Hugging Face's logo

CodeIsAbstract
/
HybridTimeScaleModel-Specialist-0.1B

Safetensors
hybrid_fourier_lm
custom_code
Model card Files
xet
 Community
CodeIsAbstract commited on
Commit
c19bc03
·
verified ·
1 Parent(s): 197508c

Upload folder using huggingface_hub

Browse files
Files changed (8) hide show
  1. chat_template.jinja +1 -1
  2. model.safetensors +2 -2
  3. optimizer.pt +1 -1
  4. rng_state.pth +1 -1
  5. scheduler.pt +1 -1
  6. tokenizer.json +1 -6
  7. trainer_state.json +3906 -6
  8. training_args.bin +1 -1
chat_template.jinja CHANGED
@@ -1 +1 @@
1
- {{ bos_token }}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if message['role'] == 'user' %}{{ '[INST] ' + message['content'] + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ ' ' + message['content'] + eos_token }}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ ' ' }}{% endif %}
 
1
+ {{ bos_token }}{% for message in messages %}{% if message['role'] == 'system' %}{{ '[INST] System: ' + message['content'] + '\n\n' }}{% elif message['role'] == 'user' %}{% if loop.index0 == 0 or (loop.index0 == 1 and messages[0]['role'] == 'system') %}{{ message['content'] + ' [/INST]' }}{% else %}{{ '[INST] ' + message['content'] + ' [/INST]' }}{% endif %}{% elif message['role'] == 'assistant' %}{{ ' ' + message['content'] + eos_token }}{% else %}{{ raise_exception('Only system, user, and assistant roles are supported!') }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ ' ' }}{% endif %}
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2989225b566d339103b420ea4d3078d89b77f81c03349d3c7fbc4753166ec49d
3
- size 234681136
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:baa02591aab05ff9bad05fc7381edea799063962632ac48b1987d2eca172c192
3
+ size 234683688
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f3b73014300d86c2a2f52052dbdf1091e2258cb3d9789111c1779c6709dbfde3
3
  size 469516363
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b6a1df18722e12e76f3bd2db06d96d71a15a1dae7ca4076f9a398ec089e8f3cc
3
  size 469516363
rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f88e21bc7f577d6a08f1a19b17a9c2fed0f4dddc3a152748fb34da132e9f98ed
3
  size 14645
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e130043fc26edfb2a98fa587872aeafd67df05e350574fb32438728408f9490c
3
  size 14645
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:33b1731f6f172e685db8110e11ac90faa1e42e5311e23a5ac694b5d79a258fbe
3
  size 1465
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02a24dfa5a92512b727034864335e93049f5bf286f1eb5d10531ac4b92fdbcdb
3
  size 1465
tokenizer.json CHANGED
@@ -1,11 +1,6 @@
1
  {
2
  "version": "1.0",
3
- "truncation": {
4
- "direction": "Right",
5
- "max_length": 1024,
6
- "strategy": "LongestFirst",
7
- "stride": 0
8
- },
9
  "padding": null,
10
  "added_tokens": [
11
  {
 
1
  {
2
  "version": "1.0",
3
+ "truncation": null,
 
 
 
 
 
4
  "padding": null,
5
  "added_tokens": [
6
  {
trainer_state.json CHANGED
@@ -2,17 +2,3917 @@
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
- "epoch": 2.0,
6
  "eval_steps": 500,
7
- "global_step": 2,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
11
- "log_history": [],
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
12
  "logging_steps": 5,
13
- "max_steps": 2,
14
  "num_input_tokens_seen": 0,
15
- "num_train_epochs": 2,
16
  "save_steps": 500,
17
  "stateful_callbacks": {
18
  "TrainerControl": {
@@ -26,7 +3926,7 @@
26
  "attributes": {}
27
  }
28
  },
29
- "total_flos": 1734111461376.0,
30
  "train_batch_size": 16,
31
  "trial_name": null,
32
  "trial_params": null
 
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
+ "epoch": 3.0,
6
  "eval_steps": 500,
7
+ "global_step": 2787,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.005382131324004306,
14
+ "grad_norm": 19.25,
15
+ "learning_rate": 5.7142857142857145e-06,
16
+ "loss": 7.5826,
17
+ "step": 5
18
+ },
19
+ {
20
+ "epoch": 0.010764262648008612,
21
+ "grad_norm": 24.125,
22
+ "learning_rate": 1.2857142857142857e-05,
23
+ "loss": 7.7389,
24
+ "step": 10
25
+ },
26
+ {
27
+ "epoch": 0.016146393972012917,
28
+ "grad_norm": 20.375,
29
+ "learning_rate": 2e-05,
30
+ "loss": 8.0474,
31
+ "step": 15
32
+ },
33
+ {
34
+ "epoch": 0.021528525296017224,
35
+ "grad_norm": 17.375,
36
+ "learning_rate": 2.714285714285714e-05,
37
+ "loss": 7.2989,
38
+ "step": 20
39
+ },
40
+ {
41
+ "epoch": 0.02691065662002153,
42
+ "grad_norm": 15.0625,
43
+ "learning_rate": 3.428571428571429e-05,
44
+ "loss": 7.8579,
45
+ "step": 25
46
+ },
47
+ {
48
+ "epoch": 0.03229278794402583,
49
+ "grad_norm": 12.875,
50
+ "learning_rate": 4.1428571428571437e-05,
51
+ "loss": 7.3271,
52
+ "step": 30
53
+ },
54
+ {
55
+ "epoch": 0.03767491926803014,
56
+ "grad_norm": 8.5,
57
+ "learning_rate": 4.8571428571428576e-05,
58
+ "loss": 7.0945,
59
+ "step": 35
60
+ },
61
+ {
62
+ "epoch": 0.04305705059203445,
63
+ "grad_norm": 7.46875,
64
+ "learning_rate": 5.571428571428572e-05,
65
+ "loss": 6.3577,
66
+ "step": 40
67
+ },
68
+ {
69
+ "epoch": 0.04843918191603875,
70
+ "grad_norm": 7.40625,
71
+ "learning_rate": 6.285714285714286e-05,
72
+ "loss": 6.2656,
73
+ "step": 45
74
+ },
75
+ {
76
+ "epoch": 0.05382131324004306,
77
+ "grad_norm": 3.53125,
78
+ "learning_rate": 7e-05,
79
+ "loss": 6.2146,
80
+ "step": 50
81
+ },
82
+ {
83
+ "epoch": 0.059203444564047365,
84
+ "grad_norm": 3.671875,
85
+ "learning_rate": 7.714285714285715e-05,
86
+ "loss": 5.7147,
87
+ "step": 55
88
+ },
89
+ {
90
+ "epoch": 0.06458557588805167,
91
+ "grad_norm": 2.6875,
92
+ "learning_rate": 8.428571428571429e-05,
93
+ "loss": 5.2684,
94
+ "step": 60
95
+ },
96
+ {
97
+ "epoch": 0.06996770721205597,
98
+ "grad_norm": 2.203125,
99
+ "learning_rate": 9.142857142857143e-05,
100
+ "loss": 4.914,
101
+ "step": 65
102
+ },
103
+ {
104
+ "epoch": 0.07534983853606028,
105
+ "grad_norm": 2.28125,
106
+ "learning_rate": 9.857142857142858e-05,
107
+ "loss": 5.5474,
108
+ "step": 70
109
+ },
110
+ {
111
+ "epoch": 0.08073196986006459,
112
+ "grad_norm": 3.03125,
113
+ "learning_rate": 0.00010571428571428572,
114
+ "loss": 4.0901,
115
+ "step": 75
116
+ },
117
+ {
118
+ "epoch": 0.0861141011840689,
119
+ "grad_norm": 2.40625,
120
+ "learning_rate": 0.00011285714285714286,
121
+ "loss": 4.1259,
122
+ "step": 80
123
+ },
124
+ {
125
+ "epoch": 0.09149623250807319,
126
+ "grad_norm": 2.328125,
127
+ "learning_rate": 0.00012,
128
+ "loss": 3.7235,
129
+ "step": 85
130
+ },
131
+ {
132
+ "epoch": 0.0968783638320775,
133
+ "grad_norm": 2.46875,
134
+ "learning_rate": 0.00012714285714285714,
135
+ "loss": 3.1657,
136
+ "step": 90
137
+ },
138
+ {
139
+ "epoch": 0.10226049515608181,
140
+ "grad_norm": 3.171875,
141
+ "learning_rate": 0.00013428571428571428,
142
+ "loss": 3.4388,
143
+ "step": 95
144
+ },
145
+ {
146
+ "epoch": 0.10764262648008611,
147
+ "grad_norm": 2.34375,
148
+ "learning_rate": 0.00014142857142857145,
149
+ "loss": 3.637,
150
+ "step": 100
151
+ },
152
+ {
153
+ "epoch": 0.11302475780409042,
154
+ "grad_norm": 1.9375,
155
+ "learning_rate": 0.00014857142857142857,
156
+ "loss": 2.8895,
157
+ "step": 105
158
+ },
159
+ {
160
+ "epoch": 0.11840688912809473,
161
+ "grad_norm": 2.40625,
162
+ "learning_rate": 0.00015571428571428572,
163
+ "loss": 3.011,
164
+ "step": 110
165
+ },
166
+ {
167
+ "epoch": 0.12378902045209902,
168
+ "grad_norm": 1.6796875,
169
+ "learning_rate": 0.00016285714285714287,
170
+ "loss": 2.5217,
171
+ "step": 115
172
+ },
173
+ {
174
+ "epoch": 0.12917115177610333,
175
+ "grad_norm": 1.609375,
176
+ "learning_rate": 0.00017,
177
+ "loss": 2.3684,
178
+ "step": 120
179
+ },
180
+ {
181
+ "epoch": 0.13455328310010764,
182
+ "grad_norm": 1.3203125,
183
+ "learning_rate": 0.00017714285714285713,
184
+ "loss": 2.6268,
185
+ "step": 125
186
+ },
187
+ {
188
+ "epoch": 0.13993541442411195,
189
+ "grad_norm": 1.6953125,
190
+ "learning_rate": 0.00018428571428571428,
191
+ "loss": 2.2101,
192
+ "step": 130
193
+ },
194
+ {
195
+ "epoch": 0.14531754574811626,
196
+ "grad_norm": 1.71875,
197
+ "learning_rate": 0.00019142857142857145,
198
+ "loss": 2.1823,
199
+ "step": 135
200
+ },
201
+ {
202
+ "epoch": 0.15069967707212056,
203
+ "grad_norm": 0.9375,
204
+ "learning_rate": 0.0001985714285714286,
205
+ "loss": 1.8939,
206
+ "step": 140
207
+ },
208
+ {
209
+ "epoch": 0.15608180839612487,
210
+ "grad_norm": 1.3125,
211
+ "learning_rate": 0.00019999887311127372,
212
+ "loss": 2.6186,
213
+ "step": 145
214
+ },
215
+ {
216
+ "epoch": 0.16146393972012918,
217
+ "grad_norm": 1.15625,
218
+ "learning_rate": 0.0001999942951693512,
219
+ "loss": 2.3448,
220
+ "step": 150
221
+ },
222
+ {
223
+ "epoch": 0.1668460710441335,
224
+ "grad_norm": 1.53125,
225
+ "learning_rate": 0.00019998619590477638,
226
+ "loss": 2.0453,
227
+ "step": 155
228
+ },
229
+ {
230
+ "epoch": 0.1722282023681378,
231
+ "grad_norm": 1.1328125,
232
+ "learning_rate": 0.00019997457560276666,
233
+ "loss": 1.9159,
234
+ "step": 160
235
+ },
236
+ {
237
+ "epoch": 0.17761033369214208,
238
+ "grad_norm": 1.3359375,
239
+ "learning_rate": 0.00019995943467253385,
240
+ "loss": 2.2564,
241
+ "step": 165
242
+ },
243
+ {
244
+ "epoch": 0.18299246501614638,
245
+ "grad_norm": 1.453125,
246
+ "learning_rate": 0.00019994077364726925,
247
+ "loss": 2.1459,
248
+ "step": 170
249
+ },
250
+ {
251
+ "epoch": 0.1883745963401507,
252
+ "grad_norm": 1.15625,
253
+ "learning_rate": 0.00019991859318412537,
254
+ "loss": 2.3652,
255
+ "step": 175
256
+ },
257
+ {
258
+ "epoch": 0.193756727664155,
259
+ "grad_norm": 1.703125,
260
+ "learning_rate": 0.00019989289406419238,
261
+ "loss": 2.0883,
262
+ "step": 180
263
+ },
264
+ {
265
+ "epoch": 0.1991388589881593,
266
+ "grad_norm": 1.046875,
267
+ "learning_rate": 0.00019986367719247082,
268
+ "loss": 2.0466,
269
+ "step": 185
270
+ },
271
+ {
272
+ "epoch": 0.20452099031216361,
273
+ "grad_norm": 1.9765625,
274
+ "learning_rate": 0.00019983094359783977,
275
+ "loss": 2.0306,
276
+ "step": 190
277
+ },
278
+ {
279
+ "epoch": 0.20990312163616792,
280
+ "grad_norm": 0.93359375,
281
+ "learning_rate": 0.00019979469443302046,
282
+ "loss": 1.73,
283
+ "step": 195
284
+ },
285
+ {
286
+ "epoch": 0.21528525296017223,
287
+ "grad_norm": 1.203125,
288
+ "learning_rate": 0.0001997549309745357,
289
+ "loss": 2.1835,
290
+ "step": 200
291
+ },
292
+ {
293
+ "epoch": 0.22066738428417654,
294
+ "grad_norm": 0.80859375,
295
+ "learning_rate": 0.00019971165462266512,
296
+ "loss": 1.8594,
297
+ "step": 205
298
+ },
299
+ {
300
+ "epoch": 0.22604951560818085,
301
+ "grad_norm": 0.98828125,
302
+ "learning_rate": 0.00019966486690139564,
303
+ "loss": 1.8338,
304
+ "step": 210
305
+ },
306
+ {
307
+ "epoch": 0.23143164693218515,
308
+ "grad_norm": 1.15625,
309
+ "learning_rate": 0.0001996145694583678,
310
+ "loss": 2.0899,
311
+ "step": 215
312
+ },
313
+ {
314
+ "epoch": 0.23681377825618946,
315
+ "grad_norm": 1.296875,
316
+ "learning_rate": 0.000199560764064818,
317
+ "loss": 2.0116,
318
+ "step": 220
319
+ },
320
+ {
321
+ "epoch": 0.24219590958019377,
322
+ "grad_norm": 1.4375,
323
+ "learning_rate": 0.00019950345261551583,
324
+ "loss": 1.9816,
325
+ "step": 225
326
+ },
327
+ {
328
+ "epoch": 0.24757804090419805,
329
+ "grad_norm": 0.89453125,
330
+ "learning_rate": 0.0001994426371286974,
331
+ "loss": 1.4429,
332
+ "step": 230
333
+ },
334
+ {
335
+ "epoch": 0.2529601722282024,
336
+ "grad_norm": 1.171875,
337
+ "learning_rate": 0.00019937831974599447,
338
+ "loss": 1.8191,
339
+ "step": 235
340
+ },
341
+ {
342
+ "epoch": 0.25834230355220666,
343
+ "grad_norm": 1.0625,
344
+ "learning_rate": 0.0001993105027323588,
345
+ "loss": 2.0092,
346
+ "step": 240
347
+ },
348
+ {
349
+ "epoch": 0.263724434876211,
350
+ "grad_norm": 0.95703125,
351
+ "learning_rate": 0.00019923918847598247,
352
+ "loss": 1.6442,
353
+ "step": 245
354
+ },
355
+ {
356
+ "epoch": 0.2691065662002153,
357
+ "grad_norm": 1.1953125,
358
+ "learning_rate": 0.00019916437948821382,
359
+ "loss": 2.29,
360
+ "step": 250
361
+ },
362
+ {
363
+ "epoch": 0.2744886975242196,
364
+ "grad_norm": 1.2734375,
365
+ "learning_rate": 0.00019908607840346903,
366
+ "loss": 1.7178,
367
+ "step": 255
368
+ },
369
+ {
370
+ "epoch": 0.2798708288482239,
371
+ "grad_norm": 0.91015625,
372
+ "learning_rate": 0.00019900428797913915,
373
+ "loss": 1.8539,
374
+ "step": 260
375
+ },
376
+ {
377
+ "epoch": 0.2852529601722282,
378
+ "grad_norm": 1.21875,
379
+ "learning_rate": 0.0001989190110954933,
380
+ "loss": 1.9345,
381
+ "step": 265
382
+ },
383
+ {
384
+ "epoch": 0.2906350914962325,
385
+ "grad_norm": 1.3515625,
386
+ "learning_rate": 0.000198830250755577,
387
+ "loss": 2.1381,
388
+ "step": 270
389
+ },
390
+ {
391
+ "epoch": 0.2960172228202368,
392
+ "grad_norm": 1.015625,
393
+ "learning_rate": 0.0001987380100851065,
394
+ "loss": 2.0466,
395
+ "step": 275
396
+ },
397
+ {
398
+ "epoch": 0.3013993541442411,
399
+ "grad_norm": 0.953125,
400
+ "learning_rate": 0.00019864229233235875,
401
+ "loss": 2.1104,
402
+ "step": 280
403
+ },
404
+ {
405
+ "epoch": 0.3067814854682454,
406
+ "grad_norm": 0.984375,
407
+ "learning_rate": 0.000198543100868057,
408
+ "loss": 2.0808,
409
+ "step": 285
410
+ },
411
+ {
412
+ "epoch": 0.31216361679224974,
413
+ "grad_norm": 1.390625,
414
+ "learning_rate": 0.00019844043918525195,
415
+ "loss": 1.9378,
416
+ "step": 290
417
+ },
418
+ {
419
+ "epoch": 0.317545748116254,
420
+ "grad_norm": 0.78515625,
421
+ "learning_rate": 0.00019833431089919898,
422
+ "loss": 1.565,
423
+ "step": 295
424
+ },
425
+ {
426
+ "epoch": 0.32292787944025836,
427
+ "grad_norm": 1.7109375,
428
+ "learning_rate": 0.00019822471974723064,
429
+ "loss": 1.8316,
430
+ "step": 300
431
+ },
432
+ {
433
+ "epoch": 0.32831001076426264,
434
+ "grad_norm": 1.0234375,
435
+ "learning_rate": 0.0001981116695886252,
436
+ "loss": 1.9357,
437
+ "step": 305
438
+ },
439
+ {
440
+ "epoch": 0.333692142088267,
441
+ "grad_norm": 0.99609375,
442
+ "learning_rate": 0.00019799516440447058,
443
+ "loss": 1.7745,
444
+ "step": 310
445
+ },
446
+ {
447
+ "epoch": 0.33907427341227125,
448
+ "grad_norm": 0.97265625,
449
+ "learning_rate": 0.00019787520829752427,
450
+ "loss": 1.8947,
451
+ "step": 315
452
+ },
453
+ {
454
+ "epoch": 0.3444564047362756,
455
+ "grad_norm": 0.8046875,
456
+ "learning_rate": 0.00019775180549206888,
457
+ "loss": 1.6128,
458
+ "step": 320
459
+ },
460
+ {
461
+ "epoch": 0.34983853606027987,
462
+ "grad_norm": 0.921875,
463
+ "learning_rate": 0.0001976249603337632,
464
+ "loss": 1.6003,
465
+ "step": 325
466
+ },
467
+ {
468
+ "epoch": 0.35522066738428415,
469
+ "grad_norm": 0.7421875,
470
+ "learning_rate": 0.00019749467728948943,
471
+ "loss": 1.7543,
472
+ "step": 330
473
+ },
474
+ {
475
+ "epoch": 0.3606027987082885,
476
+ "grad_norm": 1.46875,
477
+ "learning_rate": 0.0001973609609471956,
478
+ "loss": 1.8543,
479
+ "step": 335
480
+ },
481
+ {
482
+ "epoch": 0.36598493003229277,
483
+ "grad_norm": 0.953125,
484
+ "learning_rate": 0.00019722381601573413,
485
+ "loss": 1.7436,
486
+ "step": 340
487
+ },
488
+ {
489
+ "epoch": 0.3713670613562971,
490
+ "grad_norm": 0.9765625,
491
+ "learning_rate": 0.0001970832473246962,
492
+ "loss": 2.1039,
493
+ "step": 345
494
+ },
495
+ {
496
+ "epoch": 0.3767491926803014,
497
+ "grad_norm": 0.76171875,
498
+ "learning_rate": 0.0001969392598242413,
499
+ "loss": 1.9624,
500
+ "step": 350
501
+ },
502
+ {
503
+ "epoch": 0.3821313240043057,
504
+ "grad_norm": 0.6796875,
505
+ "learning_rate": 0.0001967918585849232,
506
+ "loss": 1.4343,
507
+ "step": 355
508
+ },
509
+ {
510
+ "epoch": 0.38751345532831,
511
+ "grad_norm": 1.2421875,
512
+ "learning_rate": 0.00019664104879751123,
513
+ "loss": 1.5505,
514
+ "step": 360
515
+ },
516
+ {
517
+ "epoch": 0.39289558665231433,
518
+ "grad_norm": 0.86328125,
519
+ "learning_rate": 0.00019648683577280757,
520
+ "loss": 1.6731,
521
+ "step": 365
522
+ },
523
+ {
524
+ "epoch": 0.3982777179763186,
525
+ "grad_norm": 0.90234375,
526
+ "learning_rate": 0.0001963292249414602,
527
+ "loss": 1.6294,
528
+ "step": 370
529
+ },
530
+ {
531
+ "epoch": 0.40365984930032295,
532
+ "grad_norm": 0.80859375,
533
+ "learning_rate": 0.0001961682218537717,
534
+ "loss": 1.7152,
535
+ "step": 375
536
+ },
537
+ {
538
+ "epoch": 0.40904198062432723,
539
+ "grad_norm": 1.2890625,
540
+ "learning_rate": 0.00019600383217950365,
541
+ "loss": 2.0493,
542
+ "step": 380
543
+ },
544
+ {
545
+ "epoch": 0.41442411194833156,
546
+ "grad_norm": 1.2890625,
547
+ "learning_rate": 0.00019583606170767718,
548
+ "loss": 1.809,
549
+ "step": 385
550
+ },
551
+ {
552
+ "epoch": 0.41980624327233584,
553
+ "grad_norm": 0.93359375,
554
+ "learning_rate": 0.00019566491634636897,
555
+ "loss": 2.0224,
556
+ "step": 390
557
+ },
558
+ {
559
+ "epoch": 0.4251883745963401,
560
+ "grad_norm": 1.1484375,
561
+ "learning_rate": 0.0001954904021225032,
562
+ "loss": 1.8455,
563
+ "step": 395
564
+ },
565
+ {
566
+ "epoch": 0.43057050592034446,
567
+ "grad_norm": 0.83984375,
568
+ "learning_rate": 0.0001953125251816394,
569
+ "loss": 1.5119,
570
+ "step": 400
571
+ },
572
+ {
573
+ "epoch": 0.43595263724434874,
574
+ "grad_norm": 0.95703125,
575
+ "learning_rate": 0.00019513129178775588,
576
+ "loss": 1.6183,
577
+ "step": 405
578
+ },
579
+ {
580
+ "epoch": 0.4413347685683531,
581
+ "grad_norm": 0.97265625,
582
+ "learning_rate": 0.0001949467083230293,
583
+ "loss": 1.5994,
584
+ "step": 410
585
+ },
586
+ {
587
+ "epoch": 0.44671689989235736,
588
+ "grad_norm": 0.92578125,
589
+ "learning_rate": 0.0001947587812876099,
590
+ "loss": 1.9839,
591
+ "step": 415
592
+ },
593
+ {
594
+ "epoch": 0.4520990312163617,
595
+ "grad_norm": 0.95703125,
596
+ "learning_rate": 0.00019456751729939238,
597
+ "loss": 1.7201,
598
+ "step": 420
599
+ },
600
+ {
601
+ "epoch": 0.45748116254036597,
602
+ "grad_norm": 0.74609375,
603
+ "learning_rate": 0.00019437292309378324,
604
+ "loss": 1.8024,
605
+ "step": 425
606
+ },
607
+ {
608
+ "epoch": 0.4628632938643703,
609
+ "grad_norm": 1.609375,
610
+ "learning_rate": 0.00019417500552346316,
611
+ "loss": 1.421,
612
+ "step": 430
613
+ },
614
+ {
615
+ "epoch": 0.4682454251883746,
616
+ "grad_norm": 0.98046875,
617
+ "learning_rate": 0.000193973771558146,
618
+ "loss": 1.5199,
619
+ "step": 435
620
+ },
621
+ {
622
+ "epoch": 0.4736275565123789,
623
+ "grad_norm": 0.97265625,
624
+ "learning_rate": 0.0001937692282843333,
625
+ "loss": 1.8052,
626
+ "step": 440
627
+ },
628
+ {
629
+ "epoch": 0.4790096878363832,
630
+ "grad_norm": 1.078125,
631
+ "learning_rate": 0.00019356138290506461,
632
+ "loss": 2.0869,
633
+ "step": 445
634
+ },
635
+ {
636
+ "epoch": 0.48439181916038754,
637
+ "grad_norm": 0.99609375,
638
+ "learning_rate": 0.00019335024273966383,
639
+ "loss": 1.5393,
640
+ "step": 450
641
+ },
642
+ {
643
+ "epoch": 0.4897739504843918,
644
+ "grad_norm": 1.171875,
645
+ "learning_rate": 0.00019313581522348164,
646
+ "loss": 1.6464,
647
+ "step": 455
648
+ },
649
+ {
650
+ "epoch": 0.4951560818083961,
651
+ "grad_norm": 0.83203125,
652
+ "learning_rate": 0.00019291810790763355,
653
+ "loss": 1.8133,
654
+ "step": 460
655
+ },
656
+ {
657
+ "epoch": 0.5005382131324004,
658
+ "grad_norm": 1.453125,
659
+ "learning_rate": 0.00019269712845873397,
660
+ "loss": 2.0506,
661
+ "step": 465
662
+ },
663
+ {
664
+ "epoch": 0.5059203444564048,
665
+ "grad_norm": 0.84375,
666
+ "learning_rate": 0.00019247288465862615,
667
+ "loss": 1.95,
668
+ "step": 470
669
+ },
670
+ {
671
+ "epoch": 0.511302475780409,
672
+ "grad_norm": 0.91015625,
673
+ "learning_rate": 0.0001922453844041084,
674
+ "loss": 1.565,
675
+ "step": 475
676
+ },
677
+ {
678
+ "epoch": 0.5166846071044133,
679
+ "grad_norm": 0.71875,
680
+ "learning_rate": 0.00019201463570665573,
681
+ "loss": 1.7664,
682
+ "step": 480
683
+ },
684
+ {
685
+ "epoch": 0.5220667384284177,
686
+ "grad_norm": 0.6015625,
687
+ "learning_rate": 0.0001917806466921378,
688
+ "loss": 1.6515,
689
+ "step": 485
690
+ },
691
+ {
692
+ "epoch": 0.527448869752422,
693
+ "grad_norm": 0.69140625,
694
+ "learning_rate": 0.00019154342560053297,
695
+ "loss": 1.654,
696
+ "step": 490
697
+ },
698
+ {
699
+ "epoch": 0.5328310010764262,
700
+ "grad_norm": 0.62109375,
701
+ "learning_rate": 0.0001913029807856378,
702
+ "loss": 1.5976,
703
+ "step": 495
704
+ },
705
+ {
706
+ "epoch": 0.5382131324004306,
707
+ "grad_norm": 1.0859375,
708
+ "learning_rate": 0.00019105932071477303,
709
+ "loss": 1.497,
710
+ "step": 500
711
+ },
712
+ {
713
+ "epoch": 0.5435952637244349,
714
+ "grad_norm": 0.87890625,
715
+ "learning_rate": 0.00019081245396848547,
716
+ "loss": 1.6065,
717
+ "step": 505
718
+ },
719
+ {
720
+ "epoch": 0.5489773950484392,
721
+ "grad_norm": 0.78125,
722
+ "learning_rate": 0.00019056238924024574,
723
+ "loss": 1.3089,
724
+ "step": 510
725
+ },
726
+ {
727
+ "epoch": 0.5543595263724435,
728
+ "grad_norm": 1.0546875,
729
+ "learning_rate": 0.00019030913533614208,
730
+ "loss": 1.8274,
731
+ "step": 515
732
+ },
733
+ {
734
+ "epoch": 0.5597416576964478,
735
+ "grad_norm": 1.1953125,
736
+ "learning_rate": 0.00019005270117457043,
737
+ "loss": 1.6165,
738
+ "step": 520
739
+ },
740
+ {
741
+ "epoch": 0.5651237890204521,
742
+ "grad_norm": 1.0625,
743
+ "learning_rate": 0.00018979309578592016,
744
+ "loss": 1.578,
745
+ "step": 525
746
+ },
747
+ {
748
+ "epoch": 0.5705059203444564,
749
+ "grad_norm": 1.7109375,
750
+ "learning_rate": 0.0001895303283122561,
751
+ "loss": 1.7197,
752
+ "step": 530
753
+ },
754
+ {
755
+ "epoch": 0.5758880516684607,
756
+ "grad_norm": 1.0390625,
757
+ "learning_rate": 0.00018926440800699678,
758
+ "loss": 1.7235,
759
+ "step": 535
760
+ },
761
+ {
762
+ "epoch": 0.581270182992465,
763
+ "grad_norm": 0.92578125,
764
+ "learning_rate": 0.0001889953442345884,
765
+ "loss": 1.536,
766
+ "step": 540
767
+ },
768
+ {
769
+ "epoch": 0.5866523143164694,
770
+ "grad_norm": 0.8125,
771
+ "learning_rate": 0.000188723146470175,
772
+ "loss": 1.8374,
773
+ "step": 545
774
+ },
775
+ {
776
+ "epoch": 0.5920344456404736,
777
+ "grad_norm": 0.96484375,
778
+ "learning_rate": 0.00018844782429926495,
779
+ "loss": 1.4953,
780
+ "step": 550
781
+ },
782
+ {
783
+ "epoch": 0.5974165769644779,
784
+ "grad_norm": 0.7734375,
785
+ "learning_rate": 0.0001881693874173934,
786
+ "loss": 1.3424,
787
+ "step": 555
788
+ },
789
+ {
790
+ "epoch": 0.6027987082884823,
791
+ "grad_norm": 0.7890625,
792
+ "learning_rate": 0.0001878878456297807,
793
+ "loss": 1.4368,
794
+ "step": 560
795
+ },
796
+ {
797
+ "epoch": 0.6081808396124866,
798
+ "grad_norm": 1.09375,
799
+ "learning_rate": 0.00018760320885098715,
800
+ "loss": 1.7983,
801
+ "step": 565
802
+ },
803
+ {
804
+ "epoch": 0.6135629709364908,
805
+ "grad_norm": 1.1015625,
806
+ "learning_rate": 0.00018731548710456398,
807
+ "loss": 1.8513,
808
+ "step": 570
809
+ },
810
+ {
811
+ "epoch": 0.6189451022604952,
812
+ "grad_norm": 2.125,
813
+ "learning_rate": 0.00018702469052270023,
814
+ "loss": 2.1254,
815
+ "step": 575
816
+ },
817
+ {
818
+ "epoch": 0.6243272335844995,
819
+ "grad_norm": 0.7421875,
820
+ "learning_rate": 0.00018673082934586607,
821
+ "loss": 1.6763,
822
+ "step": 580
823
+ },
824
+ {
825
+ "epoch": 0.6297093649085038,
826
+ "grad_norm": 1.140625,
827
+ "learning_rate": 0.00018643391392245197,
828
+ "loss": 1.6641,
829
+ "step": 585
830
+ },
831
+ {
832
+ "epoch": 0.635091496232508,
833
+ "grad_norm": 0.890625,
834
+ "learning_rate": 0.00018613395470840454,
835
+ "loss": 2.0576,
836
+ "step": 590
837
+ },
838
+ {
839
+ "epoch": 0.6404736275565124,
840
+ "grad_norm": 0.74609375,
841
+ "learning_rate": 0.0001858309622668581,
842
+ "loss": 1.643,
843
+ "step": 595
844
+ },
845
+ {
846
+ "epoch": 0.6458557588805167,
847
+ "grad_norm": 1.3671875,
848
+ "learning_rate": 0.00018552494726776283,
849
+ "loss": 1.8816,
850
+ "step": 600
851
+ },
852
+ {
853
+ "epoch": 0.6512378902045209,
854
+ "grad_norm": 0.8515625,
855
+ "learning_rate": 0.00018521592048750907,
856
+ "loss": 1.4728,
857
+ "step": 605
858
+ },
859
+ {
860
+ "epoch": 0.6566200215285253,
861
+ "grad_norm": 0.94921875,
862
+ "learning_rate": 0.00018490389280854754,
863
+ "loss": 1.6078,
864
+ "step": 610
865
+ },
866
+ {
867
+ "epoch": 0.6620021528525296,
868
+ "grad_norm": 0.96484375,
869
+ "learning_rate": 0.0001845888752190065,
870
+ "loss": 1.6509,
871
+ "step": 615
872
+ },
873
+ {
874
+ "epoch": 0.667384284176534,
875
+ "grad_norm": 0.73046875,
876
+ "learning_rate": 0.00018427087881230455,
877
+ "loss": 1.1777,
878
+ "step": 620
879
+ },
880
+ {
881
+ "epoch": 0.6727664155005382,
882
+ "grad_norm": 1.09375,
883
+ "learning_rate": 0.00018394991478676005,
884
+ "loss": 1.4283,
885
+ "step": 625
886
+ },
887
+ {
888
+ "epoch": 0.6781485468245425,
889
+ "grad_norm": 0.859375,
890
+ "learning_rate": 0.0001836259944451967,
891
+ "loss": 1.6932,
892
+ "step": 630
893
+ },
894
+ {
895
+ "epoch": 0.6835306781485468,
896
+ "grad_norm": 1.03125,
897
+ "learning_rate": 0.00018329912919454565,
898
+ "loss": 1.4168,
899
+ "step": 635
900
+ },
901
+ {
902
+ "epoch": 0.6889128094725512,
903
+ "grad_norm": 0.77734375,
904
+ "learning_rate": 0.00018296933054544367,
905
+ "loss": 1.7149,
906
+ "step": 640
907
+ },
908
+ {
909
+ "epoch": 0.6942949407965554,
910
+ "grad_norm": 0.921875,
911
+ "learning_rate": 0.00018263661011182783,
912
+ "loss": 1.5492,
913
+ "step": 645
914
+ },
915
+ {
916
+ "epoch": 0.6996770721205597,
917
+ "grad_norm": 0.8828125,
918
+ "learning_rate": 0.00018230097961052658,
919
+ "loss": 1.8432,
920
+ "step": 650
921
+ },
922
+ {
923
+ "epoch": 0.7050592034445641,
924
+ "grad_norm": 0.8359375,
925
+ "learning_rate": 0.00018196245086084703,
926
+ "loss": 1.5463,
927
+ "step": 655
928
+ },
929
+ {
930
+ "epoch": 0.7104413347685683,
931
+ "grad_norm": 1.03125,
932
+ "learning_rate": 0.00018162103578415884,
933
+ "loss": 1.6883,
934
+ "step": 660
935
+ },
936
+ {
937
+ "epoch": 0.7158234660925726,
938
+ "grad_norm": 0.8515625,
939
+ "learning_rate": 0.0001812767464034743,
940
+ "loss": 1.5512,
941
+ "step": 665
942
+ },
943
+ {
944
+ "epoch": 0.721205597416577,
945
+ "grad_norm": 0.91796875,
946
+ "learning_rate": 0.00018092959484302515,
947
+ "loss": 1.387,
948
+ "step": 670
949
+ },
950
+ {
951
+ "epoch": 0.7265877287405813,
952
+ "grad_norm": 0.7421875,
953
+ "learning_rate": 0.00018057959332783518,
954
+ "loss": 1.2412,
955
+ "step": 675
956
+ },
957
+ {
958
+ "epoch": 0.7319698600645855,
959
+ "grad_norm": 0.6875,
960
+ "learning_rate": 0.0001802267541832903,
961
+ "loss": 1.4303,
962
+ "step": 680
963
+ },
964
+ {
965
+ "epoch": 0.7373519913885899,
966
+ "grad_norm": 1.359375,
967
+ "learning_rate": 0.00017987108983470403,
968
+ "loss": 1.9346,
969
+ "step": 685
970
+ },
971
+ {
972
+ "epoch": 0.7427341227125942,
973
+ "grad_norm": 1.890625,
974
+ "learning_rate": 0.0001795126128068801,
975
+ "loss": 1.5194,
976
+ "step": 690
977
+ },
978
+ {
979
+ "epoch": 0.7481162540365985,
980
+ "grad_norm": 1.1796875,
981
+ "learning_rate": 0.00017915133572367155,
982
+ "loss": 1.9612,
983
+ "step": 695
984
+ },
985
+ {
986
+ "epoch": 0.7534983853606028,
987
+ "grad_norm": 1.15625,
988
+ "learning_rate": 0.00017878727130753592,
989
+ "loss": 1.8595,
990
+ "step": 700
991
+ },
992
+ {
993
+ "epoch": 0.7588805166846071,
994
+ "grad_norm": 1.1640625,
995
+ "learning_rate": 0.00017842043237908733,
996
+ "loss": 1.4777,
997
+ "step": 705
998
+ },
999
+ {
1000
+ "epoch": 0.7642626480086114,
1001
+ "grad_norm": 1.1171875,
1002
+ "learning_rate": 0.00017805083185664508,
1003
+ "loss": 1.6248,
1004
+ "step": 710
1005
+ },
1006
+ {
1007
+ "epoch": 0.7696447793326158,
1008
+ "grad_norm": 0.82421875,
1009
+ "learning_rate": 0.00017767848275577856,
1010
+ "loss": 1.7387,
1011
+ "step": 715
1012
+ },
1013
+ {
1014
+ "epoch": 0.77502691065662,
1015
+ "grad_norm": 0.83984375,
1016
+ "learning_rate": 0.0001773033981888491,
1017
+ "loss": 1.7641,
1018
+ "step": 720
1019
+ },
1020
+ {
1021
+ "epoch": 0.7804090419806243,
1022
+ "grad_norm": 0.8359375,
1023
+ "learning_rate": 0.000176925591364548,
1024
+ "loss": 1.853,
1025
+ "step": 725
1026
+ },
1027
+ {
1028
+ "epoch": 0.7857911733046287,
1029
+ "grad_norm": 1.046875,
1030
+ "learning_rate": 0.00017654507558743153,
1031
+ "loss": 1.3802,
1032
+ "step": 730
1033
+ },
1034
+ {
1035
+ "epoch": 0.7911733046286329,
1036
+ "grad_norm": 0.87890625,
1037
+ "learning_rate": 0.00017616186425745248,
1038
+ "loss": 1.8941,
1039
+ "step": 735
1040
+ },
1041
+ {
1042
+ "epoch": 0.7965554359526372,
1043
+ "grad_norm": 0.80078125,
1044
+ "learning_rate": 0.00017577597086948797,
1045
+ "loss": 1.6458,
1046
+ "step": 740
1047
+ },
1048
+ {
1049
+ "epoch": 0.8019375672766416,
1050
+ "grad_norm": 1.0078125,
1051
+ "learning_rate": 0.00017538740901286464,
1052
+ "loss": 1.6792,
1053
+ "step": 745
1054
+ },
1055
+ {
1056
+ "epoch": 0.8073196986006459,
1057
+ "grad_norm": 0.640625,
1058
+ "learning_rate": 0.00017499619237087969,
1059
+ "loss": 1.389,
1060
+ "step": 750
1061
+ },
1062
+ {
1063
+ "epoch": 0.8127018299246501,
1064
+ "grad_norm": 1.015625,
1065
+ "learning_rate": 0.00017460233472031935,
1066
+ "loss": 2.0551,
1067
+ "step": 755
1068
+ },
1069
+ {
1070
+ "epoch": 0.8180839612486545,
1071
+ "grad_norm": 1.0625,
1072
+ "learning_rate": 0.0001742058499309735,
1073
+ "loss": 1.6721,
1074
+ "step": 760
1075
+ },
1076
+ {
1077
+ "epoch": 0.8234660925726588,
1078
+ "grad_norm": 0.83203125,
1079
+ "learning_rate": 0.00017380675196514739,
1080
+ "loss": 1.3685,
1081
+ "step": 765
1082
+ },
1083
+ {
1084
+ "epoch": 0.8288482238966631,
1085
+ "grad_norm": 1.09375,
1086
+ "learning_rate": 0.00017340505487716985,
1087
+ "loss": 1.7255,
1088
+ "step": 770
1089
+ },
1090
+ {
1091
+ "epoch": 0.8342303552206674,
1092
+ "grad_norm": 1.0546875,
1093
+ "learning_rate": 0.00017300077281289845,
1094
+ "loss": 1.5643,
1095
+ "step": 775
1096
+ },
1097
+ {
1098
+ "epoch": 0.8396124865446717,
1099
+ "grad_norm": 0.8203125,
1100
+ "learning_rate": 0.00017259392000922125,
1101
+ "loss": 1.516,
1102
+ "step": 780
1103
+ },
1104
+ {
1105
+ "epoch": 0.844994617868676,
1106
+ "grad_norm": 1.1328125,
1107
+ "learning_rate": 0.0001721845107935556,
1108
+ "loss": 1.6089,
1109
+ "step": 785
1110
+ },
1111
+ {
1112
+ "epoch": 0.8503767491926802,
1113
+ "grad_norm": 0.67578125,
1114
+ "learning_rate": 0.00017177255958334342,
1115
+ "loss": 1.5203,
1116
+ "step": 790
1117
+ },
1118
+ {
1119
+ "epoch": 0.8557588805166846,
1120
+ "grad_norm": 0.796875,
1121
+ "learning_rate": 0.00017135808088554358,
1122
+ "loss": 2.1079,
1123
+ "step": 795
1124
+ },
1125
+ {
1126
+ "epoch": 0.8611410118406889,
1127
+ "grad_norm": 0.87890625,
1128
+ "learning_rate": 0.000170941089296121,
1129
+ "loss": 1.8979,
1130
+ "step": 800
1131
+ },
1132
+ {
1133
+ "epoch": 0.8665231431646933,
1134
+ "grad_norm": 0.90625,
1135
+ "learning_rate": 0.00017052159949953278,
1136
+ "loss": 1.4225,
1137
+ "step": 805
1138
+ },
1139
+ {
1140
+ "epoch": 0.8719052744886975,
1141
+ "grad_norm": 1.03125,
1142
+ "learning_rate": 0.00017009962626821082,
1143
+ "loss": 1.7616,
1144
+ "step": 810
1145
+ },
1146
+ {
1147
+ "epoch": 0.8772874058127018,
1148
+ "grad_norm": 0.87890625,
1149
+ "learning_rate": 0.0001696751844620419,
1150
+ "loss": 1.8864,
1151
+ "step": 815
1152
+ },
1153
+ {
1154
+ "epoch": 0.8826695371367062,
1155
+ "grad_norm": 1.234375,
1156
+ "learning_rate": 0.00016924828902784407,
1157
+ "loss": 1.6568,
1158
+ "step": 820
1159
+ },
1160
+ {
1161
+ "epoch": 0.8880516684607105,
1162
+ "grad_norm": 0.8046875,
1163
+ "learning_rate": 0.00016881895499884072,
1164
+ "loss": 1.6826,
1165
+ "step": 825
1166
+ },
1167
+ {
1168
+ "epoch": 0.8934337997847147,
1169
+ "grad_norm": 0.95703125,
1170
+ "learning_rate": 0.00016838719749413063,
1171
+ "loss": 1.3967,
1172
+ "step": 830
1173
+ },
1174
+ {
1175
+ "epoch": 0.898815931108719,
1176
+ "grad_norm": 1.1953125,
1177
+ "learning_rate": 0.00016795303171815616,
1178
+ "loss": 1.7213,
1179
+ "step": 835
1180
+ },
1181
+ {
1182
+ "epoch": 0.9041980624327234,
1183
+ "grad_norm": 0.8125,
1184
+ "learning_rate": 0.00016751647296016725,
1185
+ "loss": 1.6946,
1186
+ "step": 840
1187
+ },
1188
+ {
1189
+ "epoch": 0.9095801937567277,
1190
+ "grad_norm": 0.71484375,
1191
+ "learning_rate": 0.00016707753659368337,
1192
+ "loss": 1.8869,
1193
+ "step": 845
1194
+ },
1195
+ {
1196
+ "epoch": 0.9149623250807319,
1197
+ "grad_norm": 1.0,
1198
+ "learning_rate": 0.0001666362380759521,
1199
+ "loss": 1.5258,
1200
+ "step": 850
1201
+ },
1202
+ {
1203
+ "epoch": 0.9203444564047363,
1204
+ "grad_norm": 1.28125,
1205
+ "learning_rate": 0.0001661925929474046,
1206
+ "loss": 1.8925,
1207
+ "step": 855
1208
+ },
1209
+ {
1210
+ "epoch": 0.9257265877287406,
1211
+ "grad_norm": 0.84765625,
1212
+ "learning_rate": 0.00016574661683110858,
1213
+ "loss": 1.7461,
1214
+ "step": 860
1215
+ },
1216
+ {
1217
+ "epoch": 0.9311087190527448,
1218
+ "grad_norm": 1.0546875,
1219
+ "learning_rate": 0.00016529832543221796,
1220
+ "loss": 1.6321,
1221
+ "step": 865
1222
+ },
1223
+ {
1224
+ "epoch": 0.9364908503767492,
1225
+ "grad_norm": 0.83203125,
1226
+ "learning_rate": 0.00016484773453741999,
1227
+ "loss": 1.4394,
1228
+ "step": 870
1229
+ },
1230
+ {
1231
+ "epoch": 0.9418729817007535,
1232
+ "grad_norm": 0.84375,
1233
+ "learning_rate": 0.0001643948600143791,
1234
+ "loss": 1.7612,
1235
+ "step": 875
1236
+ },
1237
+ {
1238
+ "epoch": 0.9472551130247578,
1239
+ "grad_norm": 1.2890625,
1240
+ "learning_rate": 0.00016393971781117827,
1241
+ "loss": 1.5663,
1242
+ "step": 880
1243
+ },
1244
+ {
1245
+ "epoch": 0.9526372443487621,
1246
+ "grad_norm": 0.9375,
1247
+ "learning_rate": 0.00016348232395575738,
1248
+ "loss": 1.6009,
1249
+ "step": 885
1250
+ },
1251
+ {
1252
+ "epoch": 0.9580193756727664,
1253
+ "grad_norm": 0.796875,
1254
+ "learning_rate": 0.0001630226945553487,
1255
+ "loss": 1.5051,
1256
+ "step": 890
1257
+ },
1258
+ {
1259
+ "epoch": 0.9634015069967707,
1260
+ "grad_norm": 0.58984375,
1261
+ "learning_rate": 0.00016256084579590991,
1262
+ "loss": 1.3798,
1263
+ "step": 895
1264
+ },
1265
+ {
1266
+ "epoch": 0.9687836383207751,
1267
+ "grad_norm": 1.1328125,
1268
+ "learning_rate": 0.00016209679394155378,
1269
+ "loss": 1.542,
1270
+ "step": 900
1271
+ },
1272
+ {
1273
+ "epoch": 0.9741657696447793,
1274
+ "grad_norm": 0.875,
1275
+ "learning_rate": 0.0001616305553339756,
1276
+ "loss": 1.7239,
1277
+ "step": 905
1278
+ },
1279
+ {
1280
+ "epoch": 0.9795479009687836,
1281
+ "grad_norm": 1.0078125,
1282
+ "learning_rate": 0.0001611621463918778,
1283
+ "loss": 1.6649,
1284
+ "step": 910
1285
+ },
1286
+ {
1287
+ "epoch": 0.984930032292788,
1288
+ "grad_norm": 0.91796875,
1289
+ "learning_rate": 0.0001606915836103916,
1290
+ "loss": 1.5315,
1291
+ "step": 915
1292
+ },
1293
+ {
1294
+ "epoch": 0.9903121636167922,
1295
+ "grad_norm": 0.7734375,
1296
+ "learning_rate": 0.00016021888356049607,
1297
+ "loss": 1.6008,
1298
+ "step": 920
1299
+ },
1300
+ {
1301
+ "epoch": 0.9956942949407965,
1302
+ "grad_norm": 0.875,
1303
+ "learning_rate": 0.00015974406288843485,
1304
+ "loss": 1.4304,
1305
+ "step": 925
1306
+ },
1307
+ {
1308
+ "epoch": 1.0010764262648009,
1309
+ "grad_norm": 0.76953125,
1310
+ "learning_rate": 0.0001592671383151297,
1311
+ "loss": 1.2241,
1312
+ "step": 930
1313
+ },
1314
+ {
1315
+ "epoch": 1.0064585575888052,
1316
+ "grad_norm": 0.6640625,
1317
+ "learning_rate": 0.0001587881266355918,
1318
+ "loss": 1.2942,
1319
+ "step": 935
1320
+ },
1321
+ {
1322
+ "epoch": 1.0118406889128095,
1323
+ "grad_norm": 0.51953125,
1324
+ "learning_rate": 0.00015830704471833028,
1325
+ "loss": 1.4488,
1326
+ "step": 940
1327
+ },
1328
+ {
1329
+ "epoch": 1.0172228202368139,
1330
+ "grad_norm": 1.03125,
1331
+ "learning_rate": 0.00015782390950475804,
1332
+ "loss": 1.3599,
1333
+ "step": 945
1334
+ },
1335
+ {
1336
+ "epoch": 1.022604951560818,
1337
+ "grad_norm": 0.95703125,
1338
+ "learning_rate": 0.0001573387380085955,
1339
+ "loss": 1.4847,
1340
+ "step": 950
1341
+ },
1342
+ {
1343
+ "epoch": 1.0279870828848223,
1344
+ "grad_norm": 0.69921875,
1345
+ "learning_rate": 0.00015685154731527113,
1346
+ "loss": 1.7192,
1347
+ "step": 955
1348
+ },
1349
+ {
1350
+ "epoch": 1.0333692142088267,
1351
+ "grad_norm": 0.69140625,
1352
+ "learning_rate": 0.0001563623545813199,
1353
+ "loss": 1.3389,
1354
+ "step": 960
1355
+ },
1356
+ {
1357
+ "epoch": 1.038751345532831,
1358
+ "grad_norm": 0.83984375,
1359
+ "learning_rate": 0.0001558711770337792,
1360
+ "loss": 1.625,
1361
+ "step": 965
1362
+ },
1363
+ {
1364
+ "epoch": 1.0441334768568353,
1365
+ "grad_norm": 0.9453125,
1366
+ "learning_rate": 0.00015537803196958195,
1367
+ "loss": 1.5407,
1368
+ "step": 970
1369
+ },
1370
+ {
1371
+ "epoch": 1.0495156081808397,
1372
+ "grad_norm": 0.6328125,
1373
+ "learning_rate": 0.00015488293675494777,
1374
+ "loss": 1.3429,
1375
+ "step": 975
1376
+ },
1377
+ {
1378
+ "epoch": 1.054897739504844,
1379
+ "grad_norm": 0.953125,
1380
+ "learning_rate": 0.00015438590882477118,
1381
+ "loss": 1.016,
1382
+ "step": 980
1383
+ },
1384
+ {
1385
+ "epoch": 1.0602798708288481,
1386
+ "grad_norm": 1.0,
1387
+ "learning_rate": 0.00015388696568200788,
1388
+ "loss": 1.428,
1389
+ "step": 985
1390
+ },
1391
+ {
1392
+ "epoch": 1.0656620021528525,
1393
+ "grad_norm": 0.84765625,
1394
+ "learning_rate": 0.00015338612489705807,
1395
+ "loss": 1.2952,
1396
+ "step": 990
1397
+ },
1398
+ {
1399
+ "epoch": 1.0710441334768568,
1400
+ "grad_norm": 0.73828125,
1401
+ "learning_rate": 0.00015288340410714798,
1402
+ "loss": 1.4856,
1403
+ "step": 995
1404
+ },
1405
+ {
1406
+ "epoch": 1.0764262648008611,
1407
+ "grad_norm": 0.80078125,
1408
+ "learning_rate": 0.00015237882101570854,
1409
+ "loss": 1.7727,
1410
+ "step": 1000
1411
+ },
1412
+ {
1413
+ "epoch": 1.0818083961248655,
1414
+ "grad_norm": 0.796875,
1415
+ "learning_rate": 0.00015187239339175225,
1416
+ "loss": 1.6775,
1417
+ "step": 1005
1418
+ },
1419
+ {
1420
+ "epoch": 1.0871905274488698,
1421
+ "grad_norm": 0.91796875,
1422
+ "learning_rate": 0.00015136413906924708,
1423
+ "loss": 1.4311,
1424
+ "step": 1010
1425
+ },
1426
+ {
1427
+ "epoch": 1.0925726587728741,
1428
+ "grad_norm": 1.2109375,
1429
+ "learning_rate": 0.00015085407594648864,
1430
+ "loss": 1.4255,
1431
+ "step": 1015
1432
+ },
1433
+ {
1434
+ "epoch": 1.0979547900968785,
1435
+ "grad_norm": 0.84765625,
1436
+ "learning_rate": 0.00015034222198547002,
1437
+ "loss": 1.4475,
1438
+ "step": 1020
1439
+ },
1440
+ {
1441
+ "epoch": 1.1033369214208826,
1442
+ "grad_norm": 0.61328125,
1443
+ "learning_rate": 0.00014982859521124893,
1444
+ "loss": 1.6184,
1445
+ "step": 1025
1446
+ },
1447
+ {
1448
+ "epoch": 1.108719052744887,
1449
+ "grad_norm": 0.85546875,
1450
+ "learning_rate": 0.00014931321371131323,
1451
+ "loss": 1.403,
1452
+ "step": 1030
1453
+ },
1454
+ {
1455
+ "epoch": 1.1141011840688912,
1456
+ "grad_norm": 0.61328125,
1457
+ "learning_rate": 0.00014879609563494373,
1458
+ "loss": 1.3015,
1459
+ "step": 1035
1460
+ },
1461
+ {
1462
+ "epoch": 1.1194833153928956,
1463
+ "grad_norm": 0.7734375,
1464
+ "learning_rate": 0.00014827725919257532,
1465
+ "loss": 1.3636,
1466
+ "step": 1040
1467
+ },
1468
+ {
1469
+ "epoch": 1.1248654467169,
1470
+ "grad_norm": 0.74609375,
1471
+ "learning_rate": 0.0001477567226551555,
1472
+ "loss": 1.318,
1473
+ "step": 1045
1474
+ },
1475
+ {
1476
+ "epoch": 1.1302475780409043,
1477
+ "grad_norm": 0.578125,
1478
+ "learning_rate": 0.00014723450435350101,
1479
+ "loss": 1.4797,
1480
+ "step": 1050
1481
+ },
1482
+ {
1483
+ "epoch": 1.1356297093649086,
1484
+ "grad_norm": 1.703125,
1485
+ "learning_rate": 0.00014671062267765232,
1486
+ "loss": 1.7593,
1487
+ "step": 1055
1488
+ },
1489
+ {
1490
+ "epoch": 1.141011840688913,
1491
+ "grad_norm": 0.6328125,
1492
+ "learning_rate": 0.000146185096076226,
1493
+ "loss": 1.4441,
1494
+ "step": 1060
1495
+ },
1496
+ {
1497
+ "epoch": 1.146393972012917,
1498
+ "grad_norm": 0.8828125,
1499
+ "learning_rate": 0.00014565794305576512,
1500
+ "loss": 1.3185,
1501
+ "step": 1065
1502
+ },
1503
+ {
1504
+ "epoch": 1.1517761033369214,
1505
+ "grad_norm": 1.0859375,
1506
+ "learning_rate": 0.0001451291821800874,
1507
+ "loss": 1.335,
1508
+ "step": 1070
1509
+ },
1510
+ {
1511
+ "epoch": 1.1571582346609257,
1512
+ "grad_norm": 0.92578125,
1513
+ "learning_rate": 0.00014459883206963158,
1514
+ "loss": 1.9123,
1515
+ "step": 1075
1516
+ },
1517
+ {
1518
+ "epoch": 1.16254036598493,
1519
+ "grad_norm": 1.1640625,
1520
+ "learning_rate": 0.00014406691140080174,
1521
+ "loss": 1.5474,
1522
+ "step": 1080
1523
+ },
1524
+ {
1525
+ "epoch": 1.1679224973089344,
1526
+ "grad_norm": 0.78125,
1527
+ "learning_rate": 0.00014353343890530947,
1528
+ "loss": 1.7054,
1529
+ "step": 1085
1530
+ },
1531
+ {
1532
+ "epoch": 1.1733046286329387,
1533
+ "grad_norm": 1.1015625,
1534
+ "learning_rate": 0.00014299843336951435,
1535
+ "loss": 1.7048,
1536
+ "step": 1090
1537
+ },
1538
+ {
1539
+ "epoch": 1.178686759956943,
1540
+ "grad_norm": 0.91015625,
1541
+ "learning_rate": 0.0001424619136337623,
1542
+ "loss": 1.4729,
1543
+ "step": 1095
1544
+ },
1545
+ {
1546
+ "epoch": 1.1840688912809472,
1547
+ "grad_norm": 0.8046875,
1548
+ "learning_rate": 0.00014192389859172215,
1549
+ "loss": 1.5441,
1550
+ "step": 1100
1551
+ },
1552
+ {
1553
+ "epoch": 1.1894510226049515,
1554
+ "grad_norm": 1.0234375,
1555
+ "learning_rate": 0.0001413844071897203,
1556
+ "loss": 1.3415,
1557
+ "step": 1105
1558
+ },
1559
+ {
1560
+ "epoch": 1.1948331539289558,
1561
+ "grad_norm": 0.6796875,
1562
+ "learning_rate": 0.00014084345842607353,
1563
+ "loss": 1.6552,
1564
+ "step": 1110
1565
+ },
1566
+ {
1567
+ "epoch": 1.2002152852529602,
1568
+ "grad_norm": 1.015625,
1569
+ "learning_rate": 0.00014030107135041994,
1570
+ "loss": 1.6165,
1571
+ "step": 1115
1572
+ },
1573
+ {
1574
+ "epoch": 1.2055974165769645,
1575
+ "grad_norm": 0.9453125,
1576
+ "learning_rate": 0.00013975726506304815,
1577
+ "loss": 1.7135,
1578
+ "step": 1120
1579
+ },
1580
+ {
1581
+ "epoch": 1.2109795479009688,
1582
+ "grad_norm": 0.9765625,
1583
+ "learning_rate": 0.0001392120587142246,
1584
+ "loss": 1.3601,
1585
+ "step": 1125
1586
+ },
1587
+ {
1588
+ "epoch": 1.2163616792249732,
1589
+ "grad_norm": 0.86328125,
1590
+ "learning_rate": 0.00013866547150351925,
1591
+ "loss": 1.5728,
1592
+ "step": 1130
1593
+ },
1594
+ {
1595
+ "epoch": 1.2217438105489773,
1596
+ "grad_norm": 1.0390625,
1597
+ "learning_rate": 0.00013811752267912946,
1598
+ "loss": 1.5137,
1599
+ "step": 1135
1600
+ },
1601
+ {
1602
+ "epoch": 1.2271259418729816,
1603
+ "grad_norm": 0.7578125,
1604
+ "learning_rate": 0.00013756823153720214,
1605
+ "loss": 1.5428,
1606
+ "step": 1140
1607
+ },
1608
+ {
1609
+ "epoch": 1.232508073196986,
1610
+ "grad_norm": 1.03125,
1611
+ "learning_rate": 0.0001370176174211542,
1612
+ "loss": 1.5014,
1613
+ "step": 1145
1614
+ },
1615
+ {
1616
+ "epoch": 1.2378902045209903,
1617
+ "grad_norm": 1.0546875,
1618
+ "learning_rate": 0.00013646569972099136,
1619
+ "loss": 1.6073,
1620
+ "step": 1150
1621
+ },
1622
+ {
1623
+ "epoch": 1.2432723358449946,
1624
+ "grad_norm": 0.79296875,
1625
+ "learning_rate": 0.00013591249787262542,
1626
+ "loss": 1.4086,
1627
+ "step": 1155
1628
+ },
1629
+ {
1630
+ "epoch": 1.248654467168999,
1631
+ "grad_norm": 1.4921875,
1632
+ "learning_rate": 0.0001353580313571898,
1633
+ "loss": 1.633,
1634
+ "step": 1160
1635
+ },
1636
+ {
1637
+ "epoch": 1.2540365984930033,
1638
+ "grad_norm": 0.68359375,
1639
+ "learning_rate": 0.00013480231970035335,
1640
+ "loss": 1.7106,
1641
+ "step": 1165
1642
+ },
1643
+ {
1644
+ "epoch": 1.2594187298170074,
1645
+ "grad_norm": 0.74609375,
1646
+ "learning_rate": 0.00013424538247163294,
1647
+ "loss": 1.5289,
1648
+ "step": 1170
1649
+ },
1650
+ {
1651
+ "epoch": 1.264800861141012,
1652
+ "grad_norm": 0.703125,
1653
+ "learning_rate": 0.00013368723928370427,
1654
+ "loss": 1.5076,
1655
+ "step": 1175
1656
+ },
1657
+ {
1658
+ "epoch": 1.270182992465016,
1659
+ "grad_norm": 0.6875,
1660
+ "learning_rate": 0.00013312790979171107,
1661
+ "loss": 1.3272,
1662
+ "step": 1180
1663
+ },
1664
+ {
1665
+ "epoch": 1.2755651237890204,
1666
+ "grad_norm": 0.7578125,
1667
+ "learning_rate": 0.0001325674136925732,
1668
+ "loss": 1.3299,
1669
+ "step": 1185
1670
+ },
1671
+ {
1672
+ "epoch": 1.2809472551130248,
1673
+ "grad_norm": 1.0078125,
1674
+ "learning_rate": 0.00013200577072429277,
1675
+ "loss": 1.6876,
1676
+ "step": 1190
1677
+ },
1678
+ {
1679
+ "epoch": 1.286329386437029,
1680
+ "grad_norm": 0.6328125,
1681
+ "learning_rate": 0.00013144300066525913,
1682
+ "loss": 1.26,
1683
+ "step": 1195
1684
+ },
1685
+ {
1686
+ "epoch": 1.2917115177610334,
1687
+ "grad_norm": 0.8984375,
1688
+ "learning_rate": 0.00013087912333355247,
1689
+ "loss": 1.5801,
1690
+ "step": 1200
1691
+ },
1692
+ {
1693
+ "epoch": 1.2970936490850375,
1694
+ "grad_norm": 0.6875,
1695
+ "learning_rate": 0.00013031415858624578,
1696
+ "loss": 1.5057,
1697
+ "step": 1205
1698
+ },
1699
+ {
1700
+ "epoch": 1.302475780409042,
1701
+ "grad_norm": 0.5546875,
1702
+ "learning_rate": 0.00012974812631870574,
1703
+ "loss": 1.2602,
1704
+ "step": 1210
1705
+ },
1706
+ {
1707
+ "epoch": 1.3078579117330462,
1708
+ "grad_norm": 0.8125,
1709
+ "learning_rate": 0.0001291810464638919,
1710
+ "loss": 1.2763,
1711
+ "step": 1215
1712
+ },
1713
+ {
1714
+ "epoch": 1.3132400430570506,
1715
+ "grad_norm": 0.8203125,
1716
+ "learning_rate": 0.0001286129389916549,
1717
+ "loss": 1.6902,
1718
+ "step": 1220
1719
+ },
1720
+ {
1721
+ "epoch": 1.318622174381055,
1722
+ "grad_norm": 0.8203125,
1723
+ "learning_rate": 0.00012804382390803315,
1724
+ "loss": 1.537,
1725
+ "step": 1225
1726
+ },
1727
+ {
1728
+ "epoch": 1.3240043057050592,
1729
+ "grad_norm": 0.8671875,
1730
+ "learning_rate": 0.0001274737212545484,
1731
+ "loss": 1.7872,
1732
+ "step": 1230
1733
+ },
1734
+ {
1735
+ "epoch": 1.3293864370290636,
1736
+ "grad_norm": 0.84375,
1737
+ "learning_rate": 0.00012690265110749978,
1738
+ "loss": 1.4077,
1739
+ "step": 1235
1740
+ },
1741
+ {
1742
+ "epoch": 1.334768568353068,
1743
+ "grad_norm": 0.76171875,
1744
+ "learning_rate": 0.00012633063357725703,
1745
+ "loss": 1.4193,
1746
+ "step": 1240
1747
+ },
1748
+ {
1749
+ "epoch": 1.3401506996770722,
1750
+ "grad_norm": 1.2578125,
1751
+ "learning_rate": 0.00012575768880755217,
1752
+ "loss": 1.9034,
1753
+ "step": 1245
1754
+ },
1755
+ {
1756
+ "epoch": 1.3455328310010763,
1757
+ "grad_norm": 0.82421875,
1758
+ "learning_rate": 0.00012518383697477018,
1759
+ "loss": 1.5607,
1760
+ "step": 1250
1761
+ },
1762
+ {
1763
+ "epoch": 1.3509149623250807,
1764
+ "grad_norm": 1.0234375,
1765
+ "learning_rate": 0.00012460909828723854,
1766
+ "loss": 1.4859,
1767
+ "step": 1255
1768
+ },
1769
+ {
1770
+ "epoch": 1.356297093649085,
1771
+ "grad_norm": 0.734375,
1772
+ "learning_rate": 0.00012403349298451544,
1773
+ "loss": 1.2861,
1774
+ "step": 1260
1775
+ },
1776
+ {
1777
+ "epoch": 1.3616792249730894,
1778
+ "grad_norm": 0.6953125,
1779
+ "learning_rate": 0.00012345704133667716,
1780
+ "loss": 1.336,
1781
+ "step": 1265
1782
+ },
1783
+ {
1784
+ "epoch": 1.3670613562970937,
1785
+ "grad_norm": 0.80859375,
1786
+ "learning_rate": 0.00012287976364360427,
1787
+ "loss": 1.4538,
1788
+ "step": 1270
1789
+ },
1790
+ {
1791
+ "epoch": 1.372443487621098,
1792
+ "grad_norm": 1.1640625,
1793
+ "learning_rate": 0.00012230168023426665,
1794
+ "loss": 1.8344,
1795
+ "step": 1275
1796
+ },
1797
+ {
1798
+ "epoch": 1.3778256189451024,
1799
+ "grad_norm": 0.61328125,
1800
+ "learning_rate": 0.00012172281146600768,
1801
+ "loss": 1.1302,
1802
+ "step": 1280
1803
+ },
1804
+ {
1805
+ "epoch": 1.3832077502691065,
1806
+ "grad_norm": 0.77734375,
1807
+ "learning_rate": 0.00012114317772382736,
1808
+ "loss": 1.8846,
1809
+ "step": 1285
1810
+ },
1811
+ {
1812
+ "epoch": 1.3885898815931108,
1813
+ "grad_norm": 0.71484375,
1814
+ "learning_rate": 0.00012056279941966445,
1815
+ "loss": 1.4932,
1816
+ "step": 1290
1817
+ },
1818
+ {
1819
+ "epoch": 1.3939720129171151,
1820
+ "grad_norm": 1.0234375,
1821
+ "learning_rate": 0.00011998169699167754,
1822
+ "loss": 1.5204,
1823
+ "step": 1295
1824
+ },
1825
+ {
1826
+ "epoch": 1.3993541442411195,
1827
+ "grad_norm": 0.66796875,
1828
+ "learning_rate": 0.00011939989090352553,
1829
+ "loss": 1.6152,
1830
+ "step": 1300
1831
+ },
1832
+ {
1833
+ "epoch": 1.4047362755651238,
1834
+ "grad_norm": 0.7421875,
1835
+ "learning_rate": 0.00011881740164364683,
1836
+ "loss": 1.3754,
1837
+ "step": 1305
1838
+ },
1839
+ {
1840
+ "epoch": 1.4101184068891282,
1841
+ "grad_norm": 0.85546875,
1842
+ "learning_rate": 0.00011823424972453789,
1843
+ "loss": 1.5163,
1844
+ "step": 1310
1845
+ },
1846
+ {
1847
+ "epoch": 1.4155005382131325,
1848
+ "grad_norm": 1.453125,
1849
+ "learning_rate": 0.00011765045568203083,
1850
+ "loss": 1.7636,
1851
+ "step": 1315
1852
+ },
1853
+ {
1854
+ "epoch": 1.4208826695371366,
1855
+ "grad_norm": 0.84765625,
1856
+ "learning_rate": 0.00011706604007457043,
1857
+ "loss": 1.5983,
1858
+ "step": 1320
1859
+ },
1860
+ {
1861
+ "epoch": 1.426264800861141,
1862
+ "grad_norm": 0.83203125,
1863
+ "learning_rate": 0.00011648102348248984,
1864
+ "loss": 1.5605,
1865
+ "step": 1325
1866
+ },
1867
+ {
1868
+ "epoch": 1.4316469321851453,
1869
+ "grad_norm": 0.6953125,
1870
+ "learning_rate": 0.00011589542650728615,
1871
+ "loss": 1.3519,
1872
+ "step": 1330
1873
+ },
1874
+ {
1875
+ "epoch": 1.4370290635091496,
1876
+ "grad_norm": 0.79296875,
1877
+ "learning_rate": 0.00011530926977089475,
1878
+ "loss": 1.4494,
1879
+ "step": 1335
1880
+ },
1881
+ {
1882
+ "epoch": 1.442411194833154,
1883
+ "grad_norm": 0.796875,
1884
+ "learning_rate": 0.00011472257391496312,
1885
+ "loss": 1.4084,
1886
+ "step": 1340
1887
+ },
1888
+ {
1889
+ "epoch": 1.4477933261571583,
1890
+ "grad_norm": 0.6484375,
1891
+ "learning_rate": 0.00011413535960012404,
1892
+ "loss": 1.6266,
1893
+ "step": 1345
1894
+ },
1895
+ {
1896
+ "epoch": 1.4531754574811626,
1897
+ "grad_norm": 0.77734375,
1898
+ "learning_rate": 0.00011354764750526786,
1899
+ "loss": 1.4008,
1900
+ "step": 1350
1901
+ },
1902
+ {
1903
+ "epoch": 1.4585575888051667,
1904
+ "grad_norm": 0.83203125,
1905
+ "learning_rate": 0.00011295945832681441,
1906
+ "loss": 1.2117,
1907
+ "step": 1355
1908
+ },
1909
+ {
1910
+ "epoch": 1.4639397201291713,
1911
+ "grad_norm": 0.75,
1912
+ "learning_rate": 0.00011237081277798409,
1913
+ "loss": 1.5329,
1914
+ "step": 1360
1915
+ },
1916
+ {
1917
+ "epoch": 1.4693218514531754,
1918
+ "grad_norm": 1.515625,
1919
+ "learning_rate": 0.0001117817315880686,
1920
+ "loss": 1.2452,
1921
+ "step": 1365
1922
+ },
1923
+ {
1924
+ "epoch": 1.4747039827771797,
1925
+ "grad_norm": 1.359375,
1926
+ "learning_rate": 0.0001111922355017007,
1927
+ "loss": 1.7004,
1928
+ "step": 1370
1929
+ },
1930
+ {
1931
+ "epoch": 1.480086114101184,
1932
+ "grad_norm": 0.7578125,
1933
+ "learning_rate": 0.00011060234527812392,
1934
+ "loss": 1.5766,
1935
+ "step": 1375
1936
+ },
1937
+ {
1938
+ "epoch": 1.4854682454251884,
1939
+ "grad_norm": 0.9296875,
1940
+ "learning_rate": 0.0001100120816904614,
1941
+ "loss": 1.5775,
1942
+ "step": 1380
1943
+ },
1944
+ {
1945
+ "epoch": 1.4908503767491927,
1946
+ "grad_norm": 0.85546875,
1947
+ "learning_rate": 0.00010942146552498436,
1948
+ "loss": 1.3943,
1949
+ "step": 1385
1950
+ },
1951
+ {
1952
+ "epoch": 1.4962325080731969,
1953
+ "grad_norm": 0.86328125,
1954
+ "learning_rate": 0.00010883051758038019,
1955
+ "loss": 1.382,
1956
+ "step": 1390
1957
+ },
1958
+ {
1959
+ "epoch": 1.5016146393972014,
1960
+ "grad_norm": 0.7421875,
1961
+ "learning_rate": 0.00010823925866701987,
1962
+ "loss": 1.4657,
1963
+ "step": 1395
1964
+ },
1965
+ {
1966
+ "epoch": 1.5069967707212055,
1967
+ "grad_norm": 0.78125,
1968
+ "learning_rate": 0.00010764770960622535,
1969
+ "loss": 1.2955,
1970
+ "step": 1400
1971
+ },
1972
+ {
1973
+ "epoch": 1.5123789020452099,
1974
+ "grad_norm": 1.1875,
1975
+ "learning_rate": 0.00010705589122953601,
1976
+ "loss": 1.4443,
1977
+ "step": 1405
1978
+ },
1979
+ {
1980
+ "epoch": 1.5177610333692142,
1981
+ "grad_norm": 0.90234375,
1982
+ "learning_rate": 0.00010646382437797544,
1983
+ "loss": 1.5391,
1984
+ "step": 1410
1985
+ },
1986
+ {
1987
+ "epoch": 1.5231431646932185,
1988
+ "grad_norm": 0.84765625,
1989
+ "learning_rate": 0.00010587152990131722,
1990
+ "loss": 1.3099,
1991
+ "step": 1415
1992
+ },
1993
+ {
1994
+ "epoch": 1.5285252960172229,
1995
+ "grad_norm": 0.96484375,
1996
+ "learning_rate": 0.0001052790286573508,
1997
+ "loss": 1.616,
1998
+ "step": 1420
1999
+ },
2000
+ {
2001
+ "epoch": 1.533907427341227,
2002
+ "grad_norm": 0.75,
2003
+ "learning_rate": 0.00010468634151114707,
2004
+ "loss": 1.5043,
2005
+ "step": 1425
2006
+ },
2007
+ {
2008
+ "epoch": 1.5392895586652315,
2009
+ "grad_norm": 0.86328125,
2010
+ "learning_rate": 0.0001040934893343234,
2011
+ "loss": 1.498,
2012
+ "step": 1430
2013
+ },
2014
+ {
2015
+ "epoch": 1.5446716899892357,
2016
+ "grad_norm": 1.125,
2017
+ "learning_rate": 0.00010350049300430884,
2018
+ "loss": 1.7188,
2019
+ "step": 1435
2020
+ },
2021
+ {
2022
+ "epoch": 1.55005382131324,
2023
+ "grad_norm": 0.9296875,
2024
+ "learning_rate": 0.00010290737340360878,
2025
+ "loss": 1.7336,
2026
+ "step": 1440
2027
+ },
2028
+ {
2029
+ "epoch": 1.5554359526372443,
2030
+ "grad_norm": 0.75,
2031
+ "learning_rate": 0.00010231415141906968,
2032
+ "loss": 1.5862,
2033
+ "step": 1445
2034
+ },
2035
+ {
2036
+ "epoch": 1.5608180839612487,
2037
+ "grad_norm": 0.828125,
2038
+ "learning_rate": 0.00010172084794114336,
2039
+ "loss": 1.5656,
2040
+ "step": 1450
2041
+ },
2042
+ {
2043
+ "epoch": 1.566200215285253,
2044
+ "grad_norm": 0.90234375,
2045
+ "learning_rate": 0.00010112748386315157,
2046
+ "loss": 1.6676,
2047
+ "step": 1455
2048
+ },
2049
+ {
2050
+ "epoch": 1.571582346609257,
2051
+ "grad_norm": 0.66015625,
2052
+ "learning_rate": 0.00010053408008054999,
2053
+ "loss": 1.136,
2054
+ "step": 1460
2055
+ },
2056
+ {
2057
+ "epoch": 1.5769644779332617,
2058
+ "grad_norm": 0.63671875,
2059
+ "learning_rate": 9.994065749019263e-05,
2060
+ "loss": 1.321,
2061
+ "step": 1465
2062
+ },
2063
+ {
2064
+ "epoch": 1.5823466092572658,
2065
+ "grad_norm": 0.92578125,
2066
+ "learning_rate": 9.934723698959571e-05,
2067
+ "loss": 1.4646,
2068
+ "step": 1470
2069
+ },
2070
+ {
2071
+ "epoch": 1.5877287405812703,
2072
+ "grad_norm": 0.80859375,
2073
+ "learning_rate": 9.875383947620195e-05,
2074
+ "loss": 1.4613,
2075
+ "step": 1475
2076
+ },
2077
+ {
2078
+ "epoch": 1.5931108719052745,
2079
+ "grad_norm": 0.80078125,
2080
+ "learning_rate": 9.816048584664445e-05,
2081
+ "loss": 1.3651,
2082
+ "step": 1480
2083
+ },
2084
+ {
2085
+ "epoch": 1.5984930032292788,
2086
+ "grad_norm": 0.75390625,
2087
+ "learning_rate": 9.756719699601109e-05,
2088
+ "loss": 1.4559,
2089
+ "step": 1485
2090
+ },
2091
+ {
2092
+ "epoch": 1.6038751345532831,
2093
+ "grad_norm": 0.953125,
2094
+ "learning_rate": 9.697399381710844e-05,
2095
+ "loss": 1.6278,
2096
+ "step": 1490
2097
+ },
2098
+ {
2099
+ "epoch": 1.6092572658772875,
2100
+ "grad_norm": 0.9140625,
2101
+ "learning_rate": 9.638089719972609e-05,
2102
+ "loss": 1.7128,
2103
+ "step": 1495
2104
+ },
2105
+ {
2106
+ "epoch": 1.6146393972012918,
2107
+ "grad_norm": 0.78515625,
2108
+ "learning_rate": 9.578792802990116e-05,
2109
+ "loss": 1.5033,
2110
+ "step": 1500
2111
+ },
2112
+ {
2113
+ "epoch": 1.620021528525296,
2114
+ "grad_norm": 0.80859375,
2115
+ "learning_rate": 9.519510718918251e-05,
2116
+ "loss": 1.5971,
2117
+ "step": 1505
2118
+ },
2119
+ {
2120
+ "epoch": 1.6254036598493005,
2121
+ "grad_norm": 0.734375,
2122
+ "learning_rate": 9.460245555389575e-05,
2123
+ "loss": 1.982,
2124
+ "step": 1510
2125
+ },
2126
+ {
2127
+ "epoch": 1.6307857911733046,
2128
+ "grad_norm": 0.65234375,
2129
+ "learning_rate": 9.400999399440775e-05,
2130
+ "loss": 1.7542,
2131
+ "step": 1515
2132
+ },
2133
+ {
2134
+ "epoch": 1.636167922497309,
2135
+ "grad_norm": 0.73046875,
2136
+ "learning_rate": 9.341774337439182e-05,
2137
+ "loss": 1.7067,
2138
+ "step": 1520
2139
+ },
2140
+ {
2141
+ "epoch": 1.6415500538213132,
2142
+ "grad_norm": 1.0390625,
2143
+ "learning_rate": 9.282572455009304e-05,
2144
+ "loss": 1.8912,
2145
+ "step": 1525
2146
+ },
2147
+ {
2148
+ "epoch": 1.6469321851453176,
2149
+ "grad_norm": 1.15625,
2150
+ "learning_rate": 9.223395836959371e-05,
2151
+ "loss": 1.3994,
2152
+ "step": 1530
2153
+ },
2154
+ {
2155
+ "epoch": 1.652314316469322,
2156
+ "grad_norm": 1.0703125,
2157
+ "learning_rate": 9.164246567207927e-05,
2158
+ "loss": 1.2383,
2159
+ "step": 1535
2160
+ },
2161
+ {
2162
+ "epoch": 1.657696447793326,
2163
+ "grad_norm": 0.7265625,
2164
+ "learning_rate": 9.105126728710432e-05,
2165
+ "loss": 1.7729,
2166
+ "step": 1540
2167
+ },
2168
+ {
2169
+ "epoch": 1.6630785791173306,
2170
+ "grad_norm": 0.8046875,
2171
+ "learning_rate": 9.046038403385922e-05,
2172
+ "loss": 1.2427,
2173
+ "step": 1545
2174
+ },
2175
+ {
2176
+ "epoch": 1.6684607104413347,
2177
+ "grad_norm": 0.70703125,
2178
+ "learning_rate": 8.986983672043684e-05,
2179
+ "loss": 1.3331,
2180
+ "step": 1550
2181
+ },
2182
+ {
2183
+ "epoch": 1.673842841765339,
2184
+ "grad_norm": 0.65625,
2185
+ "learning_rate": 8.927964614309999e-05,
2186
+ "loss": 1.4483,
2187
+ "step": 1555
2188
+ },
2189
+ {
2190
+ "epoch": 1.6792249730893434,
2191
+ "grad_norm": 0.70703125,
2192
+ "learning_rate": 8.868983308554875e-05,
2193
+ "loss": 1.718,
2194
+ "step": 1560
2195
+ },
2196
+ {
2197
+ "epoch": 1.6846071044133477,
2198
+ "grad_norm": 0.77734375,
2199
+ "learning_rate": 8.810041831818889e-05,
2200
+ "loss": 1.5577,
2201
+ "step": 1565
2202
+ },
2203
+ {
2204
+ "epoch": 1.689989235737352,
2205
+ "grad_norm": 0.84375,
2206
+ "learning_rate": 8.751142259740014e-05,
2207
+ "loss": 1.4279,
2208
+ "step": 1570
2209
+ },
2210
+ {
2211
+ "epoch": 1.6953713670613562,
2212
+ "grad_norm": 1.3984375,
2213
+ "learning_rate": 8.692286666480568e-05,
2214
+ "loss": 1.7283,
2215
+ "step": 1575
2216
+ },
2217
+ {
2218
+ "epoch": 1.7007534983853607,
2219
+ "grad_norm": 0.93359375,
2220
+ "learning_rate": 8.633477124654116e-05,
2221
+ "loss": 1.596,
2222
+ "step": 1580
2223
+ },
2224
+ {
2225
+ "epoch": 1.7061356297093648,
2226
+ "grad_norm": 0.71484375,
2227
+ "learning_rate": 8.574715705252527e-05,
2228
+ "loss": 1.8197,
2229
+ "step": 1585
2230
+ },
2231
+ {
2232
+ "epoch": 1.7115177610333692,
2233
+ "grad_norm": 0.9140625,
2234
+ "learning_rate": 8.516004477573022e-05,
2235
+ "loss": 1.4374,
2236
+ "step": 1590
2237
+ },
2238
+ {
2239
+ "epoch": 1.7168998923573735,
2240
+ "grad_norm": 0.80859375,
2241
+ "learning_rate": 8.457345509145305e-05,
2242
+ "loss": 1.5237,
2243
+ "step": 1595
2244
+ },
2245
+ {
2246
+ "epoch": 1.7222820236813778,
2247
+ "grad_norm": 0.7109375,
2248
+ "learning_rate": 8.398740865658773e-05,
2249
+ "loss": 1.4578,
2250
+ "step": 1600
2251
+ },
2252
+ {
2253
+ "epoch": 1.7276641550053822,
2254
+ "grad_norm": 0.78125,
2255
+ "learning_rate": 8.340192610889741e-05,
2256
+ "loss": 1.4532,
2257
+ "step": 1605
2258
+ },
2259
+ {
2260
+ "epoch": 1.7330462863293863,
2261
+ "grad_norm": 0.80859375,
2262
+ "learning_rate": 8.28170280662879e-05,
2263
+ "loss": 1.5097,
2264
+ "step": 1610
2265
+ },
2266
+ {
2267
+ "epoch": 1.7384284176533908,
2268
+ "grad_norm": 0.9140625,
2269
+ "learning_rate": 8.223273512608146e-05,
2270
+ "loss": 1.2694,
2271
+ "step": 1615
2272
+ },
2273
+ {
2274
+ "epoch": 1.743810548977395,
2275
+ "grad_norm": 1.015625,
2276
+ "learning_rate": 8.164906786429166e-05,
2277
+ "loss": 1.5997,
2278
+ "step": 1620
2279
+ },
2280
+ {
2281
+ "epoch": 1.7491926803013995,
2282
+ "grad_norm": 0.85546875,
2283
+ "learning_rate": 8.106604683489853e-05,
2284
+ "loss": 1.6204,
2285
+ "step": 1625
2286
+ },
2287
+ {
2288
+ "epoch": 1.7545748116254036,
2289
+ "grad_norm": 1.2734375,
2290
+ "learning_rate": 8.048369256912491e-05,
2291
+ "loss": 1.5516,
2292
+ "step": 1630
2293
+ },
2294
+ {
2295
+ "epoch": 1.759956942949408,
2296
+ "grad_norm": 0.62890625,
2297
+ "learning_rate": 7.990202557471343e-05,
2298
+ "loss": 1.1211,
2299
+ "step": 1635
2300
+ },
2301
+ {
2302
+ "epoch": 1.7653390742734123,
2303
+ "grad_norm": 0.625,
2304
+ "learning_rate": 7.932106633520427e-05,
2305
+ "loss": 1.5284,
2306
+ "step": 1640
2307
+ },
2308
+ {
2309
+ "epoch": 1.7707212055974164,
2310
+ "grad_norm": 0.7890625,
2311
+ "learning_rate": 7.874083530921383e-05,
2312
+ "loss": 1.7255,
2313
+ "step": 1645
2314
+ },
2315
+ {
2316
+ "epoch": 1.776103336921421,
2317
+ "grad_norm": 0.78125,
2318
+ "learning_rate": 7.816135292971438e-05,
2319
+ "loss": 1.7802,
2320
+ "step": 1650
2321
+ },
2322
+ {
2323
+ "epoch": 1.781485468245425,
2324
+ "grad_norm": 1.2890625,
2325
+ "learning_rate": 7.758263960331438e-05,
2326
+ "loss": 1.395,
2327
+ "step": 1655
2328
+ },
2329
+ {
2330
+ "epoch": 1.7868675995694296,
2331
+ "grad_norm": 0.84375,
2332
+ "learning_rate": 7.700471570953981e-05,
2333
+ "loss": 1.4316,
2334
+ "step": 1660
2335
+ },
2336
+ {
2337
+ "epoch": 1.7922497308934338,
2338
+ "grad_norm": 0.76171875,
2339
+ "learning_rate": 7.642760160011681e-05,
2340
+ "loss": 1.7198,
2341
+ "step": 1665
2342
+ },
2343
+ {
2344
+ "epoch": 1.797631862217438,
2345
+ "grad_norm": 0.66015625,
2346
+ "learning_rate": 7.58513175982546e-05,
2347
+ "loss": 1.3174,
2348
+ "step": 1670
2349
+ },
2350
+ {
2351
+ "epoch": 1.8030139935414424,
2352
+ "grad_norm": 0.81640625,
2353
+ "learning_rate": 7.527588399792999e-05,
2354
+ "loss": 1.3408,
2355
+ "step": 1675
2356
+ },
2357
+ {
2358
+ "epoch": 1.8083961248654468,
2359
+ "grad_norm": 0.7890625,
2360
+ "learning_rate": 7.470132106317275e-05,
2361
+ "loss": 1.4783,
2362
+ "step": 1680
2363
+ },
2364
+ {
2365
+ "epoch": 1.813778256189451,
2366
+ "grad_norm": 1.0625,
2367
+ "learning_rate": 7.412764902735193e-05,
2368
+ "loss": 1.0823,
2369
+ "step": 1685
2370
+ },
2371
+ {
2372
+ "epoch": 1.8191603875134552,
2373
+ "grad_norm": 0.73046875,
2374
+ "learning_rate": 7.355488809246336e-05,
2375
+ "loss": 1.5913,
2376
+ "step": 1690
2377
+ },
2378
+ {
2379
+ "epoch": 1.8245425188374598,
2380
+ "grad_norm": 0.6953125,
2381
+ "learning_rate": 7.298305842841826e-05,
2382
+ "loss": 1.5883,
2383
+ "step": 1695
2384
+ },
2385
+ {
2386
+ "epoch": 1.8299246501614639,
2387
+ "grad_norm": 0.83984375,
2388
+ "learning_rate": 7.24121801723329e-05,
2389
+ "loss": 1.4392,
2390
+ "step": 1700
2391
+ },
2392
+ {
2393
+ "epoch": 1.8353067814854682,
2394
+ "grad_norm": 0.74609375,
2395
+ "learning_rate": 7.184227342781948e-05,
2396
+ "loss": 1.6952,
2397
+ "step": 1705
2398
+ },
2399
+ {
2400
+ "epoch": 1.8406889128094726,
2401
+ "grad_norm": 0.83203125,
2402
+ "learning_rate": 7.127335826427828e-05,
2403
+ "loss": 1.6377,
2404
+ "step": 1710
2405
+ },
2406
+ {
2407
+ "epoch": 1.846071044133477,
2408
+ "grad_norm": 1.1171875,
2409
+ "learning_rate": 7.070545471619072e-05,
2410
+ "loss": 1.5065,
2411
+ "step": 1715
2412
+ },
2413
+ {
2414
+ "epoch": 1.8514531754574812,
2415
+ "grad_norm": 0.66796875,
2416
+ "learning_rate": 7.013858278241399e-05,
2417
+ "loss": 1.3861,
2418
+ "step": 1720
2419
+ },
2420
+ {
2421
+ "epoch": 1.8568353067814853,
2422
+ "grad_norm": 0.86328125,
2423
+ "learning_rate": 6.957276242547672e-05,
2424
+ "loss": 1.2102,
2425
+ "step": 1725
2426
+ },
2427
+ {
2428
+ "epoch": 1.86221743810549,
2429
+ "grad_norm": 1.234375,
2430
+ "learning_rate": 6.9008013570876e-05,
2431
+ "loss": 1.5305,
2432
+ "step": 1730
2433
+ },
2434
+ {
2435
+ "epoch": 1.867599569429494,
2436
+ "grad_norm": 0.7109375,
2437
+ "learning_rate": 6.844435610637583e-05,
2438
+ "loss": 1.2648,
2439
+ "step": 1735
2440
+ },
2441
+ {
2442
+ "epoch": 1.8729817007534983,
2443
+ "grad_norm": 0.79296875,
2444
+ "learning_rate": 6.78818098813065e-05,
2445
+ "loss": 1.3824,
2446
+ "step": 1740
2447
+ },
2448
+ {
2449
+ "epoch": 1.8783638320775027,
2450
+ "grad_norm": 0.7421875,
2451
+ "learning_rate": 6.732039470586577e-05,
2452
+ "loss": 1.7781,
2453
+ "step": 1745
2454
+ },
2455
+ {
2456
+ "epoch": 1.883745963401507,
2457
+ "grad_norm": 0.7421875,
2458
+ "learning_rate": 6.67601303504213e-05,
2459
+ "loss": 1.2321,
2460
+ "step": 1750
2461
+ },
2462
+ {
2463
+ "epoch": 1.8891280947255114,
2464
+ "grad_norm": 1.1015625,
2465
+ "learning_rate": 6.620103654481429e-05,
2466
+ "loss": 1.4159,
2467
+ "step": 1755
2468
+ },
2469
+ {
2470
+ "epoch": 1.8945102260495155,
2471
+ "grad_norm": 0.96484375,
2472
+ "learning_rate": 6.56431329776648e-05,
2473
+ "loss": 1.4128,
2474
+ "step": 1760
2475
+ },
2476
+ {
2477
+ "epoch": 1.89989235737352,
2478
+ "grad_norm": 0.9140625,
2479
+ "learning_rate": 6.508643929567828e-05,
2480
+ "loss": 1.4293,
2481
+ "step": 1765
2482
+ },
2483
+ {
2484
+ "epoch": 1.9052744886975241,
2485
+ "grad_norm": 0.94140625,
2486
+ "learning_rate": 6.453097510295387e-05,
2487
+ "loss": 2.0126,
2488
+ "step": 1770
2489
+ },
2490
+ {
2491
+ "epoch": 1.9106566200215285,
2492
+ "grad_norm": 0.73828125,
2493
+ "learning_rate": 6.397675996029388e-05,
2494
+ "loss": 1.3171,
2495
+ "step": 1775
2496
+ },
2497
+ {
2498
+ "epoch": 1.9160387513455328,
2499
+ "grad_norm": 0.875,
2500
+ "learning_rate": 6.342381338451513e-05,
2501
+ "loss": 1.5463,
2502
+ "step": 1780
2503
+ },
2504
+ {
2505
+ "epoch": 1.9214208826695371,
2506
+ "grad_norm": 0.85546875,
2507
+ "learning_rate": 6.28721548477615e-05,
2508
+ "loss": 1.4422,
2509
+ "step": 1785
2510
+ },
2511
+ {
2512
+ "epoch": 1.9268030139935415,
2513
+ "grad_norm": 0.6796875,
2514
+ "learning_rate": 6.232180377681822e-05,
2515
+ "loss": 1.703,
2516
+ "step": 1790
2517
+ },
2518
+ {
2519
+ "epoch": 1.9321851453175456,
2520
+ "grad_norm": 0.6953125,
2521
+ "learning_rate": 6.177277955242786e-05,
2522
+ "loss": 1.5379,
2523
+ "step": 1795
2524
+ },
2525
+ {
2526
+ "epoch": 1.9375672766415502,
2527
+ "grad_norm": 1.40625,
2528
+ "learning_rate": 6.122510150860785e-05,
2529
+ "loss": 1.3108,
2530
+ "step": 1800
2531
+ },
2532
+ {
2533
+ "epoch": 1.9429494079655543,
2534
+ "grad_norm": 0.97265625,
2535
+ "learning_rate": 6.067878893196941e-05,
2536
+ "loss": 1.4381,
2537
+ "step": 1805
2538
+ },
2539
+ {
2540
+ "epoch": 1.9483315392895588,
2541
+ "grad_norm": 0.859375,
2542
+ "learning_rate": 6.01338610610386e-05,
2543
+ "loss": 1.5073,
2544
+ "step": 1810
2545
+ },
2546
+ {
2547
+ "epoch": 1.953713670613563,
2548
+ "grad_norm": 0.94921875,
2549
+ "learning_rate": 5.9590337085578726e-05,
2550
+ "loss": 1.5662,
2551
+ "step": 1815
2552
+ },
2553
+ {
2554
+ "epoch": 1.9590958019375673,
2555
+ "grad_norm": 0.91015625,
2556
+ "learning_rate": 5.9048236145914524e-05,
2557
+ "loss": 1.5529,
2558
+ "step": 1820
2559
+ },
2560
+ {
2561
+ "epoch": 1.9644779332615716,
2562
+ "grad_norm": 1.859375,
2563
+ "learning_rate": 5.8507577332258356e-05,
2564
+ "loss": 1.4747,
2565
+ "step": 1825
2566
+ },
2567
+ {
2568
+ "epoch": 1.969860064585576,
2569
+ "grad_norm": 0.8359375,
2570
+ "learning_rate": 5.7968379684037565e-05,
2571
+ "loss": 1.641,
2572
+ "step": 1830
2573
+ },
2574
+ {
2575
+ "epoch": 1.9752421959095803,
2576
+ "grad_norm": 1.078125,
2577
+ "learning_rate": 5.743066218922439e-05,
2578
+ "loss": 1.6829,
2579
+ "step": 1835
2580
+ },
2581
+ {
2582
+ "epoch": 1.9806243272335844,
2583
+ "grad_norm": 1.0,
2584
+ "learning_rate": 5.689444378366696e-05,
2585
+ "loss": 2.097,
2586
+ "step": 1840
2587
+ },
2588
+ {
2589
+ "epoch": 1.986006458557589,
2590
+ "grad_norm": 0.7421875,
2591
+ "learning_rate": 5.635974335042275e-05,
2592
+ "loss": 1.2503,
2593
+ "step": 1845
2594
+ },
2595
+ {
2596
+ "epoch": 1.991388589881593,
2597
+ "grad_norm": 0.99609375,
2598
+ "learning_rate": 5.582657971909342e-05,
2599
+ "loss": 1.3495,
2600
+ "step": 1850
2601
+ },
2602
+ {
2603
+ "epoch": 1.9967707212055974,
2604
+ "grad_norm": 0.84765625,
2605
+ "learning_rate": 5.529497166516184e-05,
2606
+ "loss": 1.3844,
2607
+ "step": 1855
2608
+ },
2609
+ {
2610
+ "epoch": 2.0021528525296017,
2611
+ "grad_norm": 0.6015625,
2612
+ "learning_rate": 5.4764937909330746e-05,
2613
+ "loss": 1.4768,
2614
+ "step": 1860
2615
+ },
2616
+ {
2617
+ "epoch": 2.007534983853606,
2618
+ "grad_norm": 0.703125,
2619
+ "learning_rate": 5.4236497116863726e-05,
2620
+ "loss": 1.4405,
2621
+ "step": 1865
2622
+ },
2623
+ {
2624
+ "epoch": 2.0129171151776104,
2625
+ "grad_norm": 0.9765625,
2626
+ "learning_rate": 5.370966789692771e-05,
2627
+ "loss": 1.5051,
2628
+ "step": 1870
2629
+ },
2630
+ {
2631
+ "epoch": 2.0182992465016145,
2632
+ "grad_norm": 0.79296875,
2633
+ "learning_rate": 5.318446880193779e-05,
2634
+ "loss": 1.2863,
2635
+ "step": 1875
2636
+ },
2637
+ {
2638
+ "epoch": 2.023681377825619,
2639
+ "grad_norm": 1.1328125,
2640
+ "learning_rate": 5.26609183269037e-05,
2641
+ "loss": 1.6646,
2642
+ "step": 1880
2643
+ },
2644
+ {
2645
+ "epoch": 2.029063509149623,
2646
+ "grad_norm": 0.9375,
2647
+ "learning_rate": 5.213903490877874e-05,
2648
+ "loss": 1.5516,
2649
+ "step": 1885
2650
+ },
2651
+ {
2652
+ "epoch": 2.0344456404736277,
2653
+ "grad_norm": 0.796875,
2654
+ "learning_rate": 5.161883692581038e-05,
2655
+ "loss": 1.5062,
2656
+ "step": 1890
2657
+ },
2658
+ {
2659
+ "epoch": 2.039827771797632,
2660
+ "grad_norm": 0.94921875,
2661
+ "learning_rate": 5.11003426968931e-05,
2662
+ "loss": 1.2587,
2663
+ "step": 1895
2664
+ },
2665
+ {
2666
+ "epoch": 2.045209903121636,
2667
+ "grad_norm": 1.015625,
2668
+ "learning_rate": 5.0583570480923325e-05,
2669
+ "loss": 1.2944,
2670
+ "step": 1900
2671
+ },
2672
+ {
2673
+ "epoch": 2.0505920344456405,
2674
+ "grad_norm": 0.89453125,
2675
+ "learning_rate": 5.0068538476156225e-05,
2676
+ "loss": 1.3802,
2677
+ "step": 1905
2678
+ },
2679
+ {
2680
+ "epoch": 2.0559741657696446,
2681
+ "grad_norm": 0.75,
2682
+ "learning_rate": 4.9555264819565205e-05,
2683
+ "loss": 1.4541,
2684
+ "step": 1910
2685
+ },
2686
+ {
2687
+ "epoch": 2.061356297093649,
2688
+ "grad_norm": 0.5234375,
2689
+ "learning_rate": 4.904376758620293e-05,
2690
+ "loss": 1.4703,
2691
+ "step": 1915
2692
+ },
2693
+ {
2694
+ "epoch": 2.0667384284176533,
2695
+ "grad_norm": 1.1171875,
2696
+ "learning_rate": 4.853406478856498e-05,
2697
+ "loss": 1.5858,
2698
+ "step": 1920
2699
+ },
2700
+ {
2701
+ "epoch": 2.072120559741658,
2702
+ "grad_norm": 1.1875,
2703
+ "learning_rate": 4.8026174375955336e-05,
2704
+ "loss": 1.3889,
2705
+ "step": 1925
2706
+ },
2707
+ {
2708
+ "epoch": 2.077502691065662,
2709
+ "grad_norm": 0.5703125,
2710
+ "learning_rate": 4.75201142338545e-05,
2711
+ "loss": 1.2628,
2712
+ "step": 1930
2713
+ },
2714
+ {
2715
+ "epoch": 2.082884822389666,
2716
+ "grad_norm": 0.9140625,
2717
+ "learning_rate": 4.701590218328955e-05,
2718
+ "loss": 1.9168,
2719
+ "step": 1935
2720
+ },
2721
+ {
2722
+ "epoch": 2.0882669537136707,
2723
+ "grad_norm": 0.9609375,
2724
+ "learning_rate": 4.6513555980206643e-05,
2725
+ "loss": 1.6785,
2726
+ "step": 1940
2727
+ },
2728
+ {
2729
+ "epoch": 2.0936490850376748,
2730
+ "grad_norm": 0.70703125,
2731
+ "learning_rate": 4.601309331484554e-05,
2732
+ "loss": 1.6131,
2733
+ "step": 1945
2734
+ },
2735
+ {
2736
+ "epoch": 2.0990312163616793,
2737
+ "grad_norm": 1.265625,
2738
+ "learning_rate": 4.551453181111688e-05,
2739
+ "loss": 1.6658,
2740
+ "step": 1950
2741
+ },
2742
+ {
2743
+ "epoch": 2.1044133476856834,
2744
+ "grad_norm": 0.75,
2745
+ "learning_rate": 4.501788902598143e-05,
2746
+ "loss": 1.6256,
2747
+ "step": 1955
2748
+ },
2749
+ {
2750
+ "epoch": 2.109795479009688,
2751
+ "grad_norm": 0.96875,
2752
+ "learning_rate": 4.4523182448831834e-05,
2753
+ "loss": 1.8338,
2754
+ "step": 1960
2755
+ },
2756
+ {
2757
+ "epoch": 2.115177610333692,
2758
+ "grad_norm": 0.69140625,
2759
+ "learning_rate": 4.4030429500876735e-05,
2760
+ "loss": 1.5877,
2761
+ "step": 1965
2762
+ },
2763
+ {
2764
+ "epoch": 2.1205597416576962,
2765
+ "grad_norm": 0.7734375,
2766
+ "learning_rate": 4.353964753452716e-05,
2767
+ "loss": 1.3338,
2768
+ "step": 1970
2769
+ },
2770
+ {
2771
+ "epoch": 2.125941872981701,
2772
+ "grad_norm": 0.82421875,
2773
+ "learning_rate": 4.305085383278567e-05,
2774
+ "loss": 1.4618,
2775
+ "step": 1975
2776
+ },
2777
+ {
2778
+ "epoch": 2.131324004305705,
2779
+ "grad_norm": 0.75,
2780
+ "learning_rate": 4.256406560863758e-05,
2781
+ "loss": 1.5282,
2782
+ "step": 1980
2783
+ },
2784
+ {
2785
+ "epoch": 2.1367061356297095,
2786
+ "grad_norm": 0.78125,
2787
+ "learning_rate": 4.2079300004444865e-05,
2788
+ "loss": 1.4044,
2789
+ "step": 1985
2790
+ },
2791
+ {
2792
+ "epoch": 2.1420882669537136,
2793
+ "grad_norm": 0.62890625,
2794
+ "learning_rate": 4.1596574091342376e-05,
2795
+ "loss": 1.4315,
2796
+ "step": 1990
2797
+ },
2798
+ {
2799
+ "epoch": 2.147470398277718,
2800
+ "grad_norm": 0.98046875,
2801
+ "learning_rate": 4.1115904868636856e-05,
2802
+ "loss": 1.4956,
2803
+ "step": 1995
2804
+ },
2805
+ {
2806
+ "epoch": 2.1528525296017222,
2807
+ "grad_norm": 0.73828125,
2808
+ "learning_rate": 4.063730926320818e-05,
2809
+ "loss": 1.4915,
2810
+ "step": 2000
2811
+ },
2812
+ {
2813
+ "epoch": 2.158234660925727,
2814
+ "grad_norm": 0.77734375,
2815
+ "learning_rate": 4.016080412891328e-05,
2816
+ "loss": 1.3758,
2817
+ "step": 2005
2818
+ },
2819
+ {
2820
+ "epoch": 2.163616792249731,
2821
+ "grad_norm": 0.67578125,
2822
+ "learning_rate": 3.968640624599272e-05,
2823
+ "loss": 1.4847,
2824
+ "step": 2010
2825
+ },
2826
+ {
2827
+ "epoch": 2.168998923573735,
2828
+ "grad_norm": 1.0,
2829
+ "learning_rate": 3.9214132320479614e-05,
2830
+ "loss": 1.6858,
2831
+ "step": 2015
2832
+ },
2833
+ {
2834
+ "epoch": 2.1743810548977396,
2835
+ "grad_norm": 0.6328125,
2836
+ "learning_rate": 3.874399898361146e-05,
2837
+ "loss": 1.3079,
2838
+ "step": 2020
2839
+ },
2840
+ {
2841
+ "epoch": 2.1797631862217437,
2842
+ "grad_norm": 0.99609375,
2843
+ "learning_rate": 3.827602279124447e-05,
2844
+ "loss": 1.6033,
2845
+ "step": 2025
2846
+ },
2847
+ {
2848
+ "epoch": 2.1851453175457483,
2849
+ "grad_norm": 0.80859375,
2850
+ "learning_rate": 3.7810220223270466e-05,
2851
+ "loss": 1.3613,
2852
+ "step": 2030
2853
+ },
2854
+ {
2855
+ "epoch": 2.1905274488697524,
2856
+ "grad_norm": 0.77734375,
2857
+ "learning_rate": 3.7346607683036525e-05,
2858
+ "loss": 1.3334,
2859
+ "step": 2035
2860
+ },
2861
+ {
2862
+ "epoch": 2.195909580193757,
2863
+ "grad_norm": 0.9609375,
2864
+ "learning_rate": 3.688520149676748e-05,
2865
+ "loss": 1.2178,
2866
+ "step": 2040
2867
+ },
2868
+ {
2869
+ "epoch": 2.201291711517761,
2870
+ "grad_norm": 1.28125,
2871
+ "learning_rate": 3.642601791299076e-05,
2872
+ "loss": 1.4418,
2873
+ "step": 2045
2874
+ },
2875
+ {
2876
+ "epoch": 2.206673842841765,
2877
+ "grad_norm": 0.65234375,
2878
+ "learning_rate": 3.596907310196451e-05,
2879
+ "loss": 1.2051,
2880
+ "step": 2050
2881
+ },
2882
+ {
2883
+ "epoch": 2.2120559741657697,
2884
+ "grad_norm": 0.78125,
2885
+ "learning_rate": 3.551438315510778e-05,
2886
+ "loss": 1.6727,
2887
+ "step": 2055
2888
+ },
2889
+ {
2890
+ "epoch": 2.217438105489774,
2891
+ "grad_norm": 0.75,
2892
+ "learning_rate": 3.506196408443414e-05,
2893
+ "loss": 1.6572,
2894
+ "step": 2060
2895
+ },
2896
+ {
2897
+ "epoch": 2.2228202368137784,
2898
+ "grad_norm": 0.77734375,
2899
+ "learning_rate": 3.4611831821987726e-05,
2900
+ "loss": 1.2514,
2901
+ "step": 2065
2902
+ },
2903
+ {
2904
+ "epoch": 2.2282023681377825,
2905
+ "grad_norm": 1.4375,
2906
+ "learning_rate": 3.4164002219282145e-05,
2907
+ "loss": 1.8706,
2908
+ "step": 2070
2909
+ },
2910
+ {
2911
+ "epoch": 2.233584499461787,
2912
+ "grad_norm": 0.625,
2913
+ "learning_rate": 3.371849104674234e-05,
2914
+ "loss": 1.4124,
2915
+ "step": 2075
2916
+ },
2917
+ {
2918
+ "epoch": 2.238966630785791,
2919
+ "grad_norm": 1.0703125,
2920
+ "learning_rate": 3.327531399314914e-05,
2921
+ "loss": 1.528,
2922
+ "step": 2080
2923
+ },
2924
+ {
2925
+ "epoch": 2.2443487621097953,
2926
+ "grad_norm": 0.8984375,
2927
+ "learning_rate": 3.283448666508687e-05,
2928
+ "loss": 1.4631,
2929
+ "step": 2085
2930
+ },
2931
+ {
2932
+ "epoch": 2.2497308934338,
2933
+ "grad_norm": 0.89453125,
2934
+ "learning_rate": 3.239602458639364e-05,
2935
+ "loss": 1.0781,
2936
+ "step": 2090
2937
+ },
2938
+ {
2939
+ "epoch": 2.255113024757804,
2940
+ "grad_norm": 1.140625,
2941
+ "learning_rate": 3.195994319761492e-05,
2942
+ "loss": 1.6776,
2943
+ "step": 2095
2944
+ },
2945
+ {
2946
+ "epoch": 2.2604951560818085,
2947
+ "grad_norm": 0.8125,
2948
+ "learning_rate": 3.152625785545942e-05,
2949
+ "loss": 1.3804,
2950
+ "step": 2100
2951
+ },
2952
+ {
2953
+ "epoch": 2.2658772874058126,
2954
+ "grad_norm": 0.7421875,
2955
+ "learning_rate": 3.1094983832258605e-05,
2956
+ "loss": 1.6414,
2957
+ "step": 2105
2958
+ },
2959
+ {
2960
+ "epoch": 2.271259418729817,
2961
+ "grad_norm": 0.7890625,
2962
+ "learning_rate": 3.0666136315428794e-05,
2963
+ "loss": 1.3944,
2964
+ "step": 2110
2965
+ },
2966
+ {
2967
+ "epoch": 2.2766415500538213,
2968
+ "grad_norm": 0.96484375,
2969
+ "learning_rate": 3.0239730406936295e-05,
2970
+ "loss": 1.5675,
2971
+ "step": 2115
2972
+ },
2973
+ {
2974
+ "epoch": 2.282023681377826,
2975
+ "grad_norm": 0.87890625,
2976
+ "learning_rate": 2.9815781122765528e-05,
2977
+ "loss": 1.2669,
2978
+ "step": 2120
2979
+ },
2980
+ {
2981
+ "epoch": 2.28740581270183,
2982
+ "grad_norm": 1.546875,
2983
+ "learning_rate": 2.939430339239041e-05,
2984
+ "loss": 1.491,
2985
+ "step": 2125
2986
+ },
2987
+ {
2988
+ "epoch": 2.292787944025834,
2989
+ "grad_norm": 0.68359375,
2990
+ "learning_rate": 2.8975312058248504e-05,
2991
+ "loss": 1.7579,
2992
+ "step": 2130
2993
+ },
2994
+ {
2995
+ "epoch": 2.2981700753498386,
2996
+ "grad_norm": 0.64453125,
2997
+ "learning_rate": 2.855882187521821e-05,
2998
+ "loss": 1.3731,
2999
+ "step": 2135
3000
+ },
3001
+ {
3002
+ "epoch": 2.3035522066738428,
3003
+ "grad_norm": 0.57421875,
3004
+ "learning_rate": 2.8144847510099538e-05,
3005
+ "loss": 1.465,
3006
+ "step": 2140
3007
+ },
3008
+ {
3009
+ "epoch": 2.3089343379978473,
3010
+ "grad_norm": 0.859375,
3011
+ "learning_rate": 2.7733403541097148e-05,
3012
+ "loss": 1.5397,
3013
+ "step": 2145
3014
+ },
3015
+ {
3016
+ "epoch": 2.3143164693218514,
3017
+ "grad_norm": 1.4140625,
3018
+ "learning_rate": 2.7324504457307355e-05,
3019
+ "loss": 1.5433,
3020
+ "step": 2150
3021
+ },
3022
+ {
3023
+ "epoch": 2.3196986006458555,
3024
+ "grad_norm": 0.62890625,
3025
+ "learning_rate": 2.69181646582076e-05,
3026
+ "loss": 1.4736,
3027
+ "step": 2155
3028
+ },
3029
+ {
3030
+ "epoch": 2.32508073196986,
3031
+ "grad_norm": 0.68359375,
3032
+ "learning_rate": 2.6514398453149714e-05,
3033
+ "loss": 1.482,
3034
+ "step": 2160
3035
+ },
3036
+ {
3037
+ "epoch": 2.330462863293864,
3038
+ "grad_norm": 0.75390625,
3039
+ "learning_rate": 2.6113220060855615e-05,
3040
+ "loss": 1.1509,
3041
+ "step": 2165
3042
+ },
3043
+ {
3044
+ "epoch": 2.3358449946178688,
3045
+ "grad_norm": 0.5625,
3046
+ "learning_rate": 2.571464360891689e-05,
3047
+ "loss": 1.4047,
3048
+ "step": 2170
3049
+ },
3050
+ {
3051
+ "epoch": 2.341227125941873,
3052
+ "grad_norm": 0.8046875,
3053
+ "learning_rate": 2.531868313329717e-05,
3054
+ "loss": 1.4702,
3055
+ "step": 2175
3056
+ },
3057
+ {
3058
+ "epoch": 2.3466092572658774,
3059
+ "grad_norm": 0.671875,
3060
+ "learning_rate": 2.4925352577837778e-05,
3061
+ "loss": 1.3064,
3062
+ "step": 2180
3063
+ },
3064
+ {
3065
+ "epoch": 2.3519913885898815,
3066
+ "grad_norm": 0.63671875,
3067
+ "learning_rate": 2.4534665793766964e-05,
3068
+ "loss": 1.2348,
3069
+ "step": 2185
3070
+ },
3071
+ {
3072
+ "epoch": 2.357373519913886,
3073
+ "grad_norm": 0.81640625,
3074
+ "learning_rate": 2.4146636539211754e-05,
3075
+ "loss": 1.6612,
3076
+ "step": 2190
3077
+ },
3078
+ {
3079
+ "epoch": 2.36275565123789,
3080
+ "grad_norm": 0.6875,
3081
+ "learning_rate": 2.3761278478713754e-05,
3082
+ "loss": 1.3195,
3083
+ "step": 2195
3084
+ },
3085
+ {
3086
+ "epoch": 2.3681377825618943,
3087
+ "grad_norm": 0.77734375,
3088
+ "learning_rate": 2.337860518274776e-05,
3089
+ "loss": 1.3896,
3090
+ "step": 2200
3091
+ },
3092
+ {
3093
+ "epoch": 2.373519913885899,
3094
+ "grad_norm": 0.84765625,
3095
+ "learning_rate": 2.2998630127244104e-05,
3096
+ "loss": 1.7125,
3097
+ "step": 2205
3098
+ },
3099
+ {
3100
+ "epoch": 2.378902045209903,
3101
+ "grad_norm": 0.640625,
3102
+ "learning_rate": 2.2621366693113766e-05,
3103
+ "loss": 1.3499,
3104
+ "step": 2210
3105
+ },
3106
+ {
3107
+ "epoch": 2.3842841765339076,
3108
+ "grad_norm": 0.984375,
3109
+ "learning_rate": 2.224682816577747e-05,
3110
+ "loss": 1.2244,
3111
+ "step": 2215
3112
+ },
3113
+ {
3114
+ "epoch": 2.3896663078579117,
3115
+ "grad_norm": 0.9765625,
3116
+ "learning_rate": 2.187502773469762e-05,
3117
+ "loss": 1.8846,
3118
+ "step": 2220
3119
+ },
3120
+ {
3121
+ "epoch": 2.395048439181916,
3122
+ "grad_norm": 0.84765625,
3123
+ "learning_rate": 2.1505978492913958e-05,
3124
+ "loss": 1.4663,
3125
+ "step": 2225
3126
+ },
3127
+ {
3128
+ "epoch": 2.4004305705059203,
3129
+ "grad_norm": 1.3203125,
3130
+ "learning_rate": 2.1139693436582453e-05,
3131
+ "loss": 1.7524,
3132
+ "step": 2230
3133
+ },
3134
+ {
3135
+ "epoch": 2.4058127018299245,
3136
+ "grad_norm": 0.8671875,
3137
+ "learning_rate": 2.0776185464517627e-05,
3138
+ "loss": 1.5157,
3139
+ "step": 2235
3140
+ },
3141
+ {
3142
+ "epoch": 2.411194833153929,
3143
+ "grad_norm": 0.7734375,
3144
+ "learning_rate": 2.041546737773834e-05,
3145
+ "loss": 1.1478,
3146
+ "step": 2240
3147
+ },
3148
+ {
3149
+ "epoch": 2.416576964477933,
3150
+ "grad_norm": 0.80078125,
3151
+ "learning_rate": 2.005755187901689e-05,
3152
+ "loss": 1.5565,
3153
+ "step": 2245
3154
+ },
3155
+ {
3156
+ "epoch": 2.4219590958019377,
3157
+ "grad_norm": 0.6328125,
3158
+ "learning_rate": 1.9702451572431946e-05,
3159
+ "loss": 1.1969,
3160
+ "step": 2250
3161
+ },
3162
+ {
3163
+ "epoch": 2.427341227125942,
3164
+ "grad_norm": 0.609375,
3165
+ "learning_rate": 1.935017896292436e-05,
3166
+ "loss": 1.1778,
3167
+ "step": 2255
3168
+ },
3169
+ {
3170
+ "epoch": 2.4327233584499464,
3171
+ "grad_norm": 0.84765625,
3172
+ "learning_rate": 1.9000746455857078e-05,
3173
+ "loss": 1.3546,
3174
+ "step": 2260
3175
+ },
3176
+ {
3177
+ "epoch": 2.4381054897739505,
3178
+ "grad_norm": 0.921875,
3179
+ "learning_rate": 1.8654166356578074e-05,
3180
+ "loss": 1.173,
3181
+ "step": 2265
3182
+ },
3183
+ {
3184
+ "epoch": 2.4434876210979546,
3185
+ "grad_norm": 0.59765625,
3186
+ "learning_rate": 1.8310450869987195e-05,
3187
+ "loss": 1.2932,
3188
+ "step": 2270
3189
+ },
3190
+ {
3191
+ "epoch": 2.448869752421959,
3192
+ "grad_norm": 1.0078125,
3193
+ "learning_rate": 1.7969612100106238e-05,
3194
+ "loss": 1.3395,
3195
+ "step": 2275
3196
+ },
3197
+ {
3198
+ "epoch": 2.4542518837459633,
3199
+ "grad_norm": 1.0234375,
3200
+ "learning_rate": 1.763166204965275e-05,
3201
+ "loss": 1.3869,
3202
+ "step": 2280
3203
+ },
3204
+ {
3205
+ "epoch": 2.459634015069968,
3206
+ "grad_norm": 1.8203125,
3207
+ "learning_rate": 1.729661261961737e-05,
3208
+ "loss": 1.5892,
3209
+ "step": 2285
3210
+ },
3211
+ {
3212
+ "epoch": 2.465016146393972,
3213
+ "grad_norm": 0.65625,
3214
+ "learning_rate": 1.696447560884462e-05,
3215
+ "loss": 1.1645,
3216
+ "step": 2290
3217
+ },
3218
+ {
3219
+ "epoch": 2.4703982777179765,
3220
+ "grad_norm": 0.5859375,
3221
+ "learning_rate": 1.6635262713617583e-05,
3222
+ "loss": 1.5684,
3223
+ "step": 2295
3224
+ },
3225
+ {
3226
+ "epoch": 2.4757804090419806,
3227
+ "grad_norm": 0.921875,
3228
+ "learning_rate": 1.6308985527245892e-05,
3229
+ "loss": 1.5329,
3230
+ "step": 2300
3231
+ },
3232
+ {
3233
+ "epoch": 2.481162540365985,
3234
+ "grad_norm": 1.0625,
3235
+ "learning_rate": 1.5985655539657517e-05,
3236
+ "loss": 1.6929,
3237
+ "step": 2305
3238
+ },
3239
+ {
3240
+ "epoch": 2.4865446716899893,
3241
+ "grad_norm": 1.109375,
3242
+ "learning_rate": 1.5665284136994084e-05,
3243
+ "loss": 1.7067,
3244
+ "step": 2310
3245
+ },
3246
+ {
3247
+ "epoch": 2.4919268030139934,
3248
+ "grad_norm": 1.0546875,
3249
+ "learning_rate": 1.5347882601209983e-05,
3250
+ "loss": 1.4725,
3251
+ "step": 2315
3252
+ },
3253
+ {
3254
+ "epoch": 2.497308934337998,
3255
+ "grad_norm": 0.61328125,
3256
+ "learning_rate": 1.5033462109675056e-05,
3257
+ "loss": 1.4677,
3258
+ "step": 2320
3259
+ },
3260
+ {
3261
+ "epoch": 2.502691065662002,
3262
+ "grad_norm": 0.55078125,
3263
+ "learning_rate": 1.4722033734780983e-05,
3264
+ "loss": 1.2087,
3265
+ "step": 2325
3266
+ },
3267
+ {
3268
+ "epoch": 2.5080731969860066,
3269
+ "grad_norm": 0.86328125,
3270
+ "learning_rate": 1.4413608443551263e-05,
3271
+ "loss": 1.1739,
3272
+ "step": 2330
3273
+ },
3274
+ {
3275
+ "epoch": 2.5134553283100107,
3276
+ "grad_norm": 0.71875,
3277
+ "learning_rate": 1.4108197097255193e-05,
3278
+ "loss": 1.2955,
3279
+ "step": 2335
3280
+ },
3281
+ {
3282
+ "epoch": 2.518837459634015,
3283
+ "grad_norm": 0.78515625,
3284
+ "learning_rate": 1.3805810451025258e-05,
3285
+ "loss": 1.2485,
3286
+ "step": 2340
3287
+ },
3288
+ {
3289
+ "epoch": 2.5242195909580194,
3290
+ "grad_norm": 0.84765625,
3291
+ "learning_rate": 1.3506459153478412e-05,
3292
+ "loss": 1.4608,
3293
+ "step": 2345
3294
+ },
3295
+ {
3296
+ "epoch": 2.529601722282024,
3297
+ "grad_norm": 0.796875,
3298
+ "learning_rate": 1.3210153746341103e-05,
3299
+ "loss": 1.2995,
3300
+ "step": 2350
3301
+ },
3302
+ {
3303
+ "epoch": 2.534983853606028,
3304
+ "grad_norm": 0.609375,
3305
+ "learning_rate": 1.2916904664078e-05,
3306
+ "loss": 1.2907,
3307
+ "step": 2355
3308
+ },
3309
+ {
3310
+ "epoch": 2.540365984930032,
3311
+ "grad_norm": 0.66015625,
3312
+ "learning_rate": 1.262672223352459e-05,
3313
+ "loss": 1.6971,
3314
+ "step": 2360
3315
+ },
3316
+ {
3317
+ "epoch": 2.5457481162540367,
3318
+ "grad_norm": 0.8125,
3319
+ "learning_rate": 1.2339616673523523e-05,
3320
+ "loss": 1.4335,
3321
+ "step": 2365
3322
+ },
3323
+ {
3324
+ "epoch": 2.551130247578041,
3325
+ "grad_norm": 0.71484375,
3326
+ "learning_rate": 1.2055598094564713e-05,
3327
+ "loss": 1.5192,
3328
+ "step": 2370
3329
+ },
3330
+ {
3331
+ "epoch": 2.5565123789020454,
3332
+ "grad_norm": 0.79296875,
3333
+ "learning_rate": 1.177467649842927e-05,
3334
+ "loss": 1.8432,
3335
+ "step": 2375
3336
+ },
3337
+ {
3338
+ "epoch": 2.5618945102260495,
3339
+ "grad_norm": 0.77734375,
3340
+ "learning_rate": 1.1496861777837364e-05,
3341
+ "loss": 1.6651,
3342
+ "step": 2380
3343
+ },
3344
+ {
3345
+ "epoch": 2.5672766415500536,
3346
+ "grad_norm": 0.83203125,
3347
+ "learning_rate": 1.122216371609982e-05,
3348
+ "loss": 1.5468,
3349
+ "step": 2385
3350
+ },
3351
+ {
3352
+ "epoch": 2.572658772874058,
3353
+ "grad_norm": 0.88671875,
3354
+ "learning_rate": 1.0950591986773584e-05,
3355
+ "loss": 1.6867,
3356
+ "step": 2390
3357
+ },
3358
+ {
3359
+ "epoch": 2.5780409041980623,
3360
+ "grad_norm": 0.90234375,
3361
+ "learning_rate": 1.068215615332101e-05,
3362
+ "loss": 1.3598,
3363
+ "step": 2395
3364
+ },
3365
+ {
3366
+ "epoch": 2.583423035522067,
3367
+ "grad_norm": 0.640625,
3368
+ "learning_rate": 1.0416865668773223e-05,
3369
+ "loss": 1.6012,
3370
+ "step": 2400
3371
+ },
3372
+ {
3373
+ "epoch": 2.588805166846071,
3374
+ "grad_norm": 0.8359375,
3375
+ "learning_rate": 1.0154729875397096e-05,
3376
+ "loss": 1.5018,
3377
+ "step": 2405
3378
+ },
3379
+ {
3380
+ "epoch": 2.594187298170075,
3381
+ "grad_norm": 0.85546875,
3382
+ "learning_rate": 9.895758004366329e-06,
3383
+ "loss": 1.517,
3384
+ "step": 2410
3385
+ },
3386
+ {
3387
+ "epoch": 2.5995694294940797,
3388
+ "grad_norm": 0.625,
3389
+ "learning_rate": 9.639959175436375e-06,
3390
+ "loss": 1.4521,
3391
+ "step": 2415
3392
+ },
3393
+ {
3394
+ "epoch": 2.604951560818084,
3395
+ "grad_norm": 0.78515625,
3396
+ "learning_rate": 9.387342396623188e-06,
3397
+ "loss": 1.3317,
3398
+ "step": 2420
3399
+ },
3400
+ {
3401
+ "epoch": 2.6103336921420883,
3402
+ "grad_norm": 0.93359375,
3403
+ "learning_rate": 9.13791656388614e-06,
3404
+ "loss": 1.9211,
3405
+ "step": 2425
3406
+ },
3407
+ {
3408
+ "epoch": 2.6157158234660924,
3409
+ "grad_norm": 0.7578125,
3410
+ "learning_rate": 8.89169046081466e-06,
3411
+ "loss": 1.5994,
3412
+ "step": 2430
3413
+ },
3414
+ {
3415
+ "epoch": 2.621097954790097,
3416
+ "grad_norm": 0.6328125,
3417
+ "learning_rate": 8.648672758318998e-06,
3418
+ "loss": 1.4924,
3419
+ "step": 2435
3420
+ },
3421
+ {
3422
+ "epoch": 2.626480086114101,
3423
+ "grad_norm": 0.76953125,
3424
+ "learning_rate": 8.40887201432471e-06,
3425
+ "loss": 1.3793,
3426
+ "step": 2440
3427
+ },
3428
+ {
3429
+ "epoch": 2.6318622174381057,
3430
+ "grad_norm": 0.8203125,
3431
+ "learning_rate": 8.172296673471502e-06,
3432
+ "loss": 1.5989,
3433
+ "step": 2445
3434
+ },
3435
+ {
3436
+ "epoch": 2.63724434876211,
3437
+ "grad_norm": 0.7890625,
3438
+ "learning_rate": 7.938955066815712e-06,
3439
+ "loss": 1.6884,
3440
+ "step": 2450
3441
+ },
3442
+ {
3443
+ "epoch": 2.642626480086114,
3444
+ "grad_norm": 0.7265625,
3445
+ "learning_rate": 7.708855411536975e-06,
3446
+ "loss": 1.5167,
3447
+ "step": 2455
3448
+ },
3449
+ {
3450
+ "epoch": 2.6480086114101185,
3451
+ "grad_norm": 1.1328125,
3452
+ "learning_rate": 7.482005810648829e-06,
3453
+ "loss": 1.489,
3454
+ "step": 2460
3455
+ },
3456
+ {
3457
+ "epoch": 2.6533907427341226,
3458
+ "grad_norm": 0.78125,
3459
+ "learning_rate": 7.258414252713386e-06,
3460
+ "loss": 1.4912,
3461
+ "step": 2465
3462
+ },
3463
+ {
3464
+ "epoch": 2.658772874058127,
3465
+ "grad_norm": 0.7578125,
3466
+ "learning_rate": 7.038088611560034e-06,
3467
+ "loss": 1.3837,
3468
+ "step": 2470
3469
+ },
3470
+ {
3471
+ "epoch": 2.6641550053821312,
3472
+ "grad_norm": 0.73046875,
3473
+ "learning_rate": 6.821036646008072e-06,
3474
+ "loss": 1.4303,
3475
+ "step": 2475
3476
+ },
3477
+ {
3478
+ "epoch": 2.669537136706136,
3479
+ "grad_norm": 0.91015625,
3480
+ "learning_rate": 6.6072659995936255e-06,
3481
+ "loss": 1.3774,
3482
+ "step": 2480
3483
+ },
3484
+ {
3485
+ "epoch": 2.67491926803014,
3486
+ "grad_norm": 0.78125,
3487
+ "learning_rate": 6.396784200300321e-06,
3488
+ "loss": 1.3987,
3489
+ "step": 2485
3490
+ },
3491
+ {
3492
+ "epoch": 2.6803013993541445,
3493
+ "grad_norm": 0.6953125,
3494
+ "learning_rate": 6.189598660294305e-06,
3495
+ "loss": 2.0529,
3496
+ "step": 2490
3497
+ },
3498
+ {
3499
+ "epoch": 2.6856835306781486,
3500
+ "grad_norm": 0.64453125,
3501
+ "learning_rate": 5.985716675663133e-06,
3502
+ "loss": 1.3103,
3503
+ "step": 2495
3504
+ },
3505
+ {
3506
+ "epoch": 2.6910656620021527,
3507
+ "grad_norm": 0.67578125,
3508
+ "learning_rate": 5.785145426158944e-06,
3509
+ "loss": 1.5402,
3510
+ "step": 2500
3511
+ },
3512
+ {
3513
+ "epoch": 2.6964477933261572,
3514
+ "grad_norm": 0.474609375,
3515
+ "learning_rate": 5.5878919749454914e-06,
3516
+ "loss": 1.3674,
3517
+ "step": 2505
3518
+ },
3519
+ {
3520
+ "epoch": 2.7018299246501614,
3521
+ "grad_norm": 0.76953125,
3522
+ "learning_rate": 5.393963268349511e-06,
3523
+ "loss": 1.1132,
3524
+ "step": 2510
3525
+ },
3526
+ {
3527
+ "epoch": 2.707212055974166,
3528
+ "grad_norm": 0.97265625,
3529
+ "learning_rate": 5.203366135616117e-06,
3530
+ "loss": 1.7984,
3531
+ "step": 2515
3532
+ },
3533
+ {
3534
+ "epoch": 2.71259418729817,
3535
+ "grad_norm": 0.96875,
3536
+ "learning_rate": 5.016107288668148e-06,
3537
+ "loss": 1.4965,
3538
+ "step": 2520
3539
+ },
3540
+ {
3541
+ "epoch": 2.717976318622174,
3542
+ "grad_norm": 0.71484375,
3543
+ "learning_rate": 4.832193321870082e-06,
3544
+ "loss": 1.4131,
3545
+ "step": 2525
3546
+ },
3547
+ {
3548
+ "epoch": 2.7233584499461787,
3549
+ "grad_norm": 1.21875,
3550
+ "learning_rate": 4.651630711795507e-06,
3551
+ "loss": 1.742,
3552
+ "step": 2530
3553
+ },
3554
+ {
3555
+ "epoch": 2.7287405812701833,
3556
+ "grad_norm": 1.5703125,
3557
+ "learning_rate": 4.474425816999306e-06,
3558
+ "loss": 1.7317,
3559
+ "step": 2535
3560
+ },
3561
+ {
3562
+ "epoch": 2.7341227125941874,
3563
+ "grad_norm": 0.7734375,
3564
+ "learning_rate": 4.300584877793523e-06,
3565
+ "loss": 1.4012,
3566
+ "step": 2540
3567
+ },
3568
+ {
3569
+ "epoch": 2.7395048439181915,
3570
+ "grad_norm": 0.78515625,
3571
+ "learning_rate": 4.13011401602782e-06,
3572
+ "loss": 1.269,
3573
+ "step": 2545
3574
+ },
3575
+ {
3576
+ "epoch": 2.744886975242196,
3577
+ "grad_norm": 0.69140625,
3578
+ "learning_rate": 3.96301923487371e-06,
3579
+ "loss": 1.4719,
3580
+ "step": 2550
3581
+ },
3582
+ {
3583
+ "epoch": 2.7502691065662,
3584
+ "grad_norm": 0.98828125,
3585
+ "learning_rate": 3.799306418613291e-06,
3586
+ "loss": 1.424,
3587
+ "step": 2555
3588
+ },
3589
+ {
3590
+ "epoch": 2.7556512378902047,
3591
+ "grad_norm": 1.0,
3592
+ "learning_rate": 3.638981332431901e-06,
3593
+ "loss": 1.5294,
3594
+ "step": 2560
3595
+ },
3596
+ {
3597
+ "epoch": 2.761033369214209,
3598
+ "grad_norm": 0.94140625,
3599
+ "learning_rate": 3.4820496222152333e-06,
3600
+ "loss": 1.5446,
3601
+ "step": 2565
3602
+ },
3603
+ {
3604
+ "epoch": 2.766415500538213,
3605
+ "grad_norm": 0.73046875,
3606
+ "learning_rate": 3.328516814350402e-06,
3607
+ "loss": 1.5216,
3608
+ "step": 2570
3609
+ },
3610
+ {
3611
+ "epoch": 2.7717976318622175,
3612
+ "grad_norm": 0.9375,
3613
+ "learning_rate": 3.178388315531411e-06,
3614
+ "loss": 1.5401,
3615
+ "step": 2575
3616
+ },
3617
+ {
3618
+ "epoch": 2.7771797631862216,
3619
+ "grad_norm": 0.703125,
3620
+ "learning_rate": 3.031669412568694e-06,
3621
+ "loss": 1.4231,
3622
+ "step": 2580
3623
+ },
3624
+ {
3625
+ "epoch": 2.782561894510226,
3626
+ "grad_norm": 1.65625,
3627
+ "learning_rate": 2.888365272202953e-06,
3628
+ "loss": 1.2964,
3629
+ "step": 2585
3630
+ },
3631
+ {
3632
+ "epoch": 2.7879440258342303,
3633
+ "grad_norm": 0.7109375,
3634
+ "learning_rate": 2.7484809409232705e-06,
3635
+ "loss": 1.2211,
3636
+ "step": 2590
3637
+ },
3638
+ {
3639
+ "epoch": 2.7933261571582344,
3640
+ "grad_norm": 0.86328125,
3641
+ "learning_rate": 2.612021344789284e-06,
3642
+ "loss": 1.4812,
3643
+ "step": 2595
3644
+ },
3645
+ {
3646
+ "epoch": 2.798708288482239,
3647
+ "grad_norm": 0.6953125,
3648
+ "learning_rate": 2.4789912892578258e-06,
3649
+ "loss": 1.7189,
3650
+ "step": 2600
3651
+ },
3652
+ {
3653
+ "epoch": 2.8040904198062435,
3654
+ "grad_norm": 0.65625,
3655
+ "learning_rate": 2.349395459013626e-06,
3656
+ "loss": 1.3196,
3657
+ "step": 2605
3658
+ },
3659
+ {
3660
+ "epoch": 2.8094725511302476,
3661
+ "grad_norm": 1.046875,
3662
+ "learning_rate": 2.223238417804352e-06,
3663
+ "loss": 1.4144,
3664
+ "step": 2610
3665
+ },
3666
+ {
3667
+ "epoch": 2.8148546824542517,
3668
+ "grad_norm": 0.71875,
3669
+ "learning_rate": 2.100524608279919e-06,
3670
+ "loss": 1.3011,
3671
+ "step": 2615
3672
+ },
3673
+ {
3674
+ "epoch": 2.8202368137782563,
3675
+ "grad_norm": 0.85546875,
3676
+ "learning_rate": 1.9812583518360573e-06,
3677
+ "loss": 1.4989,
3678
+ "step": 2620
3679
+ },
3680
+ {
3681
+ "epoch": 2.8256189451022604,
3682
+ "grad_norm": 1.1953125,
3683
+ "learning_rate": 1.865443848462045e-06,
3684
+ "loss": 1.6008,
3685
+ "step": 2625
3686
+ },
3687
+ {
3688
+ "epoch": 2.831001076426265,
3689
+ "grad_norm": 0.671875,
3690
+ "learning_rate": 1.7530851765928947e-06,
3691
+ "loss": 1.3724,
3692
+ "step": 2630
3693
+ },
3694
+ {
3695
+ "epoch": 2.836383207750269,
3696
+ "grad_norm": 0.8359375,
3697
+ "learning_rate": 1.6441862929657104e-06,
3698
+ "loss": 1.4647,
3699
+ "step": 2635
3700
+ },
3701
+ {
3702
+ "epoch": 2.841765339074273,
3703
+ "grad_norm": 0.85546875,
3704
+ "learning_rate": 1.5387510324803122e-06,
3705
+ "loss": 1.5235,
3706
+ "step": 2640
3707
+ },
3708
+ {
3709
+ "epoch": 2.8471474703982778,
3710
+ "grad_norm": 0.6328125,
3711
+ "learning_rate": 1.4367831080642435e-06,
3712
+ "loss": 1.3849,
3713
+ "step": 2645
3714
+ },
3715
+ {
3716
+ "epoch": 2.852529601722282,
3717
+ "grad_norm": 0.6796875,
3718
+ "learning_rate": 1.338286110541942e-06,
3719
+ "loss": 1.4417,
3720
+ "step": 2650
3721
+ },
3722
+ {
3723
+ "epoch": 2.8579117330462864,
3724
+ "grad_norm": 1.140625,
3725
+ "learning_rate": 1.2432635085084076e-06,
3726
+ "loss": 1.2763,
3727
+ "step": 2655
3728
+ },
3729
+ {
3730
+ "epoch": 2.8632938643702905,
3731
+ "grad_norm": 0.703125,
3732
+ "learning_rate": 1.1517186482069232e-06,
3733
+ "loss": 1.4491,
3734
+ "step": 2660
3735
+ },
3736
+ {
3737
+ "epoch": 2.868675995694295,
3738
+ "grad_norm": 0.875,
3739
+ "learning_rate": 1.0636547534113139e-06,
3740
+ "loss": 1.6601,
3741
+ "step": 2665
3742
+ },
3743
+ {
3744
+ "epoch": 2.874058127018299,
3745
+ "grad_norm": 0.65234375,
3746
+ "learning_rate": 9.79074925312362e-07,
3747
+ "loss": 1.2571,
3748
+ "step": 2670
3749
+ },
3750
+ {
3751
+ "epoch": 2.8794402583423038,
3752
+ "grad_norm": 0.7734375,
3753
+ "learning_rate": 8.979821424086488e-07,
3754
+ "loss": 1.5522,
3755
+ "step": 2675
3756
+ },
3757
+ {
3758
+ "epoch": 2.884822389666308,
3759
+ "grad_norm": 0.71875,
3760
+ "learning_rate": 8.203792604016159e-07,
3761
+ "loss": 1.5063,
3762
+ "step": 2680
3763
+ },
3764
+ {
3765
+ "epoch": 2.890204520990312,
3766
+ "grad_norm": 0.80078125,
3767
+ "learning_rate": 7.462690120950134e-07,
3768
+ "loss": 1.4617,
3769
+ "step": 2685
3770
+ },
3771
+ {
3772
+ "epoch": 2.8955866523143166,
3773
+ "grad_norm": 0.78515625,
3774
+ "learning_rate": 6.756540072987094e-07,
3775
+ "loss": 1.2957,
3776
+ "step": 2690
3777
+ },
3778
+ {
3779
+ "epoch": 2.9009687836383207,
3780
+ "grad_norm": 1.3125,
3781
+ "learning_rate": 6.085367327367197e-07,
3782
+ "loss": 1.443,
3783
+ "step": 2695
3784
+ },
3785
+ {
3786
+ "epoch": 2.9063509149623252,
3787
+ "grad_norm": 0.6875,
3788
+ "learning_rate": 5.449195519596661e-07,
3789
+ "loss": 1.4819,
3790
+ "step": 2700
3791
+ },
3792
+ {
3793
+ "epoch": 2.9117330462863293,
3794
+ "grad_norm": 0.796875,
3795
+ "learning_rate": 4.848047052615545e-07,
3796
+ "loss": 1.6234,
3797
+ "step": 2705
3798
+ },
3799
+ {
3800
+ "epoch": 2.9171151776103335,
3801
+ "grad_norm": 0.60546875,
3802
+ "learning_rate": 4.2819430960087113e-07,
3803
+ "loss": 1.5406,
3804
+ "step": 2710
3805
+ },
3806
+ {
3807
+ "epoch": 2.922497308934338,
3808
+ "grad_norm": 0.66015625,
3809
+ "learning_rate": 3.7509035852602015e-07,
3810
+ "loss": 1.5775,
3811
+ "step": 2715
3812
+ },
3813
+ {
3814
+ "epoch": 2.9278794402583426,
3815
+ "grad_norm": 0.77734375,
3816
+ "learning_rate": 3.254947221051685e-07,
3817
+ "loss": 1.5152,
3818
+ "step": 2720
3819
+ },
3820
+ {
3821
+ "epoch": 2.9332615715823467,
3822
+ "grad_norm": 0.90625,
3823
+ "learning_rate": 2.794091468602988e-07,
3824
+ "loss": 1.57,
3825
+ "step": 2725
3826
+ },
3827
+ {
3828
+ "epoch": 2.938643702906351,
3829
+ "grad_norm": 0.69140625,
3830
+ "learning_rate": 2.368352557058362e-07,
3831
+ "loss": 1.6458,
3832
+ "step": 2730
3833
+ },
3834
+ {
3835
+ "epoch": 2.9440258342303554,
3836
+ "grad_norm": 0.78515625,
3837
+ "learning_rate": 1.9777454789134953e-07,
3838
+ "loss": 1.4748,
3839
+ "step": 2735
3840
+ },
3841
+ {
3842
+ "epoch": 2.9494079655543595,
3843
+ "grad_norm": 1.421875,
3844
+ "learning_rate": 1.6222839894887154e-07,
3845
+ "loss": 1.6024,
3846
+ "step": 2740
3847
+ },
3848
+ {
3849
+ "epoch": 2.954790096878364,
3850
+ "grad_norm": 0.6796875,
3851
+ "learning_rate": 1.301980606444153e-07,
3852
+ "loss": 1.0898,
3853
+ "step": 2745
3854
+ },
3855
+ {
3856
+ "epoch": 2.960172228202368,
3857
+ "grad_norm": 0.71484375,
3858
+ "learning_rate": 1.0168466093387619e-07,
3859
+ "loss": 1.8948,
3860
+ "step": 2750
3861
+ },
3862
+ {
3863
+ "epoch": 2.9655543595263723,
3864
+ "grad_norm": 0.7734375,
3865
+ "learning_rate": 7.668920392335244e-08,
3866
+ "loss": 1.5294,
3867
+ "step": 2755
3868
+ },
3869
+ {
3870
+ "epoch": 2.970936490850377,
3871
+ "grad_norm": 0.55859375,
3872
+ "learning_rate": 5.521256983374024e-08,
3873
+ "loss": 1.0666,
3874
+ "step": 2760
3875
+ },
3876
+ {
3877
+ "epoch": 2.976318622174381,
3878
+ "grad_norm": 0.6796875,
3879
+ "learning_rate": 3.725551496978064e-08,
3880
+ "loss": 1.6953,
3881
+ "step": 2765
3882
+ },
3883
+ {
3884
+ "epoch": 2.9817007534983855,
3885
+ "grad_norm": 0.6796875,
3886
+ "learning_rate": 2.281867169336982e-08,
3887
+ "loss": 1.5101,
3888
+ "step": 2770
3889
+ },
3890
+ {
3891
+ "epoch": 2.9870828848223896,
3892
+ "grad_norm": 0.79296875,
3893
+ "learning_rate": 1.190254840135463e-08,
3894
+ "loss": 1.4111,
3895
+ "step": 2775
3896
+ },
3897
+ {
3898
+ "epoch": 2.9924650161463937,
3899
+ "grad_norm": 0.8359375,
3900
+ "learning_rate": 4.507529507580266e-09,
3901
+ "loss": 1.4808,
3902
+ "step": 2780
3903
+ },
3904
+ {
3905
+ "epoch": 2.9978471474703983,
3906
+ "grad_norm": 0.7734375,
3907
+ "learning_rate": 6.33875429389974e-10,
3908
+ "loss": 1.6763,
3909
+ "step": 2785
3910
+ }
3911
+ ],
3912
  "logging_steps": 5,
3913
+ "max_steps": 2787,
3914
  "num_input_tokens_seen": 0,
3915
+ "num_train_epochs": 3,
3916
  "save_steps": 500,
3917
  "stateful_callbacks": {
3918
  "TrainerControl": {
 
3926
  "attributes": {}
3927
  }
3928
  },
3929
+ "total_flos": 5.04910971100201e+16,
3930
  "train_batch_size": 16,
3931
  "trial_name": null,
3932
  "trial_params": null
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:aac91bec2e12f16beb7db12d9d3c37fc06c2ba1193e5bd6365cb71cd8280232f
3
  size 5777
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c04b39bd61cbf2c402d59875f6b4d7fb59ebbd0fca703456d167d41db414a80b
3
  size 5777