system_prompt": false
84
+ }
trainer_state.json ADDED
@@ -0,0 +1,3114 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 20,
6
+ "global_step": 8786,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0022765430693491933,
13
+ "grad_norm": 0.469247430562973,
14
+ "learning_rate": 0.0002,
15
+ "loss": 1.9469,
16
+ "step": 20
17
+ },
18
+ {
19
+ "epoch": 0.004553086138698387,
20
+ "grad_norm": 0.6239348649978638,
21
+ "learning_rate": 0.0002,
22
+ "loss": 1.556,
23
+ "step": 40
24
+ },
25
+ {
26
+ "epoch": 0.006829629208047579,
27
+ "grad_norm": 0.4587397277355194,
28
+ "learning_rate": 0.0002,
29
+ "loss": 1.4108,
30
+ "step": 60
31
+ },
32
+ {
33
+ "epoch": 0.009106172277396773,
34
+ "grad_norm": 0.42919760942459106,
35
+ "learning_rate": 0.0002,
36
+ "loss": 1.3352,
37
+ "step": 80
38
+ },
39
+ {
40
+ "epoch": 0.011382715346745967,
41
+ "grad_norm": 0.46492573618888855,
42
+ "learning_rate": 0.0002,
43
+ "loss": 1.3388,
44
+ "step": 100
45
+ },
46
+ {
47
+ "epoch": 0.013659258416095159,
48
+ "grad_norm": 0.453070729970932,
49
+ "learning_rate": 0.0002,
50
+ "loss": 1.2295,
51
+ "step": 120
52
+ },
53
+ {
54
+ "epoch": 0.015935801485444354,
55
+ "grad_norm": 0.4760678708553314,
56
+ "learning_rate": 0.0002,
57
+ "loss": 1.2493,
58
+ "step": 140
59
+ },
60
+ {
61
+ "epoch": 0.018212344554793546,
62
+ "grad_norm": 0.4545675814151764,
63
+ "learning_rate": 0.0002,
64
+ "loss": 1.215,
65
+ "step": 160
66
+ },
67
+ {
68
+ "epoch": 0.020488887624142738,
69
+ "grad_norm": 0.4772235155105591,
70
+ "learning_rate": 0.0002,
71
+ "loss": 1.2173,
72
+ "step": 180
73
+ },
74
+ {
75
+ "epoch": 0.022765430693491934,
76
+ "grad_norm": 0.4403541088104248,
77
+ "learning_rate": 0.0002,
78
+ "loss": 1.1058,
79
+ "step": 200
80
+ },
81
+ {
82
+ "epoch": 0.025041973762841126,
83
+ "grad_norm": 0.511401355266571,
84
+ "learning_rate": 0.0002,
85
+ "loss": 1.1049,
86
+ "step": 220
87
+ },
88
+ {
89
+ "epoch": 0.027318516832190318,
90
+ "grad_norm": 0.3809013366699219,
91
+ "learning_rate": 0.0002,
92
+ "loss": 1.0498,
93
+ "step": 240
94
+ },
95
+ {
96
+ "epoch": 0.029595059901539513,
97
+ "grad_norm": 0.3980010449886322,
98
+ "learning_rate": 0.0002,
99
+ "loss": 0.9842,
100
+ "step": 260
101
+ },
102
+ {
103
+ "epoch": 0.03187160297088871,
104
+ "grad_norm": 0.5747793316841125,
105
+ "learning_rate": 0.0002,
106
+ "loss": 1.0988,
107
+ "step": 280
108
+ },
109
+ {
110
+ "epoch": 0.0341481460402379,
111
+ "grad_norm": 0.46827971935272217,
112
+ "learning_rate": 0.0002,
113
+ "loss": 1.0367,
114
+ "step": 300
115
+ },
116
+ {
117
+ "epoch": 0.03642468910958709,
118
+ "grad_norm": 0.4702209532260895,
119
+ "learning_rate": 0.0002,
120
+ "loss": 1.066,
121
+ "step": 320
122
+ },
123
+ {
124
+ "epoch": 0.038701232178936285,
125
+ "grad_norm": 0.5084996223449707,
126
+ "learning_rate": 0.0002,
127
+ "loss": 1.0652,
128
+ "step": 340
129
+ },
130
+ {
131
+ "epoch": 0.040977775248285477,
132
+ "grad_norm": 0.3944012522697449,
133
+ "learning_rate": 0.0002,
134
+ "loss": 0.9642,
135
+ "step": 360
136
+ },
137
+ {
138
+ "epoch": 0.04325431831763467,
139
+ "grad_norm": 0.40287718176841736,
140
+ "learning_rate": 0.0002,
141
+ "loss": 0.9431,
142
+ "step": 380
143
+ },
144
+ {
145
+ "epoch": 0.04553086138698387,
146
+ "grad_norm": 0.4629077613353729,
147
+ "learning_rate": 0.0002,
148
+ "loss": 0.9615,
149
+ "step": 400
150
+ },
151
+ {
152
+ "epoch": 0.04780740445633306,
153
+ "grad_norm": 0.44827452301979065,
154
+ "learning_rate": 0.0002,
155
+ "loss": 0.9434,
156
+ "step": 420
157
+ },
158
+ {
159
+ "epoch": 0.05008394752568225,
160
+ "grad_norm": 0.41644710302352905,
161
+ "learning_rate": 0.0002,
162
+ "loss": 0.9241,
163
+ "step": 440
164
+ },
165
+ {
166
+ "epoch": 0.05236049059503144,
167
+ "grad_norm": 0.4760611057281494,
168
+ "learning_rate": 0.0002,
169
+ "loss": 0.8475,
170
+ "step": 460
171
+ },
172
+ {
173
+ "epoch": 0.054637033664380635,
174
+ "grad_norm": 0.45987364649772644,
175
+ "learning_rate": 0.0002,
176
+ "loss": 0.898,
177
+ "step": 480
178
+ },
179
+ {
180
+ "epoch": 0.056913576733729834,
181
+ "grad_norm": 0.4840068817138672,
182
+ "learning_rate": 0.0002,
183
+ "loss": 0.9611,
184
+ "step": 500
185
+ },
186
+ {
187
+ "epoch": 0.059190119803079026,
188
+ "grad_norm": 0.40314286947250366,
189
+ "learning_rate": 0.0002,
190
+ "loss": 0.8884,
191
+ "step": 520
192
+ },
193
+ {
194
+ "epoch": 0.06146666287242822,
195
+ "grad_norm": 0.5458106398582458,
196
+ "learning_rate": 0.0002,
197
+ "loss": 0.8939,
198
+ "step": 540
199
+ },
200
+ {
201
+ "epoch": 0.06374320594177742,
202
+ "grad_norm": 0.5420896410942078,
203
+ "learning_rate": 0.0002,
204
+ "loss": 0.8265,
205
+ "step": 560
206
+ },
207
+ {
208
+ "epoch": 0.0660197490111266,
209
+ "grad_norm": 0.5356529355049133,
210
+ "learning_rate": 0.0002,
211
+ "loss": 0.8432,
212
+ "step": 580
213
+ },
214
+ {
215
+ "epoch": 0.0682962920804758,
216
+ "grad_norm": 0.5064826011657715,
217
+ "learning_rate": 0.0002,
218
+ "loss": 0.8272,
219
+ "step": 600
220
+ },
221
+ {
222
+ "epoch": 0.07057283514982499,
223
+ "grad_norm": 0.4143005311489105,
224
+ "learning_rate": 0.0002,
225
+ "loss": 0.7854,
226
+ "step": 620
227
+ },
228
+ {
229
+ "epoch": 0.07284937821917419,
230
+ "grad_norm": 0.3817225396633148,
231
+ "learning_rate": 0.0002,
232
+ "loss": 0.8219,
233
+ "step": 640
234
+ },
235
+ {
236
+ "epoch": 0.07512592128852338,
237
+ "grad_norm": 0.5336936712265015,
238
+ "learning_rate": 0.0002,
239
+ "loss": 0.7977,
240
+ "step": 660
241
+ },
242
+ {
243
+ "epoch": 0.07740246435787257,
244
+ "grad_norm": 0.5397001504898071,
245
+ "learning_rate": 0.0002,
246
+ "loss": 0.8117,
247
+ "step": 680
248
+ },
249
+ {
250
+ "epoch": 0.07967900742722177,
251
+ "grad_norm": 0.4968530535697937,
252
+ "learning_rate": 0.0002,
253
+ "loss": 0.7527,
254
+ "step": 700
255
+ },
256
+ {
257
+ "epoch": 0.08195555049657095,
258
+ "grad_norm": 0.4084935784339905,
259
+ "learning_rate": 0.0002,
260
+ "loss": 0.651,
261
+ "step": 720
262
+ },
263
+ {
264
+ "epoch": 0.08423209356592015,
265
+ "grad_norm": 0.48406732082366943,
266
+ "learning_rate": 0.0002,
267
+ "loss": 0.7352,
268
+ "step": 740
269
+ },
270
+ {
271
+ "epoch": 0.08650863663526934,
272
+ "grad_norm": 0.5246301293373108,
273
+ "learning_rate": 0.0002,
274
+ "loss": 0.7785,
275
+ "step": 760
276
+ },
277
+ {
278
+ "epoch": 0.08878517970461854,
279
+ "grad_norm": 0.5729619264602661,
280
+ "learning_rate": 0.0002,
281
+ "loss": 0.7646,
282
+ "step": 780
283
+ },
284
+ {
285
+ "epoch": 0.09106172277396773,
286
+ "grad_norm": 0.5675190687179565,
287
+ "learning_rate": 0.0002,
288
+ "loss": 0.7784,
289
+ "step": 800
290
+ },
291
+ {
292
+ "epoch": 0.09333826584331692,
293
+ "grad_norm": 0.4682878255844116,
294
+ "learning_rate": 0.0002,
295
+ "loss": 0.7284,
296
+ "step": 820
297
+ },
298
+ {
299
+ "epoch": 0.09561480891266612,
300
+ "grad_norm": 0.5388545393943787,
301
+ "learning_rate": 0.0002,
302
+ "loss": 0.6959,
303
+ "step": 840
304
+ },
305
+ {
306
+ "epoch": 0.0978913519820153,
307
+ "grad_norm": 0.48806509375572205,
308
+ "learning_rate": 0.0002,
309
+ "loss": 0.7585,
310
+ "step": 860
311
+ },
312
+ {
313
+ "epoch": 0.1001678950513645,
314
+ "grad_norm": 0.4149261713027954,
315
+ "learning_rate": 0.0002,
316
+ "loss": 0.6978,
317
+ "step": 880
318
+ },
319
+ {
320
+ "epoch": 0.1024444381207137,
321
+ "grad_norm": 0.4971105754375458,
322
+ "learning_rate": 0.0002,
323
+ "loss": 0.7103,
324
+ "step": 900
325
+ },
326
+ {
327
+ "epoch": 0.10472098119006289,
328
+ "grad_norm": 0.5066735744476318,
329
+ "learning_rate": 0.0002,
330
+ "loss": 0.6854,
331
+ "step": 920
332
+ },
333
+ {
334
+ "epoch": 0.10699752425941209,
335
+ "grad_norm": 0.4922661781311035,
336
+ "learning_rate": 0.0002,
337
+ "loss": 0.6231,
338
+ "step": 940
339
+ },
340
+ {
341
+ "epoch": 0.10927406732876127,
342
+ "grad_norm": 0.5949555039405823,
343
+ "learning_rate": 0.0002,
344
+ "loss": 0.6813,
345
+ "step": 960
346
+ },
347
+ {
348
+ "epoch": 0.11155061039811047,
349
+ "grad_norm": 0.581446647644043,
350
+ "learning_rate": 0.0002,
351
+ "loss": 0.6174,
352
+ "step": 980
353
+ },
354
+ {
355
+ "epoch": 0.11382715346745967,
356
+ "grad_norm": 0.6152529716491699,
357
+ "learning_rate": 0.0002,
358
+ "loss": 0.6405,
359
+ "step": 1000
360
+ },
361
+ {
362
+ "epoch": 0.11610369653680885,
363
+ "grad_norm": 0.5986836552619934,
364
+ "learning_rate": 0.0002,
365
+ "loss": 0.5776,
366
+ "step": 1020
367
+ },
368
+ {
369
+ "epoch": 0.11838023960615805,
370
+ "grad_norm": 0.4255094528198242,
371
+ "learning_rate": 0.0002,
372
+ "loss": 0.6576,
373
+ "step": 1040
374
+ },
375
+ {
376
+ "epoch": 0.12065678267550724,
377
+ "grad_norm": 0.4563849866390228,
378
+ "learning_rate": 0.0002,
379
+ "loss": 0.6647,
380
+ "step": 1060
381
+ },
382
+ {
383
+ "epoch": 0.12293332574485644,
384
+ "grad_norm": 0.593227744102478,
385
+ "learning_rate": 0.0002,
386
+ "loss": 0.6043,
387
+ "step": 1080
388
+ },
389
+ {
390
+ "epoch": 0.12520986881420562,
391
+ "grad_norm": 0.47059598565101624,
392
+ "learning_rate": 0.0002,
393
+ "loss": 0.591,
394
+ "step": 1100
395
+ },
396
+ {
397
+ "epoch": 0.12748641188355483,
398
+ "grad_norm": 0.5013225674629211,
399
+ "learning_rate": 0.0002,
400
+ "loss": 0.5947,
401
+ "step": 1120
402
+ },
403
+ {
404
+ "epoch": 0.12976295495290402,
405
+ "grad_norm": 0.46772757172584534,
406
+ "learning_rate": 0.0002,
407
+ "loss": 0.6292,
408
+ "step": 1140
409
+ },
410
+ {
411
+ "epoch": 0.1320394980222532,
412
+ "grad_norm": 0.5844313502311707,
413
+ "learning_rate": 0.0002,
414
+ "loss": 0.6128,
415
+ "step": 1160
416
+ },
417
+ {
418
+ "epoch": 0.1343160410916024,
419
+ "grad_norm": 0.5295489430427551,
420
+ "learning_rate": 0.0002,
421
+ "loss": 0.6064,
422
+ "step": 1180
423
+ },
424
+ {
425
+ "epoch": 0.1365925841609516,
426
+ "grad_norm": 0.4482004642486572,
427
+ "learning_rate": 0.0002,
428
+ "loss": 0.5899,
429
+ "step": 1200
430
+ },
431
+ {
432
+ "epoch": 0.1388691272303008,
433
+ "grad_norm": 0.6281692981719971,
434
+ "learning_rate": 0.0002,
435
+ "loss": 0.6109,
436
+ "step": 1220
437
+ },
438
+ {
439
+ "epoch": 0.14114567029964997,
440
+ "grad_norm": 0.4718242585659027,
441
+ "learning_rate": 0.0002,
442
+ "loss": 0.5857,
443
+ "step": 1240
444
+ },
445
+ {
446
+ "epoch": 0.14342221336899919,
447
+ "grad_norm": 0.5219341516494751,
448
+ "learning_rate": 0.0002,
449
+ "loss": 0.5581,
450
+ "step": 1260
451
+ },
452
+ {
453
+ "epoch": 0.14569875643834837,
454
+ "grad_norm": 0.47050580382347107,
455
+ "learning_rate": 0.0002,
456
+ "loss": 0.6368,
457
+ "step": 1280
458
+ },
459
+ {
460
+ "epoch": 0.14797529950769756,
461
+ "grad_norm": 0.5425338745117188,
462
+ "learning_rate": 0.0002,
463
+ "loss": 0.5626,
464
+ "step": 1300
465
+ },
466
+ {
467
+ "epoch": 0.15025184257704677,
468
+ "grad_norm": 0.4944934844970703,
469
+ "learning_rate": 0.0002,
470
+ "loss": 0.5337,
471
+ "step": 1320
472
+ },
473
+ {
474
+ "epoch": 0.15252838564639595,
475
+ "grad_norm": 0.5921599864959717,
476
+ "learning_rate": 0.0002,
477
+ "loss": 0.5672,
478
+ "step": 1340
479
+ },
480
+ {
481
+ "epoch": 0.15480492871574514,
482
+ "grad_norm": 0.4866751730442047,
483
+ "learning_rate": 0.0002,
484
+ "loss": 0.5305,
485
+ "step": 1360
486
+ },
487
+ {
488
+ "epoch": 0.15708147178509432,
489
+ "grad_norm": 0.62166827917099,
490
+ "learning_rate": 0.0002,
491
+ "loss": 0.5737,
492
+ "step": 1380
493
+ },
494
+ {
495
+ "epoch": 0.15935801485444354,
496
+ "grad_norm": 0.5006982684135437,
497
+ "learning_rate": 0.0002,
498
+ "loss": 0.5542,
499
+ "step": 1400
500
+ },
501
+ {
502
+ "epoch": 0.16163455792379272,
503
+ "grad_norm": 0.6090095043182373,
504
+ "learning_rate": 0.0002,
505
+ "loss": 0.5215,
506
+ "step": 1420
507
+ },
508
+ {
509
+ "epoch": 0.1639111009931419,
510
+ "grad_norm": 0.4260309636592865,
511
+ "learning_rate": 0.0002,
512
+ "loss": 0.5535,
513
+ "step": 1440
514
+ },
515
+ {
516
+ "epoch": 0.16618764406249112,
517
+ "grad_norm": 0.48657718300819397,
518
+ "learning_rate": 0.0002,
519
+ "loss": 0.5441,
520
+ "step": 1460
521
+ },
522
+ {
523
+ "epoch": 0.1684641871318403,
524
+ "grad_norm": 0.43275007605552673,
525
+ "learning_rate": 0.0002,
526
+ "loss": 0.5161,
527
+ "step": 1480
528
+ },
529
+ {
530
+ "epoch": 0.1707407302011895,
531
+ "grad_norm": 0.4225006699562073,
532
+ "learning_rate": 0.0002,
533
+ "loss": 0.512,
534
+ "step": 1500
535
+ },
536
+ {
537
+ "epoch": 0.17301727327053867,
538
+ "grad_norm": 0.5176346302032471,
539
+ "learning_rate": 0.0002,
540
+ "loss": 0.5384,
541
+ "step": 1520
542
+ },
543
+ {
544
+ "epoch": 0.1752938163398879,
545
+ "grad_norm": 0.6492679715156555,
546
+ "learning_rate": 0.0002,
547
+ "loss": 0.4981,
548
+ "step": 1540
549
+ },
550
+ {
551
+ "epoch": 0.17757035940923707,
552
+ "grad_norm": 0.5511758327484131,
553
+ "learning_rate": 0.0002,
554
+ "loss": 0.5289,
555
+ "step": 1560
556
+ },
557
+ {
558
+ "epoch": 0.17984690247858626,
559
+ "grad_norm": 0.5211341977119446,
560
+ "learning_rate": 0.0002,
561
+ "loss": 0.5002,
562
+ "step": 1580
563
+ },
564
+ {
565
+ "epoch": 0.18212344554793547,
566
+ "grad_norm": 0.5488260984420776,
567
+ "learning_rate": 0.0002,
568
+ "loss": 0.5178,
569
+ "step": 1600
570
+ },
571
+ {
572
+ "epoch": 0.18439998861728465,
573
+ "grad_norm": 0.6779264211654663,
574
+ "learning_rate": 0.0002,
575
+ "loss": 0.5155,
576
+ "step": 1620
577
+ },
578
+ {
579
+ "epoch": 0.18667653168663384,
580
+ "grad_norm": 0.502919614315033,
581
+ "learning_rate": 0.0002,
582
+ "loss": 0.4923,
583
+ "step": 1640
584
+ },
585
+ {
586
+ "epoch": 0.18895307475598305,
587
+ "grad_norm": 0.4989205300807953,
588
+ "learning_rate": 0.0002,
589
+ "loss": 0.4825,
590
+ "step": 1660
591
+ },
592
+ {
593
+ "epoch": 0.19122961782533224,
594
+ "grad_norm": 0.5155315399169922,
595
+ "learning_rate": 0.0002,
596
+ "loss": 0.4796,
597
+ "step": 1680
598
+ },
599
+ {
600
+ "epoch": 0.19350616089468142,
601
+ "grad_norm": 0.5648865699768066,
602
+ "learning_rate": 0.0002,
603
+ "loss": 0.4985,
604
+ "step": 1700
605
+ },
606
+ {
607
+ "epoch": 0.1957827039640306,
608
+ "grad_norm": 0.606176495552063,
609
+ "learning_rate": 0.0002,
610
+ "loss": 0.4819,
611
+ "step": 1720
612
+ },
613
+ {
614
+ "epoch": 0.19805924703337982,
615
+ "grad_norm": 0.5440786480903625,
616
+ "learning_rate": 0.0002,
617
+ "loss": 0.5213,
618
+ "step": 1740
619
+ },
620
+ {
621
+ "epoch": 0.200335790102729,
622
+ "grad_norm": 0.43152502179145813,
623
+ "learning_rate": 0.0002,
624
+ "loss": 0.4429,
625
+ "step": 1760
626
+ },
627
+ {
628
+ "epoch": 0.2026123331720782,
629
+ "grad_norm": 0.5701313614845276,
630
+ "learning_rate": 0.0002,
631
+ "loss": 0.4486,
632
+ "step": 1780
633
+ },
634
+ {
635
+ "epoch": 0.2048888762414274,
636
+ "grad_norm": 0.565666913986206,
637
+ "learning_rate": 0.0002,
638
+ "loss": 0.4561,
639
+ "step": 1800
640
+ },
641
+ {
642
+ "epoch": 0.2071654193107766,
643
+ "grad_norm": 0.5725598931312561,
644
+ "learning_rate": 0.0002,
645
+ "loss": 0.4757,
646
+ "step": 1820
647
+ },
648
+ {
649
+ "epoch": 0.20944196238012577,
650
+ "grad_norm": 0.4642520248889923,
651
+ "learning_rate": 0.0002,
652
+ "loss": 0.438,
653
+ "step": 1840
654
+ },
655
+ {
656
+ "epoch": 0.21171850544947496,
657
+ "grad_norm": 0.6077229976654053,
658
+ "learning_rate": 0.0002,
659
+ "loss": 0.4295,
660
+ "step": 1860
661
+ },
662
+ {
663
+ "epoch": 0.21399504851882417,
664
+ "grad_norm": 0.6314090490341187,
665
+ "learning_rate": 0.0002,
666
+ "loss": 0.449,
667
+ "step": 1880
668
+ },
669
+ {
670
+ "epoch": 0.21627159158817336,
671
+ "grad_norm": 0.4416756331920624,
672
+ "learning_rate": 0.0002,
673
+ "loss": 0.4554,
674
+ "step": 1900
675
+ },
676
+ {
677
+ "epoch": 0.21854813465752254,
678
+ "grad_norm": 0.5278882384300232,
679
+ "learning_rate": 0.0002,
680
+ "loss": 0.4554,
681
+ "step": 1920
682
+ },
683
+ {
684
+ "epoch": 0.22082467772687175,
685
+ "grad_norm": 0.45619043707847595,
686
+ "learning_rate": 0.0002,
687
+ "loss": 0.4868,
688
+ "step": 1940
689
+ },
690
+ {
691
+ "epoch": 0.22310122079622094,
692
+ "grad_norm": 0.5881581902503967,
693
+ "learning_rate": 0.0002,
694
+ "loss": 0.4672,
695
+ "step": 1960
696
+ },
697
+ {
698
+ "epoch": 0.22537776386557012,
699
+ "grad_norm": 0.5379284024238586,
700
+ "learning_rate": 0.0002,
701
+ "loss": 0.4531,
702
+ "step": 1980
703
+ },
704
+ {
705
+ "epoch": 0.22765430693491934,
706
+ "grad_norm": 0.5562624931335449,
707
+ "learning_rate": 0.0002,
708
+ "loss": 0.464,
709
+ "step": 2000
710
+ },
711
+ {
712
+ "epoch": 0.22993085000426852,
713
+ "grad_norm": 0.554499626159668,
714
+ "learning_rate": 0.0002,
715
+ "loss": 0.446,
716
+ "step": 2020
717
+ },
718
+ {
719
+ "epoch": 0.2322073930736177,
720
+ "grad_norm": 0.509219229221344,
721
+ "learning_rate": 0.0002,
722
+ "loss": 0.4417,
723
+ "step": 2040
724
+ },
725
+ {
726
+ "epoch": 0.2344839361429669,
727
+ "grad_norm": 0.5206849575042725,
728
+ "learning_rate": 0.0002,
729
+ "loss": 0.4118,
730
+ "step": 2060
731
+ },
732
+ {
733
+ "epoch": 0.2367604792123161,
734
+ "grad_norm": 0.548729658126831,
735
+ "learning_rate": 0.0002,
736
+ "loss": 0.4067,
737
+ "step": 2080
738
+ },
739
+ {
740
+ "epoch": 0.2390370222816653,
741
+ "grad_norm": 0.4220084846019745,
742
+ "learning_rate": 0.0002,
743
+ "loss": 0.428,
744
+ "step": 2100
745
+ },
746
+ {
747
+ "epoch": 0.24131356535101448,
748
+ "grad_norm": 0.5507292747497559,
749
+ "learning_rate": 0.0002,
750
+ "loss": 0.4176,
751
+ "step": 2120
752
+ },
753
+ {
754
+ "epoch": 0.2435901084203637,
755
+ "grad_norm": 0.5605701208114624,
756
+ "learning_rate": 0.0002,
757
+ "loss": 0.4661,
758
+ "step": 2140
759
+ },
760
+ {
761
+ "epoch": 0.24586665148971287,
762
+ "grad_norm": 0.43142881989479065,
763
+ "learning_rate": 0.0002,
764
+ "loss": 0.4197,
765
+ "step": 2160
766
+ },
767
+ {
768
+ "epoch": 0.24814319455906206,
769
+ "grad_norm": 0.47790080308914185,
770
+ "learning_rate": 0.0002,
771
+ "loss": 0.4568,
772
+ "step": 2180
773
+ },
774
+ {
775
+ "epoch": 0.25041973762841124,
776
+ "grad_norm": 0.6048968434333801,
777
+ "learning_rate": 0.0002,
778
+ "loss": 0.4199,
779
+ "step": 2200
780
+ },
781
+ {
782
+ "epoch": 0.25269628069776046,
783
+ "grad_norm": 0.4925907850265503,
784
+ "learning_rate": 0.0002,
785
+ "loss": 0.4325,
786
+ "step": 2220
787
+ },
788
+ {
789
+ "epoch": 0.25497282376710967,
790
+ "grad_norm": 0.5463051199913025,
791
+ "learning_rate": 0.0002,
792
+ "loss": 0.4549,
793
+ "step": 2240
794
+ },
795
+ {
796
+ "epoch": 0.2572493668364588,
797
+ "grad_norm": 0.4631319046020508,
798
+ "learning_rate": 0.0002,
799
+ "loss": 0.3977,
800
+ "step": 2260
801
+ },
802
+ {
803
+ "epoch": 0.25952590990580804,
804
+ "grad_norm": 0.4965234398841858,
805
+ "learning_rate": 0.0002,
806
+ "loss": 0.4285,
807
+ "step": 2280
808
+ },
809
+ {
810
+ "epoch": 0.2618024529751572,
811
+ "grad_norm": 0.5436238646507263,
812
+ "learning_rate": 0.0002,
813
+ "loss": 0.4039,
814
+ "step": 2300
815
+ },
816
+ {
817
+ "epoch": 0.2640789960445064,
818
+ "grad_norm": 0.5218191742897034,
819
+ "learning_rate": 0.0002,
820
+ "loss": 0.4092,
821
+ "step": 2320
822
+ },
823
+ {
824
+ "epoch": 0.2663555391138556,
825
+ "grad_norm": 0.5417261719703674,
826
+ "learning_rate": 0.0002,
827
+ "loss": 0.3825,
828
+ "step": 2340
829
+ },
830
+ {
831
+ "epoch": 0.2686320821832048,
832
+ "grad_norm": 0.6126281023025513,
833
+ "learning_rate": 0.0002,
834
+ "loss": 0.4391,
835
+ "step": 2360
836
+ },
837
+ {
838
+ "epoch": 0.270908625252554,
839
+ "grad_norm": 0.4734433889389038,
840
+ "learning_rate": 0.0002,
841
+ "loss": 0.4151,
842
+ "step": 2380
843
+ },
844
+ {
845
+ "epoch": 0.2731851683219032,
846
+ "grad_norm": 0.4501429796218872,
847
+ "learning_rate": 0.0002,
848
+ "loss": 0.4178,
849
+ "step": 2400
850
+ },
851
+ {
852
+ "epoch": 0.27546171139125236,
853
+ "grad_norm": 0.5258509516716003,
854
+ "learning_rate": 0.0002,
855
+ "loss": 0.4007,
856
+ "step": 2420
857
+ },
858
+ {
859
+ "epoch": 0.2777382544606016,
860
+ "grad_norm": 0.47874951362609863,
861
+ "learning_rate": 0.0002,
862
+ "loss": 0.4245,
863
+ "step": 2440
864
+ },
865
+ {
866
+ "epoch": 0.2800147975299508,
867
+ "grad_norm": 0.528533399105072,
868
+ "learning_rate": 0.0002,
869
+ "loss": 0.3794,
870
+ "step": 2460
871
+ },
872
+ {
873
+ "epoch": 0.28229134059929994,
874
+ "grad_norm": 0.46465063095092773,
875
+ "learning_rate": 0.0002,
876
+ "loss": 0.4019,
877
+ "step": 2480
878
+ },
879
+ {
880
+ "epoch": 0.28456788366864916,
881
+ "grad_norm": 0.5217177867889404,
882
+ "learning_rate": 0.0002,
883
+ "loss": 0.4104,
884
+ "step": 2500
885
+ },
886
+ {
887
+ "epoch": 0.28684442673799837,
888
+ "grad_norm": 0.510036289691925,
889
+ "learning_rate": 0.0002,
890
+ "loss": 0.389,
891
+ "step": 2520
892
+ },
893
+ {
894
+ "epoch": 0.2891209698073475,
895
+ "grad_norm": 0.6968228220939636,
896
+ "learning_rate": 0.0002,
897
+ "loss": 0.4152,
898
+ "step": 2540
899
+ },
900
+ {
901
+ "epoch": 0.29139751287669674,
902
+ "grad_norm": 0.4529867470264435,
903
+ "learning_rate": 0.0002,
904
+ "loss": 0.3987,
905
+ "step": 2560
906
+ },
907
+ {
908
+ "epoch": 0.29367405594604595,
909
+ "grad_norm": 0.5680263638496399,
910
+ "learning_rate": 0.0002,
911
+ "loss": 0.3828,
912
+ "step": 2580
913
+ },
914
+ {
915
+ "epoch": 0.2959505990153951,
916
+ "grad_norm": 0.4892405867576599,
917
+ "learning_rate": 0.0002,
918
+ "loss": 0.4006,
919
+ "step": 2600
920
+ },
921
+ {
922
+ "epoch": 0.2982271420847443,
923
+ "grad_norm": 0.47588276863098145,
924
+ "learning_rate": 0.0002,
925
+ "loss": 0.4197,
926
+ "step": 2620
927
+ },
928
+ {
929
+ "epoch": 0.30050368515409354,
930
+ "grad_norm": 0.5624070167541504,
931
+ "learning_rate": 0.0002,
932
+ "loss": 0.3997,
933
+ "step": 2640
934
+ },
935
+ {
936
+ "epoch": 0.3027802282234427,
937
+ "grad_norm": 0.5434039831161499,
938
+ "learning_rate": 0.0002,
939
+ "loss": 0.3977,
940
+ "step": 2660
941
+ },
942
+ {
943
+ "epoch": 0.3050567712927919,
944
+ "grad_norm": 0.5572277903556824,
945
+ "learning_rate": 0.0002,
946
+ "loss": 0.3966,
947
+ "step": 2680
948
+ },
949
+ {
950
+ "epoch": 0.30733331436214106,
951
+ "grad_norm": 0.5533374547958374,
952
+ "learning_rate": 0.0002,
953
+ "loss": 0.3803,
954
+ "step": 2700
955
+ },
956
+ {
957
+ "epoch": 0.3096098574314903,
958
+ "grad_norm": 0.40596967935562134,
959
+ "learning_rate": 0.0002,
960
+ "loss": 0.3682,
961
+ "step": 2720
962
+ },
963
+ {
964
+ "epoch": 0.3118864005008395,
965
+ "grad_norm": 0.4737823009490967,
966
+ "learning_rate": 0.0002,
967
+ "loss": 0.3761,
968
+ "step": 2740
969
+ },
970
+ {
971
+ "epoch": 0.31416294357018865,
972
+ "grad_norm": 0.4295174777507782,
973
+ "learning_rate": 0.0002,
974
+ "loss": 0.4035,
975
+ "step": 2760
976
+ },
977
+ {
978
+ "epoch": 0.31643948663953786,
979
+ "grad_norm": 0.5348454713821411,
980
+ "learning_rate": 0.0002,
981
+ "loss": 0.404,
982
+ "step": 2780
983
+ },
984
+ {
985
+ "epoch": 0.31871602970888707,
986
+ "grad_norm": 0.4819965362548828,
987
+ "learning_rate": 0.0002,
988
+ "loss": 0.3929,
989
+ "step": 2800
990
+ },
991
+ {
992
+ "epoch": 0.32099257277823623,
993
+ "grad_norm": 0.5920088291168213,
994
+ "learning_rate": 0.0002,
995
+ "loss": 0.3798,
996
+ "step": 2820
997
+ },
998
+ {
999
+ "epoch": 0.32326911584758544,
1000
+ "grad_norm": 0.4936531186103821,
1001
+ "learning_rate": 0.0002,
1002
+ "loss": 0.3995,
1003
+ "step": 2840
1004
+ },
1005
+ {
1006
+ "epoch": 0.32554565891693465,
1007
+ "grad_norm": 0.5252315998077393,
1008
+ "learning_rate": 0.0002,
1009
+ "loss": 0.3842,
1010
+ "step": 2860
1011
+ },
1012
+ {
1013
+ "epoch": 0.3278222019862838,
1014
+ "grad_norm": 0.5818414688110352,
1015
+ "learning_rate": 0.0002,
1016
+ "loss": 0.3533,
1017
+ "step": 2880
1018
+ },
1019
+ {
1020
+ "epoch": 0.330098745055633,
1021
+ "grad_norm": 0.44053876399993896,
1022
+ "learning_rate": 0.0002,
1023
+ "loss": 0.3402,
1024
+ "step": 2900
1025
+ },
1026
+ {
1027
+ "epoch": 0.33237528812498224,
1028
+ "grad_norm": 0.5421345233917236,
1029
+ "learning_rate": 0.0002,
1030
+ "loss": 0.3542,
1031
+ "step": 2920
1032
+ },
1033
+ {
1034
+ "epoch": 0.3346518311943314,
1035
+ "grad_norm": 0.4642751216888428,
1036
+ "learning_rate": 0.0002,
1037
+ "loss": 0.3755,
1038
+ "step": 2940
1039
+ },
1040
+ {
1041
+ "epoch": 0.3369283742636806,
1042
+ "grad_norm": 0.5137833952903748,
1043
+ "learning_rate": 0.0002,
1044
+ "loss": 0.3602,
1045
+ "step": 2960
1046
+ },
1047
+ {
1048
+ "epoch": 0.3392049173330298,
1049
+ "grad_norm": 0.5032792687416077,
1050
+ "learning_rate": 0.0002,
1051
+ "loss": 0.3451,
1052
+ "step": 2980
1053
+ },
1054
+ {
1055
+ "epoch": 0.341481460402379,
1056
+ "grad_norm": 0.4932720363140106,
1057
+ "learning_rate": 0.0002,
1058
+ "loss": 0.384,
1059
+ "step": 3000
1060
+ },
1061
+ {
1062
+ "epoch": 0.3437580034717282,
1063
+ "grad_norm": 0.49986231327056885,
1064
+ "learning_rate": 0.0002,
1065
+ "loss": 0.3826,
1066
+ "step": 3020
1067
+ },
1068
+ {
1069
+ "epoch": 0.34603454654107735,
1070
+ "grad_norm": 0.6325618624687195,
1071
+ "learning_rate": 0.0002,
1072
+ "loss": 0.3582,
1073
+ "step": 3040
1074
+ },
1075
+ {
1076
+ "epoch": 0.34831108961042656,
1077
+ "grad_norm": 0.5402369499206543,
1078
+ "learning_rate": 0.0002,
1079
+ "loss": 0.3706,
1080
+ "step": 3060
1081
+ },
1082
+ {
1083
+ "epoch": 0.3505876326797758,
1084
+ "grad_norm": 0.4967012107372284,
1085
+ "learning_rate": 0.0002,
1086
+ "loss": 0.3456,
1087
+ "step": 3080
1088
+ },
1089
+ {
1090
+ "epoch": 0.35286417574912493,
1091
+ "grad_norm": 0.4491735100746155,
1092
+ "learning_rate": 0.0002,
1093
+ "loss": 0.347,
1094
+ "step": 3100
1095
+ },
1096
+ {
1097
+ "epoch": 0.35514071881847414,
1098
+ "grad_norm": 0.9062516093254089,
1099
+ "learning_rate": 0.0002,
1100
+ "loss": 0.3617,
1101
+ "step": 3120
1102
+ },
1103
+ {
1104
+ "epoch": 0.35741726188782336,
1105
+ "grad_norm": 0.5253359079360962,
1106
+ "learning_rate": 0.0002,
1107
+ "loss": 0.3512,
1108
+ "step": 3140
1109
+ },
1110
+ {
1111
+ "epoch": 0.3596938049571725,
1112
+ "grad_norm": 0.4836867153644562,
1113
+ "learning_rate": 0.0002,
1114
+ "loss": 0.3585,
1115
+ "step": 3160
1116
+ },
1117
+ {
1118
+ "epoch": 0.3619703480265217,
1119
+ "grad_norm": 0.49537473917007446,
1120
+ "learning_rate": 0.0002,
1121
+ "loss": 0.364,
1122
+ "step": 3180
1123
+ },
1124
+ {
1125
+ "epoch": 0.36424689109587094,
1126
+ "grad_norm": 0.6098095178604126,
1127
+ "learning_rate": 0.0002,
1128
+ "loss": 0.3455,
1129
+ "step": 3200
1130
+ },
1131
+ {
1132
+ "epoch": 0.3665234341652201,
1133
+ "grad_norm": 0.5926884412765503,
1134
+ "learning_rate": 0.0002,
1135
+ "loss": 0.3406,
1136
+ "step": 3220
1137
+ },
1138
+ {
1139
+ "epoch": 0.3687999772345693,
1140
+ "grad_norm": 0.5868669152259827,
1141
+ "learning_rate": 0.0002,
1142
+ "loss": 0.3643,
1143
+ "step": 3240
1144
+ },
1145
+ {
1146
+ "epoch": 0.3710765203039185,
1147
+ "grad_norm": 0.42670106887817383,
1148
+ "learning_rate": 0.0002,
1149
+ "loss": 0.344,
1150
+ "step": 3260
1151
+ },
1152
+ {
1153
+ "epoch": 0.3733530633732677,
1154
+ "grad_norm": 0.5992838740348816,
1155
+ "learning_rate": 0.0002,
1156
+ "loss": 0.3588,
1157
+ "step": 3280
1158
+ },
1159
+ {
1160
+ "epoch": 0.3756296064426169,
1161
+ "grad_norm": 0.4388341009616852,
1162
+ "learning_rate": 0.0002,
1163
+ "loss": 0.3375,
1164
+ "step": 3300
1165
+ },
1166
+ {
1167
+ "epoch": 0.3779061495119661,
1168
+ "grad_norm": 0.596488893032074,
1169
+ "learning_rate": 0.0002,
1170
+ "loss": 0.3425,
1171
+ "step": 3320
1172
+ },
1173
+ {
1174
+ "epoch": 0.38018269258131526,
1175
+ "grad_norm": 0.4572538137435913,
1176
+ "learning_rate": 0.0002,
1177
+ "loss": 0.3711,
1178
+ "step": 3340
1179
+ },
1180
+ {
1181
+ "epoch": 0.3824592356506645,
1182
+ "grad_norm": 0.5661656856536865,
1183
+ "learning_rate": 0.0002,
1184
+ "loss": 0.3415,
1185
+ "step": 3360
1186
+ },
1187
+ {
1188
+ "epoch": 0.38473577872001363,
1189
+ "grad_norm": 0.45082923769950867,
1190
+ "learning_rate": 0.0002,
1191
+ "loss": 0.3495,
1192
+ "step": 3380
1193
+ },
1194
+ {
1195
+ "epoch": 0.38701232178936285,
1196
+ "grad_norm": 0.4995211660861969,
1197
+ "learning_rate": 0.0002,
1198
+ "loss": 0.3311,
1199
+ "step": 3400
1200
+ },
1201
+ {
1202
+ "epoch": 0.38928886485871206,
1203
+ "grad_norm": 0.5004004240036011,
1204
+ "learning_rate": 0.0002,
1205
+ "loss": 0.3506,
1206
+ "step": 3420
1207
+ },
1208
+ {
1209
+ "epoch": 0.3915654079280612,
1210
+ "grad_norm": 0.5676460266113281,
1211
+ "learning_rate": 0.0002,
1212
+ "loss": 0.3383,
1213
+ "step": 3440
1214
+ },
1215
+ {
1216
+ "epoch": 0.39384195099741043,
1217
+ "grad_norm": 0.4805515706539154,
1218
+ "learning_rate": 0.0002,
1219
+ "loss": 0.3382,
1220
+ "step": 3460
1221
+ },
1222
+ {
1223
+ "epoch": 0.39611849406675964,
1224
+ "grad_norm": 0.47675764560699463,
1225
+ "learning_rate": 0.0002,
1226
+ "loss": 0.3021,
1227
+ "step": 3480
1228
+ },
1229
+ {
1230
+ "epoch": 0.3983950371361088,
1231
+ "grad_norm": 0.6285260915756226,
1232
+ "learning_rate": 0.0002,
1233
+ "loss": 0.3467,
1234
+ "step": 3500
1235
+ },
1236
+ {
1237
+ "epoch": 0.400671580205458,
1238
+ "grad_norm": 0.5657575130462646,
1239
+ "learning_rate": 0.0002,
1240
+ "loss": 0.3382,
1241
+ "step": 3520
1242
+ },
1243
+ {
1244
+ "epoch": 0.4029481232748072,
1245
+ "grad_norm": 0.6148316860198975,
1246
+ "learning_rate": 0.0002,
1247
+ "loss": 0.3396,
1248
+ "step": 3540
1249
+ },
1250
+ {
1251
+ "epoch": 0.4052246663441564,
1252
+ "grad_norm": 0.5819992423057556,
1253
+ "learning_rate": 0.0002,
1254
+ "loss": 0.3373,
1255
+ "step": 3560
1256
+ },
1257
+ {
1258
+ "epoch": 0.4075012094135056,
1259
+ "grad_norm": 0.6080338954925537,
1260
+ "learning_rate": 0.0002,
1261
+ "loss": 0.3463,
1262
+ "step": 3580
1263
+ },
1264
+ {
1265
+ "epoch": 0.4097777524828548,
1266
+ "grad_norm": 0.6103864312171936,
1267
+ "learning_rate": 0.0002,
1268
+ "loss": 0.3441,
1269
+ "step": 3600
1270
+ },
1271
+ {
1272
+ "epoch": 0.41205429555220396,
1273
+ "grad_norm": 0.5234800577163696,
1274
+ "learning_rate": 0.0002,
1275
+ "loss": 0.3272,
1276
+ "step": 3620
1277
+ },
1278
+ {
1279
+ "epoch": 0.4143308386215532,
1280
+ "grad_norm": 0.5393822193145752,
1281
+ "learning_rate": 0.0002,
1282
+ "loss": 0.3308,
1283
+ "step": 3640
1284
+ },
1285
+ {
1286
+ "epoch": 0.4166073816909024,
1287
+ "grad_norm": 0.4853431284427643,
1288
+ "learning_rate": 0.0002,
1289
+ "loss": 0.3152,
1290
+ "step": 3660
1291
+ },
1292
+ {
1293
+ "epoch": 0.41888392476025155,
1294
+ "grad_norm": 0.5507264733314514,
1295
+ "learning_rate": 0.0002,
1296
+ "loss": 0.3229,
1297
+ "step": 3680
1298
+ },
1299
+ {
1300
+ "epoch": 0.42116046782960076,
1301
+ "grad_norm": 0.44306129217147827,
1302
+ "learning_rate": 0.0002,
1303
+ "loss": 0.3389,
1304
+ "step": 3700
1305
+ },
1306
+ {
1307
+ "epoch": 0.4234370108989499,
1308
+ "grad_norm": 0.4574294984340668,
1309
+ "learning_rate": 0.0002,
1310
+ "loss": 0.3516,
1311
+ "step": 3720
1312
+ },
1313
+ {
1314
+ "epoch": 0.42571355396829913,
1315
+ "grad_norm": 0.5367994904518127,
1316
+ "learning_rate": 0.0002,
1317
+ "loss": 0.3576,
1318
+ "step": 3740
1319
+ },
1320
+ {
1321
+ "epoch": 0.42799009703764834,
1322
+ "grad_norm": 0.5044491291046143,
1323
+ "learning_rate": 0.0002,
1324
+ "loss": 0.3449,
1325
+ "step": 3760
1326
+ },
1327
+ {
1328
+ "epoch": 0.4302666401069975,
1329
+ "grad_norm": 0.41715556383132935,
1330
+ "learning_rate": 0.0002,
1331
+ "loss": 0.3128,
1332
+ "step": 3780
1333
+ },
1334
+ {
1335
+ "epoch": 0.4325431831763467,
1336
+ "grad_norm": 0.4355817437171936,
1337
+ "learning_rate": 0.0002,
1338
+ "loss": 0.3131,
1339
+ "step": 3800
1340
+ },
1341
+ {
1342
+ "epoch": 0.4348197262456959,
1343
+ "grad_norm": 0.5237382650375366,
1344
+ "learning_rate": 0.0002,
1345
+ "loss": 0.3281,
1346
+ "step": 3820
1347
+ },
1348
+ {
1349
+ "epoch": 0.4370962693150451,
1350
+ "grad_norm": 0.6210081577301025,
1351
+ "learning_rate": 0.0002,
1352
+ "loss": 0.3195,
1353
+ "step": 3840
1354
+ },
1355
+ {
1356
+ "epoch": 0.4393728123843943,
1357
+ "grad_norm": 0.5145352482795715,
1358
+ "learning_rate": 0.0002,
1359
+ "loss": 0.3107,
1360
+ "step": 3860
1361
+ },
1362
+ {
1363
+ "epoch": 0.4416493554537435,
1364
+ "grad_norm": 0.5554608106613159,
1365
+ "learning_rate": 0.0002,
1366
+ "loss": 0.3418,
1367
+ "step": 3880
1368
+ },
1369
+ {
1370
+ "epoch": 0.44392589852309267,
1371
+ "grad_norm": 0.4971628487110138,
1372
+ "learning_rate": 0.0002,
1373
+ "loss": 0.3293,
1374
+ "step": 3900
1375
+ },
1376
+ {
1377
+ "epoch": 0.4462024415924419,
1378
+ "grad_norm": 0.49732130765914917,
1379
+ "learning_rate": 0.0002,
1380
+ "loss": 0.3138,
1381
+ "step": 3920
1382
+ },
1383
+ {
1384
+ "epoch": 0.4484789846617911,
1385
+ "grad_norm": 0.5883257985115051,
1386
+ "learning_rate": 0.0002,
1387
+ "loss": 0.3357,
1388
+ "step": 3940
1389
+ },
1390
+ {
1391
+ "epoch": 0.45075552773114025,
1392
+ "grad_norm": 0.5349528193473816,
1393
+ "learning_rate": 0.0002,
1394
+ "loss": 0.3381,
1395
+ "step": 3960
1396
+ },
1397
+ {
1398
+ "epoch": 0.45303207080048946,
1399
+ "grad_norm": 0.5360047221183777,
1400
+ "learning_rate": 0.0002,
1401
+ "loss": 0.3116,
1402
+ "step": 3980
1403
+ },
1404
+ {
1405
+ "epoch": 0.4553086138698387,
1406
+ "grad_norm": 0.4889732003211975,
1407
+ "learning_rate": 0.0002,
1408
+ "loss": 0.3154,
1409
+ "step": 4000
1410
+ },
1411
+ {
1412
+ "epoch": 0.45758515693918783,
1413
+ "grad_norm": 0.4912421703338623,
1414
+ "learning_rate": 0.0002,
1415
+ "loss": 0.3054,
1416
+ "step": 4020
1417
+ },
1418
+ {
1419
+ "epoch": 0.45986170000853704,
1420
+ "grad_norm": 0.4449983835220337,
1421
+ "learning_rate": 0.0002,
1422
+ "loss": 0.3079,
1423
+ "step": 4040
1424
+ },
1425
+ {
1426
+ "epoch": 0.46213824307788626,
1427
+ "grad_norm": 0.4488675892353058,
1428
+ "learning_rate": 0.0002,
1429
+ "loss": 0.3027,
1430
+ "step": 4060
1431
+ },
1432
+ {
1433
+ "epoch": 0.4644147861472354,
1434
+ "grad_norm": 0.5412561893463135,
1435
+ "learning_rate": 0.0002,
1436
+ "loss": 0.2932,
1437
+ "step": 4080
1438
+ },
1439
+ {
1440
+ "epoch": 0.4666913292165846,
1441
+ "grad_norm": 0.41218650341033936,
1442
+ "learning_rate": 0.0002,
1443
+ "loss": 0.3087,
1444
+ "step": 4100
1445
+ },
1446
+ {
1447
+ "epoch": 0.4689678722859338,
1448
+ "grad_norm": 0.5233949422836304,
1449
+ "learning_rate": 0.0002,
1450
+ "loss": 0.3157,
1451
+ "step": 4120
1452
+ },
1453
+ {
1454
+ "epoch": 0.471244415355283,
1455
+ "grad_norm": 0.5676075220108032,
1456
+ "learning_rate": 0.0002,
1457
+ "loss": 0.3267,
1458
+ "step": 4140
1459
+ },
1460
+ {
1461
+ "epoch": 0.4735209584246322,
1462
+ "grad_norm": 0.5336834788322449,
1463
+ "learning_rate": 0.0002,
1464
+ "loss": 0.3185,
1465
+ "step": 4160
1466
+ },
1467
+ {
1468
+ "epoch": 0.47579750149398137,
1469
+ "grad_norm": 0.5505925416946411,
1470
+ "learning_rate": 0.0002,
1471
+ "loss": 0.3116,
1472
+ "step": 4180
1473
+ },
1474
+ {
1475
+ "epoch": 0.4780740445633306,
1476
+ "grad_norm": 0.5440223813056946,
1477
+ "learning_rate": 0.0002,
1478
+ "loss": 0.3234,
1479
+ "step": 4200
1480
+ },
1481
+ {
1482
+ "epoch": 0.4803505876326798,
1483
+ "grad_norm": 0.46334293484687805,
1484
+ "learning_rate": 0.0002,
1485
+ "loss": 0.3209,
1486
+ "step": 4220
1487
+ },
1488
+ {
1489
+ "epoch": 0.48262713070202895,
1490
+ "grad_norm": 0.452364444732666,
1491
+ "learning_rate": 0.0002,
1492
+ "loss": 0.3056,
1493
+ "step": 4240
1494
+ },
1495
+ {
1496
+ "epoch": 0.48490367377137816,
1497
+ "grad_norm": 0.5037956833839417,
1498
+ "learning_rate": 0.0002,
1499
+ "loss": 0.3141,
1500
+ "step": 4260
1501
+ },
1502
+ {
1503
+ "epoch": 0.4871802168407274,
1504
+ "grad_norm": 0.4308939278125763,
1505
+ "learning_rate": 0.0002,
1506
+ "loss": 0.2948,
1507
+ "step": 4280
1508
+ },
1509
+ {
1510
+ "epoch": 0.48945675991007653,
1511
+ "grad_norm": 0.45019960403442383,
1512
+ "learning_rate": 0.0002,
1513
+ "loss": 0.3142,
1514
+ "step": 4300
1515
+ },
1516
+ {
1517
+ "epoch": 0.49173330297942575,
1518
+ "grad_norm": 0.4351404011249542,
1519
+ "learning_rate": 0.0002,
1520
+ "loss": 0.31,
1521
+ "step": 4320
1522
+ },
1523
+ {
1524
+ "epoch": 0.49400984604877496,
1525
+ "grad_norm": 0.38306841254234314,
1526
+ "learning_rate": 0.0002,
1527
+ "loss": 0.2889,
1528
+ "step": 4340
1529
+ },
1530
+ {
1531
+ "epoch": 0.4962863891181241,
1532
+ "grad_norm": 0.545360803604126,
1533
+ "learning_rate": 0.0002,
1534
+ "loss": 0.311,
1535
+ "step": 4360
1536
+ },
1537
+ {
1538
+ "epoch": 0.49856293218747333,
1539
+ "grad_norm": 0.44942232966423035,
1540
+ "learning_rate": 0.0002,
1541
+ "loss": 0.2899,
1542
+ "step": 4380
1543
+ },
1544
+ {
1545
+ "epoch": 0.5008394752568225,
1546
+ "grad_norm": 0.46564239263534546,
1547
+ "learning_rate": 0.0002,
1548
+ "loss": 0.3013,
1549
+ "step": 4400
1550
+ },
1551
+ {
1552
+ "epoch": 0.5031160183261717,
1553
+ "grad_norm": 0.5398554801940918,
1554
+ "learning_rate": 0.0002,
1555
+ "loss": 0.3104,
1556
+ "step": 4420
1557
+ },
1558
+ {
1559
+ "epoch": 0.5053925613955209,
1560
+ "grad_norm": 0.47367504239082336,
1561
+ "learning_rate": 0.0002,
1562
+ "loss": 0.2945,
1563
+ "step": 4440
1564
+ },
1565
+ {
1566
+ "epoch": 0.5076691044648701,
1567
+ "grad_norm": 0.45659711956977844,
1568
+ "learning_rate": 0.0002,
1569
+ "loss": 0.304,
1570
+ "step": 4460
1571
+ },
1572
+ {
1573
+ "epoch": 0.5099456475342193,
1574
+ "grad_norm": 0.4942033290863037,
1575
+ "learning_rate": 0.0002,
1576
+ "loss": 0.2969,
1577
+ "step": 4480
1578
+ },
1579
+ {
1580
+ "epoch": 0.5122221906035684,
1581
+ "grad_norm": 0.46578243374824524,
1582
+ "learning_rate": 0.0002,
1583
+ "loss": 0.2935,
1584
+ "step": 4500
1585
+ },
1586
+ {
1587
+ "epoch": 0.5144987336729177,
1588
+ "grad_norm": 0.6523891687393188,
1589
+ "learning_rate": 0.0002,
1590
+ "loss": 0.2823,
1591
+ "step": 4520
1592
+ },
1593
+ {
1594
+ "epoch": 0.5167752767422669,
1595
+ "grad_norm": 0.4787238538265228,
1596
+ "learning_rate": 0.0002,
1597
+ "loss": 0.3148,
1598
+ "step": 4540
1599
+ },
1600
+ {
1601
+ "epoch": 0.5190518198116161,
1602
+ "grad_norm": 0.46825891733169556,
1603
+ "learning_rate": 0.0002,
1604
+ "loss": 0.3089,
1605
+ "step": 4560
1606
+ },
1607
+ {
1608
+ "epoch": 0.5213283628809653,
1609
+ "grad_norm": 0.46605536341667175,
1610
+ "learning_rate": 0.0002,
1611
+ "loss": 0.3012,
1612
+ "step": 4580
1613
+ },
1614
+ {
1615
+ "epoch": 0.5236049059503144,
1616
+ "grad_norm": 0.5826888680458069,
1617
+ "learning_rate": 0.0002,
1618
+ "loss": 0.3043,
1619
+ "step": 4600
1620
+ },
1621
+ {
1622
+ "epoch": 0.5258814490196636,
1623
+ "grad_norm": 0.48641151189804077,
1624
+ "learning_rate": 0.0002,
1625
+ "loss": 0.2952,
1626
+ "step": 4620
1627
+ },
1628
+ {
1629
+ "epoch": 0.5281579920890128,
1630
+ "grad_norm": 0.5396175384521484,
1631
+ "learning_rate": 0.0002,
1632
+ "loss": 0.2926,
1633
+ "step": 4640
1634
+ },
1635
+ {
1636
+ "epoch": 0.530434535158362,
1637
+ "grad_norm": 0.5584241151809692,
1638
+ "learning_rate": 0.0002,
1639
+ "loss": 0.3048,
1640
+ "step": 4660
1641
+ },
1642
+ {
1643
+ "epoch": 0.5327110782277112,
1644
+ "grad_norm": 0.5832685232162476,
1645
+ "learning_rate": 0.0002,
1646
+ "loss": 0.2948,
1647
+ "step": 4680
1648
+ },
1649
+ {
1650
+ "epoch": 0.5349876212970605,
1651
+ "grad_norm": 0.4676337242126465,
1652
+ "learning_rate": 0.0002,
1653
+ "loss": 0.3043,
1654
+ "step": 4700
1655
+ },
1656
+ {
1657
+ "epoch": 0.5372641643664096,
1658
+ "grad_norm": 0.4440428614616394,
1659
+ "learning_rate": 0.0002,
1660
+ "loss": 0.288,
1661
+ "step": 4720
1662
+ },
1663
+ {
1664
+ "epoch": 0.5395407074357588,
1665
+ "grad_norm": 0.49934279918670654,
1666
+ "learning_rate": 0.0002,
1667
+ "loss": 0.2882,
1668
+ "step": 4740
1669
+ },
1670
+ {
1671
+ "epoch": 0.541817250505108,
1672
+ "grad_norm": 0.5172054171562195,
1673
+ "learning_rate": 0.0002,
1674
+ "loss": 0.3225,
1675
+ "step": 4760
1676
+ },
1677
+ {
1678
+ "epoch": 0.5440937935744572,
1679
+ "grad_norm": 0.4527619183063507,
1680
+ "learning_rate": 0.0002,
1681
+ "loss": 0.2869,
1682
+ "step": 4780
1683
+ },
1684
+ {
1685
+ "epoch": 0.5463703366438064,
1686
+ "grad_norm": 0.548918604850769,
1687
+ "learning_rate": 0.0002,
1688
+ "loss": 0.3105,
1689
+ "step": 4800
1690
+ },
1691
+ {
1692
+ "epoch": 0.5486468797131556,
1693
+ "grad_norm": 0.48801419138908386,
1694
+ "learning_rate": 0.0002,
1695
+ "loss": 0.2835,
1696
+ "step": 4820
1697
+ },
1698
+ {
1699
+ "epoch": 0.5509234227825047,
1700
+ "grad_norm": 0.49810609221458435,
1701
+ "learning_rate": 0.0002,
1702
+ "loss": 0.3227,
1703
+ "step": 4840
1704
+ },
1705
+ {
1706
+ "epoch": 0.5531999658518539,
1707
+ "grad_norm": 0.49763086438179016,
1708
+ "learning_rate": 0.0002,
1709
+ "loss": 0.2786,
1710
+ "step": 4860
1711
+ },
1712
+ {
1713
+ "epoch": 0.5554765089212031,
1714
+ "grad_norm": 0.48815059661865234,
1715
+ "learning_rate": 0.0002,
1716
+ "loss": 0.2802,
1717
+ "step": 4880
1718
+ },
1719
+ {
1720
+ "epoch": 0.5577530519905524,
1721
+ "grad_norm": 0.3571115732192993,
1722
+ "learning_rate": 0.0002,
1723
+ "loss": 0.2796,
1724
+ "step": 4900
1725
+ },
1726
+ {
1727
+ "epoch": 0.5600295950599016,
1728
+ "grad_norm": 0.6448425650596619,
1729
+ "learning_rate": 0.0002,
1730
+ "loss": 0.2844,
1731
+ "step": 4920
1732
+ },
1733
+ {
1734
+ "epoch": 0.5623061381292508,
1735
+ "grad_norm": 0.49660468101501465,
1736
+ "learning_rate": 0.0002,
1737
+ "loss": 0.2892,
1738
+ "step": 4940
1739
+ },
1740
+ {
1741
+ "epoch": 0.5645826811985999,
1742
+ "grad_norm": 0.47702720761299133,
1743
+ "learning_rate": 0.0002,
1744
+ "loss": 0.3111,
1745
+ "step": 4960
1746
+ },
1747
+ {
1748
+ "epoch": 0.5668592242679491,
1749
+ "grad_norm": 0.5281921029090881,
1750
+ "learning_rate": 0.0002,
1751
+ "loss": 0.2908,
1752
+ "step": 4980
1753
+ },
1754
+ {
1755
+ "epoch": 0.5691357673372983,
1756
+ "grad_norm": 0.6427987813949585,
1757
+ "learning_rate": 0.0002,
1758
+ "loss": 0.2848,
1759
+ "step": 5000
1760
+ },
1761
+ {
1762
+ "epoch": 0.5714123104066475,
1763
+ "grad_norm": 0.5437233448028564,
1764
+ "learning_rate": 0.0002,
1765
+ "loss": 0.3023,
1766
+ "step": 5020
1767
+ },
1768
+ {
1769
+ "epoch": 0.5736888534759967,
1770
+ "grad_norm": 0.517444372177124,
1771
+ "learning_rate": 0.0002,
1772
+ "loss": 0.2876,
1773
+ "step": 5040
1774
+ },
1775
+ {
1776
+ "epoch": 0.5759653965453458,
1777
+ "grad_norm": 0.5197298526763916,
1778
+ "learning_rate": 0.0002,
1779
+ "loss": 0.304,
1780
+ "step": 5060
1781
+ },
1782
+ {
1783
+ "epoch": 0.578241939614695,
1784
+ "grad_norm": 0.3452152907848358,
1785
+ "learning_rate": 0.0002,
1786
+ "loss": 0.2794,
1787
+ "step": 5080
1788
+ },
1789
+ {
1790
+ "epoch": 0.5805184826840443,
1791
+ "grad_norm": 0.5630306601524353,
1792
+ "learning_rate": 0.0002,
1793
+ "loss": 0.2979,
1794
+ "step": 5100
1795
+ },
1796
+ {
1797
+ "epoch": 0.5827950257533935,
1798
+ "grad_norm": 0.5696737170219421,
1799
+ "learning_rate": 0.0002,
1800
+ "loss": 0.3035,
1801
+ "step": 5120
1802
+ },
1803
+ {
1804
+ "epoch": 0.5850715688227427,
1805
+ "grad_norm": 0.5024551153182983,
1806
+ "learning_rate": 0.0002,
1807
+ "loss": 0.2717,
1808
+ "step": 5140
1809
+ },
1810
+ {
1811
+ "epoch": 0.5873481118920919,
1812
+ "grad_norm": 0.4166383147239685,
1813
+ "learning_rate": 0.0002,
1814
+ "loss": 0.3065,
1815
+ "step": 5160
1816
+ },
1817
+ {
1818
+ "epoch": 0.589624654961441,
1819
+ "grad_norm": 0.36780408024787903,
1820
+ "learning_rate": 0.0002,
1821
+ "loss": 0.2864,
1822
+ "step": 5180
1823
+ },
1824
+ {
1825
+ "epoch": 0.5919011980307902,
1826
+ "grad_norm": 0.436526894569397,
1827
+ "learning_rate": 0.0002,
1828
+ "loss": 0.2764,
1829
+ "step": 5200
1830
+ },
1831
+ {
1832
+ "epoch": 0.5941777411001394,
1833
+ "grad_norm": 0.43115249276161194,
1834
+ "learning_rate": 0.0002,
1835
+ "loss": 0.2791,
1836
+ "step": 5220
1837
+ },
1838
+ {
1839
+ "epoch": 0.5964542841694886,
1840
+ "grad_norm": 0.359739750623703,
1841
+ "learning_rate": 0.0002,
1842
+ "loss": 0.3108,
1843
+ "step": 5240
1844
+ },
1845
+ {
1846
+ "epoch": 0.5987308272388379,
1847
+ "grad_norm": 0.4555259644985199,
1848
+ "learning_rate": 0.0002,
1849
+ "loss": 0.2623,
1850
+ "step": 5260
1851
+ },
1852
+ {
1853
+ "epoch": 0.6010073703081871,
1854
+ "grad_norm": 0.4587076008319855,
1855
+ "learning_rate": 0.0002,
1856
+ "loss": 0.293,
1857
+ "step": 5280
1858
+ },
1859
+ {
1860
+ "epoch": 0.6032839133775362,
1861
+ "grad_norm": 0.5236973166465759,
1862
+ "learning_rate": 0.0002,
1863
+ "loss": 0.2888,
1864
+ "step": 5300
1865
+ },
1866
+ {
1867
+ "epoch": 0.6055604564468854,
1868
+ "grad_norm": 0.46685513854026794,
1869
+ "learning_rate": 0.0002,
1870
+ "loss": 0.2731,
1871
+ "step": 5320
1872
+ },
1873
+ {
1874
+ "epoch": 0.6078369995162346,
1875
+ "grad_norm": 0.5701884627342224,
1876
+ "learning_rate": 0.0002,
1877
+ "loss": 0.28,
1878
+ "step": 5340
1879
+ },
1880
+ {
1881
+ "epoch": 0.6101135425855838,
1882
+ "grad_norm": 0.5002717971801758,
1883
+ "learning_rate": 0.0002,
1884
+ "loss": 0.2777,
1885
+ "step": 5360
1886
+ },
1887
+ {
1888
+ "epoch": 0.612390085654933,
1889
+ "grad_norm": 0.5896885395050049,
1890
+ "learning_rate": 0.0002,
1891
+ "loss": 0.3048,
1892
+ "step": 5380
1893
+ },
1894
+ {
1895
+ "epoch": 0.6146666287242821,
1896
+ "grad_norm": 0.49014943838119507,
1897
+ "learning_rate": 0.0002,
1898
+ "loss": 0.2642,
1899
+ "step": 5400
1900
+ },
1901
+ {
1902
+ "epoch": 0.6169431717936313,
1903
+ "grad_norm": 0.5924846529960632,
1904
+ "learning_rate": 0.0002,
1905
+ "loss": 0.2943,
1906
+ "step": 5420
1907
+ },
1908
+ {
1909
+ "epoch": 0.6192197148629806,
1910
+ "grad_norm": 0.49827829003334045,
1911
+ "learning_rate": 0.0002,
1912
+ "loss": 0.2879,
1913
+ "step": 5440
1914
+ },
1915
+ {
1916
+ "epoch": 0.6214962579323298,
1917
+ "grad_norm": 0.45312178134918213,
1918
+ "learning_rate": 0.0002,
1919
+ "loss": 0.2728,
1920
+ "step": 5460
1921
+ },
1922
+ {
1923
+ "epoch": 0.623772801001679,
1924
+ "grad_norm": 0.3595191538333893,
1925
+ "learning_rate": 0.0002,
1926
+ "loss": 0.2713,
1927
+ "step": 5480
1928
+ },
1929
+ {
1930
+ "epoch": 0.6260493440710282,
1931
+ "grad_norm": 0.6547619104385376,
1932
+ "learning_rate": 0.0002,
1933
+ "loss": 0.2855,
1934
+ "step": 5500
1935
+ },
1936
+ {
1937
+ "epoch": 0.6283258871403773,
1938
+ "grad_norm": 0.4659534692764282,
1939
+ "learning_rate": 0.0002,
1940
+ "loss": 0.2908,
1941
+ "step": 5520
1942
+ },
1943
+ {
1944
+ "epoch": 0.6306024302097265,
1945
+ "grad_norm": 0.4027460813522339,
1946
+ "learning_rate": 0.0002,
1947
+ "loss": 0.2651,
1948
+ "step": 5540
1949
+ },
1950
+ {
1951
+ "epoch": 0.6328789732790757,
1952
+ "grad_norm": 0.36129653453826904,
1953
+ "learning_rate": 0.0002,
1954
+ "loss": 0.2915,
1955
+ "step": 5560
1956
+ },
1957
+ {
1958
+ "epoch": 0.6351555163484249,
1959
+ "grad_norm": 0.5963912010192871,
1960
+ "learning_rate": 0.0002,
1961
+ "loss": 0.2968,
1962
+ "step": 5580
1963
+ },
1964
+ {
1965
+ "epoch": 0.6374320594177741,
1966
+ "grad_norm": 0.49669450521469116,
1967
+ "learning_rate": 0.0002,
1968
+ "loss": 0.2965,
1969
+ "step": 5600
1970
+ },
1971
+ {
1972
+ "epoch": 0.6397086024871234,
1973
+ "grad_norm": 0.5784302353858948,
1974
+ "learning_rate": 0.0002,
1975
+ "loss": 0.2626,
1976
+ "step": 5620
1977
+ },
1978
+ {
1979
+ "epoch": 0.6419851455564725,
1980
+ "grad_norm": 0.5651645660400391,
1981
+ "learning_rate": 0.0002,
1982
+ "loss": 0.2738,
1983
+ "step": 5640
1984
+ },
1985
+ {
1986
+ "epoch": 0.6442616886258217,
1987
+ "grad_norm": 0.45475292205810547,
1988
+ "learning_rate": 0.0002,
1989
+ "loss": 0.2653,
1990
+ "step": 5660
1991
+ },
1992
+ {
1993
+ "epoch": 0.6465382316951709,
1994
+ "grad_norm": 0.4691898822784424,
1995
+ "learning_rate": 0.0002,
1996
+ "loss": 0.2634,
1997
+ "step": 5680
1998
+ },
1999
+ {
2000
+ "epoch": 0.6488147747645201,
2001
+ "grad_norm": 0.4604431092739105,
2002
+ "learning_rate": 0.0002,
2003
+ "loss": 0.2838,
2004
+ "step": 5700
2005
+ },
2006
+ {
2007
+ "epoch": 0.6510913178338693,
2008
+ "grad_norm": 0.506804883480072,
2009
+ "learning_rate": 0.0002,
2010
+ "loss": 0.2657,
2011
+ "step": 5720
2012
+ },
2013
+ {
2014
+ "epoch": 0.6533678609032184,
2015
+ "grad_norm": 0.5051881670951843,
2016
+ "learning_rate": 0.0002,
2017
+ "loss": 0.2976,
2018
+ "step": 5740
2019
+ },
2020
+ {
2021
+ "epoch": 0.6556444039725676,
2022
+ "grad_norm": 0.4780672788619995,
2023
+ "learning_rate": 0.0002,
2024
+ "loss": 0.2828,
2025
+ "step": 5760
2026
+ },
2027
+ {
2028
+ "epoch": 0.6579209470419168,
2029
+ "grad_norm": 0.4695095121860504,
2030
+ "learning_rate": 0.0002,
2031
+ "loss": 0.2685,
2032
+ "step": 5780
2033
+ },
2034
+ {
2035
+ "epoch": 0.660197490111266,
2036
+ "grad_norm": 0.4259052276611328,
2037
+ "learning_rate": 0.0002,
2038
+ "loss": 0.2635,
2039
+ "step": 5800
2040
+ },
2041
+ {
2042
+ "epoch": 0.6624740331806153,
2043
+ "grad_norm": 0.5684182643890381,
2044
+ "learning_rate": 0.0002,
2045
+ "loss": 0.2879,
2046
+ "step": 5820
2047
+ },
2048
+ {
2049
+ "epoch": 0.6647505762499645,
2050
+ "grad_norm": 0.42193594574928284,
2051
+ "learning_rate": 0.0002,
2052
+ "loss": 0.2678,
2053
+ "step": 5840
2054
+ },
2055
+ {
2056
+ "epoch": 0.6670271193193136,
2057
+ "grad_norm": 0.5095034241676331,
2058
+ "learning_rate": 0.0002,
2059
+ "loss": 0.2677,
2060
+ "step": 5860
2061
+ },
2062
+ {
2063
+ "epoch": 0.6693036623886628,
2064
+ "grad_norm": 0.46626052260398865,
2065
+ "learning_rate": 0.0002,
2066
+ "loss": 0.2906,
2067
+ "step": 5880
2068
+ },
2069
+ {
2070
+ "epoch": 0.671580205458012,
2071
+ "grad_norm": 0.5086765289306641,
2072
+ "learning_rate": 0.0002,
2073
+ "loss": 0.2775,
2074
+ "step": 5900
2075
+ },
2076
+ {
2077
+ "epoch": 0.6738567485273612,
2078
+ "grad_norm": 0.44444966316223145,
2079
+ "learning_rate": 0.0002,
2080
+ "loss": 0.2764,
2081
+ "step": 5920
2082
+ },
2083
+ {
2084
+ "epoch": 0.6761332915967104,
2085
+ "grad_norm": 0.4477381706237793,
2086
+ "learning_rate": 0.0002,
2087
+ "loss": 0.2729,
2088
+ "step": 5940
2089
+ },
2090
+ {
2091
+ "epoch": 0.6784098346660596,
2092
+ "grad_norm": 0.46984028816223145,
2093
+ "learning_rate": 0.0002,
2094
+ "loss": 0.273,
2095
+ "step": 5960
2096
+ },
2097
+ {
2098
+ "epoch": 0.6806863777354087,
2099
+ "grad_norm": 0.417084276676178,
2100
+ "learning_rate": 0.0002,
2101
+ "loss": 0.2744,
2102
+ "step": 5980
2103
+ },
2104
+ {
2105
+ "epoch": 0.682962920804758,
2106
+ "grad_norm": 0.4144213795661926,
2107
+ "learning_rate": 0.0002,
2108
+ "loss": 0.2704,
2109
+ "step": 6000
2110
+ },
2111
+ {
2112
+ "epoch": 0.6852394638741072,
2113
+ "grad_norm": 0.5844799876213074,
2114
+ "learning_rate": 0.0002,
2115
+ "loss": 0.2635,
2116
+ "step": 6020
2117
+ },
2118
+ {
2119
+ "epoch": 0.6875160069434564,
2120
+ "grad_norm": 0.39512693881988525,
2121
+ "learning_rate": 0.0002,
2122
+ "loss": 0.2471,
2123
+ "step": 6040
2124
+ },
2125
+ {
2126
+ "epoch": 0.6897925500128056,
2127
+ "grad_norm": 0.5299990773200989,
2128
+ "learning_rate": 0.0002,
2129
+ "loss": 0.2648,
2130
+ "step": 6060
2131
+ },
2132
+ {
2133
+ "epoch": 0.6920690930821547,
2134
+ "grad_norm": 0.4980265498161316,
2135
+ "learning_rate": 0.0002,
2136
+ "loss": 0.2725,
2137
+ "step": 6080
2138
+ },
2139
+ {
2140
+ "epoch": 0.6943456361515039,
2141
+ "grad_norm": 0.4003869891166687,
2142
+ "learning_rate": 0.0002,
2143
+ "loss": 0.2768,
2144
+ "step": 6100
2145
+ },
2146
+ {
2147
+ "epoch": 0.6966221792208531,
2148
+ "grad_norm": 0.5103460550308228,
2149
+ "learning_rate": 0.0002,
2150
+ "loss": 0.2638,
2151
+ "step": 6120
2152
+ },
2153
+ {
2154
+ "epoch": 0.6988987222902023,
2155
+ "grad_norm": 0.737101137638092,
2156
+ "learning_rate": 0.0002,
2157
+ "loss": 0.2779,
2158
+ "step": 6140
2159
+ },
2160
+ {
2161
+ "epoch": 0.7011752653595515,
2162
+ "grad_norm": 0.4731826186180115,
2163
+ "learning_rate": 0.0002,
2164
+ "loss": 0.2691,
2165
+ "step": 6160
2166
+ },
2167
+ {
2168
+ "epoch": 0.7034518084289008,
2169
+ "grad_norm": 0.5234053730964661,
2170
+ "learning_rate": 0.0002,
2171
+ "loss": 0.2739,
2172
+ "step": 6180
2173
+ },
2174
+ {
2175
+ "epoch": 0.7057283514982499,
2176
+ "grad_norm": 0.5235525369644165,
2177
+ "learning_rate": 0.0002,
2178
+ "loss": 0.2754,
2179
+ "step": 6200
2180
+ },
2181
+ {
2182
+ "epoch": 0.7080048945675991,
2183
+ "grad_norm": 0.4453619122505188,
2184
+ "learning_rate": 0.0002,
2185
+ "loss": 0.2833,
2186
+ "step": 6220
2187
+ },
2188
+ {
2189
+ "epoch": 0.7102814376369483,
2190
+ "grad_norm": 0.4025666117668152,
2191
+ "learning_rate": 0.0002,
2192
+ "loss": 0.2713,
2193
+ "step": 6240
2194
+ },
2195
+ {
2196
+ "epoch": 0.7125579807062975,
2197
+ "grad_norm": 0.35240331292152405,
2198
+ "learning_rate": 0.0002,
2199
+ "loss": 0.2786,
2200
+ "step": 6260
2201
+ },
2202
+ {
2203
+ "epoch": 0.7148345237756467,
2204
+ "grad_norm": 0.4521905779838562,
2205
+ "learning_rate": 0.0002,
2206
+ "loss": 0.2639,
2207
+ "step": 6280
2208
+ },
2209
+ {
2210
+ "epoch": 0.7171110668449959,
2211
+ "grad_norm": 0.5230519771575928,
2212
+ "learning_rate": 0.0002,
2213
+ "loss": 0.2517,
2214
+ "step": 6300
2215
+ },
2216
+ {
2217
+ "epoch": 0.719387609914345,
2218
+ "grad_norm": 0.5415637493133545,
2219
+ "learning_rate": 0.0002,
2220
+ "loss": 0.2739,
2221
+ "step": 6320
2222
+ },
2223
+ {
2224
+ "epoch": 0.7216641529836942,
2225
+ "grad_norm": 0.4067966341972351,
2226
+ "learning_rate": 0.0002,
2227
+ "loss": 0.2751,
2228
+ "step": 6340
2229
+ },
2230
+ {
2231
+ "epoch": 0.7239406960530435,
2232
+ "grad_norm": 0.4670214354991913,
2233
+ "learning_rate": 0.0002,
2234
+ "loss": 0.2644,
2235
+ "step": 6360
2236
+ },
2237
+ {
2238
+ "epoch": 0.7262172391223927,
2239
+ "grad_norm": 0.5316203236579895,
2240
+ "learning_rate": 0.0002,
2241
+ "loss": 0.2746,
2242
+ "step": 6380
2243
+ },
2244
+ {
2245
+ "epoch": 0.7284937821917419,
2246
+ "grad_norm": 0.46312493085861206,
2247
+ "learning_rate": 0.0002,
2248
+ "loss": 0.2539,
2249
+ "step": 6400
2250
+ },
2251
+ {
2252
+ "epoch": 0.730770325261091,
2253
+ "grad_norm": 0.465279221534729,
2254
+ "learning_rate": 0.0002,
2255
+ "loss": 0.2742,
2256
+ "step": 6420
2257
+ },
2258
+ {
2259
+ "epoch": 0.7330468683304402,
2260
+ "grad_norm": 0.5096962451934814,
2261
+ "learning_rate": 0.0002,
2262
+ "loss": 0.2546,
2263
+ "step": 6440
2264
+ },
2265
+ {
2266
+ "epoch": 0.7353234113997894,
2267
+ "grad_norm": 0.4525590240955353,
2268
+ "learning_rate": 0.0002,
2269
+ "loss": 0.2694,
2270
+ "step": 6460
2271
+ },
2272
+ {
2273
+ "epoch": 0.7375999544691386,
2274
+ "grad_norm": 0.5033881664276123,
2275
+ "learning_rate": 0.0002,
2276
+ "loss": 0.2627,
2277
+ "step": 6480
2278
+ },
2279
+ {
2280
+ "epoch": 0.7398764975384878,
2281
+ "grad_norm": 0.44053900241851807,
2282
+ "learning_rate": 0.0002,
2283
+ "loss": 0.258,
2284
+ "step": 6500
2285
+ },
2286
+ {
2287
+ "epoch": 0.742153040607837,
2288
+ "grad_norm": 0.4677462875843048,
2289
+ "learning_rate": 0.0002,
2290
+ "loss": 0.2659,
2291
+ "step": 6520
2292
+ },
2293
+ {
2294
+ "epoch": 0.7444295836771861,
2295
+ "grad_norm": 0.5687553882598877,
2296
+ "learning_rate": 0.0002,
2297
+ "loss": 0.271,
2298
+ "step": 6540
2299
+ },
2300
+ {
2301
+ "epoch": 0.7467061267465354,
2302
+ "grad_norm": 0.4980468451976776,
2303
+ "learning_rate": 0.0002,
2304
+ "loss": 0.265,
2305
+ "step": 6560
2306
+ },
2307
+ {
2308
+ "epoch": 0.7489826698158846,
2309
+ "grad_norm": 0.5155619382858276,
2310
+ "learning_rate": 0.0002,
2311
+ "loss": 0.2491,
2312
+ "step": 6580
2313
+ },
2314
+ {
2315
+ "epoch": 0.7512592128852338,
2316
+ "grad_norm": 0.5364673733711243,
2317
+ "learning_rate": 0.0002,
2318
+ "loss": 0.2564,
2319
+ "step": 6600
2320
+ },
2321
+ {
2322
+ "epoch": 0.753535755954583,
2323
+ "grad_norm": 0.421838641166687,
2324
+ "learning_rate": 0.0002,
2325
+ "loss": 0.267,
2326
+ "step": 6620
2327
+ },
2328
+ {
2329
+ "epoch": 0.7558122990239322,
2330
+ "grad_norm": 0.46299833059310913,
2331
+ "learning_rate": 0.0002,
2332
+ "loss": 0.2461,
2333
+ "step": 6640
2334
+ },
2335
+ {
2336
+ "epoch": 0.7580888420932813,
2337
+ "grad_norm": 0.3832832872867584,
2338
+ "learning_rate": 0.0002,
2339
+ "loss": 0.265,
2340
+ "step": 6660
2341
+ },
2342
+ {
2343
+ "epoch": 0.7603653851626305,
2344
+ "grad_norm": 0.5560947060585022,
2345
+ "learning_rate": 0.0002,
2346
+ "loss": 0.253,
2347
+ "step": 6680
2348
+ },
2349
+ {
2350
+ "epoch": 0.7626419282319797,
2351
+ "grad_norm": 0.4832628667354584,
2352
+ "learning_rate": 0.0002,
2353
+ "loss": 0.2515,
2354
+ "step": 6700
2355
+ },
2356
+ {
2357
+ "epoch": 0.764918471301329,
2358
+ "grad_norm": 0.44354599714279175,
2359
+ "learning_rate": 0.0002,
2360
+ "loss": 0.2687,
2361
+ "step": 6720
2362
+ },
2363
+ {
2364
+ "epoch": 0.7671950143706782,
2365
+ "grad_norm": 0.3746070861816406,
2366
+ "learning_rate": 0.0002,
2367
+ "loss": 0.2481,
2368
+ "step": 6740
2369
+ },
2370
+ {
2371
+ "epoch": 0.7694715574400273,
2372
+ "grad_norm": 0.3048388659954071,
2373
+ "learning_rate": 0.0002,
2374
+ "loss": 0.269,
2375
+ "step": 6760
2376
+ },
2377
+ {
2378
+ "epoch": 0.7717481005093765,
2379
+ "grad_norm": 0.46471843123435974,
2380
+ "learning_rate": 0.0002,
2381
+ "loss": 0.2642,
2382
+ "step": 6780
2383
+ },
2384
+ {
2385
+ "epoch": 0.7740246435787257,
2386
+ "grad_norm": 0.44309428334236145,
2387
+ "learning_rate": 0.0002,
2388
+ "loss": 0.2565,
2389
+ "step": 6800
2390
+ },
2391
+ {
2392
+ "epoch": 0.7763011866480749,
2393
+ "grad_norm": 0.4174291789531708,
2394
+ "learning_rate": 0.0002,
2395
+ "loss": 0.262,
2396
+ "step": 6820
2397
+ },
2398
+ {
2399
+ "epoch": 0.7785777297174241,
2400
+ "grad_norm": 0.42592549324035645,
2401
+ "learning_rate": 0.0002,
2402
+ "loss": 0.2608,
2403
+ "step": 6840
2404
+ },
2405
+ {
2406
+ "epoch": 0.7808542727867733,
2407
+ "grad_norm": 0.4378054141998291,
2408
+ "learning_rate": 0.0002,
2409
+ "loss": 0.2765,
2410
+ "step": 6860
2411
+ },
2412
+ {
2413
+ "epoch": 0.7831308158561224,
2414
+ "grad_norm": 0.4560708701610565,
2415
+ "learning_rate": 0.0002,
2416
+ "loss": 0.2381,
2417
+ "step": 6880
2418
+ },
2419
+ {
2420
+ "epoch": 0.7854073589254716,
2421
+ "grad_norm": 0.4595545828342438,
2422
+ "learning_rate": 0.0002,
2423
+ "loss": 0.2561,
2424
+ "step": 6900
2425
+ },
2426
+ {
2427
+ "epoch": 0.7876839019948209,
2428
+ "grad_norm": 0.45213592052459717,
2429
+ "learning_rate": 0.0002,
2430
+ "loss": 0.2645,
2431
+ "step": 6920
2432
+ },
2433
+ {
2434
+ "epoch": 0.7899604450641701,
2435
+ "grad_norm": 0.4857342839241028,
2436
+ "learning_rate": 0.0002,
2437
+ "loss": 0.2687,
2438
+ "step": 6940
2439
+ },
2440
+ {
2441
+ "epoch": 0.7922369881335193,
2442
+ "grad_norm": 0.4939437508583069,
2443
+ "learning_rate": 0.0002,
2444
+ "loss": 0.2642,
2445
+ "step": 6960
2446
+ },
2447
+ {
2448
+ "epoch": 0.7945135312028685,
2449
+ "grad_norm": 0.46244382858276367,
2450
+ "learning_rate": 0.0002,
2451
+ "loss": 0.2536,
2452
+ "step": 6980
2453
+ },
2454
+ {
2455
+ "epoch": 0.7967900742722176,
2456
+ "grad_norm": 0.5876993536949158,
2457
+ "learning_rate": 0.0002,
2458
+ "loss": 0.2492,
2459
+ "step": 7000
2460
+ },
2461
+ {
2462
+ "epoch": 0.7990666173415668,
2463
+ "grad_norm": 0.5170072913169861,
2464
+ "learning_rate": 0.0002,
2465
+ "loss": 0.2548,
2466
+ "step": 7020
2467
+ },
2468
+ {
2469
+ "epoch": 0.801343160410916,
2470
+ "grad_norm": 0.394380658864975,
2471
+ "learning_rate": 0.0002,
2472
+ "loss": 0.2524,
2473
+ "step": 7040
2474
+ },
2475
+ {
2476
+ "epoch": 0.8036197034802652,
2477
+ "grad_norm": 0.4716455340385437,
2478
+ "learning_rate": 0.0002,
2479
+ "loss": 0.2573,
2480
+ "step": 7060
2481
+ },
2482
+ {
2483
+ "epoch": 0.8058962465496144,
2484
+ "grad_norm": 0.34525179862976074,
2485
+ "learning_rate": 0.0002,
2486
+ "loss": 0.246,
2487
+ "step": 7080
2488
+ },
2489
+ {
2490
+ "epoch": 0.8081727896189635,
2491
+ "grad_norm": 0.5030418038368225,
2492
+ "learning_rate": 0.0002,
2493
+ "loss": 0.2596,
2494
+ "step": 7100
2495
+ },
2496
+ {
2497
+ "epoch": 0.8104493326883128,
2498
+ "grad_norm": 0.5586132407188416,
2499
+ "learning_rate": 0.0002,
2500
+ "loss": 0.2568,
2501
+ "step": 7120
2502
+ },
2503
+ {
2504
+ "epoch": 0.812725875757662,
2505
+ "grad_norm": 0.47025129199028015,
2506
+ "learning_rate": 0.0002,
2507
+ "loss": 0.265,
2508
+ "step": 7140
2509
+ },
2510
+ {
2511
+ "epoch": 0.8150024188270112,
2512
+ "grad_norm": 0.5654832720756531,
2513
+ "learning_rate": 0.0002,
2514
+ "loss": 0.2468,
2515
+ "step": 7160
2516
+ },
2517
+ {
2518
+ "epoch": 0.8172789618963604,
2519
+ "grad_norm": 0.4701017141342163,
2520
+ "learning_rate": 0.0002,
2521
+ "loss": 0.2538,
2522
+ "step": 7180
2523
+ },
2524
+ {
2525
+ "epoch": 0.8195555049657096,
2526
+ "grad_norm": 0.47270438075065613,
2527
+ "learning_rate": 0.0002,
2528
+ "loss": 0.2529,
2529
+ "step": 7200
2530
+ },
2531
+ {
2532
+ "epoch": 0.8218320480350587,
2533
+ "grad_norm": 0.39433714747428894,
2534
+ "learning_rate": 0.0002,
2535
+ "loss": 0.2445,
2536
+ "step": 7220
2537
+ },
2538
+ {
2539
+ "epoch": 0.8241085911044079,
2540
+ "grad_norm": 0.4521467685699463,
2541
+ "learning_rate": 0.0002,
2542
+ "loss": 0.2556,
2543
+ "step": 7240
2544
+ },
2545
+ {
2546
+ "epoch": 0.8263851341737571,
2547
+ "grad_norm": 0.28483667969703674,
2548
+ "learning_rate": 0.0002,
2549
+ "loss": 0.2451,
2550
+ "step": 7260
2551
+ },
2552
+ {
2553
+ "epoch": 0.8286616772431064,
2554
+ "grad_norm": 0.4298310875892639,
2555
+ "learning_rate": 0.0002,
2556
+ "loss": 0.2599,
2557
+ "step": 7280
2558
+ },
2559
+ {
2560
+ "epoch": 0.8309382203124556,
2561
+ "grad_norm": 0.39677906036376953,
2562
+ "learning_rate": 0.0002,
2563
+ "loss": 0.2539,
2564
+ "step": 7300
2565
+ },
2566
+ {
2567
+ "epoch": 0.8332147633818048,
2568
+ "grad_norm": 0.5800175666809082,
2569
+ "learning_rate": 0.0002,
2570
+ "loss": 0.2463,
2571
+ "step": 7320
2572
+ },
2573
+ {
2574
+ "epoch": 0.8354913064511539,
2575
+ "grad_norm": 0.42742472887039185,
2576
+ "learning_rate": 0.0002,
2577
+ "loss": 0.2593,
2578
+ "step": 7340
2579
+ },
2580
+ {
2581
+ "epoch": 0.8377678495205031,
2582
+ "grad_norm": 0.5521807670593262,
2583
+ "learning_rate": 0.0002,
2584
+ "loss": 0.253,
2585
+ "step": 7360
2586
+ },
2587
+ {
2588
+ "epoch": 0.8400443925898523,
2589
+ "grad_norm": 0.5068047046661377,
2590
+ "learning_rate": 0.0002,
2591
+ "loss": 0.2503,
2592
+ "step": 7380
2593
+ },
2594
+ {
2595
+ "epoch": 0.8423209356592015,
2596
+ "grad_norm": 0.4325120151042938,
2597
+ "learning_rate": 0.0002,
2598
+ "loss": 0.2466,
2599
+ "step": 7400
2600
+ },
2601
+ {
2602
+ "epoch": 0.8445974787285507,
2603
+ "grad_norm": 0.5130394101142883,
2604
+ "learning_rate": 0.0002,
2605
+ "loss": 0.2521,
2606
+ "step": 7420
2607
+ },
2608
+ {
2609
+ "epoch": 0.8468740217978998,
2610
+ "grad_norm": 0.5091120600700378,
2611
+ "learning_rate": 0.0002,
2612
+ "loss": 0.2429,
2613
+ "step": 7440
2614
+ },
2615
+ {
2616
+ "epoch": 0.849150564867249,
2617
+ "grad_norm": 0.4635036289691925,
2618
+ "learning_rate": 0.0002,
2619
+ "loss": 0.235,
2620
+ "step": 7460
2621
+ },
2622
+ {
2623
+ "epoch": 0.8514271079365983,
2624
+ "grad_norm": 0.3827108144760132,
2625
+ "learning_rate": 0.0002,
2626
+ "loss": 0.2487,
2627
+ "step": 7480
2628
+ },
2629
+ {
2630
+ "epoch": 0.8537036510059475,
2631
+ "grad_norm": 0.3880899250507355,
2632
+ "learning_rate": 0.0002,
2633
+ "loss": 0.2469,
2634
+ "step": 7500
2635
+ },
2636
+ {
2637
+ "epoch": 0.8559801940752967,
2638
+ "grad_norm": 0.408933162689209,
2639
+ "learning_rate": 0.0002,
2640
+ "loss": 0.2499,
2641
+ "step": 7520
2642
+ },
2643
+ {
2644
+ "epoch": 0.8582567371446459,
2645
+ "grad_norm": 0.5049706101417542,
2646
+ "learning_rate": 0.0002,
2647
+ "loss": 0.2418,
2648
+ "step": 7540
2649
+ },
2650
+ {
2651
+ "epoch": 0.860533280213995,
2652
+ "grad_norm": 0.43551701307296753,
2653
+ "learning_rate": 0.0002,
2654
+ "loss": 0.2478,
2655
+ "step": 7560
2656
+ },
2657
+ {
2658
+ "epoch": 0.8628098232833442,
2659
+ "grad_norm": 0.5024411678314209,
2660
+ "learning_rate": 0.0002,
2661
+ "loss": 0.2538,
2662
+ "step": 7580
2663
+ },
2664
+ {
2665
+ "epoch": 0.8650863663526934,
2666
+ "grad_norm": 0.36361223459243774,
2667
+ "learning_rate": 0.0002,
2668
+ "loss": 0.2536,
2669
+ "step": 7600
2670
+ },
2671
+ {
2672
+ "epoch": 0.8673629094220426,
2673
+ "grad_norm": 0.4526277482509613,
2674
+ "learning_rate": 0.0002,
2675
+ "loss": 0.242,
2676
+ "step": 7620
2677
+ },
2678
+ {
2679
+ "epoch": 0.8696394524913919,
2680
+ "grad_norm": 0.5677676200866699,
2681
+ "learning_rate": 0.0002,
2682
+ "loss": 0.2572,
2683
+ "step": 7640
2684
+ },
2685
+ {
2686
+ "epoch": 0.8719159955607411,
2687
+ "grad_norm": 0.4915711283683777,
2688
+ "learning_rate": 0.0002,
2689
+ "loss": 0.2562,
2690
+ "step": 7660
2691
+ },
2692
+ {
2693
+ "epoch": 0.8741925386300902,
2694
+ "grad_norm": 0.36850452423095703,
2695
+ "learning_rate": 0.0002,
2696
+ "loss": 0.2523,
2697
+ "step": 7680
2698
+ },
2699
+ {
2700
+ "epoch": 0.8764690816994394,
2701
+ "grad_norm": 0.38313761353492737,
2702
+ "learning_rate": 0.0002,
2703
+ "loss": 0.2596,
2704
+ "step": 7700
2705
+ },
2706
+ {
2707
+ "epoch": 0.8787456247687886,
2708
+ "grad_norm": 0.5384640097618103,
2709
+ "learning_rate": 0.0002,
2710
+ "loss": 0.2455,
2711
+ "step": 7720
2712
+ },
2713
+ {
2714
+ "epoch": 0.8810221678381378,
2715
+ "grad_norm": 0.5308900475502014,
2716
+ "learning_rate": 0.0002,
2717
+ "loss": 0.2439,
2718
+ "step": 7740
2719
+ },
2720
+ {
2721
+ "epoch": 0.883298710907487,
2722
+ "grad_norm": 0.5488154292106628,
2723
+ "learning_rate": 0.0002,
2724
+ "loss": 0.2428,
2725
+ "step": 7760
2726
+ },
2727
+ {
2728
+ "epoch": 0.8855752539768362,
2729
+ "grad_norm": 0.5271242260932922,
2730
+ "learning_rate": 0.0002,
2731
+ "loss": 0.2372,
2732
+ "step": 7780
2733
+ },
2734
+ {
2735
+ "epoch": 0.8878517970461853,
2736
+ "grad_norm": 0.46171802282333374,
2737
+ "learning_rate": 0.0002,
2738
+ "loss": 0.2506,
2739
+ "step": 7800
2740
+ },
2741
+ {
2742
+ "epoch": 0.8901283401155345,
2743
+ "grad_norm": 0.45436665415763855,
2744
+ "learning_rate": 0.0002,
2745
+ "loss": 0.2414,
2746
+ "step": 7820
2747
+ },
2748
+ {
2749
+ "epoch": 0.8924048831848838,
2750
+ "grad_norm": 0.4920847415924072,
2751
+ "learning_rate": 0.0002,
2752
+ "loss": 0.2669,
2753
+ "step": 7840
2754
+ },
2755
+ {
2756
+ "epoch": 0.894681426254233,
2757
+ "grad_norm": 0.5913518071174622,
2758
+ "learning_rate": 0.0002,
2759
+ "loss": 0.2552,
2760
+ "step": 7860
2761
+ },
2762
+ {
2763
+ "epoch": 0.8969579693235822,
2764
+ "grad_norm": 0.6011972427368164,
2765
+ "learning_rate": 0.0002,
2766
+ "loss": 0.2533,
2767
+ "step": 7880
2768
+ },
2769
+ {
2770
+ "epoch": 0.8992345123929313,
2771
+ "grad_norm": 0.4650927186012268,
2772
+ "learning_rate": 0.0002,
2773
+ "loss": 0.2448,
2774
+ "step": 7900
2775
+ },
2776
+ {
2777
+ "epoch": 0.9015110554622805,
2778
+ "grad_norm": 0.5828790664672852,
2779
+ "learning_rate": 0.0002,
2780
+ "loss": 0.2381,
2781
+ "step": 7920
2782
+ },
2783
+ {
2784
+ "epoch": 0.9037875985316297,
2785
+ "grad_norm": 0.5178338885307312,
2786
+ "learning_rate": 0.0002,
2787
+ "loss": 0.2619,
2788
+ "step": 7940
2789
+ },
2790
+ {
2791
+ "epoch": 0.9060641416009789,
2792
+ "grad_norm": 0.5147708058357239,
2793
+ "learning_rate": 0.0002,
2794
+ "loss": 0.258,
2795
+ "step": 7960
2796
+ },
2797
+ {
2798
+ "epoch": 0.9083406846703281,
2799
+ "grad_norm": 0.45790836215019226,
2800
+ "learning_rate": 0.0002,
2801
+ "loss": 0.2474,
2802
+ "step": 7980
2803
+ },
2804
+ {
2805
+ "epoch": 0.9106172277396773,
2806
+ "grad_norm": 0.3837074935436249,
2807
+ "learning_rate": 0.0002,
2808
+ "loss": 0.2356,
2809
+ "step": 8000
2810
+ },
2811
+ {
2812
+ "epoch": 0.9128937708090265,
2813
+ "grad_norm": 0.4466090500354767,
2814
+ "learning_rate": 0.0002,
2815
+ "loss": 0.237,
2816
+ "step": 8020
2817
+ },
2818
+ {
2819
+ "epoch": 0.9151703138783757,
2820
+ "grad_norm": 0.5893344283103943,
2821
+ "learning_rate": 0.0002,
2822
+ "loss": 0.2399,
2823
+ "step": 8040
2824
+ },
2825
+ {
2826
+ "epoch": 0.9174468569477249,
2827
+ "grad_norm": 0.49547362327575684,
2828
+ "learning_rate": 0.0002,
2829
+ "loss": 0.2526,
2830
+ "step": 8060
2831
+ },
2832
+ {
2833
+ "epoch": 0.9197234000170741,
2834
+ "grad_norm": 0.47068551182746887,
2835
+ "learning_rate": 0.0002,
2836
+ "loss": 0.2631,
2837
+ "step": 8080
2838
+ },
2839
+ {
2840
+ "epoch": 0.9219999430864233,
2841
+ "grad_norm": 0.3512951135635376,
2842
+ "learning_rate": 0.0002,
2843
+ "loss": 0.2395,
2844
+ "step": 8100
2845
+ },
2846
+ {
2847
+ "epoch": 0.9242764861557725,
2848
+ "grad_norm": 0.3996793031692505,
2849
+ "learning_rate": 0.0002,
2850
+ "loss": 0.2424,
2851
+ "step": 8120
2852
+ },
2853
+ {
2854
+ "epoch": 0.9265530292251216,
2855
+ "grad_norm": 0.5782022476196289,
2856
+ "learning_rate": 0.0002,
2857
+ "loss": 0.2549,
2858
+ "step": 8140
2859
+ },
2860
+ {
2861
+ "epoch": 0.9288295722944708,
2862
+ "grad_norm": 0.450860857963562,
2863
+ "learning_rate": 0.0002,
2864
+ "loss": 0.2465,
2865
+ "step": 8160
2866
+ },
2867
+ {
2868
+ "epoch": 0.93110611536382,
2869
+ "grad_norm": 0.4679816663265228,
2870
+ "learning_rate": 0.0002,
2871
+ "loss": 0.2326,
2872
+ "step": 8180
2873
+ },
2874
+ {
2875
+ "epoch": 0.9333826584331693,
2876
+ "grad_norm": 0.5497337579727173,
2877
+ "learning_rate": 0.0002,
2878
+ "loss": 0.2457,
2879
+ "step": 8200
2880
+ },
2881
+ {
2882
+ "epoch": 0.9356592015025185,
2883
+ "grad_norm": 0.3775748312473297,
2884
+ "learning_rate": 0.0002,
2885
+ "loss": 0.2331,
2886
+ "step": 8220
2887
+ },
2888
+ {
2889
+ "epoch": 0.9379357445718676,
2890
+ "grad_norm": 0.5428327918052673,
2891
+ "learning_rate": 0.0002,
2892
+ "loss": 0.2399,
2893
+ "step": 8240
2894
+ },
2895
+ {
2896
+ "epoch": 0.9402122876412168,
2897
+ "grad_norm": 0.4089830219745636,
2898
+ "learning_rate": 0.0002,
2899
+ "loss": 0.246,
2900
+ "step": 8260
2901
+ },
2902
+ {
2903
+ "epoch": 0.942488830710566,
2904
+ "grad_norm": 0.5781340003013611,
2905
+ "learning_rate": 0.0002,
2906
+ "loss": 0.2451,
2907
+ "step": 8280
2908
+ },
2909
+ {
2910
+ "epoch": 0.9447653737799152,
2911
+ "grad_norm": 0.5869989395141602,
2912
+ "learning_rate": 0.0002,
2913
+ "loss": 0.2541,
2914
+ "step": 8300
2915
+ },
2916
+ {
2917
+ "epoch": 0.9470419168492644,
2918
+ "grad_norm": 0.47708019614219666,
2919
+ "learning_rate": 0.0002,
2920
+ "loss": 0.2559,
2921
+ "step": 8320
2922
+ },
2923
+ {
2924
+ "epoch": 0.9493184599186136,
2925
+ "grad_norm": 0.5445525050163269,
2926
+ "learning_rate": 0.0002,
2927
+ "loss": 0.2466,
2928
+ "step": 8340
2929
+ },
2930
+ {
2931
+ "epoch": 0.9515950029879627,
2932
+ "grad_norm": 0.480214387178421,
2933
+ "learning_rate": 0.0002,
2934
+ "loss": 0.236,
2935
+ "step": 8360
2936
+ },
2937
+ {
2938
+ "epoch": 0.953871546057312,
2939
+ "grad_norm": 0.5392053127288818,
2940
+ "learning_rate": 0.0002,
2941
+ "loss": 0.2383,
2942
+ "step": 8380
2943
+ },
2944
+ {
2945
+ "epoch": 0.9561480891266612,
2946
+ "grad_norm": 0.4515858292579651,
2947
+ "learning_rate": 0.0002,
2948
+ "loss": 0.238,
2949
+ "step": 8400
2950
+ },
2951
+ {
2952
+ "epoch": 0.9584246321960104,
2953
+ "grad_norm": 0.5461826324462891,
2954
+ "learning_rate": 0.0002,
2955
+ "loss": 0.2442,
2956
+ "step": 8420
2957
+ },
2958
+ {
2959
+ "epoch": 0.9607011752653596,
2960
+ "grad_norm": 0.44309332966804504,
2961
+ "learning_rate": 0.0002,
2962
+ "loss": 0.2622,
2963
+ "step": 8440
2964
+ },
2965
+ {
2966
+ "epoch": 0.9629777183347088,
2967
+ "grad_norm": 0.5409505367279053,
2968
+ "learning_rate": 0.0002,
2969
+ "loss": 0.2303,
2970
+ "step": 8460
2971
+ },
2972
+ {
2973
+ "epoch": 0.9652542614040579,
2974
+ "grad_norm": 0.3868342638015747,
2975
+ "learning_rate": 0.0002,
2976
+ "loss": 0.2624,
2977
+ "step": 8480
2978
+ },
2979
+ {
2980
+ "epoch": 0.9675308044734071,
2981
+ "grad_norm": 0.38888975977897644,
2982
+ "learning_rate": 0.0002,
2983
+ "loss": 0.246,
2984
+ "step": 8500
2985
+ },
2986
+ {
2987
+ "epoch": 0.9698073475427563,
2988
+ "grad_norm": 0.38946032524108887,
2989
+ "learning_rate": 0.0002,
2990
+ "loss": 0.2503,
2991
+ "step": 8520
2992
+ },
2993
+ {
2994
+ "epoch": 0.9720838906121055,
2995
+ "grad_norm": 0.42425817251205444,
2996
+ "learning_rate": 0.0002,
2997
+ "loss": 0.2556,
2998
+ "step": 8540
2999
+ },
3000
+ {
3001
+ "epoch": 0.9743604336814548,
3002
+ "grad_norm": 0.41515296697616577,
3003
+ "learning_rate": 0.0002,
3004
+ "loss": 0.2437,
3005
+ "step": 8560
3006
+ },
3007
+ {
3008
+ "epoch": 0.9766369767508039,
3009
+ "grad_norm": 0.4085826575756073,
3010
+ "learning_rate": 0.0002,
3011
+ "loss": 0.2293,
3012
+ "step": 8580
3013
+ },
3014
+ {
3015
+ "epoch": 0.9789135198201531,
3016
+ "grad_norm": 0.3404542803764343,
3017
+ "learning_rate": 0.0002,
3018
+ "loss": 0.242,
3019
+ "step": 8600
3020
+ },
3021
+ {
3022
+ "epoch": 0.9811900628895023,
3023
+ "grad_norm": 0.43266579508781433,
3024
+ "learning_rate": 0.0002,
3025
+ "loss": 0.2513,
3026
+ "step": 8620
3027
+ },
3028
+ {
3029
+ "epoch": 0.9834666059588515,
3030
+ "grad_norm": 0.42724549770355225,
3031
+ "learning_rate": 0.0002,
3032
+ "loss": 0.2384,
3033
+ "step": 8640
3034
+ },
3035
+ {
3036
+ "epoch": 0.9857431490282007,
3037
+ "grad_norm": 0.5089221596717834,
3038
+ "learning_rate": 0.0002,
3039
+ "loss": 0.2409,
3040
+ "step": 8660
3041
+ },
3042
+ {
3043
+ "epoch": 0.9880196920975499,
3044
+ "grad_norm": 0.519223690032959,
3045
+ "learning_rate": 0.0002,
3046
+ "loss": 0.2353,
3047
+ "step": 8680
3048
+ },
3049
+ {
3050
+ "epoch": 0.990296235166899,
3051
+ "grad_norm": 0.5701056122779846,
3052
+ "learning_rate": 0.0002,
3053
+ "loss": 0.2486,
3054
+ "step": 8700
3055
+ },
3056
+ {
3057
+ "epoch": 0.9925727782362482,
3058
+ "grad_norm": 0.4519595503807068,
3059
+ "learning_rate": 0.0002,
3060
+ "loss": 0.2374,
3061
+ "step": 8720
3062
+ },
3063
+ {
3064
+ "epoch": 0.9948493213055974,
3065
+ "grad_norm": 0.4883946180343628,
3066
+ "learning_rate": 0.0002,
3067
+ "loss": 0.2441,
3068
+ "step": 8740
3069
+ },
3070
+ {
3071
+ "epoch": 0.9971258643749467,
3072
+ "grad_norm": 0.6918900012969971,
3073
+ "learning_rate": 0.0002,
3074
+ "loss": 0.2403,
3075
+ "step": 8760
3076
+ },
3077
+ {
3078
+ "epoch": 0.9994024074442959,
3079
+ "grad_norm": 0.4810091555118561,
3080
+ "learning_rate": 0.0002,
3081
+ "loss": 0.2334,
3082
+ "step": 8780
3083
+ },
3084
+ {
3085
+ "epoch": 1.0,
3086
+ "eval_loss": 0.30941203236579895,
3087
+ "eval_runtime": 408.7196,
3088
+ "eval_samples_per_second": 7.083,
3089
+ "eval_steps_per_second": 0.886,
3090
+ "step": 8786
3091
+ }
3092
+ ],
3093
+ "logging_steps": 20,
3094
+ "max_steps": 13000,
3095
+ "num_input_tokens_seen": 0,
3096
+ "num_train_epochs": 2,
3097
+ "save_steps": 77,
3098
+ "stateful_callbacks": {
3099
+ "TrainerControl": {
3100
+ "args": {
3101
+ "should_epoch_stop": false,
3102
+ "should_evaluate": false,
3103
+ "should_log": false,
3104
+ "should_save": true,
3105
+ "should_training_stop": false
3106
+ },
3107
+ "attributes": {}
3108
+ }
3109
+ },
3110
+ "total_flos": 2.923169198364426e+18,
3111
+ "train_batch_size": 2,
3112
+ "trial_name": null,
3113
+ "trial_params": null
3114
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e81583ff738f437f9b8ba61f8cd63306401c4b51ce22ce038811cf0a2a0f493e
3
+ size 5816