Hugging Face's logo

mlfoundations-dev
/
DCFT-Stratos-Verified-114k-qwen-2.5-72B

Text Generation
Transformers
Safetensors
qwen2
llama-factory
full
Generated from Trainer
conversational
text-generation-inference
Model card Files
xet
Community
1
marianna13 commited on
Commit
ec9de84
·
verified ·
1 Parent(s): 23dd165

Upload folder using huggingface_hub

Browse files
Files changed (47) hide show
  1. .gitattributes +1 -0
  2. README.md +60 -0
  3. added_tokens.json +24 -0
  4. all_results.json +8 -0
  5. config.json +29 -0
  6. generation_config.json +14 -0
  7. merges.txt +0 -0
  8. model-00001-of-00031.safetensors +3 -0
  9. model-00002-of-00031.safetensors +3 -0
  10. model-00003-of-00031.safetensors +3 -0
  11. model-00004-of-00031.safetensors +3 -0
  12. model-00005-of-00031.safetensors +3 -0
  13. model-00006-of-00031.safetensors +3 -0
  14. model-00007-of-00031.safetensors +3 -0
  15. model-00008-of-00031.safetensors +3 -0
  16. model-00009-of-00031.safetensors +3 -0
  17. model-00010-of-00031.safetensors +3 -0
  18. model-00011-of-00031.safetensors +3 -0
  19. model-00012-of-00031.safetensors +3 -0
  20. model-00013-of-00031.safetensors +3 -0
  21. model-00014-of-00031.safetensors +3 -0
  22. model-00015-of-00031.safetensors +3 -0
  23. model-00016-of-00031.safetensors +3 -0
  24. model-00017-of-00031.safetensors +3 -0
  25. model-00018-of-00031.safetensors +3 -0
  26. model-00019-of-00031.safetensors +3 -0
  27. model-00020-of-00031.safetensors +3 -0
  28. model-00021-of-00031.safetensors +3 -0
  29. model-00022-of-00031.safetensors +3 -0
  30. model-00023-of-00031.safetensors +3 -0
  31. model-00024-of-00031.safetensors +3 -0
  32. model-00025-of-00031.safetensors +3 -0
  33. model-00026-of-00031.safetensors +3 -0
  34. model-00027-of-00031.safetensors +3 -0
  35. model-00028-of-00031.safetensors +3 -0
  36. model-00029-of-00031.safetensors +3 -0
  37. model-00030-of-00031.safetensors +3 -0
  38. model-00031-of-00031.safetensors +3 -0
  39. model.safetensors.index.json +970 -0
  40. special_tokens_map.json +31 -0
  41. tokenizer.json +3 -0
  42. tokenizer_config.json +208 -0
  43. train_results.json +8 -0
  44. trainer_log.jsonl +337 -0
  45. trainer_state.json +2394 -0
  46. training_args.bin +3 -0
  47. vocab.json +0 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: other
4
+ base_model: Qwen/Qwen2.5-72B-Instruct
5
+ tags:
6
+ - llama-factory
7
+ - full
8
+ - generated_from_trainer
9
+ model-index:
10
+ - name: DCFT-Stratos-Verified-114k-qwen-2.5-72B
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # DCFT-Stratos-Verified-114k-qwen-2.5-72B
18
+
19
+ This model is a fine-tuned version of [Qwen/Qwen2.5-72B-Instruct](https://huggingface.co/Qwen/Qwen2.5-72B-Instruct) on the mlfoundations-dev/stratos_verified_mix dataset.
20
+
21
+ ## Model description
22
+
23
+ More information needed
24
+
25
+ ## Intended uses & limitations
26
+
27
+ More information needed
28
+
29
+ ## Training and evaluation data
30
+
31
+ More information needed
32
+
33
+ ## Training procedure
34
+
35
+ ### Training hyperparameters
36
+
37
+ The following hyperparameters were used during training:
38
+ - learning_rate: 1e-05
39
+ - train_batch_size: 1
40
+ - eval_batch_size: 8
41
+ - seed: 42
42
+ - distributed_type: multi-GPU
43
+ - num_devices: 1024
44
+ - total_train_batch_size: 1024
45
+ - total_eval_batch_size: 8192
46
+ - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
47
+ - lr_scheduler_type: cosine
48
+ - lr_scheduler_warmup_ratio: 0.1
49
+ - num_epochs: 3.0
50
+
51
+ ### Training results
52
+
53
+
54
+
55
+ ### Framework versions
56
+
57
+ - Transformers 4.46.1
58
+ - Pytorch 2.5.0a0+b465a5843b.nv24.09
59
+ - Datasets 3.0.2
60
+ - Tokenizers 0.20.3
added_tokens.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</tool_call>": 151658,
3
+ "<tool_call>": 151657,
4
+ "<|box_end|>": 151649,
5
+ "<|box_start|>": 151648,
6
+ "<|endoftext|>": 151643,
7
+ "<|file_sep|>": 151664,
8
+ "<|fim_middle|>": 151660,
9
+ "<|fim_pad|>": 151662,
10
+ "<|fim_prefix|>": 151659,
11
+ "<|fim_suffix|>": 151661,
12
+ "<|im_end|>": 151645,
13
+ "<|im_start|>": 151644,
14
+ "<|image_pad|>": 151655,
15
+ "<|object_ref_end|>": 151647,
16
+ "<|object_ref_start|>": 151646,
17
+ "<|quad_end|>": 151651,
18
+ "<|quad_start|>": 151650,
19
+ "<|repo_name|>": 151663,
20
+ "<|video_pad|>": 151656,
21
+ "<|vision_end|>": 151653,
22
+ "<|vision_pad|>": 151654,
23
+ "<|vision_start|>": 151652
24
+ }
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "total_flos": 2.982866014384947e+16,
4
+ "train_loss": 0.3923433954339652,
5
+ "train_runtime": 48299.7361,
6
+ "train_samples_per_second": 7.078,
7
+ "train_steps_per_second": 0.007
8
+ }
config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/leonardo_work/EUHPC_E03_068/marianna/models/Qwen/Qwen2.5-72B-Instruct",
3
+ "architectures": [
4
+ "Qwen2ForCausalLM"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "eos_token_id": 151645,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 8192,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 29568,
13
+ "max_position_embeddings": 32768,
14
+ "max_window_layers": 70,
15
+ "model_type": "qwen2",
16
+ "num_attention_heads": 64,
17
+ "num_hidden_layers": 80,
18
+ "num_key_value_heads": 8,
19
+ "rms_norm_eps": 1e-06,
20
+ "rope_scaling": null,
21
+ "rope_theta": 1000000.0,
22
+ "sliding_window": null,
23
+ "tie_word_embeddings": false,
24
+ "torch_dtype": "bfloat16",
25
+ "transformers_version": "4.46.1",
26
+ "use_cache": false,
27
+ "use_sliding_window": false,
28
+ "vocab_size": 152064
29
+ }
generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 151643,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "repetition_penalty": 1.05,
10
+ "temperature": 0.7,
11
+ "top_k": 20,
12
+ "top_p": 0.8,
13
+ "transformers_version": "4.46.1"
14
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9bdb27cc612a16cc63a39b862c63f6b5fb45343d8b342b4366f91b4d4b3bdf2
3
+ size 4548798728
model-00002-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68ff5818be595ff1eff848bc670348f0511c09e7d352aa6301a4a6ac88230200
3
+ size 4964101384
model-00003-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db5d96f741d70485a683d6f666ced58c70e70f662148f65e1f7def7147c17ea1
3
+ size 4781637328
model-00004-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ccd56bf1f72713ca8802bf8254d98b0f3a541b1de3422ae60d86a5539ba145e
3
+ size 4781670320
model-00005-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dda77ac89e4e46ee7b7f10cb78bb18d99b4cc610c7d9545887ce383e888a629e
3
+ size 4781670360
model-00006-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99ebc8cecde4d02dbb745f3d18f2ac8d1b83d4ddeb1205c74c8b405618f6fca7
3
+ size 4964101416
model-00007-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:487b76a2263823462f7bd6d5b0c1838d1c893d2998fb0f943ce41d0c2c3859a8
3
+ size 4781637360
model-00008-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f8bc0a2e675299ceaee7b696020d1363ea9ca4074c408dab6e3c0f9f12f321e
3
+ size 4781670360
model-00009-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:327f43ac6461cf87a950b5473e06ed6a4a3691f4a2c9e49b37cb2145dab94689
3
+ size 4781670360
model-00010-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd5fbe672e2301863490bf5a6fdbc3899f8c009c7d93c7601306c7248b080853
3
+ size 4964101416
model-00011-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ce5e57aa90e5f5c11fcbf911fedb3d835ec56e92f20a582637d48819143712ab
3
+ size 4781637360
model-00012-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a2b052e6e4b2c61e2bab59da439a8840fdb7089c4c066c3cc2b6b61f28c67fd
3
+ size 4781670360
model-00013-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27ef49de075d6346f9de00e9ba1973a7eb82a26465ea0c1eaa2e467666f718c8
3
+ size 4781670360
model-00014-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3dc6c94886cf1660538dc1672a760cb2d5176f9d9f4107acf04a7cf36c2ffb63
3
+ size 4964101416
model-00015-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fd8812142bd2190095d4042fd377b8fef36eeab88d0e6f09ef37fe8c97fd940
3
+ size 4781637360
model-00016-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ababc50b0e59f270fd79492ceabaf1fd3a6ce031f13183d2d241795b42f7576
3
+ size 4781670360
model-00017-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e26dec7e8d6e385c4f02ee7d6325393959b88537f5d44879e81f55a05f75a60e
3
+ size 4781670360
model-00018-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc9a1de0afd16bf75dd4064e21e2961d4c07a846572b4d4a00b7380862e19da5
3
+ size 4964101416
model-00019-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:163e2633f049d7d033536eb652593d8e2f6a7c161a3dcd9cc5f0ca2cd532d0f5
3
+ size 4781637360
model-00020-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4238250cbc4e86b30d05acc439febec1017daa148740c31161ac6c7ee812fc06
3
+ size 4781670360
model-00021-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbbd57d3d15f079461caa303f816ccc476d9a4557ffa07e9344bc3fbd9e1e37d
3
+ size 4781670360
model-00022-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b4417b6e72ed238b92c743bec942acdb25dc44cfd20d2d0629fca5e062dca66
3
+ size 4964101416
model-00023-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c812f168e7a2e993ecc4c451a8c8fe8f8d04423b33f751a536d77065c3284633
3
+ size 4781637360
model-00024-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53636d881515eadf29ae38e6d17cb4b640589058297b45ee85e102fce522f790
3
+ size 4781670360
model-00025-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e9a2117494921569a5996e6e8dc8ce144f5ec7016b0d3dacbc6e5cd925cdfc5
3
+ size 4781670360
model-00026-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1560bfc35768fcf3772d098e420a92eaa964a7cacbef2bbce5cad37bc38ceb9
3
+ size 4964101416
model-00027-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b11de21ef0d97a96a0b45c11fca8216546724ef20bff4515b590aa7ae1a3d7d5
3
+ size 4781637360
model-00028-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc494cdd4685054cb2474502d02257ca6abc94ec059da9ee2144a34e46c48e13
3
+ size 4781670360
model-00029-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67ef57cf06c8d0f05ce1ebdfc5b092c2c29715a74073d3c56ea356ee0b69fe76
3
+ size 4781670360
model-00030-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:54a744276a59b6467b2c5e3ae20f3034b36365c828f8d159d6584316b0a1c335
3
+ size 3208747032
model-00031-of-00031.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:095b512ad81a5199c2fe1ce203fb840a34d9051d9d983ba0b80ea9980ca022a1
3
+ size 2491416704
model.safetensors.index.json ADDED
@@ -0,0 +1,970 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 145412407296
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "model-00031-of-00031.safetensors",
7
+ "model.embed_tokens.weight": "model-00001-of-00031.safetensors",
8
+ "model.layers.0.input_layernorm.weight": "model-00001-of-00031.safetensors",
9
+ "model.layers.0.mlp.down_proj.weight": "model-00001-of-00031.safetensors",
10
+ "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00031.safetensors",
11
+ "model.layers.0.mlp.up_proj.weight": "model-00001-of-00031.safetensors",
12
+ "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00031.safetensors",
13
+ "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00031.safetensors",
14
+ "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00031.safetensors",
15
+ "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00031.safetensors",
16
+ "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00031.safetensors",
17
+ "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00031.safetensors",
18
+ "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00031.safetensors",
19
+ "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00031.safetensors",
20
+ "model.layers.1.input_layernorm.weight": "model-00002-of-00031.safetensors",
21
+ "model.layers.1.mlp.down_proj.weight": "model-00002-of-00031.safetensors",
22
+ "model.layers.1.mlp.gate_proj.weight": "model-00002-of-00031.safetensors",
23
+ "model.layers.1.mlp.up_proj.weight": "model-00002-of-00031.safetensors",
24
+ "model.layers.1.post_attention_layernorm.weight": "model-00002-of-00031.safetensors",
25
+ "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00031.safetensors",
26
+ "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00031.safetensors",
27
+ "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00031.safetensors",
28
+ "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00031.safetensors",
29
+ "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00031.safetensors",
30
+ "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00031.safetensors",
31
+ "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00031.safetensors",
32
+ "model.layers.10.input_layernorm.weight": "model-00005-of-00031.safetensors",
33
+ "model.layers.10.mlp.down_proj.weight": "model-00005-of-00031.safetensors",
34
+ "model.layers.10.mlp.gate_proj.weight": "model-00005-of-00031.safetensors",
35
+ "model.layers.10.mlp.up_proj.weight": "model-00005-of-00031.safetensors",
36
+ "model.layers.10.post_attention_layernorm.weight": "model-00005-of-00031.safetensors",
37
+ "model.layers.10.self_attn.k_proj.bias": "model-00005-of-00031.safetensors",
38
+ "model.layers.10.self_attn.k_proj.weight": "model-00005-of-00031.safetensors",
39
+ "model.layers.10.self_attn.o_proj.weight": "model-00005-of-00031.safetensors",
40
+ "model.layers.10.self_attn.q_proj.bias": "model-00005-of-00031.safetensors",
41
+ "model.layers.10.self_attn.q_proj.weight": "model-00005-of-00031.safetensors",
42
+ "model.layers.10.self_attn.v_proj.bias": "model-00005-of-00031.safetensors",
43
+ "model.layers.10.self_attn.v_proj.weight": "model-00005-of-00031.safetensors",
44
+ "model.layers.11.input_layernorm.weight": "model-00005-of-00031.safetensors",
45
+ "model.layers.11.mlp.down_proj.weight": "model-00005-of-00031.safetensors",
46
+ "model.layers.11.mlp.gate_proj.weight": "model-00005-of-00031.safetensors",
47
+ "model.layers.11.mlp.up_proj.weight": "model-00005-of-00031.safetensors",
48
+ "model.layers.11.post_attention_layernorm.weight": "model-00005-of-00031.safetensors",
49
+ "model.layers.11.self_attn.k_proj.bias": "model-00005-of-00031.safetensors",
50
+ "model.layers.11.self_attn.k_proj.weight": "model-00005-of-00031.safetensors",
51
+ "model.layers.11.self_attn.o_proj.weight": "model-00005-of-00031.safetensors",
52
+ "model.layers.11.self_attn.q_proj.bias": "model-00005-of-00031.safetensors",
53
+ "model.layers.11.self_attn.q_proj.weight": "model-00005-of-00031.safetensors",
54
+ "model.layers.11.self_attn.v_proj.bias": "model-00005-of-00031.safetensors",
55
+ "model.layers.11.self_attn.v_proj.weight": "model-00005-of-00031.safetensors",
56
+ "model.layers.12.input_layernorm.weight": "model-00006-of-00031.safetensors",
57
+ "model.layers.12.mlp.down_proj.weight": "model-00006-of-00031.safetensors",
58
+ "model.layers.12.mlp.gate_proj.weight": "model-00006-of-00031.safetensors",
59
+ "model.layers.12.mlp.up_proj.weight": "model-00006-of-00031.safetensors",
60
+ "model.layers.12.post_attention_layernorm.weight": "model-00006-of-00031.safetensors",
61
+ "model.layers.12.self_attn.k_proj.bias": "model-00005-of-00031.safetensors",
62
+ "model.layers.12.self_attn.k_proj.weight": "model-00005-of-00031.safetensors",
63
+ "model.layers.12.self_attn.o_proj.weight": "model-00005-of-00031.safetensors",
64
+ "model.layers.12.self_attn.q_proj.bias": "model-00005-of-00031.safetensors",
65
+ "model.layers.12.self_attn.q_proj.weight": "model-00005-of-00031.safetensors",
66
+ "model.layers.12.self_attn.v_proj.bias": "model-00005-of-00031.safetensors",
67
+ "model.layers.12.self_attn.v_proj.weight": "model-00005-of-00031.safetensors",
68
+ "model.layers.13.input_layernorm.weight": "model-00006-of-00031.safetensors",
69
+ "model.layers.13.mlp.down_proj.weight": "model-00006-of-00031.safetensors",
70
+ "model.layers.13.mlp.gate_proj.weight": "model-00006-of-00031.safetensors",
71
+ "model.layers.13.mlp.up_proj.weight": "model-00006-of-00031.safetensors",
72
+ "model.layers.13.post_attention_layernorm.weight": "model-00006-of-00031.safetensors",
73
+ "model.layers.13.self_attn.k_proj.bias": "model-00006-of-00031.safetensors",
74
+ "model.layers.13.self_attn.k_proj.weight": "model-00006-of-00031.safetensors",
75
+ "model.layers.13.self_attn.o_proj.weight": "model-00006-of-00031.safetensors",
76
+ "model.layers.13.self_attn.q_proj.bias": "model-00006-of-00031.safetensors",
77
+ "model.layers.13.self_attn.q_proj.weight": "model-00006-of-00031.safetensors",
78
+ "model.layers.13.self_attn.v_proj.bias": "model-00006-of-00031.safetensors",
79
+ "model.layers.13.self_attn.v_proj.weight": "model-00006-of-00031.safetensors",
80
+ "model.layers.14.input_layernorm.weight": "model-00006-of-00031.safetensors",
81
+ "model.layers.14.mlp.down_proj.weight": "model-00006-of-00031.safetensors",
82
+ "model.layers.14.mlp.gate_proj.weight": "model-00006-of-00031.safetensors",
83
+ "model.layers.14.mlp.up_proj.weight": "model-00006-of-00031.safetensors",
84
+ "model.layers.14.post_attention_layernorm.weight": "model-00006-of-00031.safetensors",
85
+ "model.layers.14.self_attn.k_proj.bias": "model-00006-of-00031.safetensors",
86
+ "model.layers.14.self_attn.k_proj.weight": "model-00006-of-00031.safetensors",
87
+ "model.layers.14.self_attn.o_proj.weight": "model-00006-of-00031.safetensors",
88
+ "model.layers.14.self_attn.q_proj.bias": "model-00006-of-00031.safetensors",
89
+ "model.layers.14.self_attn.q_proj.weight": "model-00006-of-00031.safetensors",
90
+ "model.layers.14.self_attn.v_proj.bias": "model-00006-of-00031.safetensors",
91
+ "model.layers.14.self_attn.v_proj.weight": "model-00006-of-00031.safetensors",
92
+ "model.layers.15.input_layernorm.weight": "model-00007-of-00031.safetensors",
93
+ "model.layers.15.mlp.down_proj.weight": "model-00007-of-00031.safetensors",
94
+ "model.layers.15.mlp.gate_proj.weight": "model-00007-of-00031.safetensors",
95
+ "model.layers.15.mlp.up_proj.weight": "model-00007-of-00031.safetensors",
96
+ "model.layers.15.post_attention_layernorm.weight": "model-00007-of-00031.safetensors",
97
+ "model.layers.15.self_attn.k_proj.bias": "model-00007-of-00031.safetensors",
98
+ "model.layers.15.self_attn.k_proj.weight": "model-00007-of-00031.safetensors",
99
+ "model.layers.15.self_attn.o_proj.weight": "model-00007-of-00031.safetensors",
100
+ "model.layers.15.self_attn.q_proj.bias": "model-00007-of-00031.safetensors",
101
+ "model.layers.15.self_attn.q_proj.weight": "model-00007-of-00031.safetensors",
102
+ "model.layers.15.self_attn.v_proj.bias": "model-00007-of-00031.safetensors",
103
+ "model.layers.15.self_attn.v_proj.weight": "model-00007-of-00031.safetensors",
104
+ "model.layers.16.input_layernorm.weight": "model-00007-of-00031.safetensors",
105
+ "model.layers.16.mlp.down_proj.weight": "model-00007-of-00031.safetensors",
106
+ "model.layers.16.mlp.gate_proj.weight": "model-00007-of-00031.safetensors",
107
+ "model.layers.16.mlp.up_proj.weight": "model-00007-of-00031.safetensors",
108
+ "model.layers.16.post_attention_layernorm.weight": "model-00007-of-00031.safetensors",
109
+ "model.layers.16.self_attn.k_proj.bias": "model-00007-of-00031.safetensors",
110
+ "model.layers.16.self_attn.k_proj.weight": "model-00007-of-00031.safetensors",
111
+ "model.layers.16.self_attn.o_proj.weight": "model-00007-of-00031.safetensors",
112
+ "model.layers.16.self_attn.q_proj.bias": "model-00007-of-00031.safetensors",
113
+ "model.layers.16.self_attn.q_proj.weight": "model-00007-of-00031.safetensors",
114
+ "model.layers.16.self_attn.v_proj.bias": "model-00007-of-00031.safetensors",
115
+ "model.layers.16.self_attn.v_proj.weight": "model-00007-of-00031.safetensors",
116
+ "model.layers.17.input_layernorm.weight": "model-00008-of-00031.safetensors",
117
+ "model.layers.17.mlp.down_proj.weight": "model-00008-of-00031.safetensors",
118
+ "model.layers.17.mlp.gate_proj.weight": "model-00007-of-00031.safetensors",
119
+ "model.layers.17.mlp.up_proj.weight": "model-00007-of-00031.safetensors",
120
+ "model.layers.17.post_attention_layernorm.weight": "model-00008-of-00031.safetensors",
121
+ "model.layers.17.self_attn.k_proj.bias": "model-00007-of-00031.safetensors",
122
+ "model.layers.17.self_attn.k_proj.weight": "model-00007-of-00031.safetensors",
123
+ "model.layers.17.self_attn.o_proj.weight": "model-00007-of-00031.safetensors",
124
+ "model.layers.17.self_attn.q_proj.bias": "model-00007-of-00031.safetensors",
125
+ "model.layers.17.self_attn.q_proj.weight": "model-00007-of-00031.safetensors",
126
+ "model.layers.17.self_attn.v_proj.bias": "model-00007-of-00031.safetensors",
127
+ "model.layers.17.self_attn.v_proj.weight": "model-00007-of-00031.safetensors",
128
+ "model.layers.18.input_layernorm.weight": "model-00008-of-00031.safetensors",
129
+ "model.layers.18.mlp.down_proj.weight": "model-00008-of-00031.safetensors",
130
+ "model.layers.18.mlp.gate_proj.weight": "model-00008-of-00031.safetensors",
131
+ "model.layers.18.mlp.up_proj.weight": "model-00008-of-00031.safetensors",
132
+ "model.layers.18.post_attention_layernorm.weight": "model-00008-of-00031.safetensors",
133
+ "model.layers.18.self_attn.k_proj.bias": "model-00008-of-00031.safetensors",
134
+ "model.layers.18.self_attn.k_proj.weight": "model-00008-of-00031.safetensors",
135
+ "model.layers.18.self_attn.o_proj.weight": "model-00008-of-00031.safetensors",
136
+ "model.layers.18.self_attn.q_proj.bias": "model-00008-of-00031.safetensors",
137
+ "model.layers.18.self_attn.q_proj.weight": "model-00008-of-00031.safetensors",
138
+ "model.layers.18.self_attn.v_proj.bias": "model-00008-of-00031.safetensors",
139
+ "model.layers.18.self_attn.v_proj.weight": "model-00008-of-00031.safetensors",
140
+ "model.layers.19.input_layernorm.weight": "model-00008-of-00031.safetensors",
141
+ "model.layers.19.mlp.down_proj.weight": "model-00008-of-00031.safetensors",
142
+ "model.layers.19.mlp.gate_proj.weight": "model-00008-of-00031.safetensors",
143
+ "model.layers.19.mlp.up_proj.weight": "model-00008-of-00031.safetensors",
144
+ "model.layers.19.post_attention_layernorm.weight": "model-00008-of-00031.safetensors",
145
+ "model.layers.19.self_attn.k_proj.bias": "model-00008-of-00031.safetensors",
146
+ "model.layers.19.self_attn.k_proj.weight": "model-00008-of-00031.safetensors",
147
+ "model.layers.19.self_attn.o_proj.weight": "model-00008-of-00031.safetensors",
148
+ "model.layers.19.self_attn.q_proj.bias": "model-00008-of-00031.safetensors",
149
+ "model.layers.19.self_attn.q_proj.weight": "model-00008-of-00031.safetensors",
150
+ "model.layers.19.self_attn.v_proj.bias": "model-00008-of-00031.safetensors",
151
+ "model.layers.19.self_attn.v_proj.weight": "model-00008-of-00031.safetensors",
152
+ "model.layers.2.input_layernorm.weight": "model-00002-of-00031.safetensors",
153
+ "model.layers.2.mlp.down_proj.weight": "model-00002-of-00031.safetensors",
154
+ "model.layers.2.mlp.gate_proj.weight": "model-00002-of-00031.safetensors",
155
+ "model.layers.2.mlp.up_proj.weight": "model-00002-of-00031.safetensors",
156
+ "model.layers.2.post_attention_layernorm.weight": "model-00002-of-00031.safetensors",
157
+ "model.layers.2.self_attn.k_proj.bias": "model-00002-of-00031.safetensors",
158
+ "model.layers.2.self_attn.k_proj.weight": "model-00002-of-00031.safetensors",
159
+ "model.layers.2.self_attn.o_proj.weight": "model-00002-of-00031.safetensors",
160
+ "model.layers.2.self_attn.q_proj.bias": "model-00002-of-00031.safetensors",
161
+ "model.layers.2.self_attn.q_proj.weight": "model-00002-of-00031.safetensors",
162
+ "model.layers.2.self_attn.v_proj.bias": "model-00002-of-00031.safetensors",
163
+ "model.layers.2.self_attn.v_proj.weight": "model-00002-of-00031.safetensors",
164
+ "model.layers.20.input_layernorm.weight": "model-00009-of-00031.safetensors",
165
+ "model.layers.20.mlp.down_proj.weight": "model-00009-of-00031.safetensors",
166
+ "model.layers.20.mlp.gate_proj.weight": "model-00008-of-00031.safetensors",
167
+ "model.layers.20.mlp.up_proj.weight": "model-00009-of-00031.safetensors",
168
+ "model.layers.20.post_attention_layernorm.weight": "model-00009-of-00031.safetensors",
169
+ "model.layers.20.self_attn.k_proj.bias": "model-00008-of-00031.safetensors",
170
+ "model.layers.20.self_attn.k_proj.weight": "model-00008-of-00031.safetensors",
171
+ "model.layers.20.self_attn.o_proj.weight": "model-00008-of-00031.safetensors",
172
+ "model.layers.20.self_attn.q_proj.bias": "model-00008-of-00031.safetensors",
173
+ "model.layers.20.self_attn.q_proj.weight": "model-00008-of-00031.safetensors",
174
+ "model.layers.20.self_attn.v_proj.bias": "model-00008-of-00031.safetensors",
175
+ "model.layers.20.self_attn.v_proj.weight": "model-00008-of-00031.safetensors",
176
+ "model.layers.21.input_layernorm.weight": "model-00009-of-00031.safetensors",
177
+ "model.layers.21.mlp.down_proj.weight": "model-00009-of-00031.safetensors",
178
+ "model.layers.21.mlp.gate_proj.weight": "model-00009-of-00031.safetensors",
179
+ "model.layers.21.mlp.up_proj.weight": "model-00009-of-00031.safetensors",
180
+ "model.layers.21.post_attention_layernorm.weight": "model-00009-of-00031.safetensors",
181
+ "model.layers.21.self_attn.k_proj.bias": "model-00009-of-00031.safetensors",
182
+ "model.layers.21.self_attn.k_proj.weight": "model-00009-of-00031.safetensors",
183
+ "model.layers.21.self_attn.o_proj.weight": "model-00009-of-00031.safetensors",
184
+ "model.layers.21.self_attn.q_proj.bias": "model-00009-of-00031.safetensors",
185
+ "model.layers.21.self_attn.q_proj.weight": "model-00009-of-00031.safetensors",
186
+ "model.layers.21.self_attn.v_proj.bias": "model-00009-of-00031.safetensors",
187
+ "model.layers.21.self_attn.v_proj.weight": "model-00009-of-00031.safetensors",
188
+ "model.layers.22.input_layernorm.weight": "model-00009-of-00031.safetensors",
189
+ "model.layers.22.mlp.down_proj.weight": "model-00009-of-00031.safetensors",
190
+ "model.layers.22.mlp.gate_proj.weight": "model-00009-of-00031.safetensors",
191
+ "model.layers.22.mlp.up_proj.weight": "model-00009-of-00031.safetensors",
192
+ "model.layers.22.post_attention_layernorm.weight": "model-00009-of-00031.safetensors",
193
+ "model.layers.22.self_attn.k_proj.bias": "model-00009-of-00031.safetensors",
194
+ "model.layers.22.self_attn.k_proj.weight": "model-00009-of-00031.safetensors",
195
+ "model.layers.22.self_attn.o_proj.weight": "model-00009-of-00031.safetensors",
196
+ "model.layers.22.self_attn.q_proj.bias": "model-00009-of-00031.safetensors",
197
+ "model.layers.22.self_attn.q_proj.weight": "model-00009-of-00031.safetensors",
198
+ "model.layers.22.self_attn.v_proj.bias": "model-00009-of-00031.safetensors",
199
+ "model.layers.22.self_attn.v_proj.weight": "model-00009-of-00031.safetensors",
200
+ "model.layers.23.input_layernorm.weight": "model-00010-of-00031.safetensors",
201
+ "model.layers.23.mlp.down_proj.weight": "model-00010-of-00031.safetensors",
202
+ "model.layers.23.mlp.gate_proj.weight": "model-00010-of-00031.safetensors",
203
+ "model.layers.23.mlp.up_proj.weight": "model-00010-of-00031.safetensors",
204
+ "model.layers.23.post_attention_layernorm.weight": "model-00010-of-00031.safetensors",
205
+ "model.layers.23.self_attn.k_proj.bias": "model-00009-of-00031.safetensors",
206
+ "model.layers.23.self_attn.k_proj.weight": "model-00009-of-00031.safetensors",
207
+ "model.layers.23.self_attn.o_proj.weight": "model-00009-of-00031.safetensors",
208
+ "model.layers.23.self_attn.q_proj.bias": "model-00009-of-00031.safetensors",
209
+ "model.layers.23.self_attn.q_proj.weight": "model-00009-of-00031.safetensors",
210
+ "model.layers.23.self_attn.v_proj.bias": "model-00009-of-00031.safetensors",
211
+ "model.layers.23.self_attn.v_proj.weight": "model-00009-of-00031.safetensors",
212
+ "model.layers.24.input_layernorm.weight": "model-00010-of-00031.safetensors",
213
+ "model.layers.24.mlp.down_proj.weight": "model-00010-of-00031.safetensors",
214
+ "model.layers.24.mlp.gate_proj.weight": "model-00010-of-00031.safetensors",
215
+ "model.layers.24.mlp.up_proj.weight": "model-00010-of-00031.safetensors",
216
+ "model.layers.24.post_attention_layernorm.weight": "model-00010-of-00031.safetensors",
217
+ "model.layers.24.self_attn.k_proj.bias": "model-00010-of-00031.safetensors",
218
+ "model.layers.24.self_attn.k_proj.weight": "model-00010-of-00031.safetensors",
219
+ "model.layers.24.self_attn.o_proj.weight": "model-00010-of-00031.safetensors",
220
+ "model.layers.24.self_attn.q_proj.bias": "model-00010-of-00031.safetensors",
221
+ "model.layers.24.self_attn.q_proj.weight": "model-00010-of-00031.safetensors",
222
+ "model.layers.24.self_attn.v_proj.bias": "model-00010-of-00031.safetensors",
223
+ "model.layers.24.self_attn.v_proj.weight": "model-00010-of-00031.safetensors",
224
+ "model.layers.25.input_layernorm.weight": "model-00010-of-00031.safetensors",
225
+ "model.layers.25.mlp.down_proj.weight": "model-00010-of-00031.safetensors",
226
+ "model.layers.25.mlp.gate_proj.weight": "model-00010-of-00031.safetensors",
227
+ "model.layers.25.mlp.up_proj.weight": "model-00010-of-00031.safetensors",
228
+ "model.layers.25.post_attention_layernorm.weight": "model-00010-of-00031.safetensors",
229
+ "model.layers.25.self_attn.k_proj.bias": "model-00010-of-00031.safetensors",
230
+ "model.layers.25.self_attn.k_proj.weight": "model-00010-of-00031.safetensors",
231
+ "model.layers.25.self_attn.o_proj.weight": "model-00010-of-00031.safetensors",
232
+ "model.layers.25.self_attn.q_proj.bias": "model-00010-of-00031.safetensors",
233
+ "model.layers.25.self_attn.q_proj.weight": "model-00010-of-00031.safetensors",
234
+ "model.layers.25.self_attn.v_proj.bias": "model-00010-of-00031.safetensors",
235
+ "model.layers.25.self_attn.v_proj.weight": "model-00010-of-00031.safetensors",
236
+ "model.layers.26.input_layernorm.weight": "model-00011-of-00031.safetensors",
237
+ "model.layers.26.mlp.down_proj.weight": "model-00011-of-00031.safetensors",
238
+ "model.layers.26.mlp.gate_proj.weight": "model-00011-of-00031.safetensors",
239
+ "model.layers.26.mlp.up_proj.weight": "model-00011-of-00031.safetensors",
240
+ "model.layers.26.post_attention_layernorm.weight": "model-00011-of-00031.safetensors",
241
+ "model.layers.26.self_attn.k_proj.bias": "model-00011-of-00031.safetensors",
242
+ "model.layers.26.self_attn.k_proj.weight": "model-00011-of-00031.safetensors",
243
+ "model.layers.26.self_attn.o_proj.weight": "model-00011-of-00031.safetensors",
244
+ "model.layers.26.self_attn.q_proj.bias": "model-00011-of-00031.safetensors",
245
+ "model.layers.26.self_attn.q_proj.weight": "model-00011-of-00031.safetensors",
246
+ "model.layers.26.self_attn.v_proj.bias": "model-00011-of-00031.safetensors",
247
+ "model.layers.26.self_attn.v_proj.weight": "model-00011-of-00031.safetensors",
248
+ "model.layers.27.input_layernorm.weight": "model-00011-of-00031.safetensors",
249
+ "model.layers.27.mlp.down_proj.weight": "model-00011-of-00031.safetensors",
250
+ "model.layers.27.mlp.gate_proj.weight": "model-00011-of-00031.safetensors",
251
+ "model.layers.27.mlp.up_proj.weight": "model-00011-of-00031.safetensors",
252
+ "model.layers.27.post_attention_layernorm.weight": "model-00011-of-00031.safetensors",
253
+ "model.layers.27.self_attn.k_proj.bias": "model-00011-of-00031.safetensors",
254
+ "model.layers.27.self_attn.k_proj.weight": "model-00011-of-00031.safetensors",
255
+ "model.layers.27.self_attn.o_proj.weight": "model-00011-of-00031.safetensors",
256
+ "model.layers.27.self_attn.q_proj.bias": "model-00011-of-00031.safetensors",
257
+ "model.layers.27.self_attn.q_proj.weight": "model-00011-of-00031.safetensors",
258
+ "model.layers.27.self_attn.v_proj.bias": "model-00011-of-00031.safetensors",
259
+ "model.layers.27.self_attn.v_proj.weight": "model-00011-of-00031.safetensors",
260
+ "model.layers.28.input_layernorm.weight": "model-00012-of-00031.safetensors",
261
+ "model.layers.28.mlp.down_proj.weight": "model-00012-of-00031.safetensors",
262
+ "model.layers.28.mlp.gate_proj.weight": "model-00011-of-00031.safetensors",
263
+ "model.layers.28.mlp.up_proj.weight": "model-00011-of-00031.safetensors",
264
+ "model.layers.28.post_attention_layernorm.weight": "model-00012-of-00031.safetensors",
265
+ "model.layers.28.self_attn.k_proj.bias": "model-00011-of-00031.safetensors",
266
+ "model.layers.28.self_attn.k_proj.weight": "model-00011-of-00031.safetensors",
267
+ "model.layers.28.self_attn.o_proj.weight": "model-00011-of-00031.safetensors",
268
+ "model.layers.28.self_attn.q_proj.bias": "model-00011-of-00031.safetensors",
269
+ "model.layers.28.self_attn.q_proj.weight": "model-00011-of-00031.safetensors",
270
+ "model.layers.28.self_attn.v_proj.bias": "model-00011-of-00031.safetensors",
271
+ "model.layers.28.self_attn.v_proj.weight": "model-00011-of-00031.safetensors",
272
+ "model.layers.29.input_layernorm.weight": "model-00012-of-00031.safetensors",
273
+ "model.layers.29.mlp.down_proj.weight": "model-00012-of-00031.safetensors",
274
+ "model.layers.29.mlp.gate_proj.weight": "model-00012-of-00031.safetensors",
275
+ "model.layers.29.mlp.up_proj.weight": "model-00012-of-00031.safetensors",
276
+ "model.layers.29.post_attention_layernorm.weight": "model-00012-of-00031.safetensors",
277
+ "model.layers.29.self_attn.k_proj.bias": "model-00012-of-00031.safetensors",
278
+ "model.layers.29.self_attn.k_proj.weight": "model-00012-of-00031.safetensors",
279
+ "model.layers.29.self_attn.o_proj.weight": "model-00012-of-00031.safetensors",
280
+ "model.layers.29.self_attn.q_proj.bias": "model-00012-of-00031.safetensors",
281
+ "model.layers.29.self_attn.q_proj.weight": "model-00012-of-00031.safetensors",
282
+ "model.layers.29.self_attn.v_proj.bias": "model-00012-of-00031.safetensors",
283
+ "model.layers.29.self_attn.v_proj.weight": "model-00012-of-00031.safetensors",
284
+ "model.layers.3.input_layernorm.weight": "model-00002-of-00031.safetensors",
285
+ "model.layers.3.mlp.down_proj.weight": "model-00002-of-00031.safetensors",
286
+ "model.layers.3.mlp.gate_proj.weight": "model-00002-of-00031.safetensors",
287
+ "model.layers.3.mlp.up_proj.weight": "model-00002-of-00031.safetensors",
288
+ "model.layers.3.post_attention_layernorm.weight": "model-00002-of-00031.safetensors",
289
+ "model.layers.3.self_attn.k_proj.bias": "model-00002-of-00031.safetensors",
290
+ "model.layers.3.self_attn.k_proj.weight": "model-00002-of-00031.safetensors",
291
+ "model.layers.3.self_attn.o_proj.weight": "model-00002-of-00031.safetensors",
292
+ "model.layers.3.self_attn.q_proj.bias": "model-00002-of-00031.safetensors",
293
+ "model.layers.3.self_attn.q_proj.weight": "model-00002-of-00031.safetensors",
294
+ "model.layers.3.self_attn.v_proj.bias": "model-00002-of-00031.safetensors",
295
+ "model.layers.3.self_attn.v_proj.weight": "model-00002-of-00031.safetensors",
296
+ "model.layers.30.input_layernorm.weight": "model-00012-of-00031.safetensors",
297
+ "model.layers.30.mlp.down_proj.weight": "model-00012-of-00031.safetensors",
298
+ "model.layers.30.mlp.gate_proj.weight": "model-00012-of-00031.safetensors",
299
+ "model.layers.30.mlp.up_proj.weight": "model-00012-of-00031.safetensors",
300
+ "model.layers.30.post_attention_layernorm.weight": "model-00012-of-00031.safetensors",
301
+ "model.layers.30.self_attn.k_proj.bias": "model-00012-of-00031.safetensors",
302
+ "model.layers.30.self_attn.k_proj.weight": "model-00012-of-00031.safetensors",
303
+ "model.layers.30.self_attn.o_proj.weight": "model-00012-of-00031.safetensors",
304
+ "model.layers.30.self_attn.q_proj.bias": "model-00012-of-00031.safetensors",
305
+ "model.layers.30.self_attn.q_proj.weight": "model-00012-of-00031.safetensors",
306
+ "model.layers.30.self_attn.v_proj.bias": "model-00012-of-00031.safetensors",
307
+ "model.layers.30.self_attn.v_proj.weight": "model-00012-of-00031.safetensors",
308
+ "model.layers.31.input_layernorm.weight": "model-00013-of-00031.safetensors",
309
+ "model.layers.31.mlp.down_proj.weight": "model-00013-of-00031.safetensors",
310
+ "model.layers.31.mlp.gate_proj.weight": "model-00012-of-00031.safetensors",
311
+ "model.layers.31.mlp.up_proj.weight": "model-00013-of-00031.safetensors",
312
+ "model.layers.31.post_attention_layernorm.weight": "model-00013-of-00031.safetensors",
313
+ "model.layers.31.self_attn.k_proj.bias": "model-00012-of-00031.safetensors",
314
+ "model.layers.31.self_attn.k_proj.weight": "model-00012-of-00031.safetensors",
315
+ "model.layers.31.self_attn.o_proj.weight": "model-00012-of-00031.safetensors",
316
+ "model.layers.31.self_attn.q_proj.bias": "model-00012-of-00031.safetensors",
317
+ "model.layers.31.self_attn.q_proj.weight": "model-00012-of-00031.safetensors",
318
+ "model.layers.31.self_attn.v_proj.bias": "model-00012-of-00031.safetensors",
319
+ "model.layers.31.self_attn.v_proj.weight": "model-00012-of-00031.safetensors",
320
+ "model.layers.32.input_layernorm.weight": "model-00013-of-00031.safetensors",
321
+ "model.layers.32.mlp.down_proj.weight": "model-00013-of-00031.safetensors",
322
+ "model.layers.32.mlp.gate_proj.weight": "model-00013-of-00031.safetensors",
323
+ "model.layers.32.mlp.up_proj.weight": "model-00013-of-00031.safetensors",
324
+ "model.layers.32.post_attention_layernorm.weight": "model-00013-of-00031.safetensors",
325
+ "model.layers.32.self_attn.k_proj.bias": "model-00013-of-00031.safetensors",
326
+ "model.layers.32.self_attn.k_proj.weight": "model-00013-of-00031.safetensors",
327
+ "model.layers.32.self_attn.o_proj.weight": "model-00013-of-00031.safetensors",
328
+ "model.layers.32.self_attn.q_proj.bias": "model-00013-of-00031.safetensors",
329
+ "model.layers.32.self_attn.q_proj.weight": "model-00013-of-00031.safetensors",
330
+ "model.layers.32.self_attn.v_proj.bias": "model-00013-of-00031.safetensors",
331
+ "model.layers.32.self_attn.v_proj.weight": "model-00013-of-00031.safetensors",
332
+ "model.layers.33.input_layernorm.weight": "model-00013-of-00031.safetensors",
333
+ "model.layers.33.mlp.down_proj.weight": "model-00013-of-00031.safetensors",
334
+ "model.layers.33.mlp.gate_proj.weight": "model-00013-of-00031.safetensors",
335
+ "model.layers.33.mlp.up_proj.weight": "model-00013-of-00031.safetensors",
336
+ "model.layers.33.post_attention_layernorm.weight": "model-00013-of-00031.safetensors",
337
+ "model.layers.33.self_attn.k_proj.bias": "model-00013-of-00031.safetensors",
338
+ "model.layers.33.self_attn.k_proj.weight": "model-00013-of-00031.safetensors",
339
+ "model.layers.33.self_attn.o_proj.weight": "model-00013-of-00031.safetensors",
340
+ "model.layers.33.self_attn.q_proj.bias": "model-00013-of-00031.safetensors",
341
+ "model.layers.33.self_attn.q_proj.weight": "model-00013-of-00031.safetensors",
342
+ "model.layers.33.self_attn.v_proj.bias": "model-00013-of-00031.safetensors",
343
+ "model.layers.33.self_attn.v_proj.weight": "model-00013-of-00031.safetensors",
344
+ "model.layers.34.input_layernorm.weight": "model-00014-of-00031.safetensors",
345
+ "model.layers.34.mlp.down_proj.weight": "model-00014-of-00031.safetensors",
346
+ "model.layers.34.mlp.gate_proj.weight": "model-00014-of-00031.safetensors",
347
+ "model.layers.34.mlp.up_proj.weight": "model-00014-of-00031.safetensors",
348
+ "model.layers.34.post_attention_layernorm.weight": "model-00014-of-00031.safetensors",
349
+ "model.layers.34.self_attn.k_proj.bias": "model-00013-of-00031.safetensors",
350
+ "model.layers.34.self_attn.k_proj.weight": "model-00013-of-00031.safetensors",
351
+ "model.layers.34.self_attn.o_proj.weight": "model-00013-of-00031.safetensors",
352
+ "model.layers.34.self_attn.q_proj.bias": "model-00013-of-00031.safetensors",
353
+ "model.layers.34.self_attn.q_proj.weight": "model-00013-of-00031.safetensors",
354
+ "model.layers.34.self_attn.v_proj.bias": "model-00013-of-00031.safetensors",
355
+ "model.layers.34.self_attn.v_proj.weight": "model-00013-of-00031.safetensors",
356
+ "model.layers.35.input_layernorm.weight": "model-00014-of-00031.safetensors",
357
+ "model.layers.35.mlp.down_proj.weight": "model-00014-of-00031.safetensors",
358
+ "model.layers.35.mlp.gate_proj.weight": "model-00014-of-00031.safetensors",
359
+ "model.layers.35.mlp.up_proj.weight": "model-00014-of-00031.safetensors",
360
+ "model.layers.35.post_attention_layernorm.weight": "model-00014-of-00031.safetensors",
361
+ "model.layers.35.self_attn.k_proj.bias": "model-00014-of-00031.safetensors",
362
+ "model.layers.35.self_attn.k_proj.weight": "model-00014-of-00031.safetensors",
363
+ "model.layers.35.self_attn.o_proj.weight": "model-00014-of-00031.safetensors",
364
+ "model.layers.35.self_attn.q_proj.bias": "model-00014-of-00031.safetensors",
365
+ "model.layers.35.self_attn.q_proj.weight": "model-00014-of-00031.safetensors",
366
+ "model.layers.35.self_attn.v_proj.bias": "model-00014-of-00031.safetensors",
367
+ "model.layers.35.self_attn.v_proj.weight": "model-00014-of-00031.safetensors",
368
+ "model.layers.36.input_layernorm.weight": "model-00014-of-00031.safetensors",
369
+ "model.layers.36.mlp.down_proj.weight": "model-00014-of-00031.safetensors",
370
+ "model.layers.36.mlp.gate_proj.weight": "model-00014-of-00031.safetensors",
371
+ "model.layers.36.mlp.up_proj.weight": "model-00014-of-00031.safetensors",
372
+ "model.layers.36.post_attention_layernorm.weight": "model-00014-of-00031.safetensors",
373
+ "model.layers.36.self_attn.k_proj.bias": "model-00014-of-00031.safetensors",
374
+ "model.layers.36.self_attn.k_proj.weight": "model-00014-of-00031.safetensors",
375
+ "model.layers.36.self_attn.o_proj.weight": "model-00014-of-00031.safetensors",
376
+ "model.layers.36.self_attn.q_proj.bias": "model-00014-of-00031.safetensors",
377
+ "model.layers.36.self_attn.q_proj.weight": "model-00014-of-00031.safetensors",
378
+ "model.layers.36.self_attn.v_proj.bias": "model-00014-of-00031.safetensors",
379
+ "model.layers.36.self_attn.v_proj.weight": "model-00014-of-00031.safetensors",
380
+ "model.layers.37.input_layernorm.weight": "model-00015-of-00031.safetensors",
381
+ "model.layers.37.mlp.down_proj.weight": "model-00015-of-00031.safetensors",
382
+ "model.layers.37.mlp.gate_proj.weight": "model-00015-of-00031.safetensors",
383
+ "model.layers.37.mlp.up_proj.weight": "model-00015-of-00031.safetensors",
384
+ "model.layers.37.post_attention_layernorm.weight": "model-00015-of-00031.safetensors",
385
+ "model.layers.37.self_attn.k_proj.bias": "model-00015-of-00031.safetensors",
386
+ "model.layers.37.self_attn.k_proj.weight": "model-00015-of-00031.safetensors",
387
+ "model.layers.37.self_attn.o_proj.weight": "model-00015-of-00031.safetensors",
388
+ "model.layers.37.self_attn.q_proj.bias": "model-00015-of-00031.safetensors",
389
+ "model.layers.37.self_attn.q_proj.weight": "model-00015-of-00031.safetensors",
390
+ "model.layers.37.self_attn.v_proj.bias": "model-00015-of-00031.safetensors",
391
+ "model.layers.37.self_attn.v_proj.weight": "model-00015-of-00031.safetensors",
392
+ "model.layers.38.input_layernorm.weight": "model-00015-of-00031.safetensors",
393
+ "model.layers.38.mlp.down_proj.weight": "model-00015-of-00031.safetensors",
394
+ "model.layers.38.mlp.gate_proj.weight": "model-00015-of-00031.safetensors",
395
+ "model.layers.38.mlp.up_proj.weight": "model-00015-of-00031.safetensors",
396
+ "model.layers.38.post_attention_layernorm.weight": "model-00015-of-00031.safetensors",
397
+ "model.layers.38.self_attn.k_proj.bias": "model-00015-of-00031.safetensors",
398
+ "model.layers.38.self_attn.k_proj.weight": "model-00015-of-00031.safetensors",
399
+ "model.layers.38.self_attn.o_proj.weight": "model-00015-of-00031.safetensors",
400
+ "model.layers.38.self_attn.q_proj.bias": "model-00015-of-00031.safetensors",
401
+ "model.layers.38.self_attn.q_proj.weight": "model-00015-of-00031.safetensors",
402
+ "model.layers.38.self_attn.v_proj.bias": "model-00015-of-00031.safetensors",
403
+ "model.layers.38.self_attn.v_proj.weight": "model-00015-of-00031.safetensors",
404
+ "model.layers.39.input_layernorm.weight": "model-00016-of-00031.safetensors",
405
+ "model.layers.39.mlp.down_proj.weight": "model-00016-of-00031.safetensors",
406
+ "model.layers.39.mlp.gate_proj.weight": "model-00015-of-00031.safetensors",
407
+ "model.layers.39.mlp.up_proj.weight": "model-00015-of-00031.safetensors",
408
+ "model.layers.39.post_attention_layernorm.weight": "model-00016-of-00031.safetensors",
409
+ "model.layers.39.self_attn.k_proj.bias": "model-00015-of-00031.safetensors",
410
+ "model.layers.39.self_attn.k_proj.weight": "model-00015-of-00031.safetensors",
411
+ "model.layers.39.self_attn.o_proj.weight": "model-00015-of-00031.safetensors",
412
+ "model.layers.39.self_attn.q_proj.bias": "model-00015-of-00031.safetensors",
413
+ "model.layers.39.self_attn.q_proj.weight": "model-00015-of-00031.safetensors",
414
+ "model.layers.39.self_attn.v_proj.bias": "model-00015-of-00031.safetensors",
415
+ "model.layers.39.self_attn.v_proj.weight": "model-00015-of-00031.safetensors",
416
+ "model.layers.4.input_layernorm.weight": "model-00003-of-00031.safetensors",
417
+ "model.layers.4.mlp.down_proj.weight": "model-00003-of-00031.safetensors",
418
+ "model.layers.4.mlp.gate_proj.weight": "model-00003-of-00031.safetensors",
419
+ "model.layers.4.mlp.up_proj.weight": "model-00003-of-00031.safetensors",
420
+ "model.layers.4.post_attention_layernorm.weight": "model-00003-of-00031.safetensors",
421
+ "model.layers.4.self_attn.k_proj.bias": "model-00003-of-00031.safetensors",
422
+ "model.layers.4.self_attn.k_proj.weight": "model-00003-of-00031.safetensors",
423
+ "model.layers.4.self_attn.o_proj.weight": "model-00003-of-00031.safetensors",
424
+ "model.layers.4.self_attn.q_proj.bias": "model-00003-of-00031.safetensors",
425
+ "model.layers.4.self_attn.q_proj.weight": "model-00003-of-00031.safetensors",
426
+ "model.layers.4.self_attn.v_proj.bias": "model-00003-of-00031.safetensors",
427
+ "model.layers.4.self_attn.v_proj.weight": "model-00003-of-00031.safetensors",
428
+ "model.layers.40.input_layernorm.weight": "model-00016-of-00031.safetensors",
429
+ "model.layers.40.mlp.down_proj.weight": "model-00016-of-00031.safetensors",
430
+ "model.layers.40.mlp.gate_proj.weight": "model-00016-of-00031.safetensors",
431
+ "model.layers.40.mlp.up_proj.weight": "model-00016-of-00031.safetensors",
432
+ "model.layers.40.post_attention_layernorm.weight": "model-00016-of-00031.safetensors",
433
+ "model.layers.40.self_attn.k_proj.bias": "model-00016-of-00031.safetensors",
434
+ "model.layers.40.self_attn.k_proj.weight": "model-00016-of-00031.safetensors",
435
+ "model.layers.40.self_attn.o_proj.weight": "model-00016-of-00031.safetensors",
436
+ "model.layers.40.self_attn.q_proj.bias": "model-00016-of-00031.safetensors",
437
+ "model.layers.40.self_attn.q_proj.weight": "model-00016-of-00031.safetensors",
438
+ "model.layers.40.self_attn.v_proj.bias": "model-00016-of-00031.safetensors",
439
+ "model.layers.40.self_attn.v_proj.weight": "model-00016-of-00031.safetensors",
440
+ "model.layers.41.input_layernorm.weight": "model-00016-of-00031.safetensors",
441
+ "model.layers.41.mlp.down_proj.weight": "model-00016-of-00031.safetensors",
442
+ "model.layers.41.mlp.gate_proj.weight": "model-00016-of-00031.safetensors",
443
+ "model.layers.41.mlp.up_proj.weight": "model-00016-of-00031.safetensors",
444
+ "model.layers.41.post_attention_layernorm.weight": "model-00016-of-00031.safetensors",
445
+ "model.layers.41.self_attn.k_proj.bias": "model-00016-of-00031.safetensors",
446
+ "model.layers.41.self_attn.k_proj.weight": "model-00016-of-00031.safetensors",
447
+ "model.layers.41.self_attn.o_proj.weight": "model-00016-of-00031.safetensors",
448
+ "model.layers.41.self_attn.q_proj.bias": "model-00016-of-00031.safetensors",
449
+ "model.layers.41.self_attn.q_proj.weight": "model-00016-of-00031.safetensors",
450
+ "model.layers.41.self_attn.v_proj.bias": "model-00016-of-00031.safetensors",
451
+ "model.layers.41.self_attn.v_proj.weight": "model-00016-of-00031.safetensors",
452
+ "model.layers.42.input_layernorm.weight": "model-00017-of-00031.safetensors",
453
+ "model.layers.42.mlp.down_proj.weight": "model-00017-of-00031.safetensors",
454
+ "model.layers.42.mlp.gate_proj.weight": "model-00016-of-00031.safetensors",
455
+ "model.layers.42.mlp.up_proj.weight": "model-00017-of-00031.safetensors",
456
+ "model.layers.42.post_attention_layernorm.weight": "model-00017-of-00031.safetensors",
457
+ "model.layers.42.self_attn.k_proj.bias": "model-00016-of-00031.safetensors",
458
+ "model.layers.42.self_attn.k_proj.weight": "model-00016-of-00031.safetensors",
459
+ "model.layers.42.self_attn.o_proj.weight": "model-00016-of-00031.safetensors",
460
+ "model.layers.42.self_attn.q_proj.bias": "model-00016-of-00031.safetensors",
461
+ "model.layers.42.self_attn.q_proj.weight": "model-00016-of-00031.safetensors",
462
+ "model.layers.42.self_attn.v_proj.bias": "model-00016-of-00031.safetensors",
463
+ "model.layers.42.self_attn.v_proj.weight": "model-00016-of-00031.safetensors",
464
+ "model.layers.43.input_layernorm.weight": "model-00017-of-00031.safetensors",
465
+ "model.layers.43.mlp.down_proj.weight": "model-00017-of-00031.safetensors",
466
+ "model.layers.43.mlp.gate_proj.weight": "model-00017-of-00031.safetensors",
467
+ "model.layers.43.mlp.up_proj.weight": "model-00017-of-00031.safetensors",
468
+ "model.layers.43.post_attention_layernorm.weight": "model-00017-of-00031.safetensors",
469
+ "model.layers.43.self_attn.k_proj.bias": "model-00017-of-00031.safetensors",
470
+ "model.layers.43.self_attn.k_proj.weight": "model-00017-of-00031.safetensors",
471
+ "model.layers.43.self_attn.o_proj.weight": "model-00017-of-00031.safetensors",
472
+ "model.layers.43.self_attn.q_proj.bias": "model-00017-of-00031.safetensors",
473
+ "model.layers.43.self_attn.q_proj.weight": "model-00017-of-00031.safetensors",
474
+ "model.layers.43.self_attn.v_proj.bias": "model-00017-of-00031.safetensors",
475
+ "model.layers.43.self_attn.v_proj.weight": "model-00017-of-00031.safetensors",
476
+ "model.layers.44.input_layernorm.weight": "model-00017-of-00031.safetensors",
477
+ "model.layers.44.mlp.down_proj.weight": "model-00017-of-00031.safetensors",
478
+ "model.layers.44.mlp.gate_proj.weight": "model-00017-of-00031.safetensors",
479
+ "model.layers.44.mlp.up_proj.weight": "model-00017-of-00031.safetensors",
480
+ "model.layers.44.post_attention_layernorm.weight": "model-00017-of-00031.safetensors",
481
+ "model.layers.44.self_attn.k_proj.bias": "model-00017-of-00031.safetensors",
482
+ "model.layers.44.self_attn.k_proj.weight": "model-00017-of-00031.safetensors",
483
+ "model.layers.44.self_attn.o_proj.weight": "model-00017-of-00031.safetensors",
484
+ "model.layers.44.self_attn.q_proj.bias": "model-00017-of-00031.safetensors",
485
+ "model.layers.44.self_attn.q_proj.weight": "model-00017-of-00031.safetensors",
486
+ "model.layers.44.self_attn.v_proj.bias": "model-00017-of-00031.safetensors",
487
+ "model.layers.44.self_attn.v_proj.weight": "model-00017-of-00031.safetensors",
488
+ "model.layers.45.input_layernorm.weight": "model-00018-of-00031.safetensors",
489
+ "model.layers.45.mlp.down_proj.weight": "model-00018-of-00031.safetensors",
490
+ "model.layers.45.mlp.gate_proj.weight": "model-00018-of-00031.safetensors",
491
+ "model.layers.45.mlp.up_proj.weight": "model-00018-of-00031.safetensors",
492
+ "model.layers.45.post_attention_layernorm.weight": "model-00018-of-00031.safetensors",
493
+ "model.layers.45.self_attn.k_proj.bias": "model-00017-of-00031.safetensors",
494
+ "model.layers.45.self_attn.k_proj.weight": "model-00017-of-00031.safetensors",
495
+ "model.layers.45.self_attn.o_proj.weight": "model-00017-of-00031.safetensors",
496
+ "model.layers.45.self_attn.q_proj.bias": "model-00017-of-00031.safetensors",
497
+ "model.layers.45.self_attn.q_proj.weight": "model-00017-of-00031.safetensors",
498
+ "model.layers.45.self_attn.v_proj.bias": "model-00017-of-00031.safetensors",
499
+ "model.layers.45.self_attn.v_proj.weight": "model-00017-of-00031.safetensors",
500
+ "model.layers.46.input_layernorm.weight": "model-00018-of-00031.safetensors",
501
+ "model.layers.46.mlp.down_proj.weight": "model-00018-of-00031.safetensors",
502
+ "model.layers.46.mlp.gate_proj.weight": "model-00018-of-00031.safetensors",
503
+ "model.layers.46.mlp.up_proj.weight": "model-00018-of-00031.safetensors",
504
+ "model.layers.46.post_attention_layernorm.weight": "model-00018-of-00031.safetensors",
505
+ "model.layers.46.self_attn.k_proj.bias": "model-00018-of-00031.safetensors",
506
+ "model.layers.46.self_attn.k_proj.weight": "model-00018-of-00031.safetensors",
507
+ "model.layers.46.self_attn.o_proj.weight": "model-00018-of-00031.safetensors",
508
+ "model.layers.46.self_attn.q_proj.bias": "model-00018-of-00031.safetensors",
509
+ "model.layers.46.self_attn.q_proj.weight": "model-00018-of-00031.safetensors",
510
+ "model.layers.46.self_attn.v_proj.bias": "model-00018-of-00031.safetensors",
511
+ "model.layers.46.self_attn.v_proj.weight": "model-00018-of-00031.safetensors",
512
+ "model.layers.47.input_layernorm.weight": "model-00018-of-00031.safetensors",
513
+ "model.layers.47.mlp.down_proj.weight": "model-00018-of-00031.safetensors",
514
+ "model.layers.47.mlp.gate_proj.weight": "model-00018-of-00031.safetensors",
515
+ "model.layers.47.mlp.up_proj.weight": "model-00018-of-00031.safetensors",
516
+ "model.layers.47.post_attention_layernorm.weight": "model-00018-of-00031.safetensors",
517
+ "model.layers.47.self_attn.k_proj.bias": "model-00018-of-00031.safetensors",
518
+ "model.layers.47.self_attn.k_proj.weight": "model-00018-of-00031.safetensors",
519
+ "model.layers.47.self_attn.o_proj.weight": "model-00018-of-00031.safetensors",
520
+ "model.layers.47.self_attn.q_proj.bias": "model-00018-of-00031.safetensors",
521
+ "model.layers.47.self_attn.q_proj.weight": "model-00018-of-00031.safetensors",
522
+ "model.layers.47.self_attn.v_proj.bias": "model-00018-of-00031.safetensors",
523
+ "model.layers.47.self_attn.v_proj.weight": "model-00018-of-00031.safetensors",
524
+ "model.layers.48.input_layernorm.weight": "model-00019-of-00031.safetensors",
525
+ "model.layers.48.mlp.down_proj.weight": "model-00019-of-00031.safetensors",
526
+ "model.layers.48.mlp.gate_proj.weight": "model-00019-of-00031.safetensors",
527
+ "model.layers.48.mlp.up_proj.weight": "model-00019-of-00031.safetensors",
528
+ "model.layers.48.post_attention_layernorm.weight": "model-00019-of-00031.safetensors",
529
+ "model.layers.48.self_attn.k_proj.bias": "model-00019-of-00031.safetensors",
530
+ "model.layers.48.self_attn.k_proj.weight": "model-00019-of-00031.safetensors",
531
+ "model.layers.48.self_attn.o_proj.weight": "model-00019-of-00031.safetensors",
532
+ "model.layers.48.self_attn.q_proj.bias": "model-00019-of-00031.safetensors",
533
+ "model.layers.48.self_attn.q_proj.weight": "model-00019-of-00031.safetensors",
534
+ "model.layers.48.self_attn.v_proj.bias": "model-00019-of-00031.safetensors",
535
+ "model.layers.48.self_attn.v_proj.weight": "model-00019-of-00031.safetensors",
536
+ "model.layers.49.input_layernorm.weight": "model-00019-of-00031.safetensors",
537
+ "model.layers.49.mlp.down_proj.weight": "model-00019-of-00031.safetensors",
538
+ "model.layers.49.mlp.gate_proj.weight": "model-00019-of-00031.safetensors",
539
+ "model.layers.49.mlp.up_proj.weight": "model-00019-of-00031.safetensors",
540
+ "model.layers.49.post_attention_layernorm.weight": "model-00019-of-00031.safetensors",
541
+ "model.layers.49.self_attn.k_proj.bias": "model-00019-of-00031.safetensors",
542
+ "model.layers.49.self_attn.k_proj.weight": "model-00019-of-00031.safetensors",
543
+ "model.layers.49.self_attn.o_proj.weight": "model-00019-of-00031.safetensors",
544
+ "model.layers.49.self_attn.q_proj.bias": "model-00019-of-00031.safetensors",
545
+ "model.layers.49.self_attn.q_proj.weight": "model-00019-of-00031.safetensors",
546
+ "model.layers.49.self_attn.v_proj.bias": "model-00019-of-00031.safetensors",
547
+ "model.layers.49.self_attn.v_proj.weight": "model-00019-of-00031.safetensors",
548
+ "model.layers.5.input_layernorm.weight": "model-00003-of-00031.safetensors",
549
+ "model.layers.5.mlp.down_proj.weight": "model-00003-of-00031.safetensors",
550
+ "model.layers.5.mlp.gate_proj.weight": "model-00003-of-00031.safetensors",
551
+ "model.layers.5.mlp.up_proj.weight": "model-00003-of-00031.safetensors",
552
+ "model.layers.5.post_attention_layernorm.weight": "model-00003-of-00031.safetensors",
553
+ "model.layers.5.self_attn.k_proj.bias": "model-00003-of-00031.safetensors",
554
+ "model.layers.5.self_attn.k_proj.weight": "model-00003-of-00031.safetensors",
555
+ "model.layers.5.self_attn.o_proj.weight": "model-00003-of-00031.safetensors",
556
+ "model.layers.5.self_attn.q_proj.bias": "model-00003-of-00031.safetensors",
557
+ "model.layers.5.self_attn.q_proj.weight": "model-00003-of-00031.safetensors",
558
+ "model.layers.5.self_attn.v_proj.bias": "model-00003-of-00031.safetensors",
559
+ "model.layers.5.self_attn.v_proj.weight": "model-00003-of-00031.safetensors",
560
+ "model.layers.50.input_layernorm.weight": "model-00020-of-00031.safetensors",
561
+ "model.layers.50.mlp.down_proj.weight": "model-00020-of-00031.safetensors",
562
+ "model.layers.50.mlp.gate_proj.weight": "model-00019-of-00031.safetensors",
563
+ "model.layers.50.mlp.up_proj.weight": "model-00019-of-00031.safetensors",
564
+ "model.layers.50.post_attention_layernorm.weight": "model-00020-of-00031.safetensors",
565
+ "model.layers.50.self_attn.k_proj.bias": "model-00019-of-00031.safetensors",
566
+ "model.layers.50.self_attn.k_proj.weight": "model-00019-of-00031.safetensors",
567
+ "model.layers.50.self_attn.o_proj.weight": "model-00019-of-00031.safetensors",
568
+ "model.layers.50.self_attn.q_proj.bias": "model-00019-of-00031.safetensors",
569
+ "model.layers.50.self_attn.q_proj.weight": "model-00019-of-00031.safetensors",
570
+ "model.layers.50.self_attn.v_proj.bias": "model-00019-of-00031.safetensors",
571
+ "model.layers.50.self_attn.v_proj.weight": "model-00019-of-00031.safetensors",
572
+ "model.layers.51.input_layernorm.weight": "model-00020-of-00031.safetensors",
573
+ "model.layers.51.mlp.down_proj.weight": "model-00020-of-00031.safetensors",
574
+ "model.layers.51.mlp.gate_proj.weight": "model-00020-of-00031.safetensors",
575
+ "model.layers.51.mlp.up_proj.weight": "model-00020-of-00031.safetensors",
576
+ "model.layers.51.post_attention_layernorm.weight": "model-00020-of-00031.safetensors",
577
+ "model.layers.51.self_attn.k_proj.bias": "model-00020-of-00031.safetensors",
578
+ "model.layers.51.self_attn.k_proj.weight": "model-00020-of-00031.safetensors",
579
+ "model.layers.51.self_attn.o_proj.weight": "model-00020-of-00031.safetensors",
580
+ "model.layers.51.self_attn.q_proj.bias": "model-00020-of-00031.safetensors",
581
+ "model.layers.51.self_attn.q_proj.weight": "model-00020-of-00031.safetensors",
582
+ "model.layers.51.self_attn.v_proj.bias": "model-00020-of-00031.safetensors",
583
+ "model.layers.51.self_attn.v_proj.weight": "model-00020-of-00031.safetensors",
584
+ "model.layers.52.input_layernorm.weight": "model-00020-of-00031.safetensors",
585
+ "model.layers.52.mlp.down_proj.weight": "model-00020-of-00031.safetensors",
586
+ "model.layers.52.mlp.gate_proj.weight": "model-00020-of-00031.safetensors",
587
+ "model.layers.52.mlp.up_proj.weight": "model-00020-of-00031.safetensors",
588
+ "model.layers.52.post_attention_layernorm.weight": "model-00020-of-00031.safetensors",
589
+ "model.layers.52.self_attn.k_proj.bias": "model-00020-of-00031.safetensors",
590
+ "model.layers.52.self_attn.k_proj.weight": "model-00020-of-00031.safetensors",
591
+ "model.layers.52.self_attn.o_proj.weight": "model-00020-of-00031.safetensors",
592
+ "model.layers.52.self_attn.q_proj.bias": "model-00020-of-00031.safetensors",
593
+ "model.layers.52.self_attn.q_proj.weight": "model-00020-of-00031.safetensors",
594
+ "model.layers.52.self_attn.v_proj.bias": "model-00020-of-00031.safetensors",
595
+ "model.layers.52.self_attn.v_proj.weight": "model-00020-of-00031.safetensors",
596
+ "model.layers.53.input_layernorm.weight": "model-00021-of-00031.safetensors",
597
+ "model.layers.53.mlp.down_proj.weight": "model-00021-of-00031.safetensors",
598
+ "model.layers.53.mlp.gate_proj.weight": "model-00020-of-00031.safetensors",
599
+ "model.layers.53.mlp.up_proj.weight": "model-00021-of-00031.safetensors",
600
+ "model.layers.53.post_attention_layernorm.weight": "model-00021-of-00031.safetensors",
601
+ "model.layers.53.self_attn.k_proj.bias": "model-00020-of-00031.safetensors",
602
+ "model.layers.53.self_attn.k_proj.weight": "model-00020-of-00031.safetensors",
603
+ "model.layers.53.self_attn.o_proj.weight": "model-00020-of-00031.safetensors",
604
+ "model.layers.53.self_attn.q_proj.bias": "model-00020-of-00031.safetensors",
605
+ "model.layers.53.self_attn.q_proj.weight": "model-00020-of-00031.safetensors",
606
+ "model.layers.53.self_attn.v_proj.bias": "model-00020-of-00031.safetensors",
607
+ "model.layers.53.self_attn.v_proj.weight": "model-00020-of-00031.safetensors",
608
+ "model.layers.54.input_layernorm.weight": "model-00021-of-00031.safetensors",
609
+ "model.layers.54.mlp.down_proj.weight": "model-00021-of-00031.safetensors",
610
+ "model.layers.54.mlp.gate_proj.weight": "model-00021-of-00031.safetensors",
611
+ "model.layers.54.mlp.up_proj.weight": "model-00021-of-00031.safetensors",
612
+ "model.layers.54.post_attention_layernorm.weight": "model-00021-of-00031.safetensors",
613
+ "model.layers.54.self_attn.k_proj.bias": "model-00021-of-00031.safetensors",
614
+ "model.layers.54.self_attn.k_proj.weight": "model-00021-of-00031.safetensors",
615
+ "model.layers.54.self_attn.o_proj.weight": "model-00021-of-00031.safetensors",
616
+ "model.layers.54.self_attn.q_proj.bias": "model-00021-of-00031.safetensors",
617
+ "model.layers.54.self_attn.q_proj.weight": "model-00021-of-00031.safetensors",
618
+ "model.layers.54.self_attn.v_proj.bias": "model-00021-of-00031.safetensors",
619
+ "model.layers.54.self_attn.v_proj.weight": "model-00021-of-00031.safetensors",
620
+ "model.layers.55.input_layernorm.weight": "model-00021-of-00031.safetensors",
621
+ "model.layers.55.mlp.down_proj.weight": "model-00021-of-00031.safetensors",
622
+ "model.layers.55.mlp.gate_proj.weight": "model-00021-of-00031.safetensors",
623
+ "model.layers.55.mlp.up_proj.weight": "model-00021-of-00031.safetensors",
624
+ "model.layers.55.post_attention_layernorm.weight": "model-00021-of-00031.safetensors",
625
+ "model.layers.55.self_attn.k_proj.bias": "model-00021-of-00031.safetensors",
626
+ "model.layers.55.self_attn.k_proj.weight": "model-00021-of-00031.safetensors",
627
+ "model.layers.55.self_attn.o_proj.weight": "model-00021-of-00031.safetensors",
628
+ "model.layers.55.self_attn.q_proj.bias": "model-00021-of-00031.safetensors",
629
+ "model.layers.55.self_attn.q_proj.weight": "model-00021-of-00031.safetensors",
630
+ "model.layers.55.self_attn.v_proj.bias": "model-00021-of-00031.safetensors",
631
+ "model.layers.55.self_attn.v_proj.weight": "model-00021-of-00031.safetensors",
632
+ "model.layers.56.input_layernorm.weight": "model-00022-of-00031.safetensors",
633
+ "model.layers.56.mlp.down_proj.weight": "model-00022-of-00031.safetensors",
634
+ "model.layers.56.mlp.gate_proj.weight": "model-00022-of-00031.safetensors",
635
+ "model.layers.56.mlp.up_proj.weight": "model-00022-of-00031.safetensors",
636
+ "model.layers.56.post_attention_layernorm.weight": "model-00022-of-00031.safetensors",
637
+ "model.layers.56.self_attn.k_proj.bias": "model-00021-of-00031.safetensors",
638
+ "model.layers.56.self_attn.k_proj.weight": "model-00021-of-00031.safetensors",
639
+ "model.layers.56.self_attn.o_proj.weight": "model-00021-of-00031.safetensors",
640
+ "model.layers.56.self_attn.q_proj.bias": "model-00021-of-00031.safetensors",
641
+ "model.layers.56.self_attn.q_proj.weight": "model-00021-of-00031.safetensors",
642
+ "model.layers.56.self_attn.v_proj.bias": "model-00021-of-00031.safetensors",
643
+ "model.layers.56.self_attn.v_proj.weight": "model-00021-of-00031.safetensors",
644
+ "model.layers.57.input_layernorm.weight": "model-00022-of-00031.safetensors",
645
+ "model.layers.57.mlp.down_proj.weight": "model-00022-of-00031.safetensors",
646
+ "model.layers.57.mlp.gate_proj.weight": "model-00022-of-00031.safetensors",
647
+ "model.layers.57.mlp.up_proj.weight": "model-00022-of-00031.safetensors",
648
+ "model.layers.57.post_attention_layernorm.weight": "model-00022-of-00031.safetensors",
649
+ "model.layers.57.self_attn.k_proj.bias": "model-00022-of-00031.safetensors",
650
+ "model.layers.57.self_attn.k_proj.weight": "model-00022-of-00031.safetensors",
651
+ "model.layers.57.self_attn.o_proj.weight": "model-00022-of-00031.safetensors",
652
+ "model.layers.57.self_attn.q_proj.bias": "model-00022-of-00031.safetensors",
653
+ "model.layers.57.self_attn.q_proj.weight": "model-00022-of-00031.safetensors",
654
+ "model.layers.57.self_attn.v_proj.bias": "model-00022-of-00031.safetensors",
655
+ "model.layers.57.self_attn.v_proj.weight": "model-00022-of-00031.safetensors",
656
+ "model.layers.58.input_layernorm.weight": "model-00022-of-00031.safetensors",
657
+ "model.layers.58.mlp.down_proj.weight": "model-00022-of-00031.safetensors",
658
+ "model.layers.58.mlp.gate_proj.weight": "model-00022-of-00031.safetensors",
659
+ "model.layers.58.mlp.up_proj.weight": "model-00022-of-00031.safetensors",
660
+ "model.layers.58.post_attention_layernorm.weight": "model-00022-of-00031.safetensors",
661
+ "model.layers.58.self_attn.k_proj.bias": "model-00022-of-00031.safetensors",
662
+ "model.layers.58.self_attn.k_proj.weight": "model-00022-of-00031.safetensors",
663
+ "model.layers.58.self_attn.o_proj.weight": "model-00022-of-00031.safetensors",
664
+ "model.layers.58.self_attn.q_proj.bias": "model-00022-of-00031.safetensors",
665
+ "model.layers.58.self_attn.q_proj.weight": "model-00022-of-00031.safetensors",
666
+ "model.layers.58.self_attn.v_proj.bias": "model-00022-of-00031.safetensors",
667
+ "model.layers.58.self_attn.v_proj.weight": "model-00022-of-00031.safetensors",
668
+ "model.layers.59.input_layernorm.weight": "model-00023-of-00031.safetensors",
669
+ "model.layers.59.mlp.down_proj.weight": "model-00023-of-00031.safetensors",
670
+ "model.layers.59.mlp.gate_proj.weight": "model-00023-of-00031.safetensors",
671
+ "model.layers.59.mlp.up_proj.weight": "model-00023-of-00031.safetensors",
672
+ "model.layers.59.post_attention_layernorm.weight": "model-00023-of-00031.safetensors",
673
+ "model.layers.59.self_attn.k_proj.bias": "model-00023-of-00031.safetensors",
674
+ "model.layers.59.self_attn.k_proj.weight": "model-00023-of-00031.safetensors",
675
+ "model.layers.59.self_attn.o_proj.weight": "model-00023-of-00031.safetensors",
676
+ "model.layers.59.self_attn.q_proj.bias": "model-00023-of-00031.safetensors",
677
+ "model.layers.59.self_attn.q_proj.weight": "model-00023-of-00031.safetensors",
678
+ "model.layers.59.self_attn.v_proj.bias": "model-00023-of-00031.safetensors",
679
+ "model.layers.59.self_attn.v_proj.weight": "model-00023-of-00031.safetensors",
680
+ "model.layers.6.input_layernorm.weight": "model-00004-of-00031.safetensors",
681
+ "model.layers.6.mlp.down_proj.weight": "model-00004-of-00031.safetensors",
682
+ "model.layers.6.mlp.gate_proj.weight": "model-00003-of-00031.safetensors",
683
+ "model.layers.6.mlp.up_proj.weight": "model-00003-of-00031.safetensors",
684
+ "model.layers.6.post_attention_layernorm.weight": "model-00004-of-00031.safetensors",
685
+ "model.layers.6.self_attn.k_proj.bias": "model-00003-of-00031.safetensors",
686
+ "model.layers.6.self_attn.k_proj.weight": "model-00003-of-00031.safetensors",
687
+ "model.layers.6.self_attn.o_proj.weight": "model-00003-of-00031.safetensors",
688
+ "model.layers.6.self_attn.q_proj.bias": "model-00003-of-00031.safetensors",
689
+ "model.layers.6.self_attn.q_proj.weight": "model-00003-of-00031.safetensors",
690
+ "model.layers.6.self_attn.v_proj.bias": "model-00003-of-00031.safetensors",
691
+ "model.layers.6.self_attn.v_proj.weight": "model-00003-of-00031.safetensors",
692
+ "model.layers.60.input_layernorm.weight": "model-00023-of-00031.safetensors",
693
+ "model.layers.60.mlp.down_proj.weight": "model-00023-of-00031.safetensors",
694
+ "model.layers.60.mlp.gate_proj.weight": "model-00023-of-00031.safetensors",
695
+ "model.layers.60.mlp.up_proj.weight": "model-00023-of-00031.safetensors",
696
+ "model.layers.60.post_attention_layernorm.weight": "model-00023-of-00031.safetensors",
697
+ "model.layers.60.self_attn.k_proj.bias": "model-00023-of-00031.safetensors",
698
+ "model.layers.60.self_attn.k_proj.weight": "model-00023-of-00031.safetensors",
699
+ "model.layers.60.self_attn.o_proj.weight": "model-00023-of-00031.safetensors",
700
+ "model.layers.60.self_attn.q_proj.bias": "model-00023-of-00031.safetensors",
701
+ "model.layers.60.self_attn.q_proj.weight": "model-00023-of-00031.safetensors",
702
+ "model.layers.60.self_attn.v_proj.bias": "model-00023-of-00031.safetensors",
703
+ "model.layers.60.self_attn.v_proj.weight": "model-00023-of-00031.safetensors",
704
+ "model.layers.61.input_layernorm.weight": "model-00024-of-00031.safetensors",
705
+ "model.layers.61.mlp.down_proj.weight": "model-00024-of-00031.safetensors",
706
+ "model.layers.61.mlp.gate_proj.weight": "model-00023-of-00031.safetensors",
707
+ "model.layers.61.mlp.up_proj.weight": "model-00023-of-00031.safetensors",
708
+ "model.layers.61.post_attention_layernorm.weight": "model-00024-of-00031.safetensors",
709
+ "model.layers.61.self_attn.k_proj.bias": "model-00023-of-00031.safetensors",
710
+ "model.layers.61.self_attn.k_proj.weight": "model-00023-of-00031.safetensors",
711
+ "model.layers.61.self_attn.o_proj.weight": "model-00023-of-00031.safetensors",
712
+ "model.layers.61.self_attn.q_proj.bias": "model-00023-of-00031.safetensors",
713
+ "model.layers.61.self_attn.q_proj.weight": "model-00023-of-00031.safetensors",
714
+ "model.layers.61.self_attn.v_proj.bias": "model-00023-of-00031.safetensors",
715
+ "model.layers.61.self_attn.v_proj.weight": "model-00023-of-00031.safetensors",
716
+ "model.layers.62.input_layernorm.weight": "model-00024-of-00031.safetensors",
717
+ "model.layers.62.mlp.down_proj.weight": "model-00024-of-00031.safetensors",
718
+ "model.layers.62.mlp.gate_proj.weight": "model-00024-of-00031.safetensors",
719
+ "model.layers.62.mlp.up_proj.weight": "model-00024-of-00031.safetensors",
720
+ "model.layers.62.post_attention_layernorm.weight": "model-00024-of-00031.safetensors",
721
+ "model.layers.62.self_attn.k_proj.bias": "model-00024-of-00031.safetensors",
722
+ "model.layers.62.self_attn.k_proj.weight": "model-00024-of-00031.safetensors",
723
+ "model.layers.62.self_attn.o_proj.weight": "model-00024-of-00031.safetensors",
724
+ "model.layers.62.self_attn.q_proj.bias": "model-00024-of-00031.safetensors",
725
+ "model.layers.62.self_attn.q_proj.weight": "model-00024-of-00031.safetensors",
726
+ "model.layers.62.self_attn.v_proj.bias": "model-00024-of-00031.safetensors",
727
+ "model.layers.62.self_attn.v_proj.weight": "model-00024-of-00031.safetensors",
728
+ "model.layers.63.input_layernorm.weight": "model-00024-of-00031.safetensors",
729
+ "model.layers.63.mlp.down_proj.weight": "model-00024-of-00031.safetensors",
730
+ "model.layers.63.mlp.gate_proj.weight": "model-00024-of-00031.safetensors",
731
+ "model.layers.63.mlp.up_proj.weight": "model-00024-of-00031.safetensors",
732
+ "model.layers.63.post_attention_layernorm.weight": "model-00024-of-00031.safetensors",
733
+ "model.layers.63.self_attn.k_proj.bias": "model-00024-of-00031.safetensors",
734
+ "model.layers.63.self_attn.k_proj.weight": "model-00024-of-00031.safetensors",
735
+ "model.layers.63.self_attn.o_proj.weight": "model-00024-of-00031.safetensors",
736
+ "model.layers.63.self_attn.q_proj.bias": "model-00024-of-00031.safetensors",
737
+ "model.layers.63.self_attn.q_proj.weight": "model-00024-of-00031.safetensors",
738
+ "model.layers.63.self_attn.v_proj.bias": "model-00024-of-00031.safetensors",
739
+ "model.layers.63.self_attn.v_proj.weight": "model-00024-of-00031.safetensors",
740
+ "model.layers.64.input_layernorm.weight": "model-00025-of-00031.safetensors",
741
+ "model.layers.64.mlp.down_proj.weight": "model-00025-of-00031.safetensors",
742
+ "model.layers.64.mlp.gate_proj.weight": "model-00024-of-00031.safetensors",
743
+ "model.layers.64.mlp.up_proj.weight": "model-00025-of-00031.safetensors",
744
+ "model.layers.64.post_attention_layernorm.weight": "model-00025-of-00031.safetensors",
745
+ "model.layers.64.self_attn.k_proj.bias": "model-00024-of-00031.safetensors",
746
+ "model.layers.64.self_attn.k_proj.weight": "model-00024-of-00031.safetensors",
747
+ "model.layers.64.self_attn.o_proj.weight": "model-00024-of-00031.safetensors",
748
+ "model.layers.64.self_attn.q_proj.bias": "model-00024-of-00031.safetensors",
749
+ "model.layers.64.self_attn.q_proj.weight": "model-00024-of-00031.safetensors",
750
+ "model.layers.64.self_attn.v_proj.bias": "model-00024-of-00031.safetensors",
751
+ "model.layers.64.self_attn.v_proj.weight": "model-00024-of-00031.safetensors",
752
+ "model.layers.65.input_layernorm.weight": "model-00025-of-00031.safetensors",
753
+ "model.layers.65.mlp.down_proj.weight": "model-00025-of-00031.safetensors",
754
+ "model.layers.65.mlp.gate_proj.weight": "model-00025-of-00031.safetensors",
755
+ "model.layers.65.mlp.up_proj.weight": "model-00025-of-00031.safetensors",
756
+ "model.layers.65.post_attention_layernorm.weight": "model-00025-of-00031.safetensors",
757
+ "model.layers.65.self_attn.k_proj.bias": "model-00025-of-00031.safetensors",
758
+ "model.layers.65.self_attn.k_proj.weight": "model-00025-of-00031.safetensors",
759
+ "model.layers.65.self_attn.o_proj.weight": "model-00025-of-00031.safetensors",
760
+ "model.layers.65.self_attn.q_proj.bias": "model-00025-of-00031.safetensors",
761
+ "model.layers.65.self_attn.q_proj.weight": "model-00025-of-00031.safetensors",
762
+ "model.layers.65.self_attn.v_proj.bias": "model-00025-of-00031.safetensors",
763
+ "model.layers.65.self_attn.v_proj.weight": "model-00025-of-00031.safetensors",
764
+ "model.layers.66.input_layernorm.weight": "model-00025-of-00031.safetensors",
765
+ "model.layers.66.mlp.down_proj.weight": "model-00025-of-00031.safetensors",
766
+ "model.layers.66.mlp.gate_proj.weight": "model-00025-of-00031.safetensors",
767
+ "model.layers.66.mlp.up_proj.weight": "model-00025-of-00031.safetensors",
768
+ "model.layers.66.post_attention_layernorm.weight": "model-00025-of-00031.safetensors",
769
+ "model.layers.66.self_attn.k_proj.bias": "model-00025-of-00031.safetensors",
770
+ "model.layers.66.self_attn.k_proj.weight": "model-00025-of-00031.safetensors",
771
+ "model.layers.66.self_attn.o_proj.weight": "model-00025-of-00031.safetensors",
772
+ "model.layers.66.self_attn.q_proj.bias": "model-00025-of-00031.safetensors",
773
+ "model.layers.66.self_attn.q_proj.weight": "model-00025-of-00031.safetensors",
774
+ "model.layers.66.self_attn.v_proj.bias": "model-00025-of-00031.safetensors",
775
+ "model.layers.66.self_attn.v_proj.weight": "model-00025-of-00031.safetensors",
776
+ "model.layers.67.input_layernorm.weight": "model-00026-of-00031.safetensors",
777
+ "model.layers.67.mlp.down_proj.weight": "model-00026-of-00031.safetensors",
778
+ "model.layers.67.mlp.gate_proj.weight": "model-00026-of-00031.safetensors",
779
+ "model.layers.67.mlp.up_proj.weight": "model-00026-of-00031.safetensors",
780
+ "model.layers.67.post_attention_layernorm.weight": "model-00026-of-00031.safetensors",
781
+ "model.layers.67.self_attn.k_proj.bias": "model-00025-of-00031.safetensors",
782
+ "model.layers.67.self_attn.k_proj.weight": "model-00025-of-00031.safetensors",
783
+ "model.layers.67.self_attn.o_proj.weight": "model-00025-of-00031.safetensors",
784
+ "model.layers.67.self_attn.q_proj.bias": "model-00025-of-00031.safetensors",
785
+ "model.layers.67.self_attn.q_proj.weight": "model-00025-of-00031.safetensors",
786
+ "model.layers.67.self_attn.v_proj.bias": "model-00025-of-00031.safetensors",
787
+ "model.layers.67.self_attn.v_proj.weight": "model-00025-of-00031.safetensors",
788
+ "model.layers.68.input_layernorm.weight": "model-00026-of-00031.safetensors",
789
+ "model.layers.68.mlp.down_proj.weight": "model-00026-of-00031.safetensors",
790
+ "model.layers.68.mlp.gate_proj.weight": "model-00026-of-00031.safetensors",
791
+ "model.layers.68.mlp.up_proj.weight": "model-00026-of-00031.safetensors",
792
+ "model.layers.68.post_attention_layernorm.weight": "model-00026-of-00031.safetensors",
793
+ "model.layers.68.self_attn.k_proj.bias": "model-00026-of-00031.safetensors",
794
+ "model.layers.68.self_attn.k_proj.weight": "model-00026-of-00031.safetensors",
795
+ "model.layers.68.self_attn.o_proj.weight": "model-00026-of-00031.safetensors",
796
+ "model.layers.68.self_attn.q_proj.bias": "model-00026-of-00031.safetensors",
797
+ "model.layers.68.self_attn.q_proj.weight": "model-00026-of-00031.safetensors",
798
+ "model.layers.68.self_attn.v_proj.bias": "model-00026-of-00031.safetensors",
799
+ "model.layers.68.self_attn.v_proj.weight": "model-00026-of-00031.safetensors",
800
+ "model.layers.69.input_layernorm.weight": "model-00026-of-00031.safetensors",
801
+ "model.layers.69.mlp.down_proj.weight": "model-00026-of-00031.safetensors",
802
+ "model.layers.69.mlp.gate_proj.weight": "model-00026-of-00031.safetensors",
803
+ "model.layers.69.mlp.up_proj.weight": "model-00026-of-00031.safetensors",
804
+ "model.layers.69.post_attention_layernorm.weight": "model-00026-of-00031.safetensors",
805
+ "model.layers.69.self_attn.k_proj.bias": "model-00026-of-00031.safetensors",
806
+ "model.layers.69.self_attn.k_proj.weight": "model-00026-of-00031.safetensors",
807
+ "model.layers.69.self_attn.o_proj.weight": "model-00026-of-00031.safetensors",
808
+ "model.layers.69.self_attn.q_proj.bias": "model-00026-of-00031.safetensors",
809
+ "model.layers.69.self_attn.q_proj.weight": "model-00026-of-00031.safetensors",
810
+ "model.layers.69.self_attn.v_proj.bias": "model-00026-of-00031.safetensors",
811
+ "model.layers.69.self_attn.v_proj.weight": "model-00026-of-00031.safetensors",
812
+ "model.layers.7.input_layernorm.weight": "model-00004-of-00031.safetensors",
813
+ "model.layers.7.mlp.down_proj.weight": "model-00004-of-00031.safetensors",
814
+ "model.layers.7.mlp.gate_proj.weight": "model-00004-of-00031.safetensors",
815
+ "model.layers.7.mlp.up_proj.weight": "model-00004-of-00031.safetensors",
816
+ "model.layers.7.post_attention_layernorm.weight": "model-00004-of-00031.safetensors",
817
+ "model.layers.7.self_attn.k_proj.bias": "model-00004-of-00031.safetensors",
818
+ "model.layers.7.self_attn.k_proj.weight": "model-00004-of-00031.safetensors",
819
+ "model.layers.7.self_attn.o_proj.weight": "model-00004-of-00031.safetensors",
820
+ "model.layers.7.self_attn.q_proj.bias": "model-00004-of-00031.safetensors",
821
+ "model.layers.7.self_attn.q_proj.weight": "model-00004-of-00031.safetensors",
822
+ "model.layers.7.self_attn.v_proj.bias": "model-00004-of-00031.safetensors",
823
+ "model.layers.7.self_attn.v_proj.weight": "model-00004-of-00031.safetensors",
824
+ "model.layers.70.input_layernorm.weight": "model-00027-of-00031.safetensors",
825
+ "model.layers.70.mlp.down_proj.weight": "model-00027-of-00031.safetensors",
826
+ "model.layers.70.mlp.gate_proj.weight": "model-00027-of-00031.safetensors",
827
+ "model.layers.70.mlp.up_proj.weight": "model-00027-of-00031.safetensors",
828
+ "model.layers.70.post_attention_layernorm.weight": "model-00027-of-00031.safetensors",
829
+ "model.layers.70.self_attn.k_proj.bias": "model-00027-of-00031.safetensors",
830
+ "model.layers.70.self_attn.k_proj.weight": "model-00027-of-00031.safetensors",
831
+ "model.layers.70.self_attn.o_proj.weight": "model-00027-of-00031.safetensors",
832
+ "model.layers.70.self_attn.q_proj.bias": "model-00027-of-00031.safetensors",
833
+ "model.layers.70.self_attn.q_proj.weight": "model-00027-of-00031.safetensors",
834
+ "model.layers.70.self_attn.v_proj.bias": "model-00027-of-00031.safetensors",
835
+ "model.layers.70.self_attn.v_proj.weight": "model-00027-of-00031.safetensors",
836
+ "model.layers.71.input_layernorm.weight": "model-00027-of-00031.safetensors",
837
+ "model.layers.71.mlp.down_proj.weight": "model-00027-of-00031.safetensors",
838
+ "model.layers.71.mlp.gate_proj.weight": "model-00027-of-00031.safetensors",
839
+ "model.layers.71.mlp.up_proj.weight": "model-00027-of-00031.safetensors",
840
+ "model.layers.71.post_attention_layernorm.weight": "model-00027-of-00031.safetensors",
841
+ "model.layers.71.self_attn.k_proj.bias": "model-00027-of-00031.safetensors",
842
+ "model.layers.71.self_attn.k_proj.weight": "model-00027-of-00031.safetensors",
843
+ "model.layers.71.self_attn.o_proj.weight": "model-00027-of-00031.safetensors",
844
+ "model.layers.71.self_attn.q_proj.bias": "model-00027-of-00031.safetensors",
845
+ "model.layers.71.self_attn.q_proj.weight": "model-00027-of-00031.safetensors",
846
+ "model.layers.71.self_attn.v_proj.bias": "model-00027-of-00031.safetensors",
847
+ "model.layers.71.self_attn.v_proj.weight": "model-00027-of-00031.safetensors",
848
+ "model.layers.72.input_layernorm.weight": "model-00028-of-00031.safetensors",
849
+ "model.layers.72.mlp.down_proj.weight": "model-00028-of-00031.safetensors",
850
+ "model.layers.72.mlp.gate_proj.weight": "model-00027-of-00031.safetensors",
851
+ "model.layers.72.mlp.up_proj.weight": "model-00027-of-00031.safetensors",
852
+ "model.layers.72.post_attention_layernorm.weight": "model-00028-of-00031.safetensors",
853
+ "model.layers.72.self_attn.k_proj.bias": "model-00027-of-00031.safetensors",
854
+ "model.layers.72.self_attn.k_proj.weight": "model-00027-of-00031.safetensors",
855
+ "model.layers.72.self_attn.o_proj.weight": "model-00027-of-00031.safetensors",
856
+ "model.layers.72.self_attn.q_proj.bias": "model-00027-of-00031.safetensors",
857
+ "model.layers.72.self_attn.q_proj.weight": "model-00027-of-00031.safetensors",
858
+ "model.layers.72.self_attn.v_proj.bias": "model-00027-of-00031.safetensors",
859
+ "model.layers.72.self_attn.v_proj.weight": "model-00027-of-00031.safetensors",
860
+ "model.layers.73.input_layernorm.weight": "model-00028-of-00031.safetensors",
861
+ "model.layers.73.mlp.down_proj.weight": "model-00028-of-00031.safetensors",
862
+ "model.layers.73.mlp.gate_proj.weight": "model-00028-of-00031.safetensors",
863
+ "model.layers.73.mlp.up_proj.weight": "model-00028-of-00031.safetensors",
864
+ "model.layers.73.post_attention_layernorm.weight": "model-00028-of-00031.safetensors",
865
+ "model.layers.73.self_attn.k_proj.bias": "model-00028-of-00031.safetensors",
866
+ "model.layers.73.self_attn.k_proj.weight": "model-00028-of-00031.safetensors",
867
+ "model.layers.73.self_attn.o_proj.weight": "model-00028-of-00031.safetensors",
868
+ "model.layers.73.self_attn.q_proj.bias": "model-00028-of-00031.safetensors",
869
+ "model.layers.73.self_attn.q_proj.weight": "model-00028-of-00031.safetensors",
870
+ "model.layers.73.self_attn.v_proj.bias": "model-00028-of-00031.safetensors",
871
+ "model.layers.73.self_attn.v_proj.weight": "model-00028-of-00031.safetensors",
872
+ "model.layers.74.input_layernorm.weight": "model-00028-of-00031.safetensors",
873
+ "model.layers.74.mlp.down_proj.weight": "model-00028-of-00031.safetensors",
874
+ "model.layers.74.mlp.gate_proj.weight": "model-00028-of-00031.safetensors",
875
+ "model.layers.74.mlp.up_proj.weight": "model-00028-of-00031.safetensors",
876
+ "model.layers.74.post_attention_layernorm.weight": "model-00028-of-00031.safetensors",
877
+ "model.layers.74.self_attn.k_proj.bias": "model-00028-of-00031.safetensors",
878
+ "model.layers.74.self_attn.k_proj.weight": "model-00028-of-00031.safetensors",
879
+ "model.layers.74.self_attn.o_proj.weight": "model-00028-of-00031.safetensors",
880
+ "model.layers.74.self_attn.q_proj.bias": "model-00028-of-00031.safetensors",
881
+ "model.layers.74.self_attn.q_proj.weight": "model-00028-of-00031.safetensors",
882
+ "model.layers.74.self_attn.v_proj.bias": "model-00028-of-00031.safetensors",
883
+ "model.layers.74.self_attn.v_proj.weight": "model-00028-of-00031.safetensors",
884
+ "model.layers.75.input_layernorm.weight": "model-00029-of-00031.safetensors",
885
+ "model.layers.75.mlp.down_proj.weight": "model-00029-of-00031.safetensors",
886
+ "model.layers.75.mlp.gate_proj.weight": "model-00028-of-00031.safetensors",
887
+ "model.layers.75.mlp.up_proj.weight": "model-00029-of-00031.safetensors",
888
+ "model.layers.75.post_attention_layernorm.weight": "model-00029-of-00031.safetensors",
889
+ "model.layers.75.self_attn.k_proj.bias": "model-00028-of-00031.safetensors",
890
+ "model.layers.75.self_attn.k_proj.weight": "model-00028-of-00031.safetensors",
891
+ "model.layers.75.self_attn.o_proj.weight": "model-00028-of-00031.safetensors",
892
+ "model.layers.75.self_attn.q_proj.bias": "model-00028-of-00031.safetensors",
893
+ "model.layers.75.self_attn.q_proj.weight": "model-00028-of-00031.safetensors",
894
+ "model.layers.75.self_attn.v_proj.bias": "model-00028-of-00031.safetensors",
895
+ "model.layers.75.self_attn.v_proj.weight": "model-00028-of-00031.safetensors",
896
+ "model.layers.76.input_layernorm.weight": "model-00029-of-00031.safetensors",
897
+ "model.layers.76.mlp.down_proj.weight": "model-00029-of-00031.safetensors",
898
+ "model.layers.76.mlp.gate_proj.weight": "model-00029-of-00031.safetensors",
899
+ "model.layers.76.mlp.up_proj.weight": "model-00029-of-00031.safetensors",
900
+ "model.layers.76.post_attention_layernorm.weight": "model-00029-of-00031.safetensors",
901
+ "model.layers.76.self_attn.k_proj.bias": "model-00029-of-00031.safetensors",
902
+ "model.layers.76.self_attn.k_proj.weight": "model-00029-of-00031.safetensors",
903
+ "model.layers.76.self_attn.o_proj.weight": "model-00029-of-00031.safetensors",
904
+ "model.layers.76.self_attn.q_proj.bias": "model-00029-of-00031.safetensors",
905
+ "model.layers.76.self_attn.q_proj.weight": "model-00029-of-00031.safetensors",
906
+ "model.layers.76.self_attn.v_proj.bias": "model-00029-of-00031.safetensors",
907
+ "model.layers.76.self_attn.v_proj.weight": "model-00029-of-00031.safetensors",
908
+ "model.layers.77.input_layernorm.weight": "model-00029-of-00031.safetensors",
909
+ "model.layers.77.mlp.down_proj.weight": "model-00029-of-00031.safetensors",
910
+ "model.layers.77.mlp.gate_proj.weight": "model-00029-of-00031.safetensors",
911
+ "model.layers.77.mlp.up_proj.weight": "model-00029-of-00031.safetensors",
912
+ "model.layers.77.post_attention_layernorm.weight": "model-00029-of-00031.safetensors",
913
+ "model.layers.77.self_attn.k_proj.bias": "model-00029-of-00031.safetensors",
914
+ "model.layers.77.self_attn.k_proj.weight": "model-00029-of-00031.safetensors",
915
+ "model.layers.77.self_attn.o_proj.weight": "model-00029-of-00031.safetensors",
916
+ "model.layers.77.self_attn.q_proj.bias": "model-00029-of-00031.safetensors",
917
+ "model.layers.77.self_attn.q_proj.weight": "model-00029-of-00031.safetensors",
918
+ "model.layers.77.self_attn.v_proj.bias": "model-00029-of-00031.safetensors",
919
+ "model.layers.77.self_attn.v_proj.weight": "model-00029-of-00031.safetensors",
920
+ "model.layers.78.input_layernorm.weight": "model-00030-of-00031.safetensors",
921
+ "model.layers.78.mlp.down_proj.weight": "model-00030-of-00031.safetensors",
922
+ "model.layers.78.mlp.gate_proj.weight": "model-00030-of-00031.safetensors",
923
+ "model.layers.78.mlp.up_proj.weight": "model-00030-of-00031.safetensors",
924
+ "model.layers.78.post_attention_layernorm.weight": "model-00030-of-00031.safetensors",
925
+ "model.layers.78.self_attn.k_proj.bias": "model-00029-of-00031.safetensors",
926
+ "model.layers.78.self_attn.k_proj.weight": "model-00029-of-00031.safetensors",
927
+ "model.layers.78.self_attn.o_proj.weight": "model-00029-of-00031.safetensors",
928
+ "model.layers.78.self_attn.q_proj.bias": "model-00029-of-00031.safetensors",
929
+ "model.layers.78.self_attn.q_proj.weight": "model-00029-of-00031.safetensors",
930
+ "model.layers.78.self_attn.v_proj.bias": "model-00029-of-00031.safetensors",
931
+ "model.layers.78.self_attn.v_proj.weight": "model-00029-of-00031.safetensors",
932
+ "model.layers.79.input_layernorm.weight": "model-00030-of-00031.safetensors",
933
+ "model.layers.79.mlp.down_proj.weight": "model-00030-of-00031.safetensors",
934
+ "model.layers.79.mlp.gate_proj.weight": "model-00030-of-00031.safetensors",
935
+ "model.layers.79.mlp.up_proj.weight": "model-00030-of-00031.safetensors",
936
+ "model.layers.79.post_attention_layernorm.weight": "model-00030-of-00031.safetensors",
937
+ "model.layers.79.self_attn.k_proj.bias": "model-00030-of-00031.safetensors",
938
+ "model.layers.79.self_attn.k_proj.weight": "model-00030-of-00031.safetensors",
939
+ "model.layers.79.self_attn.o_proj.weight": "model-00030-of-00031.safetensors",
940
+ "model.layers.79.self_attn.q_proj.bias": "model-00030-of-00031.safetensors",
941
+ "model.layers.79.self_attn.q_proj.weight": "model-00030-of-00031.safetensors",
942
+ "model.layers.79.self_attn.v_proj.bias": "model-00030-of-00031.safetensors",
943
+ "model.layers.79.self_attn.v_proj.weight": "model-00030-of-00031.safetensors",
944
+ "model.layers.8.input_layernorm.weight": "model-00004-of-00031.safetensors",
945
+ "model.layers.8.mlp.down_proj.weight": "model-00004-of-00031.safetensors",
946
+ "model.layers.8.mlp.gate_proj.weight": "model-00004-of-00031.safetensors",
947
+ "model.layers.8.mlp.up_proj.weight": "model-00004-of-00031.safetensors",
948
+ "model.layers.8.post_attention_layernorm.weight": "model-00004-of-00031.safetensors",
949
+ "model.layers.8.self_attn.k_proj.bias": "model-00004-of-00031.safetensors",
950
+ "model.layers.8.self_attn.k_proj.weight": "model-00004-of-00031.safetensors",
951
+ "model.layers.8.self_attn.o_proj.weight": "model-00004-of-00031.safetensors",
952
+ "model.layers.8.self_attn.q_proj.bias": "model-00004-of-00031.safetensors",
953
+ "model.layers.8.self_attn.q_proj.weight": "model-00004-of-00031.safetensors",
954
+ "model.layers.8.self_attn.v_proj.bias": "model-00004-of-00031.safetensors",
955
+ "model.layers.8.self_attn.v_proj.weight": "model-00004-of-00031.safetensors",
956
+ "model.layers.9.input_layernorm.weight": "model-00005-of-00031.safetensors",
957
+ "model.layers.9.mlp.down_proj.weight": "model-00005-of-00031.safetensors",
958
+ "model.layers.9.mlp.gate_proj.weight": "model-00004-of-00031.safetensors",
959
+ "model.layers.9.mlp.up_proj.weight": "model-00005-of-00031.safetensors",
960
+ "model.layers.9.post_attention_layernorm.weight": "model-00005-of-00031.safetensors",
961
+ "model.layers.9.self_attn.k_proj.bias": "model-00004-of-00031.safetensors",
962
+ "model.layers.9.self_attn.k_proj.weight": "model-00004-of-00031.safetensors",
963
+ "model.layers.9.self_attn.o_proj.weight": "model-00004-of-00031.safetensors",
964
+ "model.layers.9.self_attn.q_proj.bias": "model-00004-of-00031.safetensors",
965
+ "model.layers.9.self_attn.q_proj.weight": "model-00004-of-00031.safetensors",
966
+ "model.layers.9.self_attn.v_proj.bias": "model-00004-of-00031.safetensors",
967
+ "model.layers.9.self_attn.v_proj.weight": "model-00004-of-00031.safetensors",
968
+ "model.norm.weight": "model-00030-of-00031.safetensors"
969
+ }
970
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|object_ref_start|>",
6
+ "<|object_ref_end|>",
7
+ "<|box_start|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_start|>",
12
+ "<|vision_end|>",
13
+ "<|vision_pad|>",
14
+ "<|image_pad|>",
15
+ "<|video_pad|>"
16
+ ],
17
+ "eos_token": {
18
+ "content": "<|endoftext|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ "pad_token": {
25
+ "content": "<|endoftext|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ }
31
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
3
+ size 11421896
tokenizer_config.json ADDED
@@ -0,0 +1,208 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "151643": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "151644": {
14
+ "content": "<|im_start|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "151645": {
22
+ "content": "<|im_end|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "151646": {
30
+ "content": "<|object_ref_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "151647": {
38
+ "content": "<|object_ref_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "151648": {
46
+ "content": "<|box_start|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "151649": {
54
+ "content": "<|box_end|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ },
61
+ "151650": {
62
+ "content": "<|quad_start|>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": false,
67
+ "special": true
68
+ },
69
+ "151651": {
70
+ "content": "<|quad_end|>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": false,
75
+ "special": true
76
+ },
77
+ "151652": {
78
+ "content": "<|vision_start|>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": false,
83
+ "special": true
84
+ },
85
+ "151653": {
86
+ "content": "<|vision_end|>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": false,
91
+ "special": true
92
+ },
93
+ "151654": {
94
+ "content": "<|vision_pad|>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": false,
99
+ "special": true
100
+ },
101
+ "151655": {
102
+ "content": "<|image_pad|>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false,
107
+ "special": true
108
+ },
109
+ "151656": {
110
+ "content": "<|video_pad|>",
111
+ "lstrip": false,
112
+ "normalized": false,
113
+ "rstrip": false,
114
+ "single_word": false,
115
+ "special": true
116
+ },
117
+ "151657": {
118
+ "content": "<tool_call>",
119
+ "lstrip": false,
120
+ "normalized": false,
121
+ "rstrip": false,
122
+ "single_word": false,
123
+ "special": false
124
+ },
125
+ "151658": {
126
+ "content": "</tool_call>",
127
+ "lstrip": false,
128
+ "normalized": false,
129
+ "rstrip": false,
130
+ "single_word": false,
131
+ "special": false
132
+ },
133
+ "151659": {
134
+ "content": "<|fim_prefix|>",
135
+ "lstrip": false,
136
+ "normalized": false,
137
+ "rstrip": false,
138
+ "single_word": false,
139
+ "special": false
140
+ },
141
+ "151660": {
142
+ "content": "<|fim_middle|>",
143
+ "lstrip": false,
144
+ "normalized": false,
145
+ "rstrip": false,
146
+ "single_word": false,
147
+ "special": false
148
+ },
149
+ "151661": {
150
+ "content": "<|fim_suffix|>",
151
+ "lstrip": false,
152
+ "normalized": false,
153
+ "rstrip": false,
154
+ "single_word": false,
155
+ "special": false
156
+ },
157
+ "151662": {
158
+ "content": "<|fim_pad|>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false,
163
+ "special": false
164
+ },
165
+ "151663": {
166
+ "content": "<|repo_name|>",
167
+ "lstrip": false,
168
+ "normalized": false,
169
+ "rstrip": false,
170
+ "single_word": false,
171
+ "special": false
172
+ },
173
+ "151664": {
174
+ "content": "<|file_sep|>",
175
+ "lstrip": false,
176
+ "normalized": false,
177
+ "rstrip": false,
178
+ "single_word": false,
179
+ "special": false
180
+ }
181
+ },
182
+ "additional_special_tokens": [
183
+ "<|im_start|>",
184
+ "<|im_end|>",
185
+ "<|object_ref_start|>",
186
+ "<|object_ref_end|>",
187
+ "<|box_start|>",
188
+ "<|box_end|>",
189
+ "<|quad_start|>",
190
+ "<|quad_end|>",
191
+ "<|vision_start|>",
192
+ "<|vision_end|>",
193
+ "<|vision_pad|>",
194
+ "<|image_pad|>",
195
+ "<|video_pad|>"
196
+ ],
197
+ "bos_token": null,
198
+ "chat_template": "{%- if tools %}\n {{- '<|im_start|>system\\n' }}\n {%- if messages[0]['role'] == 'system' %}\n {{- messages[0]['content'] }}\n {%- else %}\n {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n {%- endif %}\n {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n {%- for tool in tools %}\n {{- \"\\n\" }}\n {{- tool | tojson }}\n {%- endfor %}\n {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n {%- if messages[0]['role'] == 'system' %}\n {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n {%- else %}\n {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {{- '<|im_start|>' + message.role }}\n {%- if message.content %}\n {{- '\\n' + message.content }}\n {%- endif %}\n {%- for tool_call in message.tool_calls %}\n {%- if tool_call.function is defined %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '\\n<tool_call>\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {{- tool_call.arguments | tojson }}\n {{- '}\\n</tool_call>' }}\n {%- endfor %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- message.content }}\n {{- '\\n</tool_response>' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
199
+ "clean_up_tokenization_spaces": false,
200
+ "eos_token": "<|endoftext|>",
201
+ "errors": "replace",
202
+ "model_max_length": 131072,
203
+ "pad_token": "<|endoftext|>",
204
+ "padding_side": "right",
205
+ "split_special_tokens": false,
206
+ "tokenizer_class": "Qwen2Tokenizer",
207
+ "unk_token": null
208
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "total_flos": 2.982866014384947e+16,
4
+ "train_loss": 0.3923433954339652,
5
+ "train_runtime": 48299.7361,
6
+ "train_samples_per_second": 7.078,
7
+ "train_steps_per_second": 0.007
8
+ }
trainer_log.jsonl ADDED
@@ -0,0 +1,337 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"current_steps": 1, "total_steps": 336, "loss": 0.8158, "lr": 2.9411764705882356e-07, "epoch": 0.008928571428571428, "percentage": 0.3, "elapsed_time": "0:02:11", "remaining_time": "12:15:56"}
2
+ {"current_steps": 2, "total_steps": 336, "loss": 0.8361, "lr": 5.882352941176471e-07, "epoch": 0.017857142857142856, "percentage": 0.6, "elapsed_time": "0:04:19", "remaining_time": "12:03:31"}
3
+ {"current_steps": 3, "total_steps": 336, "loss": 0.8422, "lr": 8.823529411764707e-07, "epoch": 0.026785714285714284, "percentage": 0.89, "elapsed_time": "0:06:23", "remaining_time": "11:49:11"}
4
+ {"current_steps": 4, "total_steps": 336, "loss": 0.8212, "lr": 1.1764705882352942e-06, "epoch": 0.03571428571428571, "percentage": 1.19, "elapsed_time": "0:08:18", "remaining_time": "11:30:06"}
5
+ {"current_steps": 5, "total_steps": 336, "loss": 0.8039, "lr": 1.4705882352941177e-06, "epoch": 0.044642857142857144, "percentage": 1.49, "elapsed_time": "0:10:12", "remaining_time": "11:15:58"}
6
+ {"current_steps": 6, "total_steps": 336, "loss": 0.7598, "lr": 1.7647058823529414e-06, "epoch": 0.05357142857142857, "percentage": 1.79, "elapsed_time": "0:12:13", "remaining_time": "11:12:44"}
7
+ {"current_steps": 7, "total_steps": 336, "loss": 0.7132, "lr": 2.058823529411765e-06, "epoch": 0.0625, "percentage": 2.08, "elapsed_time": "0:14:13", "remaining_time": "11:08:44"}
8
+ {"current_steps": 8, "total_steps": 336, "loss": 0.7006, "lr": 2.3529411764705885e-06, "epoch": 0.07142857142857142, "percentage": 2.38, "elapsed_time": "0:16:10", "remaining_time": "11:02:52"}
9
+ {"current_steps": 9, "total_steps": 336, "loss": 0.6781, "lr": 2.647058823529412e-06, "epoch": 0.08035714285714286, "percentage": 2.68, "elapsed_time": "0:18:07", "remaining_time": "10:58:48"}
10
+ {"current_steps": 10, "total_steps": 336, "loss": 0.6716, "lr": 2.9411764705882355e-06, "epoch": 0.08928571428571429, "percentage": 2.98, "elapsed_time": "0:20:05", "remaining_time": "10:55:14"}
11
+ {"current_steps": 11, "total_steps": 336, "loss": 0.6524, "lr": 3.2352941176470594e-06, "epoch": 0.09821428571428571, "percentage": 3.27, "elapsed_time": "0:22:03", "remaining_time": "10:51:45"}
12
+ {"current_steps": 12, "total_steps": 336, "loss": 0.6202, "lr": 3.529411764705883e-06, "epoch": 0.10714285714285714, "percentage": 3.57, "elapsed_time": "0:23:58", "remaining_time": "10:47:13"}
13
+ {"current_steps": 13, "total_steps": 336, "loss": 0.6104, "lr": 3.8235294117647055e-06, "epoch": 0.11607142857142858, "percentage": 3.87, "elapsed_time": "0:25:52", "remaining_time": "10:42:53"}
14
+ {"current_steps": 14, "total_steps": 336, "loss": 0.6036, "lr": 4.11764705882353e-06, "epoch": 0.125, "percentage": 4.17, "elapsed_time": "0:27:47", "remaining_time": "10:39:19"}
15
+ {"current_steps": 15, "total_steps": 336, "loss": 0.5893, "lr": 4.411764705882353e-06, "epoch": 0.13392857142857142, "percentage": 4.46, "elapsed_time": "0:29:42", "remaining_time": "10:35:36"}
16
+ {"current_steps": 16, "total_steps": 336, "loss": 0.5752, "lr": 4.705882352941177e-06, "epoch": 0.14285714285714285, "percentage": 4.76, "elapsed_time": "0:31:35", "remaining_time": "10:31:56"}
17
+ {"current_steps": 17, "total_steps": 336, "loss": 0.5527, "lr": 5e-06, "epoch": 0.15178571428571427, "percentage": 5.06, "elapsed_time": "0:33:32", "remaining_time": "10:29:17"}
18
+ {"current_steps": 18, "total_steps": 336, "loss": 0.5438, "lr": 5.294117647058824e-06, "epoch": 0.16071428571428573, "percentage": 5.36, "elapsed_time": "0:35:27", "remaining_time": "10:26:28"}
19
+ {"current_steps": 19, "total_steps": 336, "loss": 0.5295, "lr": 5.588235294117647e-06, "epoch": 0.16964285714285715, "percentage": 5.65, "elapsed_time": "0:37:22", "remaining_time": "10:23:28"}
20
+ {"current_steps": 20, "total_steps": 336, "loss": 0.5227, "lr": 5.882352941176471e-06, "epoch": 0.17857142857142858, "percentage": 5.95, "elapsed_time": "0:39:21", "remaining_time": "10:21:59"}
21
+ {"current_steps": 21, "total_steps": 336, "loss": 0.5146, "lr": 6.176470588235295e-06, "epoch": 0.1875, "percentage": 6.25, "elapsed_time": "0:41:18", "remaining_time": "10:19:43"}
22
+ {"current_steps": 22, "total_steps": 336, "loss": 0.503, "lr": 6.470588235294119e-06, "epoch": 0.19642857142857142, "percentage": 6.55, "elapsed_time": "0:43:15", "remaining_time": "10:17:24"}
23
+ {"current_steps": 23, "total_steps": 336, "loss": 0.499, "lr": 6.764705882352942e-06, "epoch": 0.20535714285714285, "percentage": 6.85, "elapsed_time": "0:45:09", "remaining_time": "10:14:28"}
24
+ {"current_steps": 24, "total_steps": 336, "loss": 0.4901, "lr": 7.058823529411766e-06, "epoch": 0.21428571428571427, "percentage": 7.14, "elapsed_time": "0:47:06", "remaining_time": "10:12:28"}
25
+ {"current_steps": 25, "total_steps": 336, "loss": 0.4882, "lr": 7.352941176470589e-06, "epoch": 0.22321428571428573, "percentage": 7.44, "elapsed_time": "0:49:04", "remaining_time": "10:10:24"}
26
+ {"current_steps": 26, "total_steps": 336, "loss": 0.4789, "lr": 7.647058823529411e-06, "epoch": 0.23214285714285715, "percentage": 7.74, "elapsed_time": "0:50:59", "remaining_time": "10:08:01"}
27
+ {"current_steps": 27, "total_steps": 336, "loss": 0.4792, "lr": 7.941176470588236e-06, "epoch": 0.24107142857142858, "percentage": 8.04, "elapsed_time": "0:53:02", "remaining_time": "10:06:56"}
28
+ {"current_steps": 28, "total_steps": 336, "loss": 0.4665, "lr": 8.23529411764706e-06, "epoch": 0.25, "percentage": 8.33, "elapsed_time": "0:54:56", "remaining_time": "10:04:23"}
29
+ {"current_steps": 29, "total_steps": 336, "loss": 0.4666, "lr": 8.529411764705883e-06, "epoch": 0.25892857142857145, "percentage": 8.63, "elapsed_time": "0:56:52", "remaining_time": "10:02:06"}
30
+ {"current_steps": 30, "total_steps": 336, "loss": 0.4644, "lr": 8.823529411764707e-06, "epoch": 0.26785714285714285, "percentage": 8.93, "elapsed_time": "0:58:49", "remaining_time": "9:59:59"}
31
+ {"current_steps": 31, "total_steps": 336, "loss": 0.4552, "lr": 9.11764705882353e-06, "epoch": 0.2767857142857143, "percentage": 9.23, "elapsed_time": "1:00:48", "remaining_time": "9:58:14"}
32
+ {"current_steps": 32, "total_steps": 336, "loss": 0.4611, "lr": 9.411764705882354e-06, "epoch": 0.2857142857142857, "percentage": 9.52, "elapsed_time": "1:02:44", "remaining_time": "9:56:01"}
33
+ {"current_steps": 33, "total_steps": 336, "loss": 0.449, "lr": 9.705882352941177e-06, "epoch": 0.29464285714285715, "percentage": 9.82, "elapsed_time": "1:04:42", "remaining_time": "9:54:05"}
34
+ {"current_steps": 34, "total_steps": 336, "loss": 0.4543, "lr": 1e-05, "epoch": 0.30357142857142855, "percentage": 10.12, "elapsed_time": "1:06:44", "remaining_time": "9:52:48"}
35
+ {"current_steps": 35, "total_steps": 336, "loss": 0.4449, "lr": 9.999729465939036e-06, "epoch": 0.3125, "percentage": 10.42, "elapsed_time": "1:08:47", "remaining_time": "9:51:36"}
36
+ {"current_steps": 36, "total_steps": 336, "loss": 0.4503, "lr": 9.998917893031615e-06, "epoch": 0.32142857142857145, "percentage": 10.71, "elapsed_time": "1:10:45", "remaining_time": "9:49:39"}
37
+ {"current_steps": 37, "total_steps": 336, "loss": 0.4421, "lr": 9.997565369100983e-06, "epoch": 0.33035714285714285, "percentage": 11.01, "elapsed_time": "1:12:46", "remaining_time": "9:48:07"}
38
+ {"current_steps": 38, "total_steps": 336, "loss": 0.4453, "lr": 9.995672040508656e-06, "epoch": 0.3392857142857143, "percentage": 11.31, "elapsed_time": "1:14:47", "remaining_time": "9:46:34"}
39
+ {"current_steps": 39, "total_steps": 336, "loss": 0.4388, "lr": 9.993238112138584e-06, "epoch": 0.3482142857142857, "percentage": 11.61, "elapsed_time": "1:16:51", "remaining_time": "9:45:17"}
40
+ {"current_steps": 40, "total_steps": 336, "loss": 0.4407, "lr": 9.990263847374976e-06, "epoch": 0.35714285714285715, "percentage": 11.9, "elapsed_time": "1:18:51", "remaining_time": "9:43:30"}
41
+ {"current_steps": 41, "total_steps": 336, "loss": 0.4419, "lr": 9.986749568073804e-06, "epoch": 0.36607142857142855, "percentage": 12.2, "elapsed_time": "1:20:52", "remaining_time": "9:41:57"}
42
+ {"current_steps": 42, "total_steps": 336, "loss": 0.438, "lr": 9.982695654527966e-06, "epoch": 0.375, "percentage": 12.5, "elapsed_time": "1:22:57", "remaining_time": "9:40:44"}
43
+ {"current_steps": 43, "total_steps": 336, "loss": 0.4299, "lr": 9.97810254542614e-06, "epoch": 0.38392857142857145, "percentage": 12.8, "elapsed_time": "1:24:58", "remaining_time": "9:39:00"}
44
+ {"current_steps": 44, "total_steps": 336, "loss": 0.4248, "lr": 9.972970737805312e-06, "epoch": 0.39285714285714285, "percentage": 13.1, "elapsed_time": "1:26:58", "remaining_time": "9:37:14"}
45
+ {"current_steps": 45, "total_steps": 336, "loss": 0.4333, "lr": 9.96730078699698e-06, "epoch": 0.4017857142857143, "percentage": 13.39, "elapsed_time": "1:29:00", "remaining_time": "9:35:34"}
46
+ {"current_steps": 46, "total_steps": 336, "loss": 0.4266, "lr": 9.961093306567076e-06, "epoch": 0.4107142857142857, "percentage": 13.69, "elapsed_time": "1:31:05", "remaining_time": "9:34:16"}
47
+ {"current_steps": 47, "total_steps": 336, "loss": 0.4253, "lr": 9.954348968249552e-06, "epoch": 0.41964285714285715, "percentage": 13.99, "elapsed_time": "1:33:09", "remaining_time": "9:32:49"}
48
+ {"current_steps": 48, "total_steps": 336, "loss": 0.4229, "lr": 9.947068501873702e-06, "epoch": 0.42857142857142855, "percentage": 14.29, "elapsed_time": "1:35:14", "remaining_time": "9:31:25"}
49
+ {"current_steps": 49, "total_steps": 336, "loss": 0.423, "lr": 9.93925269528518e-06, "epoch": 0.4375, "percentage": 14.58, "elapsed_time": "1:37:21", "remaining_time": "9:30:16"}
50
+ {"current_steps": 50, "total_steps": 336, "loss": 0.4136, "lr": 9.930902394260746e-06, "epoch": 0.44642857142857145, "percentage": 14.88, "elapsed_time": "1:39:27", "remaining_time": "9:28:54"}
51
+ {"current_steps": 51, "total_steps": 336, "loss": 0.4238, "lr": 9.922018502416736e-06, "epoch": 0.45535714285714285, "percentage": 15.18, "elapsed_time": "1:46:01", "remaining_time": "9:52:29"}
52
+ {"current_steps": 52, "total_steps": 336, "loss": 0.4162, "lr": 9.912601981111287e-06, "epoch": 0.4642857142857143, "percentage": 15.48, "elapsed_time": "1:48:05", "remaining_time": "9:50:22"}
53
+ {"current_steps": 53, "total_steps": 336, "loss": 0.4115, "lr": 9.902653849340296e-06, "epoch": 0.4732142857142857, "percentage": 15.77, "elapsed_time": "1:50:08", "remaining_time": "9:48:08"}
54
+ {"current_steps": 54, "total_steps": 336, "loss": 0.4144, "lr": 9.892175183627161e-06, "epoch": 0.48214285714285715, "percentage": 16.07, "elapsed_time": "1:52:13", "remaining_time": "9:46:04"}
55
+ {"current_steps": 55, "total_steps": 336, "loss": 0.4163, "lr": 9.881167117906276e-06, "epoch": 0.49107142857142855, "percentage": 16.37, "elapsed_time": "1:54:22", "remaining_time": "9:44:19"}
56
+ {"current_steps": 56, "total_steps": 336, "loss": 0.412, "lr": 9.869630843400331e-06, "epoch": 0.5, "percentage": 16.67, "elapsed_time": "1:56:30", "remaining_time": "9:42:34"}
57
+ {"current_steps": 57, "total_steps": 336, "loss": 0.4196, "lr": 9.8575676084914e-06, "epoch": 0.5089285714285714, "percentage": 16.96, "elapsed_time": "1:58:35", "remaining_time": "9:40:30"}
58
+ {"current_steps": 58, "total_steps": 336, "loss": 0.4135, "lr": 9.844978718585855e-06, "epoch": 0.5178571428571429, "percentage": 17.26, "elapsed_time": "2:00:42", "remaining_time": "9:38:33"}
59
+ {"current_steps": 59, "total_steps": 336, "loss": 0.4135, "lr": 9.831865535973103e-06, "epoch": 0.5267857142857143, "percentage": 17.56, "elapsed_time": "2:02:47", "remaining_time": "9:36:29"}
60
+ {"current_steps": 60, "total_steps": 336, "loss": 0.4085, "lr": 9.81822947967816e-06, "epoch": 0.5357142857142857, "percentage": 17.86, "elapsed_time": "2:04:54", "remaining_time": "9:34:34"}
61
+ {"current_steps": 61, "total_steps": 336, "loss": 0.4112, "lr": 9.804072025308096e-06, "epoch": 0.5446428571428571, "percentage": 18.15, "elapsed_time": "2:06:59", "remaining_time": "9:32:32"}
62
+ {"current_steps": 62, "total_steps": 336, "loss": 0.4086, "lr": 9.789394704892364e-06, "epoch": 0.5535714285714286, "percentage": 18.45, "elapsed_time": "2:09:09", "remaining_time": "9:30:48"}
63
+ {"current_steps": 63, "total_steps": 336, "loss": 0.4077, "lr": 9.774199106717004e-06, "epoch": 0.5625, "percentage": 18.75, "elapsed_time": "2:11:18", "remaining_time": "9:29:01"}
64
+ {"current_steps": 64, "total_steps": 336, "loss": 0.4072, "lr": 9.758486875152766e-06, "epoch": 0.5714285714285714, "percentage": 19.05, "elapsed_time": "2:13:22", "remaining_time": "9:26:48"}
65
+ {"current_steps": 65, "total_steps": 336, "loss": 0.4015, "lr": 9.742259710477178e-06, "epoch": 0.5803571428571429, "percentage": 19.35, "elapsed_time": "2:15:27", "remaining_time": "9:24:45"}
66
+ {"current_steps": 66, "total_steps": 336, "loss": 0.4035, "lr": 9.725519368690539e-06, "epoch": 0.5892857142857143, "percentage": 19.64, "elapsed_time": "2:17:36", "remaining_time": "9:22:56"}
67
+ {"current_steps": 67, "total_steps": 336, "loss": 0.4061, "lr": 9.70826766132591e-06, "epoch": 0.5982142857142857, "percentage": 19.94, "elapsed_time": "2:19:40", "remaining_time": "9:20:46"}
68
+ {"current_steps": 68, "total_steps": 336, "loss": 0.4086, "lr": 9.690506455253073e-06, "epoch": 0.6071428571428571, "percentage": 20.24, "elapsed_time": "2:21:47", "remaining_time": "9:18:48"}
69
+ {"current_steps": 69, "total_steps": 336, "loss": 0.4017, "lr": 9.672237672476506e-06, "epoch": 0.6160714285714286, "percentage": 20.54, "elapsed_time": "2:23:55", "remaining_time": "9:16:56"}
70
+ {"current_steps": 70, "total_steps": 336, "loss": 0.4017, "lr": 9.65346328992741e-06, "epoch": 0.625, "percentage": 20.83, "elapsed_time": "2:26:07", "remaining_time": "9:15:16"}
71
+ {"current_steps": 71, "total_steps": 336, "loss": 0.4043, "lr": 9.634185339249766e-06, "epoch": 0.6339285714285714, "percentage": 21.13, "elapsed_time": "2:28:18", "remaining_time": "9:13:32"}
72
+ {"current_steps": 72, "total_steps": 336, "loss": 0.396, "lr": 9.614405906580486e-06, "epoch": 0.6428571428571429, "percentage": 21.43, "elapsed_time": "2:30:24", "remaining_time": "9:11:30"}
73
+ {"current_steps": 73, "total_steps": 336, "loss": 0.4062, "lr": 9.594127132323669e-06, "epoch": 0.6517857142857143, "percentage": 21.73, "elapsed_time": "2:32:32", "remaining_time": "9:09:33"}
74
+ {"current_steps": 74, "total_steps": 336, "loss": 0.3977, "lr": 9.573351210918976e-06, "epoch": 0.6607142857142857, "percentage": 22.02, "elapsed_time": "2:34:40", "remaining_time": "9:07:37"}
75
+ {"current_steps": 75, "total_steps": 336, "loss": 0.3993, "lr": 9.55208039060416e-06, "epoch": 0.6696428571428571, "percentage": 22.32, "elapsed_time": "2:36:49", "remaining_time": "9:05:43"}
76
+ {"current_steps": 76, "total_steps": 336, "loss": 0.4126, "lr": 9.53031697317178e-06, "epoch": 0.6785714285714286, "percentage": 22.62, "elapsed_time": "2:38:57", "remaining_time": "9:03:48"}
77
+ {"current_steps": 77, "total_steps": 336, "loss": 0.4068, "lr": 9.50806331372012e-06, "epoch": 0.6875, "percentage": 22.92, "elapsed_time": "2:41:11", "remaining_time": "9:02:09"}
78
+ {"current_steps": 78, "total_steps": 336, "loss": 0.4014, "lr": 9.485321820398321e-06, "epoch": 0.6964285714285714, "percentage": 23.21, "elapsed_time": "2:43:25", "remaining_time": "9:00:32"}
79
+ {"current_steps": 79, "total_steps": 336, "loss": 0.3995, "lr": 9.4620949541458e-06, "epoch": 0.7053571428571429, "percentage": 23.51, "elapsed_time": "2:45:32", "remaining_time": "8:58:30"}
80
+ {"current_steps": 80, "total_steps": 336, "loss": 0.4055, "lr": 9.43838522842594e-06, "epoch": 0.7142857142857143, "percentage": 23.81, "elapsed_time": "2:47:40", "remaining_time": "8:56:33"}
81
+ {"current_steps": 81, "total_steps": 336, "loss": 0.3976, "lr": 9.41419520895409e-06, "epoch": 0.7232142857142857, "percentage": 24.11, "elapsed_time": "2:49:51", "remaining_time": "8:54:43"}
82
+ {"current_steps": 82, "total_steps": 336, "loss": 0.3922, "lr": 9.389527513419935e-06, "epoch": 0.7321428571428571, "percentage": 24.4, "elapsed_time": "2:52:05", "remaining_time": "8:53:02"}
83
+ {"current_steps": 83, "total_steps": 336, "loss": 0.4018, "lr": 9.364384811204212e-06, "epoch": 0.7410714285714286, "percentage": 24.7, "elapsed_time": "2:54:18", "remaining_time": "8:51:19"}
84
+ {"current_steps": 84, "total_steps": 336, "loss": 0.399, "lr": 9.338769823089853e-06, "epoch": 0.75, "percentage": 25.0, "elapsed_time": "2:56:35", "remaining_time": "8:49:45"}
85
+ {"current_steps": 85, "total_steps": 336, "loss": 0.3948, "lr": 9.312685320967566e-06, "epoch": 0.7589285714285714, "percentage": 25.3, "elapsed_time": "2:58:51", "remaining_time": "8:48:08"}
86
+ {"current_steps": 86, "total_steps": 336, "loss": 0.4007, "lr": 9.286134127535859e-06, "epoch": 0.7678571428571429, "percentage": 25.6, "elapsed_time": "3:01:03", "remaining_time": "8:46:20"}
87
+ {"current_steps": 87, "total_steps": 336, "loss": 0.3944, "lr": 9.25911911599561e-06, "epoch": 0.7767857142857143, "percentage": 25.89, "elapsed_time": "3:03:13", "remaining_time": "8:44:24"}
88
+ {"current_steps": 88, "total_steps": 336, "loss": 0.4006, "lr": 9.231643209739128e-06, "epoch": 0.7857142857142857, "percentage": 26.19, "elapsed_time": "3:05:24", "remaining_time": "8:42:30"}
89
+ {"current_steps": 89, "total_steps": 336, "loss": 0.3894, "lr": 9.203709382033814e-06, "epoch": 0.7946428571428571, "percentage": 26.49, "elapsed_time": "3:07:39", "remaining_time": "8:40:47"}
90
+ {"current_steps": 90, "total_steps": 336, "loss": 0.3965, "lr": 9.175320655700407e-06, "epoch": 0.8035714285714286, "percentage": 26.79, "elapsed_time": "3:09:53", "remaining_time": "8:39:02"}
91
+ {"current_steps": 91, "total_steps": 336, "loss": 0.4021, "lr": 9.146480102785871e-06, "epoch": 0.8125, "percentage": 27.08, "elapsed_time": "3:12:11", "remaining_time": "8:37:25"}
92
+ {"current_steps": 92, "total_steps": 336, "loss": 0.3979, "lr": 9.117190844230971e-06, "epoch": 0.8214285714285714, "percentage": 27.38, "elapsed_time": "3:14:29", "remaining_time": "8:35:49"}
93
+ {"current_steps": 93, "total_steps": 336, "loss": 0.3961, "lr": 9.08745604953253e-06, "epoch": 0.8303571428571429, "percentage": 27.68, "elapsed_time": "3:16:45", "remaining_time": "8:34:07"}
94
+ {"current_steps": 94, "total_steps": 336, "loss": 0.3947, "lr": 9.057278936400453e-06, "epoch": 0.8392857142857143, "percentage": 27.98, "elapsed_time": "3:19:00", "remaining_time": "8:32:19"}
95
+ {"current_steps": 95, "total_steps": 336, "loss": 0.3993, "lr": 9.026662770409524e-06, "epoch": 0.8482142857142857, "percentage": 28.27, "elapsed_time": "3:21:16", "remaining_time": "8:30:37"}
96
+ {"current_steps": 96, "total_steps": 336, "loss": 0.3997, "lr": 8.99561086464603e-06, "epoch": 0.8571428571428571, "percentage": 28.57, "elapsed_time": "3:23:33", "remaining_time": "8:28:53"}
97
+ {"current_steps": 97, "total_steps": 336, "loss": 0.3985, "lr": 8.964126579349237e-06, "epoch": 0.8660714285714286, "percentage": 28.87, "elapsed_time": "3:25:50", "remaining_time": "8:27:10"}
98
+ {"current_steps": 98, "total_steps": 336, "loss": 0.3857, "lr": 8.932213321547769e-06, "epoch": 0.875, "percentage": 29.17, "elapsed_time": "3:28:09", "remaining_time": "8:25:31"}
99
+ {"current_steps": 99, "total_steps": 336, "loss": 0.3985, "lr": 8.899874544690921e-06, "epoch": 0.8839285714285714, "percentage": 29.46, "elapsed_time": "3:30:29", "remaining_time": "8:23:53"}
100
+ {"current_steps": 100, "total_steps": 336, "loss": 0.3912, "lr": 8.86711374827494e-06, "epoch": 0.8928571428571429, "percentage": 29.76, "elapsed_time": "3:32:50", "remaining_time": "8:22:19"}
101
+ {"current_steps": 101, "total_steps": 336, "loss": 0.3972, "lr": 8.833934477464348e-06, "epoch": 0.9017857142857143, "percentage": 30.06, "elapsed_time": "3:39:32", "remaining_time": "8:30:48"}
102
+ {"current_steps": 102, "total_steps": 336, "loss": 0.3889, "lr": 8.800340322708291e-06, "epoch": 0.9107142857142857, "percentage": 30.36, "elapsed_time": "3:41:49", "remaining_time": "8:28:54"}
103
+ {"current_steps": 103, "total_steps": 336, "loss": 0.3991, "lr": 8.766334919352018e-06, "epoch": 0.9196428571428571, "percentage": 30.65, "elapsed_time": "3:44:04", "remaining_time": "8:26:53"}
104
+ {"current_steps": 104, "total_steps": 336, "loss": 0.3879, "lr": 8.73192194724347e-06, "epoch": 0.9285714285714286, "percentage": 30.95, "elapsed_time": "3:46:24", "remaining_time": "8:25:02"}
105
+ {"current_steps": 105, "total_steps": 336, "loss": 0.3962, "lr": 8.697105130335084e-06, "epoch": 0.9375, "percentage": 31.25, "elapsed_time": "3:48:43", "remaining_time": "8:23:11"}
106
+ {"current_steps": 106, "total_steps": 336, "loss": 0.3909, "lr": 8.661888236280813e-06, "epoch": 0.9464285714285714, "percentage": 31.55, "elapsed_time": "3:51:05", "remaining_time": "8:21:24"}
107
+ {"current_steps": 107, "total_steps": 336, "loss": 0.3899, "lr": 8.626275076028397e-06, "epoch": 0.9553571428571429, "percentage": 31.85, "elapsed_time": "3:53:27", "remaining_time": "8:19:38"}
108
+ {"current_steps": 108, "total_steps": 336, "loss": 0.3846, "lr": 8.590269503406986e-06, "epoch": 0.9642857142857143, "percentage": 32.14, "elapsed_time": "3:55:43", "remaining_time": "8:17:38"}
109
+ {"current_steps": 109, "total_steps": 336, "loss": 0.3922, "lr": 8.553875414710088e-06, "epoch": 0.9732142857142857, "percentage": 32.44, "elapsed_time": "3:58:04", "remaining_time": "8:15:48"}
110
+ {"current_steps": 110, "total_steps": 336, "loss": 0.3909, "lr": 8.517096748273951e-06, "epoch": 0.9821428571428571, "percentage": 32.74, "elapsed_time": "4:00:25", "remaining_time": "8:13:58"}
111
+ {"current_steps": 111, "total_steps": 336, "loss": 0.3893, "lr": 8.479937484051368e-06, "epoch": 0.9910714285714286, "percentage": 33.04, "elapsed_time": "4:02:42", "remaining_time": "8:11:59"}
112
+ {"current_steps": 112, "total_steps": 336, "loss": 0.3836, "lr": 8.442401643181e-06, "epoch": 1.0, "percentage": 33.33, "elapsed_time": "4:04:59", "remaining_time": "8:09:59"}
113
+ {"current_steps": 113, "total_steps": 336, "loss": 0.3733, "lr": 8.404493287552232e-06, "epoch": 1.0089285714285714, "percentage": 33.63, "elapsed_time": "4:07:21", "remaining_time": "8:08:09"}
114
+ {"current_steps": 114, "total_steps": 336, "loss": 0.3687, "lr": 8.366216519365623e-06, "epoch": 1.0178571428571428, "percentage": 33.93, "elapsed_time": "4:09:43", "remaining_time": "8:06:17"}
115
+ {"current_steps": 115, "total_steps": 336, "loss": 0.3734, "lr": 8.327575480688985e-06, "epoch": 1.0267857142857142, "percentage": 34.23, "elapsed_time": "4:12:02", "remaining_time": "8:04:22"}
116
+ {"current_steps": 116, "total_steps": 336, "loss": 0.3782, "lr": 8.288574353009164e-06, "epoch": 1.0357142857142858, "percentage": 34.52, "elapsed_time": "4:14:23", "remaining_time": "8:02:27"}
117
+ {"current_steps": 117, "total_steps": 336, "loss": 0.3745, "lr": 8.249217356779544e-06, "epoch": 1.0446428571428572, "percentage": 34.82, "elapsed_time": "4:16:48", "remaining_time": "8:00:42"}
118
+ {"current_steps": 118, "total_steps": 336, "loss": 0.3731, "lr": 8.20950875096333e-06, "epoch": 1.0535714285714286, "percentage": 35.12, "elapsed_time": "4:19:23", "remaining_time": "7:59:12"}
119
+ {"current_steps": 119, "total_steps": 336, "loss": 0.3711, "lr": 8.169452832572676e-06, "epoch": 1.0625, "percentage": 35.42, "elapsed_time": "4:21:42", "remaining_time": "7:57:14"}
120
+ {"current_steps": 120, "total_steps": 336, "loss": 0.3701, "lr": 8.129053936203688e-06, "epoch": 1.0714285714285714, "percentage": 35.71, "elapsed_time": "4:24:07", "remaining_time": "7:55:25"}
121
+ {"current_steps": 121, "total_steps": 336, "loss": 0.3729, "lr": 8.088316433567369e-06, "epoch": 1.0803571428571428, "percentage": 36.01, "elapsed_time": "4:26:33", "remaining_time": "7:53:37"}
122
+ {"current_steps": 122, "total_steps": 336, "loss": 0.3725, "lr": 8.04724473301652e-06, "epoch": 1.0892857142857142, "percentage": 36.31, "elapsed_time": "4:28:56", "remaining_time": "7:51:45"}
123
+ {"current_steps": 123, "total_steps": 336, "loss": 0.3689, "lr": 8.005843279068724e-06, "epoch": 1.0982142857142858, "percentage": 36.61, "elapsed_time": "4:31:17", "remaining_time": "7:49:47"}
124
+ {"current_steps": 124, "total_steps": 336, "loss": 0.377, "lr": 7.964116551925365e-06, "epoch": 1.1071428571428572, "percentage": 36.9, "elapsed_time": "4:33:36", "remaining_time": "7:47:46"}
125
+ {"current_steps": 125, "total_steps": 336, "loss": 0.3684, "lr": 7.92206906698682e-06, "epoch": 1.1160714285714286, "percentage": 37.2, "elapsed_time": "4:35:58", "remaining_time": "7:45:50"}
126
+ {"current_steps": 126, "total_steps": 336, "loss": 0.3744, "lr": 7.879705374363831e-06, "epoch": 1.125, "percentage": 37.5, "elapsed_time": "4:38:17", "remaining_time": "7:43:48"}
127
+ {"current_steps": 127, "total_steps": 336, "loss": 0.3754, "lr": 7.837030058385117e-06, "epoch": 1.1339285714285714, "percentage": 37.8, "elapsed_time": "4:40:39", "remaining_time": "7:41:52"}
128
+ {"current_steps": 128, "total_steps": 336, "loss": 0.3712, "lr": 7.794047737101298e-06, "epoch": 1.1428571428571428, "percentage": 38.1, "elapsed_time": "4:43:04", "remaining_time": "7:39:59"}
129
+ {"current_steps": 129, "total_steps": 336, "loss": 0.3721, "lr": 7.750763061785139e-06, "epoch": 1.1517857142857142, "percentage": 38.39, "elapsed_time": "4:45:28", "remaining_time": "7:38:05"}
130
+ {"current_steps": 130, "total_steps": 336, "loss": 0.3754, "lr": 7.707180716428237e-06, "epoch": 1.1607142857142858, "percentage": 38.69, "elapsed_time": "4:47:48", "remaining_time": "7:36:04"}
131
+ {"current_steps": 131, "total_steps": 336, "loss": 0.3675, "lr": 7.663305417234146e-06, "epoch": 1.1696428571428572, "percentage": 38.99, "elapsed_time": "4:50:08", "remaining_time": "7:34:02"}
132
+ {"current_steps": 132, "total_steps": 336, "loss": 0.3786, "lr": 7.619141912108008e-06, "epoch": 1.1785714285714286, "percentage": 39.29, "elapsed_time": "4:52:45", "remaining_time": "7:32:27"}
133
+ {"current_steps": 133, "total_steps": 336, "loss": 0.3731, "lr": 7.57469498014278e-06, "epoch": 1.1875, "percentage": 39.58, "elapsed_time": "4:55:06", "remaining_time": "7:30:25"}
134
+ {"current_steps": 134, "total_steps": 336, "loss": 0.3735, "lr": 7.529969431102063e-06, "epoch": 1.1964285714285714, "percentage": 39.88, "elapsed_time": "4:57:28", "remaining_time": "7:28:25"}
135
+ {"current_steps": 135, "total_steps": 336, "loss": 0.3689, "lr": 7.484970104899624e-06, "epoch": 1.2053571428571428, "percentage": 40.18, "elapsed_time": "4:59:51", "remaining_time": "7:26:27"}
136
+ {"current_steps": 136, "total_steps": 336, "loss": 0.3704, "lr": 7.4397018710756415e-06, "epoch": 1.2142857142857142, "percentage": 40.48, "elapsed_time": "5:02:16", "remaining_time": "7:24:30"}
137
+ {"current_steps": 137, "total_steps": 336, "loss": 0.3702, "lr": 7.394169628269771e-06, "epoch": 1.2232142857142858, "percentage": 40.77, "elapsed_time": "5:04:38", "remaining_time": "7:22:29"}
138
+ {"current_steps": 138, "total_steps": 336, "loss": 0.3688, "lr": 7.34837830369103e-06, "epoch": 1.2321428571428572, "percentage": 41.07, "elapsed_time": "5:06:57", "remaining_time": "7:20:25"}
139
+ {"current_steps": 139, "total_steps": 336, "loss": 0.3731, "lr": 7.302332852584619e-06, "epoch": 1.2410714285714286, "percentage": 41.37, "elapsed_time": "5:09:21", "remaining_time": "7:18:26"}
140
+ {"current_steps": 140, "total_steps": 336, "loss": 0.3691, "lr": 7.2560382576956875e-06, "epoch": 1.25, "percentage": 41.67, "elapsed_time": "5:11:50", "remaining_time": "7:16:35"}
141
+ {"current_steps": 141, "total_steps": 336, "loss": 0.3664, "lr": 7.209499528730138e-06, "epoch": 1.2589285714285714, "percentage": 41.96, "elapsed_time": "5:14:24", "remaining_time": "7:14:49"}
142
+ {"current_steps": 142, "total_steps": 336, "loss": 0.3702, "lr": 7.162721701812506e-06, "epoch": 1.2678571428571428, "percentage": 42.26, "elapsed_time": "5:16:49", "remaining_time": "7:12:50"}
143
+ {"current_steps": 143, "total_steps": 336, "loss": 0.3722, "lr": 7.115709838940983e-06, "epoch": 1.2767857142857144, "percentage": 42.56, "elapsed_time": "5:19:17", "remaining_time": "7:10:55"}
144
+ {"current_steps": 144, "total_steps": 336, "loss": 0.3741, "lr": 7.068469027439642e-06, "epoch": 1.2857142857142856, "percentage": 42.86, "elapsed_time": "5:21:46", "remaining_time": "7:09:01"}
145
+ {"current_steps": 145, "total_steps": 336, "loss": 0.3675, "lr": 7.02100437940791e-06, "epoch": 1.2946428571428572, "percentage": 43.15, "elapsed_time": "5:24:10", "remaining_time": "7:07:00"}
146
+ {"current_steps": 146, "total_steps": 336, "loss": 0.3695, "lr": 6.9733210311673826e-06, "epoch": 1.3035714285714286, "percentage": 43.45, "elapsed_time": "5:26:30", "remaining_time": "7:04:54"}
147
+ {"current_steps": 147, "total_steps": 336, "loss": 0.3665, "lr": 6.925424142705997e-06, "epoch": 1.3125, "percentage": 43.75, "elapsed_time": "5:28:53", "remaining_time": "7:02:51"}
148
+ {"current_steps": 148, "total_steps": 336, "loss": 0.3702, "lr": 6.8773188971196515e-06, "epoch": 1.3214285714285714, "percentage": 44.05, "elapsed_time": "5:31:13", "remaining_time": "7:00:45"}
149
+ {"current_steps": 149, "total_steps": 336, "loss": 0.3739, "lr": 6.829010500051319e-06, "epoch": 1.3303571428571428, "percentage": 44.35, "elapsed_time": "5:33:37", "remaining_time": "6:58:42"}
150
+ {"current_steps": 150, "total_steps": 336, "loss": 0.3689, "lr": 6.780504179127735e-06, "epoch": 1.3392857142857144, "percentage": 44.64, "elapsed_time": "5:36:03", "remaining_time": "6:56:42"}
151
+ {"current_steps": 151, "total_steps": 336, "loss": 0.3686, "lr": 6.731805183393696e-06, "epoch": 1.3482142857142856, "percentage": 44.94, "elapsed_time": "5:42:55", "remaining_time": "7:00:08"}
152
+ {"current_steps": 152, "total_steps": 336, "loss": 0.3685, "lr": 6.682918782744033e-06, "epoch": 1.3571428571428572, "percentage": 45.24, "elapsed_time": "5:45:16", "remaining_time": "6:57:58"}
153
+ {"current_steps": 153, "total_steps": 336, "loss": 0.3626, "lr": 6.63385026735334e-06, "epoch": 1.3660714285714286, "percentage": 45.54, "elapsed_time": "5:47:38", "remaining_time": "6:55:48"}
154
+ {"current_steps": 154, "total_steps": 336, "loss": 0.3635, "lr": 6.584604947103515e-06, "epoch": 1.375, "percentage": 45.83, "elapsed_time": "5:49:58", "remaining_time": "6:53:36"}
155
+ {"current_steps": 155, "total_steps": 336, "loss": 0.3699, "lr": 6.535188151009143e-06, "epoch": 1.3839285714285714, "percentage": 46.13, "elapsed_time": "5:52:19", "remaining_time": "6:51:25"}
156
+ {"current_steps": 156, "total_steps": 336, "loss": 0.362, "lr": 6.4856052266408375e-06, "epoch": 1.3928571428571428, "percentage": 46.43, "elapsed_time": "5:54:41", "remaining_time": "6:49:15"}
157
+ {"current_steps": 157, "total_steps": 336, "loss": 0.3652, "lr": 6.4358615395465506e-06, "epoch": 1.4017857142857144, "percentage": 46.73, "elapsed_time": "5:57:05", "remaining_time": "6:47:07"}
158
+ {"current_steps": 158, "total_steps": 336, "loss": 0.3703, "lr": 6.385962472670953e-06, "epoch": 1.4107142857142856, "percentage": 47.02, "elapsed_time": "5:59:28", "remaining_time": "6:44:59"}
159
+ {"current_steps": 159, "total_steps": 336, "loss": 0.3624, "lr": 6.335913425772926e-06, "epoch": 1.4196428571428572, "percentage": 47.32, "elapsed_time": "6:01:50", "remaining_time": "6:42:48"}
160
+ {"current_steps": 160, "total_steps": 336, "loss": 0.3663, "lr": 6.28571981484123e-06, "epoch": 1.4285714285714286, "percentage": 47.62, "elapsed_time": "6:04:10", "remaining_time": "6:40:35"}
161
+ {"current_steps": 161, "total_steps": 336, "loss": 0.3708, "lr": 6.235387071508427e-06, "epoch": 1.4375, "percentage": 47.92, "elapsed_time": "6:06:31", "remaining_time": "6:38:24"}
162
+ {"current_steps": 162, "total_steps": 336, "loss": 0.3703, "lr": 6.184920642463095e-06, "epoch": 1.4464285714285714, "percentage": 48.21, "elapsed_time": "6:08:53", "remaining_time": "6:36:12"}
163
+ {"current_steps": 163, "total_steps": 336, "loss": 0.3731, "lr": 6.1343259888604335e-06, "epoch": 1.4553571428571428, "percentage": 48.51, "elapsed_time": "6:11:16", "remaining_time": "6:34:03"}
164
+ {"current_steps": 164, "total_steps": 336, "loss": 0.3677, "lr": 6.083608585731283e-06, "epoch": 1.4642857142857144, "percentage": 48.81, "elapsed_time": "6:13:41", "remaining_time": "6:31:55"}
165
+ {"current_steps": 165, "total_steps": 336, "loss": 0.3708, "lr": 6.032773921389655e-06, "epoch": 1.4732142857142856, "percentage": 49.11, "elapsed_time": "6:16:05", "remaining_time": "6:29:46"}
166
+ {"current_steps": 166, "total_steps": 336, "loss": 0.3665, "lr": 5.9818274968388225e-06, "epoch": 1.4821428571428572, "percentage": 49.4, "elapsed_time": "6:18:29", "remaining_time": "6:27:36"}
167
+ {"current_steps": 167, "total_steps": 336, "loss": 0.3625, "lr": 5.930774825176034e-06, "epoch": 1.4910714285714286, "percentage": 49.7, "elapsed_time": "6:20:49", "remaining_time": "6:25:23"}
168
+ {"current_steps": 168, "total_steps": 336, "loss": 0.3699, "lr": 5.879621430995927e-06, "epoch": 1.5, "percentage": 50.0, "elapsed_time": "6:23:06", "remaining_time": "6:23:06"}
169
+ {"current_steps": 169, "total_steps": 336, "loss": 0.3678, "lr": 5.8283728497926865e-06, "epoch": 1.5089285714285714, "percentage": 50.3, "elapsed_time": "6:25:25", "remaining_time": "6:20:51"}
170
+ {"current_steps": 170, "total_steps": 336, "loss": 0.3677, "lr": 5.777034627361025e-06, "epoch": 1.5178571428571428, "percentage": 50.6, "elapsed_time": "6:27:46", "remaining_time": "6:18:38"}
171
+ {"current_steps": 171, "total_steps": 336, "loss": 0.3697, "lr": 5.725612319196065e-06, "epoch": 1.5267857142857144, "percentage": 50.89, "elapsed_time": "6:30:04", "remaining_time": "6:16:23"}
172
+ {"current_steps": 172, "total_steps": 336, "loss": 0.3696, "lr": 5.674111489892144e-06, "epoch": 1.5357142857142856, "percentage": 51.19, "elapsed_time": "6:32:25", "remaining_time": "6:14:10"}
173
+ {"current_steps": 173, "total_steps": 336, "loss": 0.3744, "lr": 5.622537712540664e-06, "epoch": 1.5446428571428572, "percentage": 51.49, "elapsed_time": "6:34:48", "remaining_time": "6:11:59"}
174
+ {"current_steps": 174, "total_steps": 336, "loss": 0.3668, "lr": 5.570896568126994e-06, "epoch": 1.5535714285714286, "percentage": 51.79, "elapsed_time": "6:37:15", "remaining_time": "6:09:51"}
175
+ {"current_steps": 175, "total_steps": 336, "loss": 0.3649, "lr": 5.5191936449265345e-06, "epoch": 1.5625, "percentage": 52.08, "elapsed_time": "6:39:30", "remaining_time": "6:07:33"}
176
+ {"current_steps": 176, "total_steps": 336, "loss": 0.3683, "lr": 5.4674345379e-06, "epoch": 1.5714285714285714, "percentage": 52.38, "elapsed_time": "6:41:47", "remaining_time": "6:05:16"}
177
+ {"current_steps": 177, "total_steps": 336, "loss": 0.3658, "lr": 5.415624848087959e-06, "epoch": 1.5803571428571428, "percentage": 52.68, "elapsed_time": "6:44:08", "remaining_time": "6:03:02"}
178
+ {"current_steps": 178, "total_steps": 336, "loss": 0.3663, "lr": 5.36377018200472e-06, "epoch": 1.5892857142857144, "percentage": 52.98, "elapsed_time": "6:46:27", "remaining_time": "6:00:47"}
179
+ {"current_steps": 179, "total_steps": 336, "loss": 0.3618, "lr": 5.311876151031642e-06, "epoch": 1.5982142857142856, "percentage": 53.27, "elapsed_time": "6:48:48", "remaining_time": "5:58:34"}
180
+ {"current_steps": 180, "total_steps": 336, "loss": 0.3707, "lr": 5.259948370809902e-06, "epoch": 1.6071428571428572, "percentage": 53.57, "elapsed_time": "6:51:11", "remaining_time": "5:56:22"}
181
+ {"current_steps": 181, "total_steps": 336, "loss": 0.3647, "lr": 5.207992460632805e-06, "epoch": 1.6160714285714286, "percentage": 53.87, "elapsed_time": "6:53:35", "remaining_time": "5:54:10"}
182
+ {"current_steps": 182, "total_steps": 336, "loss": 0.3646, "lr": 5.156014042837696e-06, "epoch": 1.625, "percentage": 54.17, "elapsed_time": "6:55:58", "remaining_time": "5:51:58"}
183
+ {"current_steps": 183, "total_steps": 336, "loss": 0.3603, "lr": 5.104018742197557e-06, "epoch": 1.6339285714285714, "percentage": 54.46, "elapsed_time": "6:58:16", "remaining_time": "5:49:42"}
184
+ {"current_steps": 184, "total_steps": 336, "loss": 0.3596, "lr": 5.052012185312322e-06, "epoch": 1.6428571428571428, "percentage": 54.76, "elapsed_time": "7:00:36", "remaining_time": "5:47:27"}
185
+ {"current_steps": 185, "total_steps": 336, "loss": 0.3646, "lr": 5e-06, "epoch": 1.6517857142857144, "percentage": 55.06, "elapsed_time": "7:02:56", "remaining_time": "5:45:12"}
186
+ {"current_steps": 186, "total_steps": 336, "loss": 0.3616, "lr": 4.94798781468768e-06, "epoch": 1.6607142857142856, "percentage": 55.36, "elapsed_time": "7:05:17", "remaining_time": "5:42:58"}
187
+ {"current_steps": 187, "total_steps": 336, "loss": 0.366, "lr": 4.895981257802444e-06, "epoch": 1.6696428571428572, "percentage": 55.65, "elapsed_time": "7:07:39", "remaining_time": "5:40:45"}
188
+ {"current_steps": 188, "total_steps": 336, "loss": 0.3613, "lr": 4.843985957162304e-06, "epoch": 1.6785714285714286, "percentage": 55.95, "elapsed_time": "7:10:01", "remaining_time": "5:38:31"}
189
+ {"current_steps": 189, "total_steps": 336, "loss": 0.3614, "lr": 4.792007539367198e-06, "epoch": 1.6875, "percentage": 56.25, "elapsed_time": "7:12:25", "remaining_time": "5:36:19"}
190
+ {"current_steps": 190, "total_steps": 336, "loss": 0.3665, "lr": 4.740051629190099e-06, "epoch": 1.6964285714285714, "percentage": 56.55, "elapsed_time": "7:14:48", "remaining_time": "5:34:06"}
191
+ {"current_steps": 191, "total_steps": 336, "loss": 0.3605, "lr": 4.6881238489683596e-06, "epoch": 1.7053571428571428, "percentage": 56.85, "elapsed_time": "7:17:07", "remaining_time": "5:31:50"}
192
+ {"current_steps": 192, "total_steps": 336, "loss": 0.3687, "lr": 4.636229817995281e-06, "epoch": 1.7142857142857144, "percentage": 57.14, "elapsed_time": "7:19:26", "remaining_time": "5:29:34"}
193
+ {"current_steps": 193, "total_steps": 336, "loss": 0.37, "lr": 4.584375151912043e-06, "epoch": 1.7232142857142856, "percentage": 57.44, "elapsed_time": "7:21:47", "remaining_time": "5:27:20"}
194
+ {"current_steps": 194, "total_steps": 336, "loss": 0.3699, "lr": 4.532565462099999e-06, "epoch": 1.7321428571428572, "percentage": 57.74, "elapsed_time": "7:24:07", "remaining_time": "5:25:05"}
195
+ {"current_steps": 195, "total_steps": 336, "loss": 0.3634, "lr": 4.480806355073467e-06, "epoch": 1.7410714285714286, "percentage": 58.04, "elapsed_time": "7:26:30", "remaining_time": "5:22:51"}
196
+ {"current_steps": 196, "total_steps": 336, "loss": 0.3647, "lr": 4.429103431873009e-06, "epoch": 1.75, "percentage": 58.33, "elapsed_time": "7:28:53", "remaining_time": "5:20:38"}
197
+ {"current_steps": 197, "total_steps": 336, "loss": 0.3685, "lr": 4.377462287459338e-06, "epoch": 1.7589285714285714, "percentage": 58.63, "elapsed_time": "7:31:15", "remaining_time": "5:18:23"}
198
+ {"current_steps": 198, "total_steps": 336, "loss": 0.359, "lr": 4.3258885101078565e-06, "epoch": 1.7678571428571428, "percentage": 58.93, "elapsed_time": "7:33:37", "remaining_time": "5:16:09"}
199
+ {"current_steps": 199, "total_steps": 336, "loss": 0.3589, "lr": 4.274387680803936e-06, "epoch": 1.7767857142857144, "percentage": 59.23, "elapsed_time": "7:35:55", "remaining_time": "5:13:52"}
200
+ {"current_steps": 200, "total_steps": 336, "loss": 0.3657, "lr": 4.2229653726389765e-06, "epoch": 1.7857142857142856, "percentage": 59.52, "elapsed_time": "7:38:17", "remaining_time": "5:11:38"}
201
+ {"current_steps": 201, "total_steps": 336, "loss": 0.3637, "lr": 4.171627150207314e-06, "epoch": 1.7946428571428572, "percentage": 59.82, "elapsed_time": "7:44:55", "remaining_time": "5:12:16"}
202
+ {"current_steps": 202, "total_steps": 336, "loss": 0.3655, "lr": 4.120378569004074e-06, "epoch": 1.8035714285714286, "percentage": 60.12, "elapsed_time": "7:47:18", "remaining_time": "5:09:59"}
203
+ {"current_steps": 203, "total_steps": 336, "loss": 0.3662, "lr": 4.069225174823968e-06, "epoch": 1.8125, "percentage": 60.42, "elapsed_time": "7:49:42", "remaining_time": "5:07:44"}
204
+ {"current_steps": 204, "total_steps": 336, "loss": 0.3612, "lr": 4.018172503161179e-06, "epoch": 1.8214285714285714, "percentage": 60.71, "elapsed_time": "7:52:05", "remaining_time": "5:05:28"}
205
+ {"current_steps": 205, "total_steps": 336, "loss": 0.3594, "lr": 3.967226078610346e-06, "epoch": 1.8303571428571428, "percentage": 61.01, "elapsed_time": "7:54:25", "remaining_time": "5:03:10"}
206
+ {"current_steps": 206, "total_steps": 336, "loss": 0.3685, "lr": 3.9163914142687185e-06, "epoch": 1.8392857142857144, "percentage": 61.31, "elapsed_time": "7:56:45", "remaining_time": "5:00:52"}
207
+ {"current_steps": 207, "total_steps": 336, "loss": 0.3582, "lr": 3.865674011139567e-06, "epoch": 1.8482142857142856, "percentage": 61.61, "elapsed_time": "7:59:07", "remaining_time": "4:58:34"}
208
+ {"current_steps": 208, "total_steps": 336, "loss": 0.3641, "lr": 3.815079357536907e-06, "epoch": 1.8571428571428572, "percentage": 61.9, "elapsed_time": "8:01:30", "remaining_time": "4:56:18"}
209
+ {"current_steps": 209, "total_steps": 336, "loss": 0.3665, "lr": 3.7646129284915754e-06, "epoch": 1.8660714285714286, "percentage": 62.2, "elapsed_time": "8:03:52", "remaining_time": "4:54:01"}
210
+ {"current_steps": 210, "total_steps": 336, "loss": 0.3612, "lr": 3.714280185158771e-06, "epoch": 1.875, "percentage": 62.5, "elapsed_time": "8:06:19", "remaining_time": "4:51:47"}
211
+ {"current_steps": 211, "total_steps": 336, "loss": 0.3629, "lr": 3.6640865742270756e-06, "epoch": 1.8839285714285714, "percentage": 62.8, "elapsed_time": "8:08:40", "remaining_time": "4:49:30"}
212
+ {"current_steps": 212, "total_steps": 336, "loss": 0.3646, "lr": 3.614037527329048e-06, "epoch": 1.8928571428571428, "percentage": 63.1, "elapsed_time": "8:11:04", "remaining_time": "4:47:13"}
213
+ {"current_steps": 213, "total_steps": 336, "loss": 0.3607, "lr": 3.5641384604534503e-06, "epoch": 1.9017857142857144, "percentage": 63.39, "elapsed_time": "8:13:23", "remaining_time": "4:44:55"}
214
+ {"current_steps": 214, "total_steps": 336, "loss": 0.3613, "lr": 3.5143947733591633e-06, "epoch": 1.9107142857142856, "percentage": 63.69, "elapsed_time": "8:15:43", "remaining_time": "4:42:36"}
215
+ {"current_steps": 215, "total_steps": 336, "loss": 0.365, "lr": 3.4648118489908588e-06, "epoch": 1.9196428571428572, "percentage": 63.99, "elapsed_time": "8:18:09", "remaining_time": "4:40:21"}
216
+ {"current_steps": 216, "total_steps": 336, "loss": 0.3644, "lr": 3.4153950528964867e-06, "epoch": 1.9285714285714286, "percentage": 64.29, "elapsed_time": "8:20:28", "remaining_time": "4:38:02"}
217
+ {"current_steps": 217, "total_steps": 336, "loss": 0.3662, "lr": 3.366149732646661e-06, "epoch": 1.9375, "percentage": 64.58, "elapsed_time": "8:22:52", "remaining_time": "4:35:46"}
218
+ {"current_steps": 218, "total_steps": 336, "loss": 0.3598, "lr": 3.3170812172559695e-06, "epoch": 1.9464285714285714, "percentage": 64.88, "elapsed_time": "8:25:19", "remaining_time": "4:33:31"}
219
+ {"current_steps": 219, "total_steps": 336, "loss": 0.3625, "lr": 3.268194816606305e-06, "epoch": 1.9553571428571428, "percentage": 65.18, "elapsed_time": "8:27:42", "remaining_time": "4:31:14"}
220
+ {"current_steps": 220, "total_steps": 336, "loss": 0.364, "lr": 3.2194958208722656e-06, "epoch": 1.9642857142857144, "percentage": 65.48, "elapsed_time": "8:30:03", "remaining_time": "4:28:56"}
221
+ {"current_steps": 221, "total_steps": 336, "loss": 0.3618, "lr": 3.1709894999486828e-06, "epoch": 1.9732142857142856, "percentage": 65.77, "elapsed_time": "8:32:23", "remaining_time": "4:26:38"}
222
+ {"current_steps": 222, "total_steps": 336, "loss": 0.3627, "lr": 3.1226811028803514e-06, "epoch": 1.9821428571428572, "percentage": 66.07, "elapsed_time": "8:34:43", "remaining_time": "4:24:19"}
223
+ {"current_steps": 223, "total_steps": 336, "loss": 0.3661, "lr": 3.0745758572940044e-06, "epoch": 1.9910714285714286, "percentage": 66.37, "elapsed_time": "8:37:06", "remaining_time": "4:22:01"}
224
+ {"current_steps": 224, "total_steps": 336, "loss": 0.3518, "lr": 3.0266789688326187e-06, "epoch": 2.0, "percentage": 66.67, "elapsed_time": "8:39:28", "remaining_time": "4:19:44"}
225
+ {"current_steps": 225, "total_steps": 336, "loss": 0.3488, "lr": 2.978995620592092e-06, "epoch": 2.0089285714285716, "percentage": 66.96, "elapsed_time": "8:41:53", "remaining_time": "4:17:28"}
226
+ {"current_steps": 226, "total_steps": 336, "loss": 0.3498, "lr": 2.9315309725603596e-06, "epoch": 2.017857142857143, "percentage": 67.26, "elapsed_time": "8:44:20", "remaining_time": "4:15:12"}
227
+ {"current_steps": 227, "total_steps": 336, "loss": 0.3411, "lr": 2.884290161059017e-06, "epoch": 2.0267857142857144, "percentage": 67.56, "elapsed_time": "8:46:44", "remaining_time": "4:12:55"}
228
+ {"current_steps": 228, "total_steps": 336, "loss": 0.354, "lr": 2.8372782981874964e-06, "epoch": 2.0357142857142856, "percentage": 67.86, "elapsed_time": "8:49:08", "remaining_time": "4:10:38"}
229
+ {"current_steps": 229, "total_steps": 336, "loss": 0.3454, "lr": 2.7905004712698646e-06, "epoch": 2.044642857142857, "percentage": 68.15, "elapsed_time": "8:51:32", "remaining_time": "4:08:21"}
230
+ {"current_steps": 230, "total_steps": 336, "loss": 0.3528, "lr": 2.7439617423043146e-06, "epoch": 2.0535714285714284, "percentage": 68.45, "elapsed_time": "8:53:57", "remaining_time": "4:06:04"}
231
+ {"current_steps": 231, "total_steps": 336, "loss": 0.3489, "lr": 2.697667147415383e-06, "epoch": 2.0625, "percentage": 68.75, "elapsed_time": "8:56:21", "remaining_time": "4:03:48"}
232
+ {"current_steps": 232, "total_steps": 336, "loss": 0.3436, "lr": 2.6516216963089698e-06, "epoch": 2.0714285714285716, "percentage": 69.05, "elapsed_time": "8:58:49", "remaining_time": "4:01:32"}
233
+ {"current_steps": 233, "total_steps": 336, "loss": 0.3432, "lr": 2.605830371730229e-06, "epoch": 2.080357142857143, "percentage": 69.35, "elapsed_time": "9:01:16", "remaining_time": "3:59:16"}
234
+ {"current_steps": 234, "total_steps": 336, "loss": 0.3424, "lr": 2.560298128924358e-06, "epoch": 2.0892857142857144, "percentage": 69.64, "elapsed_time": "9:03:43", "remaining_time": "3:57:00"}
235
+ {"current_steps": 235, "total_steps": 336, "loss": 0.3421, "lr": 2.5150298951003783e-06, "epoch": 2.0982142857142856, "percentage": 69.94, "elapsed_time": "9:06:07", "remaining_time": "3:54:43"}
236
+ {"current_steps": 236, "total_steps": 336, "loss": 0.3458, "lr": 2.470030568897938e-06, "epoch": 2.107142857142857, "percentage": 70.24, "elapsed_time": "9:08:30", "remaining_time": "3:52:25"}
237
+ {"current_steps": 237, "total_steps": 336, "loss": 0.349, "lr": 2.425305019857222e-06, "epoch": 2.1160714285714284, "percentage": 70.54, "elapsed_time": "9:10:54", "remaining_time": "3:50:07"}
238
+ {"current_steps": 238, "total_steps": 336, "loss": 0.3507, "lr": 2.3808580878919948e-06, "epoch": 2.125, "percentage": 70.83, "elapsed_time": "9:13:19", "remaining_time": "3:47:50"}
239
+ {"current_steps": 239, "total_steps": 336, "loss": 0.3429, "lr": 2.336694582765857e-06, "epoch": 2.1339285714285716, "percentage": 71.13, "elapsed_time": "9:15:46", "remaining_time": "3:45:33"}
240
+ {"current_steps": 240, "total_steps": 336, "loss": 0.3459, "lr": 2.2928192835717642e-06, "epoch": 2.142857142857143, "percentage": 71.43, "elapsed_time": "9:18:12", "remaining_time": "3:43:17"}
241
+ {"current_steps": 241, "total_steps": 336, "loss": 0.3471, "lr": 2.2492369382148634e-06, "epoch": 2.1517857142857144, "percentage": 71.73, "elapsed_time": "9:20:41", "remaining_time": "3:41:01"}
242
+ {"current_steps": 242, "total_steps": 336, "loss": 0.3456, "lr": 2.205952262898704e-06, "epoch": 2.1607142857142856, "percentage": 72.02, "elapsed_time": "9:23:07", "remaining_time": "3:38:43"}
243
+ {"current_steps": 243, "total_steps": 336, "loss": 0.3459, "lr": 2.1629699416148832e-06, "epoch": 2.169642857142857, "percentage": 72.32, "elapsed_time": "9:25:30", "remaining_time": "3:36:25"}
244
+ {"current_steps": 244, "total_steps": 336, "loss": 0.3513, "lr": 2.120294625636171e-06, "epoch": 2.1785714285714284, "percentage": 72.62, "elapsed_time": "9:27:55", "remaining_time": "3:34:08"}
245
+ {"current_steps": 245, "total_steps": 336, "loss": 0.3423, "lr": 2.077930933013182e-06, "epoch": 2.1875, "percentage": 72.92, "elapsed_time": "9:30:19", "remaining_time": "3:31:49"}
246
+ {"current_steps": 246, "total_steps": 336, "loss": 0.3511, "lr": 2.0358834480746363e-06, "epoch": 2.1964285714285716, "percentage": 73.21, "elapsed_time": "9:32:43", "remaining_time": "3:29:32"}
247
+ {"current_steps": 247, "total_steps": 336, "loss": 0.3412, "lr": 1.9941567209312768e-06, "epoch": 2.205357142857143, "percentage": 73.51, "elapsed_time": "9:35:11", "remaining_time": "3:27:15"}
248
+ {"current_steps": 248, "total_steps": 336, "loss": 0.3439, "lr": 1.9527552669834797e-06, "epoch": 2.2142857142857144, "percentage": 73.81, "elapsed_time": "9:37:39", "remaining_time": "3:24:58"}
249
+ {"current_steps": 249, "total_steps": 336, "loss": 0.3429, "lr": 1.911683566432633e-06, "epoch": 2.2232142857142856, "percentage": 74.11, "elapsed_time": "9:40:08", "remaining_time": "3:22:42"}
250
+ {"current_steps": 250, "total_steps": 336, "loss": 0.3399, "lr": 1.8709460637963123e-06, "epoch": 2.232142857142857, "percentage": 74.4, "elapsed_time": "9:42:32", "remaining_time": "3:20:23"}
251
+ {"current_steps": 251, "total_steps": 336, "loss": 0.3506, "lr": 1.8305471674273262e-06, "epoch": 2.2410714285714284, "percentage": 74.7, "elapsed_time": "9:49:13", "remaining_time": "3:19:32"}
252
+ {"current_steps": 252, "total_steps": 336, "loss": 0.3421, "lr": 1.7904912490366723e-06, "epoch": 2.25, "percentage": 75.0, "elapsed_time": "9:51:36", "remaining_time": "3:17:12"}
253
+ {"current_steps": 253, "total_steps": 336, "loss": 0.3516, "lr": 1.7507826432204571e-06, "epoch": 2.2589285714285716, "percentage": 75.3, "elapsed_time": "9:54:00", "remaining_time": "3:14:52"}
254
+ {"current_steps": 254, "total_steps": 336, "loss": 0.3476, "lr": 1.711425646990838e-06, "epoch": 2.267857142857143, "percentage": 75.6, "elapsed_time": "9:56:28", "remaining_time": "3:12:33"}
255
+ {"current_steps": 255, "total_steps": 336, "loss": 0.3456, "lr": 1.6724245193110178e-06, "epoch": 2.2767857142857144, "percentage": 75.89, "elapsed_time": "9:58:55", "remaining_time": "3:10:14"}
256
+ {"current_steps": 256, "total_steps": 336, "loss": 0.3546, "lr": 1.6337834806343783e-06, "epoch": 2.2857142857142856, "percentage": 76.19, "elapsed_time": "10:01:22", "remaining_time": "3:07:55"}
257
+ {"current_steps": 257, "total_steps": 336, "loss": 0.3518, "lr": 1.595506712447768e-06, "epoch": 2.294642857142857, "percentage": 76.49, "elapsed_time": "10:03:48", "remaining_time": "3:05:36"}
258
+ {"current_steps": 258, "total_steps": 336, "loss": 0.3488, "lr": 1.557598356819e-06, "epoch": 2.3035714285714284, "percentage": 76.79, "elapsed_time": "10:06:11", "remaining_time": "3:03:15"}
259
+ {"current_steps": 259, "total_steps": 336, "loss": 0.3471, "lr": 1.5200625159486322e-06, "epoch": 2.3125, "percentage": 77.08, "elapsed_time": "10:08:30", "remaining_time": "3:00:54"}
260
+ {"current_steps": 260, "total_steps": 336, "loss": 0.3419, "lr": 1.482903251726049e-06, "epoch": 2.3214285714285716, "percentage": 77.38, "elapsed_time": "10:10:56", "remaining_time": "2:58:34"}
261
+ {"current_steps": 261, "total_steps": 336, "loss": 0.3458, "lr": 1.446124585289913e-06, "epoch": 2.330357142857143, "percentage": 77.68, "elapsed_time": "10:13:20", "remaining_time": "2:56:14"}
262
+ {"current_steps": 262, "total_steps": 336, "loss": 0.3493, "lr": 1.409730496593016e-06, "epoch": 2.3392857142857144, "percentage": 77.98, "elapsed_time": "10:15:47", "remaining_time": "2:53:55"}
263
+ {"current_steps": 263, "total_steps": 336, "loss": 0.3503, "lr": 1.3737249239716043e-06, "epoch": 2.3482142857142856, "percentage": 78.27, "elapsed_time": "10:18:16", "remaining_time": "2:51:36"}
264
+ {"current_steps": 264, "total_steps": 336, "loss": 0.3493, "lr": 1.3381117637191887e-06, "epoch": 2.357142857142857, "percentage": 78.57, "elapsed_time": "10:20:45", "remaining_time": "2:49:17"}
265
+ {"current_steps": 265, "total_steps": 336, "loss": 0.3467, "lr": 1.3028948696649164e-06, "epoch": 2.3660714285714284, "percentage": 78.87, "elapsed_time": "10:23:11", "remaining_time": "2:46:57"}
266
+ {"current_steps": 266, "total_steps": 336, "loss": 0.3443, "lr": 1.2680780527565313e-06, "epoch": 2.375, "percentage": 79.17, "elapsed_time": "10:25:35", "remaining_time": "2:44:37"}
267
+ {"current_steps": 267, "total_steps": 336, "loss": 0.3461, "lr": 1.2336650806479826e-06, "epoch": 2.3839285714285716, "percentage": 79.46, "elapsed_time": "10:28:00", "remaining_time": "2:42:17"}
268
+ {"current_steps": 268, "total_steps": 336, "loss": 0.3543, "lr": 1.1996596772917091e-06, "epoch": 2.392857142857143, "percentage": 79.76, "elapsed_time": "10:30:26", "remaining_time": "2:39:57"}
269
+ {"current_steps": 269, "total_steps": 336, "loss": 0.3471, "lr": 1.1660655225356533e-06, "epoch": 2.4017857142857144, "percentage": 80.06, "elapsed_time": "10:32:53", "remaining_time": "2:37:38"}
270
+ {"current_steps": 270, "total_steps": 336, "loss": 0.3474, "lr": 1.132886251725061e-06, "epoch": 2.4107142857142856, "percentage": 80.36, "elapsed_time": "10:35:21", "remaining_time": "2:35:18"}
271
+ {"current_steps": 271, "total_steps": 336, "loss": 0.3451, "lr": 1.1001254553090812e-06, "epoch": 2.419642857142857, "percentage": 80.65, "elapsed_time": "10:37:50", "remaining_time": "2:32:59"}
272
+ {"current_steps": 272, "total_steps": 336, "loss": 0.3474, "lr": 1.0677866784522317e-06, "epoch": 2.4285714285714284, "percentage": 80.95, "elapsed_time": "10:40:19", "remaining_time": "2:30:39"}
273
+ {"current_steps": 273, "total_steps": 336, "loss": 0.3459, "lr": 1.0358734206507642e-06, "epoch": 2.4375, "percentage": 81.25, "elapsed_time": "10:42:45", "remaining_time": "2:28:19"}
274
+ {"current_steps": 274, "total_steps": 336, "loss": 0.3387, "lr": 1.004389135353972e-06, "epoch": 2.4464285714285716, "percentage": 81.55, "elapsed_time": "10:45:08", "remaining_time": "2:25:58"}
275
+ {"current_steps": 275, "total_steps": 336, "loss": 0.347, "lr": 9.733372295904776e-07, "epoch": 2.455357142857143, "percentage": 81.85, "elapsed_time": "10:47:32", "remaining_time": "2:23:38"}
276
+ {"current_steps": 276, "total_steps": 336, "loss": 0.3467, "lr": 9.427210635995482e-07, "epoch": 2.4642857142857144, "percentage": 82.14, "elapsed_time": "10:49:55", "remaining_time": "2:21:17"}
277
+ {"current_steps": 277, "total_steps": 336, "loss": 0.3524, "lr": 9.125439504674699e-07, "epoch": 2.4732142857142856, "percentage": 82.44, "elapsed_time": "10:52:23", "remaining_time": "2:18:57"}
278
+ {"current_steps": 278, "total_steps": 336, "loss": 0.351, "lr": 8.828091557690288e-07, "epoch": 2.482142857142857, "percentage": 82.74, "elapsed_time": "10:54:50", "remaining_time": "2:16:37"}
279
+ {"current_steps": 279, "total_steps": 336, "loss": 0.3455, "lr": 8.535198972141295e-07, "epoch": 2.4910714285714284, "percentage": 83.04, "elapsed_time": "10:57:20", "remaining_time": "2:14:17"}
280
+ {"current_steps": 280, "total_steps": 336, "loss": 0.3441, "lr": 8.246793442995954e-07, "epoch": 2.5, "percentage": 83.33, "elapsed_time": "10:59:49", "remaining_time": "2:11:57"}
281
+ {"current_steps": 281, "total_steps": 336, "loss": 0.338, "lr": 7.962906179661872e-07, "epoch": 2.508928571428571, "percentage": 83.63, "elapsed_time": "11:02:13", "remaining_time": "2:09:37"}
282
+ {"current_steps": 282, "total_steps": 336, "loss": 0.3423, "lr": 7.68356790260873e-07, "epoch": 2.517857142857143, "percentage": 83.93, "elapsed_time": "11:04:36", "remaining_time": "2:07:16"}
283
+ {"current_steps": 283, "total_steps": 336, "loss": 0.3437, "lr": 7.408808840043913e-07, "epoch": 2.5267857142857144, "percentage": 84.23, "elapsed_time": "11:07:00", "remaining_time": "2:04:55"}
284
+ {"current_steps": 284, "total_steps": 336, "loss": 0.3533, "lr": 7.138658724641417e-07, "epoch": 2.5357142857142856, "percentage": 84.52, "elapsed_time": "11:09:27", "remaining_time": "2:02:34"}
285
+ {"current_steps": 285, "total_steps": 336, "loss": 0.3473, "lr": 6.873146790324359e-07, "epoch": 2.544642857142857, "percentage": 84.82, "elapsed_time": "11:11:54", "remaining_time": "2:00:14"}
286
+ {"current_steps": 286, "total_steps": 336, "loss": 0.3466, "lr": 6.612301769101464e-07, "epoch": 2.553571428571429, "percentage": 85.12, "elapsed_time": "11:14:23", "remaining_time": "1:57:54"}
287
+ {"current_steps": 287, "total_steps": 336, "loss": 0.3526, "lr": 6.356151887957901e-07, "epoch": 2.5625, "percentage": 85.42, "elapsed_time": "11:16:51", "remaining_time": "1:55:33"}
288
+ {"current_steps": 288, "total_steps": 336, "loss": 0.3436, "lr": 6.104724865800665e-07, "epoch": 2.571428571428571, "percentage": 85.71, "elapsed_time": "11:19:18", "remaining_time": "1:53:13"}
289
+ {"current_steps": 289, "total_steps": 336, "loss": 0.3435, "lr": 5.858047910459109e-07, "epoch": 2.580357142857143, "percentage": 86.01, "elapsed_time": "11:21:42", "remaining_time": "1:50:51"}
290
+ {"current_steps": 290, "total_steps": 336, "loss": 0.3425, "lr": 5.616147715740611e-07, "epoch": 2.5892857142857144, "percentage": 86.31, "elapsed_time": "11:24:07", "remaining_time": "1:48:30"}
291
+ {"current_steps": 291, "total_steps": 336, "loss": 0.3437, "lr": 5.379050458541995e-07, "epoch": 2.5982142857142856, "percentage": 86.61, "elapsed_time": "11:26:35", "remaining_time": "1:46:10"}
292
+ {"current_steps": 292, "total_steps": 336, "loss": 0.3477, "lr": 5.146781796016798e-07, "epoch": 2.607142857142857, "percentage": 86.9, "elapsed_time": "11:29:15", "remaining_time": "1:43:51"}
293
+ {"current_steps": 293, "total_steps": 336, "loss": 0.3452, "lr": 4.919366862798807e-07, "epoch": 2.616071428571429, "percentage": 87.2, "elapsed_time": "11:31:42", "remaining_time": "1:41:30"}
294
+ {"current_steps": 294, "total_steps": 336, "loss": 0.3487, "lr": 4.696830268282204e-07, "epoch": 2.625, "percentage": 87.5, "elapsed_time": "11:34:11", "remaining_time": "1:39:10"}
295
+ {"current_steps": 295, "total_steps": 336, "loss": 0.3458, "lr": 4.479196093958421e-07, "epoch": 2.633928571428571, "percentage": 87.8, "elapsed_time": "11:36:41", "remaining_time": "1:36:49"}
296
+ {"current_steps": 296, "total_steps": 336, "loss": 0.3467, "lr": 4.2664878908102556e-07, "epoch": 2.642857142857143, "percentage": 88.1, "elapsed_time": "11:39:07", "remaining_time": "1:34:28"}
297
+ {"current_steps": 297, "total_steps": 336, "loss": 0.3517, "lr": 4.058728676763313e-07, "epoch": 2.6517857142857144, "percentage": 88.39, "elapsed_time": "11:41:34", "remaining_time": "1:32:07"}
298
+ {"current_steps": 298, "total_steps": 336, "loss": 0.3459, "lr": 3.855940934195146e-07, "epoch": 2.6607142857142856, "percentage": 88.69, "elapsed_time": "11:43:58", "remaining_time": "1:29:46"}
299
+ {"current_steps": 299, "total_steps": 336, "loss": 0.3506, "lr": 3.6581466075023443e-07, "epoch": 2.669642857142857, "percentage": 88.99, "elapsed_time": "11:46:25", "remaining_time": "1:27:25"}
300
+ {"current_steps": 300, "total_steps": 336, "loss": 0.3445, "lr": 3.4653671007259084e-07, "epoch": 2.678571428571429, "percentage": 89.29, "elapsed_time": "11:48:55", "remaining_time": "1:25:04"}
301
+ {"current_steps": 301, "total_steps": 336, "loss": 0.3417, "lr": 3.2776232752349536e-07, "epoch": 2.6875, "percentage": 89.58, "elapsed_time": "11:55:44", "remaining_time": "1:23:13"}
302
+ {"current_steps": 302, "total_steps": 336, "loss": 0.3439, "lr": 3.0949354474692937e-07, "epoch": 2.696428571428571, "percentage": 89.88, "elapsed_time": "11:58:09", "remaining_time": "1:20:51"}
303
+ {"current_steps": 303, "total_steps": 336, "loss": 0.3439, "lr": 2.9173233867409057e-07, "epoch": 2.705357142857143, "percentage": 90.18, "elapsed_time": "12:00:34", "remaining_time": "1:18:28"}
304
+ {"current_steps": 304, "total_steps": 336, "loss": 0.3495, "lr": 2.7448063130946224e-07, "epoch": 2.7142857142857144, "percentage": 90.48, "elapsed_time": "12:03:00", "remaining_time": "1:16:06"}
305
+ {"current_steps": 305, "total_steps": 336, "loss": 0.3417, "lr": 2.577402895228243e-07, "epoch": 2.7232142857142856, "percentage": 90.77, "elapsed_time": "12:05:26", "remaining_time": "1:13:43"}
306
+ {"current_steps": 306, "total_steps": 336, "loss": 0.343, "lr": 2.4151312484723465e-07, "epoch": 2.732142857142857, "percentage": 91.07, "elapsed_time": "12:07:51", "remaining_time": "1:11:21"}
307
+ {"current_steps": 307, "total_steps": 336, "loss": 0.3462, "lr": 2.2580089328299747e-07, "epoch": 2.741071428571429, "percentage": 91.37, "elapsed_time": "12:10:16", "remaining_time": "1:08:59"}
308
+ {"current_steps": 308, "total_steps": 336, "loss": 0.3443, "lr": 2.106052951076365e-07, "epoch": 2.75, "percentage": 91.67, "elapsed_time": "12:12:42", "remaining_time": "1:06:36"}
309
+ {"current_steps": 309, "total_steps": 336, "loss": 0.345, "lr": 1.959279746919057e-07, "epoch": 2.758928571428571, "percentage": 91.96, "elapsed_time": "12:15:05", "remaining_time": "1:04:13"}
310
+ {"current_steps": 310, "total_steps": 336, "loss": 0.3458, "lr": 1.8177052032184285e-07, "epoch": 2.767857142857143, "percentage": 92.26, "elapsed_time": "12:17:28", "remaining_time": "1:01:51"}
311
+ {"current_steps": 311, "total_steps": 336, "loss": 0.3482, "lr": 1.6813446402689783e-07, "epoch": 2.7767857142857144, "percentage": 92.56, "elapsed_time": "12:19:52", "remaining_time": "0:59:28"}
312
+ {"current_steps": 312, "total_steps": 336, "loss": 0.3412, "lr": 1.5502128141414496e-07, "epoch": 2.7857142857142856, "percentage": 92.86, "elapsed_time": "12:22:14", "remaining_time": "0:57:05"}
313
+ {"current_steps": 313, "total_steps": 336, "loss": 0.3438, "lr": 1.4243239150860123e-07, "epoch": 2.794642857142857, "percentage": 93.15, "elapsed_time": "12:24:39", "remaining_time": "0:54:43"}
314
+ {"current_steps": 314, "total_steps": 336, "loss": 0.3497, "lr": 1.303691565996712e-07, "epoch": 2.803571428571429, "percentage": 93.45, "elapsed_time": "12:27:05", "remaining_time": "0:52:20"}
315
+ {"current_steps": 315, "total_steps": 336, "loss": 0.3449, "lr": 1.1883288209372512e-07, "epoch": 2.8125, "percentage": 93.75, "elapsed_time": "12:29:34", "remaining_time": "0:49:58"}
316
+ {"current_steps": 316, "total_steps": 336, "loss": 0.3448, "lr": 1.0782481637284014e-07, "epoch": 2.821428571428571, "percentage": 94.05, "elapsed_time": "12:32:01", "remaining_time": "0:47:35"}
317
+ {"current_steps": 317, "total_steps": 336, "loss": 0.3474, "lr": 9.734615065970454e-08, "epoch": 2.830357142857143, "percentage": 94.35, "elapsed_time": "12:34:29", "remaining_time": "0:45:13"}
318
+ {"current_steps": 318, "total_steps": 336, "loss": 0.3484, "lr": 8.739801888871468e-08, "epoch": 2.8392857142857144, "percentage": 94.64, "elapsed_time": "12:36:54", "remaining_time": "0:42:50"}
319
+ {"current_steps": 319, "total_steps": 336, "loss": 0.3508, "lr": 7.79814975832649e-08, "epoch": 2.8482142857142856, "percentage": 94.94, "elapsed_time": "12:39:19", "remaining_time": "0:40:27"}
320
+ {"current_steps": 320, "total_steps": 336, "loss": 0.3438, "lr": 6.909760573925561e-08, "epoch": 2.857142857142857, "percentage": 95.24, "elapsed_time": "12:41:44", "remaining_time": "0:38:05"}
321
+ {"current_steps": 321, "total_steps": 336, "loss": 0.3466, "lr": 6.074730471482049e-08, "epoch": 2.866071428571429, "percentage": 95.54, "elapsed_time": "12:44:10", "remaining_time": "0:35:42"}
322
+ {"current_steps": 322, "total_steps": 336, "loss": 0.3441, "lr": 5.2931498126298495e-08, "epoch": 2.875, "percentage": 95.83, "elapsed_time": "12:46:36", "remaining_time": "0:33:19"}
323
+ {"current_steps": 323, "total_steps": 336, "loss": 0.3475, "lr": 4.565103175044883e-08, "epoch": 2.883928571428571, "percentage": 96.13, "elapsed_time": "12:49:00", "remaining_time": "0:30:57"}
324
+ {"current_steps": 324, "total_steps": 336, "loss": 0.3478, "lr": 3.890669343292464e-08, "epoch": 2.892857142857143, "percentage": 96.43, "elapsed_time": "12:51:27", "remaining_time": "0:28:34"}
325
+ {"current_steps": 325, "total_steps": 336, "loss": 0.3431, "lr": 3.269921300301959e-08, "epoch": 2.9017857142857144, "percentage": 96.73, "elapsed_time": "12:53:53", "remaining_time": "0:26:11"}
326
+ {"current_steps": 326, "total_steps": 336, "loss": 0.3441, "lr": 2.702926219468882e-08, "epoch": 2.9107142857142856, "percentage": 97.02, "elapsed_time": "12:56:19", "remaining_time": "0:23:48"}
327
+ {"current_steps": 327, "total_steps": 336, "loss": 0.3538, "lr": 2.1897454573860388e-08, "epoch": 2.919642857142857, "percentage": 97.32, "elapsed_time": "12:58:43", "remaining_time": "0:21:25"}
328
+ {"current_steps": 328, "total_steps": 336, "loss": 0.3431, "lr": 1.7304345472035634e-08, "epoch": 2.928571428571429, "percentage": 97.62, "elapsed_time": "13:01:08", "remaining_time": "0:19:03"}
329
+ {"current_steps": 329, "total_steps": 336, "loss": 0.3412, "lr": 1.3250431926197793e-08, "epoch": 2.9375, "percentage": 97.92, "elapsed_time": "13:03:35", "remaining_time": "0:16:40"}
330
+ {"current_steps": 330, "total_steps": 336, "loss": 0.346, "lr": 9.73615262502503e-09, "epoch": 2.946428571428571, "percentage": 98.21, "elapsed_time": "13:06:02", "remaining_time": "0:14:17"}
331
+ {"current_steps": 331, "total_steps": 336, "loss": 0.3491, "lr": 6.761887861417293e-09, "epoch": 2.955357142857143, "percentage": 98.51, "elapsed_time": "13:08:28", "remaining_time": "0:11:54"}
332
+ {"current_steps": 332, "total_steps": 336, "loss": 0.3433, "lr": 4.327959491344791e-09, "epoch": 2.9642857142857144, "percentage": 98.81, "elapsed_time": "13:10:56", "remaining_time": "0:09:31"}
333
+ {"current_steps": 333, "total_steps": 336, "loss": 0.3467, "lr": 2.4346308990175205e-09, "epoch": 2.9732142857142856, "percentage": 99.11, "elapsed_time": "13:13:25", "remaining_time": "0:07:08"}
334
+ {"current_steps": 334, "total_steps": 336, "loss": 0.343, "lr": 1.082106968385288e-09, "epoch": 2.982142857142857, "percentage": 99.4, "elapsed_time": "13:15:46", "remaining_time": "0:04:45"}
335
+ {"current_steps": 335, "total_steps": 336, "loss": 0.3447, "lr": 2.7053406096433366e-10, "epoch": 2.991071428571429, "percentage": 99.7, "elapsed_time": "13:18:10", "remaining_time": "0:02:22"}
336
+ {"current_steps": 336, "total_steps": 336, "loss": 0.3407, "lr": 0.0, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "13:20:36", "remaining_time": "0:00:00"}
337
+ {"current_steps": 336, "total_steps": 336, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "13:24:55", "remaining_time": "0:00:00"}
trainer_state.json ADDED
@@ -0,0 +1,2394 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 336,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.008928571428571428,
13
+ "grad_norm": 6.749195575714111,
14
+ "learning_rate": 2.9411764705882356e-07,
15
+ "loss": 0.8158,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.017857142857142856,
20
+ "grad_norm": 6.9488749504089355,
21
+ "learning_rate": 5.882352941176471e-07,
22
+ "loss": 0.8361,
23
+ "step": 2
24
+ },
25
+ {
26
+ "epoch": 0.026785714285714284,
27
+ "grad_norm": 7.039111137390137,
28
+ "learning_rate": 8.823529411764707e-07,
29
+ "loss": 0.8422,
30
+ "step": 3
31
+ },
32
+ {
33
+ "epoch": 0.03571428571428571,
34
+ "grad_norm": 6.606884002685547,
35
+ "learning_rate": 1.1764705882352942e-06,
36
+ "loss": 0.8212,
37
+ "step": 4
38
+ },
39
+ {
40
+ "epoch": 0.044642857142857144,
41
+ "grad_norm": 5.828409194946289,
42
+ "learning_rate": 1.4705882352941177e-06,
43
+ "loss": 0.8039,
44
+ "step": 5
45
+ },
46
+ {
47
+ "epoch": 0.05357142857142857,
48
+ "grad_norm": 3.9500532150268555,
49
+ "learning_rate": 1.7647058823529414e-06,
50
+ "loss": 0.7598,
51
+ "step": 6
52
+ },
53
+ {
54
+ "epoch": 0.0625,
55
+ "grad_norm": 2.794076681137085,
56
+ "learning_rate": 2.058823529411765e-06,
57
+ "loss": 0.7132,
58
+ "step": 7
59
+ },
60
+ {
61
+ "epoch": 0.07142857142857142,
62
+ "grad_norm": 2.651289463043213,
63
+ "learning_rate": 2.3529411764705885e-06,
64
+ "loss": 0.7006,
65
+ "step": 8
66
+ },
67
+ {
68
+ "epoch": 0.08035714285714286,
69
+ "grad_norm": 1.9226497411727905,
70
+ "learning_rate": 2.647058823529412e-06,
71
+ "loss": 0.6781,
72
+ "step": 9
73
+ },
74
+ {
75
+ "epoch": 0.08928571428571429,
76
+ "grad_norm": 1.8058496713638306,
77
+ "learning_rate": 2.9411764705882355e-06,
78
+ "loss": 0.6716,
79
+ "step": 10
80
+ },
81
+ {
82
+ "epoch": 0.09821428571428571,
83
+ "grad_norm": 1.5553733110427856,
84
+ "learning_rate": 3.2352941176470594e-06,
85
+ "loss": 0.6524,
86
+ "step": 11
87
+ },
88
+ {
89
+ "epoch": 0.10714285714285714,
90
+ "grad_norm": 2.1922216415405273,
91
+ "learning_rate": 3.529411764705883e-06,
92
+ "loss": 0.6202,
93
+ "step": 12
94
+ },
95
+ {
96
+ "epoch": 0.11607142857142858,
97
+ "grad_norm": 2.5917115211486816,
98
+ "learning_rate": 3.8235294117647055e-06,
99
+ "loss": 0.6104,
100
+ "step": 13
101
+ },
102
+ {
103
+ "epoch": 0.125,
104
+ "grad_norm": 2.4120588302612305,
105
+ "learning_rate": 4.11764705882353e-06,
106
+ "loss": 0.6036,
107
+ "step": 14
108
+ },
109
+ {
110
+ "epoch": 0.13392857142857142,
111
+ "grad_norm": 2.0048017501831055,
112
+ "learning_rate": 4.411764705882353e-06,
113
+ "loss": 0.5893,
114
+ "step": 15
115
+ },
116
+ {
117
+ "epoch": 0.14285714285714285,
118
+ "grad_norm": 1.3664730787277222,
119
+ "learning_rate": 4.705882352941177e-06,
120
+ "loss": 0.5752,
121
+ "step": 16
122
+ },
123
+ {
124
+ "epoch": 0.15178571428571427,
125
+ "grad_norm": 1.232675552368164,
126
+ "learning_rate": 5e-06,
127
+ "loss": 0.5527,
128
+ "step": 17
129
+ },
130
+ {
131
+ "epoch": 0.16071428571428573,
132
+ "grad_norm": 1.2626478672027588,
133
+ "learning_rate": 5.294117647058824e-06,
134
+ "loss": 0.5438,
135
+ "step": 18
136
+ },
137
+ {
138
+ "epoch": 0.16964285714285715,
139
+ "grad_norm": 1.1723309755325317,
140
+ "learning_rate": 5.588235294117647e-06,
141
+ "loss": 0.5295,
142
+ "step": 19
143
+ },
144
+ {
145
+ "epoch": 0.17857142857142858,
146
+ "grad_norm": 0.9530285000801086,
147
+ "learning_rate": 5.882352941176471e-06,
148
+ "loss": 0.5227,
149
+ "step": 20
150
+ },
151
+ {
152
+ "epoch": 0.1875,
153
+ "grad_norm": 0.8222931027412415,
154
+ "learning_rate": 6.176470588235295e-06,
155
+ "loss": 0.5146,
156
+ "step": 21
157
+ },
158
+ {
159
+ "epoch": 0.19642857142857142,
160
+ "grad_norm": 0.8412429094314575,
161
+ "learning_rate": 6.470588235294119e-06,
162
+ "loss": 0.503,
163
+ "step": 22
164
+ },
165
+ {
166
+ "epoch": 0.20535714285714285,
167
+ "grad_norm": 0.8323157429695129,
168
+ "learning_rate": 6.764705882352942e-06,
169
+ "loss": 0.499,
170
+ "step": 23
171
+ },
172
+ {
173
+ "epoch": 0.21428571428571427,
174
+ "grad_norm": 0.6052507758140564,
175
+ "learning_rate": 7.058823529411766e-06,
176
+ "loss": 0.4901,
177
+ "step": 24
178
+ },
179
+ {
180
+ "epoch": 0.22321428571428573,
181
+ "grad_norm": 0.5770189762115479,
182
+ "learning_rate": 7.352941176470589e-06,
183
+ "loss": 0.4882,
184
+ "step": 25
185
+ },
186
+ {
187
+ "epoch": 0.23214285714285715,
188
+ "grad_norm": 0.6112756729125977,
189
+ "learning_rate": 7.647058823529411e-06,
190
+ "loss": 0.4789,
191
+ "step": 26
192
+ },
193
+ {
194
+ "epoch": 0.24107142857142858,
195
+ "grad_norm": 0.6391553282737732,
196
+ "learning_rate": 7.941176470588236e-06,
197
+ "loss": 0.4792,
198
+ "step": 27
199
+ },
200
+ {
201
+ "epoch": 0.25,
202
+ "grad_norm": 0.48618483543395996,
203
+ "learning_rate": 8.23529411764706e-06,
204
+ "loss": 0.4665,
205
+ "step": 28
206
+ },
207
+ {
208
+ "epoch": 0.25892857142857145,
209
+ "grad_norm": 0.4175974726676941,
210
+ "learning_rate": 8.529411764705883e-06,
211
+ "loss": 0.4666,
212
+ "step": 29
213
+ },
214
+ {
215
+ "epoch": 0.26785714285714285,
216
+ "grad_norm": 0.5063344836235046,
217
+ "learning_rate": 8.823529411764707e-06,
218
+ "loss": 0.4644,
219
+ "step": 30
220
+ },
221
+ {
222
+ "epoch": 0.2767857142857143,
223
+ "grad_norm": 0.4095499813556671,
224
+ "learning_rate": 9.11764705882353e-06,
225
+ "loss": 0.4552,
226
+ "step": 31
227
+ },
228
+ {
229
+ "epoch": 0.2857142857142857,
230
+ "grad_norm": 0.44540026783943176,
231
+ "learning_rate": 9.411764705882354e-06,
232
+ "loss": 0.4611,
233
+ "step": 32
234
+ },
235
+ {
236
+ "epoch": 0.29464285714285715,
237
+ "grad_norm": 0.40555301308631897,
238
+ "learning_rate": 9.705882352941177e-06,
239
+ "loss": 0.449,
240
+ "step": 33
241
+ },
242
+ {
243
+ "epoch": 0.30357142857142855,
244
+ "grad_norm": 0.3954616189002991,
245
+ "learning_rate": 1e-05,
246
+ "loss": 0.4543,
247
+ "step": 34
248
+ },
249
+ {
250
+ "epoch": 0.3125,
251
+ "grad_norm": 0.4039325416088104,
252
+ "learning_rate": 9.999729465939036e-06,
253
+ "loss": 0.4449,
254
+ "step": 35
255
+ },
256
+ {
257
+ "epoch": 0.32142857142857145,
258
+ "grad_norm": 0.4171324372291565,
259
+ "learning_rate": 9.998917893031615e-06,
260
+ "loss": 0.4503,
261
+ "step": 36
262
+ },
263
+ {
264
+ "epoch": 0.33035714285714285,
265
+ "grad_norm": 0.323896586894989,
266
+ "learning_rate": 9.997565369100983e-06,
267
+ "loss": 0.4421,
268
+ "step": 37
269
+ },
270
+ {
271
+ "epoch": 0.3392857142857143,
272
+ "grad_norm": 0.3594343364238739,
273
+ "learning_rate": 9.995672040508656e-06,
274
+ "loss": 0.4453,
275
+ "step": 38
276
+ },
277
+ {
278
+ "epoch": 0.3482142857142857,
279
+ "grad_norm": 0.33965522050857544,
280
+ "learning_rate": 9.993238112138584e-06,
281
+ "loss": 0.4388,
282
+ "step": 39
283
+ },
284
+ {
285
+ "epoch": 0.35714285714285715,
286
+ "grad_norm": 0.33565545082092285,
287
+ "learning_rate": 9.990263847374976e-06,
288
+ "loss": 0.4407,
289
+ "step": 40
290
+ },
291
+ {
292
+ "epoch": 0.36607142857142855,
293
+ "grad_norm": 0.312674343585968,
294
+ "learning_rate": 9.986749568073804e-06,
295
+ "loss": 0.4419,
296
+ "step": 41
297
+ },
298
+ {
299
+ "epoch": 0.375,
300
+ "grad_norm": 0.33573076128959656,
301
+ "learning_rate": 9.982695654527966e-06,
302
+ "loss": 0.438,
303
+ "step": 42
304
+ },
305
+ {
306
+ "epoch": 0.38392857142857145,
307
+ "grad_norm": 0.3331693708896637,
308
+ "learning_rate": 9.97810254542614e-06,
309
+ "loss": 0.4299,
310
+ "step": 43
311
+ },
312
+ {
313
+ "epoch": 0.39285714285714285,
314
+ "grad_norm": 0.3128435015678406,
315
+ "learning_rate": 9.972970737805312e-06,
316
+ "loss": 0.4248,
317
+ "step": 44
318
+ },
319
+ {
320
+ "epoch": 0.4017857142857143,
321
+ "grad_norm": 0.2929725646972656,
322
+ "learning_rate": 9.96730078699698e-06,
323
+ "loss": 0.4333,
324
+ "step": 45
325
+ },
326
+ {
327
+ "epoch": 0.4107142857142857,
328
+ "grad_norm": 0.27605733275413513,
329
+ "learning_rate": 9.961093306567076e-06,
330
+ "loss": 0.4266,
331
+ "step": 46
332
+ },
333
+ {
334
+ "epoch": 0.41964285714285715,
335
+ "grad_norm": 0.3021415174007416,
336
+ "learning_rate": 9.954348968249552e-06,
337
+ "loss": 0.4253,
338
+ "step": 47
339
+ },
340
+ {
341
+ "epoch": 0.42857142857142855,
342
+ "grad_norm": 0.26961633563041687,
343
+ "learning_rate": 9.947068501873702e-06,
344
+ "loss": 0.4229,
345
+ "step": 48
346
+ },
347
+ {
348
+ "epoch": 0.4375,
349
+ "grad_norm": 0.27512669563293457,
350
+ "learning_rate": 9.93925269528518e-06,
351
+ "loss": 0.423,
352
+ "step": 49
353
+ },
354
+ {
355
+ "epoch": 0.44642857142857145,
356
+ "grad_norm": 0.25938841700553894,
357
+ "learning_rate": 9.930902394260746e-06,
358
+ "loss": 0.4136,
359
+ "step": 50
360
+ },
361
+ {
362
+ "epoch": 0.45535714285714285,
363
+ "grad_norm": 0.24410875141620636,
364
+ "learning_rate": 9.922018502416736e-06,
365
+ "loss": 0.4238,
366
+ "step": 51
367
+ },
368
+ {
369
+ "epoch": 0.4642857142857143,
370
+ "grad_norm": 0.24541054666042328,
371
+ "learning_rate": 9.912601981111287e-06,
372
+ "loss": 0.4162,
373
+ "step": 52
374
+ },
375
+ {
376
+ "epoch": 0.4732142857142857,
377
+ "grad_norm": 0.23068013787269592,
378
+ "learning_rate": 9.902653849340296e-06,
379
+ "loss": 0.4115,
380
+ "step": 53
381
+ },
382
+ {
383
+ "epoch": 0.48214285714285715,
384
+ "grad_norm": 0.2482675462961197,
385
+ "learning_rate": 9.892175183627161e-06,
386
+ "loss": 0.4144,
387
+ "step": 54
388
+ },
389
+ {
390
+ "epoch": 0.49107142857142855,
391
+ "grad_norm": 0.23585791885852814,
392
+ "learning_rate": 9.881167117906276e-06,
393
+ "loss": 0.4163,
394
+ "step": 55
395
+ },
396
+ {
397
+ "epoch": 0.5,
398
+ "grad_norm": 0.2257727086544037,
399
+ "learning_rate": 9.869630843400331e-06,
400
+ "loss": 0.412,
401
+ "step": 56
402
+ },
403
+ {
404
+ "epoch": 0.5089285714285714,
405
+ "grad_norm": 0.2443915456533432,
406
+ "learning_rate": 9.8575676084914e-06,
407
+ "loss": 0.4196,
408
+ "step": 57
409
+ },
410
+ {
411
+ "epoch": 0.5178571428571429,
412
+ "grad_norm": 0.2699064314365387,
413
+ "learning_rate": 9.844978718585855e-06,
414
+ "loss": 0.4135,
415
+ "step": 58
416
+ },
417
+ {
418
+ "epoch": 0.5267857142857143,
419
+ "grad_norm": 0.27562785148620605,
420
+ "learning_rate": 9.831865535973103e-06,
421
+ "loss": 0.4135,
422
+ "step": 59
423
+ },
424
+ {
425
+ "epoch": 0.5357142857142857,
426
+ "grad_norm": 0.28133267164230347,
427
+ "learning_rate": 9.81822947967816e-06,
428
+ "loss": 0.4085,
429
+ "step": 60
430
+ },
431
+ {
432
+ "epoch": 0.5446428571428571,
433
+ "grad_norm": 0.2471129447221756,
434
+ "learning_rate": 9.804072025308096e-06,
435
+ "loss": 0.4112,
436
+ "step": 61
437
+ },
438
+ {
439
+ "epoch": 0.5535714285714286,
440
+ "grad_norm": 0.3286897540092468,
441
+ "learning_rate": 9.789394704892364e-06,
442
+ "loss": 0.4086,
443
+ "step": 62
444
+ },
445
+ {
446
+ "epoch": 0.5625,
447
+ "grad_norm": 0.27339598536491394,
448
+ "learning_rate": 9.774199106717004e-06,
449
+ "loss": 0.4077,
450
+ "step": 63
451
+ },
452
+ {
453
+ "epoch": 0.5714285714285714,
454
+ "grad_norm": 0.24371126294136047,
455
+ "learning_rate": 9.758486875152766e-06,
456
+ "loss": 0.4072,
457
+ "step": 64
458
+ },
459
+ {
460
+ "epoch": 0.5803571428571429,
461
+ "grad_norm": 0.26882895827293396,
462
+ "learning_rate": 9.742259710477178e-06,
463
+ "loss": 0.4015,
464
+ "step": 65
465
+ },
466
+ {
467
+ "epoch": 0.5892857142857143,
468
+ "grad_norm": 0.3289386034011841,
469
+ "learning_rate": 9.725519368690539e-06,
470
+ "loss": 0.4035,
471
+ "step": 66
472
+ },
473
+ {
474
+ "epoch": 0.5982142857142857,
475
+ "grad_norm": 0.33041179180145264,
476
+ "learning_rate": 9.70826766132591e-06,
477
+ "loss": 0.4061,
478
+ "step": 67
479
+ },
480
+ {
481
+ "epoch": 0.6071428571428571,
482
+ "grad_norm": 0.2799140512943268,
483
+ "learning_rate": 9.690506455253073e-06,
484
+ "loss": 0.4086,
485
+ "step": 68
486
+ },
487
+ {
488
+ "epoch": 0.6160714285714286,
489
+ "grad_norm": 0.33769428730010986,
490
+ "learning_rate": 9.672237672476506e-06,
491
+ "loss": 0.4017,
492
+ "step": 69
493
+ },
494
+ {
495
+ "epoch": 0.625,
496
+ "grad_norm": 0.3362935483455658,
497
+ "learning_rate": 9.65346328992741e-06,
498
+ "loss": 0.4017,
499
+ "step": 70
500
+ },
501
+ {
502
+ "epoch": 0.6339285714285714,
503
+ "grad_norm": 0.3309652507305145,
504
+ "learning_rate": 9.634185339249766e-06,
505
+ "loss": 0.4043,
506
+ "step": 71
507
+ },
508
+ {
509
+ "epoch": 0.6428571428571429,
510
+ "grad_norm": 0.3004245162010193,
511
+ "learning_rate": 9.614405906580486e-06,
512
+ "loss": 0.396,
513
+ "step": 72
514
+ },
515
+ {
516
+ "epoch": 0.6517857142857143,
517
+ "grad_norm": 0.31212133169174194,
518
+ "learning_rate": 9.594127132323669e-06,
519
+ "loss": 0.4062,
520
+ "step": 73
521
+ },
522
+ {
523
+ "epoch": 0.6607142857142857,
524
+ "grad_norm": 0.2722640335559845,
525
+ "learning_rate": 9.573351210918976e-06,
526
+ "loss": 0.3977,
527
+ "step": 74
528
+ },
529
+ {
530
+ "epoch": 0.6696428571428571,
531
+ "grad_norm": 0.3749098777770996,
532
+ "learning_rate": 9.55208039060416e-06,
533
+ "loss": 0.3993,
534
+ "step": 75
535
+ },
536
+ {
537
+ "epoch": 0.6785714285714286,
538
+ "grad_norm": 0.3566979467868805,
539
+ "learning_rate": 9.53031697317178e-06,
540
+ "loss": 0.4126,
541
+ "step": 76
542
+ },
543
+ {
544
+ "epoch": 0.6875,
545
+ "grad_norm": 0.32874590158462524,
546
+ "learning_rate": 9.50806331372012e-06,
547
+ "loss": 0.4068,
548
+ "step": 77
549
+ },
550
+ {
551
+ "epoch": 0.6964285714285714,
552
+ "grad_norm": 0.29310765862464905,
553
+ "learning_rate": 9.485321820398321e-06,
554
+ "loss": 0.4014,
555
+ "step": 78
556
+ },
557
+ {
558
+ "epoch": 0.7053571428571429,
559
+ "grad_norm": 0.31753551959991455,
560
+ "learning_rate": 9.4620949541458e-06,
561
+ "loss": 0.3995,
562
+ "step": 79
563
+ },
564
+ {
565
+ "epoch": 0.7142857142857143,
566
+ "grad_norm": 0.30886319279670715,
567
+ "learning_rate": 9.43838522842594e-06,
568
+ "loss": 0.4055,
569
+ "step": 80
570
+ },
571
+ {
572
+ "epoch": 0.7232142857142857,
573
+ "grad_norm": 0.3552889823913574,
574
+ "learning_rate": 9.41419520895409e-06,
575
+ "loss": 0.3976,
576
+ "step": 81
577
+ },
578
+ {
579
+ "epoch": 0.7321428571428571,
580
+ "grad_norm": 0.3814694285392761,
581
+ "learning_rate": 9.389527513419935e-06,
582
+ "loss": 0.3922,
583
+ "step": 82
584
+ },
585
+ {
586
+ "epoch": 0.7410714285714286,
587
+ "grad_norm": 0.4074476361274719,
588
+ "learning_rate": 9.364384811204212e-06,
589
+ "loss": 0.4018,
590
+ "step": 83
591
+ },
592
+ {
593
+ "epoch": 0.75,
594
+ "grad_norm": 0.38614606857299805,
595
+ "learning_rate": 9.338769823089853e-06,
596
+ "loss": 0.399,
597
+ "step": 84
598
+ },
599
+ {
600
+ "epoch": 0.7589285714285714,
601
+ "grad_norm": 0.3035358488559723,
602
+ "learning_rate": 9.312685320967566e-06,
603
+ "loss": 0.3948,
604
+ "step": 85
605
+ },
606
+ {
607
+ "epoch": 0.7678571428571429,
608
+ "grad_norm": 0.3059258759021759,
609
+ "learning_rate": 9.286134127535859e-06,
610
+ "loss": 0.4007,
611
+ "step": 86
612
+ },
613
+ {
614
+ "epoch": 0.7767857142857143,
615
+ "grad_norm": 0.39105063676834106,
616
+ "learning_rate": 9.25911911599561e-06,
617
+ "loss": 0.3944,
618
+ "step": 87
619
+ },
620
+ {
621
+ "epoch": 0.7857142857142857,
622
+ "grad_norm": 0.40922218561172485,
623
+ "learning_rate": 9.231643209739128e-06,
624
+ "loss": 0.4006,
625
+ "step": 88
626
+ },
627
+ {
628
+ "epoch": 0.7946428571428571,
629
+ "grad_norm": 0.35727784037590027,
630
+ "learning_rate": 9.203709382033814e-06,
631
+ "loss": 0.3894,
632
+ "step": 89
633
+ },
634
+ {
635
+ "epoch": 0.8035714285714286,
636
+ "grad_norm": 0.26759040355682373,
637
+ "learning_rate": 9.175320655700407e-06,
638
+ "loss": 0.3965,
639
+ "step": 90
640
+ },
641
+ {
642
+ "epoch": 0.8125,
643
+ "grad_norm": 0.2492348849773407,
644
+ "learning_rate": 9.146480102785871e-06,
645
+ "loss": 0.4021,
646
+ "step": 91
647
+ },
648
+ {
649
+ "epoch": 0.8214285714285714,
650
+ "grad_norm": 0.29503515362739563,
651
+ "learning_rate": 9.117190844230971e-06,
652
+ "loss": 0.3979,
653
+ "step": 92
654
+ },
655
+ {
656
+ "epoch": 0.8303571428571429,
657
+ "grad_norm": 0.28773820400238037,
658
+ "learning_rate": 9.08745604953253e-06,
659
+ "loss": 0.3961,
660
+ "step": 93
661
+ },
662
+ {
663
+ "epoch": 0.8392857142857143,
664
+ "grad_norm": 0.2832280397415161,
665
+ "learning_rate": 9.057278936400453e-06,
666
+ "loss": 0.3947,
667
+ "step": 94
668
+ },
669
+ {
670
+ "epoch": 0.8482142857142857,
671
+ "grad_norm": 0.26580309867858887,
672
+ "learning_rate": 9.026662770409524e-06,
673
+ "loss": 0.3993,
674
+ "step": 95
675
+ },
676
+ {
677
+ "epoch": 0.8571428571428571,
678
+ "grad_norm": 0.2617496848106384,
679
+ "learning_rate": 8.99561086464603e-06,
680
+ "loss": 0.3997,
681
+ "step": 96
682
+ },
683
+ {
684
+ "epoch": 0.8660714285714286,
685
+ "grad_norm": 0.3017551302909851,
686
+ "learning_rate": 8.964126579349237e-06,
687
+ "loss": 0.3985,
688
+ "step": 97
689
+ },
690
+ {
691
+ "epoch": 0.875,
692
+ "grad_norm": 0.2986689507961273,
693
+ "learning_rate": 8.932213321547769e-06,
694
+ "loss": 0.3857,
695
+ "step": 98
696
+ },
697
+ {
698
+ "epoch": 0.8839285714285714,
699
+ "grad_norm": 0.3144254982471466,
700
+ "learning_rate": 8.899874544690921e-06,
701
+ "loss": 0.3985,
702
+ "step": 99
703
+ },
704
+ {
705
+ "epoch": 0.8928571428571429,
706
+ "grad_norm": 0.25709250569343567,
707
+ "learning_rate": 8.86711374827494e-06,
708
+ "loss": 0.3912,
709
+ "step": 100
710
+ },
711
+ {
712
+ "epoch": 0.9017857142857143,
713
+ "grad_norm": 0.35801225900650024,
714
+ "learning_rate": 8.833934477464348e-06,
715
+ "loss": 0.3972,
716
+ "step": 101
717
+ },
718
+ {
719
+ "epoch": 0.9107142857142857,
720
+ "grad_norm": 0.3704852759838104,
721
+ "learning_rate": 8.800340322708291e-06,
722
+ "loss": 0.3889,
723
+ "step": 102
724
+ },
725
+ {
726
+ "epoch": 0.9196428571428571,
727
+ "grad_norm": 0.3720284700393677,
728
+ "learning_rate": 8.766334919352018e-06,
729
+ "loss": 0.3991,
730
+ "step": 103
731
+ },
732
+ {
733
+ "epoch": 0.9285714285714286,
734
+ "grad_norm": 0.3510439991950989,
735
+ "learning_rate": 8.73192194724347e-06,
736
+ "loss": 0.3879,
737
+ "step": 104
738
+ },
739
+ {
740
+ "epoch": 0.9375,
741
+ "grad_norm": 0.29455068707466125,
742
+ "learning_rate": 8.697105130335084e-06,
743
+ "loss": 0.3962,
744
+ "step": 105
745
+ },
746
+ {
747
+ "epoch": 0.9464285714285714,
748
+ "grad_norm": 0.25835826992988586,
749
+ "learning_rate": 8.661888236280813e-06,
750
+ "loss": 0.3909,
751
+ "step": 106
752
+ },
753
+ {
754
+ "epoch": 0.9553571428571429,
755
+ "grad_norm": 0.29132160544395447,
756
+ "learning_rate": 8.626275076028397e-06,
757
+ "loss": 0.3899,
758
+ "step": 107
759
+ },
760
+ {
761
+ "epoch": 0.9642857142857143,
762
+ "grad_norm": 0.3577162027359009,
763
+ "learning_rate": 8.590269503406986e-06,
764
+ "loss": 0.3846,
765
+ "step": 108
766
+ },
767
+ {
768
+ "epoch": 0.9732142857142857,
769
+ "grad_norm": 0.28617140650749207,
770
+ "learning_rate": 8.553875414710088e-06,
771
+ "loss": 0.3922,
772
+ "step": 109
773
+ },
774
+ {
775
+ "epoch": 0.9821428571428571,
776
+ "grad_norm": 0.34976303577423096,
777
+ "learning_rate": 8.517096748273951e-06,
778
+ "loss": 0.3909,
779
+ "step": 110
780
+ },
781
+ {
782
+ "epoch": 0.9910714285714286,
783
+ "grad_norm": 0.307277113199234,
784
+ "learning_rate": 8.479937484051368e-06,
785
+ "loss": 0.3893,
786
+ "step": 111
787
+ },
788
+ {
789
+ "epoch": 1.0,
790
+ "grad_norm": 0.33036214113235474,
791
+ "learning_rate": 8.442401643181e-06,
792
+ "loss": 0.3836,
793
+ "step": 112
794
+ },
795
+ {
796
+ "epoch": 1.0089285714285714,
797
+ "grad_norm": 0.3078720271587372,
798
+ "learning_rate": 8.404493287552232e-06,
799
+ "loss": 0.3733,
800
+ "step": 113
801
+ },
802
+ {
803
+ "epoch": 1.0178571428571428,
804
+ "grad_norm": 0.297649085521698,
805
+ "learning_rate": 8.366216519365623e-06,
806
+ "loss": 0.3687,
807
+ "step": 114
808
+ },
809
+ {
810
+ "epoch": 1.0267857142857142,
811
+ "grad_norm": 0.2785271406173706,
812
+ "learning_rate": 8.327575480688985e-06,
813
+ "loss": 0.3734,
814
+ "step": 115
815
+ },
816
+ {
817
+ "epoch": 1.0357142857142858,
818
+ "grad_norm": 0.35386645793914795,
819
+ "learning_rate": 8.288574353009164e-06,
820
+ "loss": 0.3782,
821
+ "step": 116
822
+ },
823
+ {
824
+ "epoch": 1.0446428571428572,
825
+ "grad_norm": 0.33655157685279846,
826
+ "learning_rate": 8.249217356779544e-06,
827
+ "loss": 0.3745,
828
+ "step": 117
829
+ },
830
+ {
831
+ "epoch": 1.0535714285714286,
832
+ "grad_norm": 0.24362798035144806,
833
+ "learning_rate": 8.20950875096333e-06,
834
+ "loss": 0.3731,
835
+ "step": 118
836
+ },
837
+ {
838
+ "epoch": 1.0625,
839
+ "grad_norm": 0.2674640119075775,
840
+ "learning_rate": 8.169452832572676e-06,
841
+ "loss": 0.3711,
842
+ "step": 119
843
+ },
844
+ {
845
+ "epoch": 1.0714285714285714,
846
+ "grad_norm": 0.26036393642425537,
847
+ "learning_rate": 8.129053936203688e-06,
848
+ "loss": 0.3701,
849
+ "step": 120
850
+ },
851
+ {
852
+ "epoch": 1.0803571428571428,
853
+ "grad_norm": 0.24254180490970612,
854
+ "learning_rate": 8.088316433567369e-06,
855
+ "loss": 0.3729,
856
+ "step": 121
857
+ },
858
+ {
859
+ "epoch": 1.0892857142857142,
860
+ "grad_norm": 0.24308055639266968,
861
+ "learning_rate": 8.04724473301652e-06,
862
+ "loss": 0.3725,
863
+ "step": 122
864
+ },
865
+ {
866
+ "epoch": 1.0982142857142858,
867
+ "grad_norm": 0.2283226102590561,
868
+ "learning_rate": 8.005843279068724e-06,
869
+ "loss": 0.3689,
870
+ "step": 123
871
+ },
872
+ {
873
+ "epoch": 1.1071428571428572,
874
+ "grad_norm": 0.2410878688097,
875
+ "learning_rate": 7.964116551925365e-06,
876
+ "loss": 0.377,
877
+ "step": 124
878
+ },
879
+ {
880
+ "epoch": 1.1160714285714286,
881
+ "grad_norm": 0.25953209400177,
882
+ "learning_rate": 7.92206906698682e-06,
883
+ "loss": 0.3684,
884
+ "step": 125
885
+ },
886
+ {
887
+ "epoch": 1.125,
888
+ "grad_norm": 0.2646959125995636,
889
+ "learning_rate": 7.879705374363831e-06,
890
+ "loss": 0.3744,
891
+ "step": 126
892
+ },
893
+ {
894
+ "epoch": 1.1339285714285714,
895
+ "grad_norm": 0.23834243416786194,
896
+ "learning_rate": 7.837030058385117e-06,
897
+ "loss": 0.3754,
898
+ "step": 127
899
+ },
900
+ {
901
+ "epoch": 1.1428571428571428,
902
+ "grad_norm": 0.25867897272109985,
903
+ "learning_rate": 7.794047737101298e-06,
904
+ "loss": 0.3712,
905
+ "step": 128
906
+ },
907
+ {
908
+ "epoch": 1.1517857142857142,
909
+ "grad_norm": 0.26810479164123535,
910
+ "learning_rate": 7.750763061785139e-06,
911
+ "loss": 0.3721,
912
+ "step": 129
913
+ },
914
+ {
915
+ "epoch": 1.1607142857142858,
916
+ "grad_norm": 0.23617470264434814,
917
+ "learning_rate": 7.707180716428237e-06,
918
+ "loss": 0.3754,
919
+ "step": 130
920
+ },
921
+ {
922
+ "epoch": 1.1696428571428572,
923
+ "grad_norm": 0.24265629053115845,
924
+ "learning_rate": 7.663305417234146e-06,
925
+ "loss": 0.3675,
926
+ "step": 131
927
+ },
928
+ {
929
+ "epoch": 1.1785714285714286,
930
+ "grad_norm": 0.2951458692550659,
931
+ "learning_rate": 7.619141912108008e-06,
932
+ "loss": 0.3786,
933
+ "step": 132
934
+ },
935
+ {
936
+ "epoch": 1.1875,
937
+ "grad_norm": 0.24244728684425354,
938
+ "learning_rate": 7.57469498014278e-06,
939
+ "loss": 0.3731,
940
+ "step": 133
941
+ },
942
+ {
943
+ "epoch": 1.1964285714285714,
944
+ "grad_norm": 0.21746976673603058,
945
+ "learning_rate": 7.529969431102063e-06,
946
+ "loss": 0.3735,
947
+ "step": 134
948
+ },
949
+ {
950
+ "epoch": 1.2053571428571428,
951
+ "grad_norm": 0.25599175691604614,
952
+ "learning_rate": 7.484970104899624e-06,
953
+ "loss": 0.3689,
954
+ "step": 135
955
+ },
956
+ {
957
+ "epoch": 1.2142857142857142,
958
+ "grad_norm": 0.2244032621383667,
959
+ "learning_rate": 7.4397018710756415e-06,
960
+ "loss": 0.3704,
961
+ "step": 136
962
+ },
963
+ {
964
+ "epoch": 1.2232142857142858,
965
+ "grad_norm": 0.25709277391433716,
966
+ "learning_rate": 7.394169628269771e-06,
967
+ "loss": 0.3702,
968
+ "step": 137
969
+ },
970
+ {
971
+ "epoch": 1.2321428571428572,
972
+ "grad_norm": 0.2774679958820343,
973
+ "learning_rate": 7.34837830369103e-06,
974
+ "loss": 0.3688,
975
+ "step": 138
976
+ },
977
+ {
978
+ "epoch": 1.2410714285714286,
979
+ "grad_norm": 0.28378257155418396,
980
+ "learning_rate": 7.302332852584619e-06,
981
+ "loss": 0.3731,
982
+ "step": 139
983
+ },
984
+ {
985
+ "epoch": 1.25,
986
+ "grad_norm": 0.24090154469013214,
987
+ "learning_rate": 7.2560382576956875e-06,
988
+ "loss": 0.3691,
989
+ "step": 140
990
+ },
991
+ {
992
+ "epoch": 1.2589285714285714,
993
+ "grad_norm": 0.2390367090702057,
994
+ "learning_rate": 7.209499528730138e-06,
995
+ "loss": 0.3664,
996
+ "step": 141
997
+ },
998
+ {
999
+ "epoch": 1.2678571428571428,
1000
+ "grad_norm": 0.2720012664794922,
1001
+ "learning_rate": 7.162721701812506e-06,
1002
+ "loss": 0.3702,
1003
+ "step": 142
1004
+ },
1005
+ {
1006
+ "epoch": 1.2767857142857144,
1007
+ "grad_norm": 0.2866012156009674,
1008
+ "learning_rate": 7.115709838940983e-06,
1009
+ "loss": 0.3722,
1010
+ "step": 143
1011
+ },
1012
+ {
1013
+ "epoch": 1.2857142857142856,
1014
+ "grad_norm": 0.23692139983177185,
1015
+ "learning_rate": 7.068469027439642e-06,
1016
+ "loss": 0.3741,
1017
+ "step": 144
1018
+ },
1019
+ {
1020
+ "epoch": 1.2946428571428572,
1021
+ "grad_norm": 0.21032477915287018,
1022
+ "learning_rate": 7.02100437940791e-06,
1023
+ "loss": 0.3675,
1024
+ "step": 145
1025
+ },
1026
+ {
1027
+ "epoch": 1.3035714285714286,
1028
+ "grad_norm": 0.2565901577472687,
1029
+ "learning_rate": 6.9733210311673826e-06,
1030
+ "loss": 0.3695,
1031
+ "step": 146
1032
+ },
1033
+ {
1034
+ "epoch": 1.3125,
1035
+ "grad_norm": 0.3054284453392029,
1036
+ "learning_rate": 6.925424142705997e-06,
1037
+ "loss": 0.3665,
1038
+ "step": 147
1039
+ },
1040
+ {
1041
+ "epoch": 1.3214285714285714,
1042
+ "grad_norm": 0.2568247318267822,
1043
+ "learning_rate": 6.8773188971196515e-06,
1044
+ "loss": 0.3702,
1045
+ "step": 148
1046
+ },
1047
+ {
1048
+ "epoch": 1.3303571428571428,
1049
+ "grad_norm": 0.2898830473423004,
1050
+ "learning_rate": 6.829010500051319e-06,
1051
+ "loss": 0.3739,
1052
+ "step": 149
1053
+ },
1054
+ {
1055
+ "epoch": 1.3392857142857144,
1056
+ "grad_norm": 0.22603343427181244,
1057
+ "learning_rate": 6.780504179127735e-06,
1058
+ "loss": 0.3689,
1059
+ "step": 150
1060
+ },
1061
+ {
1062
+ "epoch": 1.3482142857142856,
1063
+ "grad_norm": 0.26400893926620483,
1064
+ "learning_rate": 6.731805183393696e-06,
1065
+ "loss": 0.3686,
1066
+ "step": 151
1067
+ },
1068
+ {
1069
+ "epoch": 1.3571428571428572,
1070
+ "grad_norm": 0.21483619511127472,
1071
+ "learning_rate": 6.682918782744033e-06,
1072
+ "loss": 0.3685,
1073
+ "step": 152
1074
+ },
1075
+ {
1076
+ "epoch": 1.3660714285714286,
1077
+ "grad_norm": 0.28114837408065796,
1078
+ "learning_rate": 6.63385026735334e-06,
1079
+ "loss": 0.3626,
1080
+ "step": 153
1081
+ },
1082
+ {
1083
+ "epoch": 1.375,
1084
+ "grad_norm": 0.22158163785934448,
1085
+ "learning_rate": 6.584604947103515e-06,
1086
+ "loss": 0.3635,
1087
+ "step": 154
1088
+ },
1089
+ {
1090
+ "epoch": 1.3839285714285714,
1091
+ "grad_norm": 0.2537171542644501,
1092
+ "learning_rate": 6.535188151009143e-06,
1093
+ "loss": 0.3699,
1094
+ "step": 155
1095
+ },
1096
+ {
1097
+ "epoch": 1.3928571428571428,
1098
+ "grad_norm": 0.24269886314868927,
1099
+ "learning_rate": 6.4856052266408375e-06,
1100
+ "loss": 0.362,
1101
+ "step": 156
1102
+ },
1103
+ {
1104
+ "epoch": 1.4017857142857144,
1105
+ "grad_norm": 0.22118201851844788,
1106
+ "learning_rate": 6.4358615395465506e-06,
1107
+ "loss": 0.3652,
1108
+ "step": 157
1109
+ },
1110
+ {
1111
+ "epoch": 1.4107142857142856,
1112
+ "grad_norm": 0.2311353087425232,
1113
+ "learning_rate": 6.385962472670953e-06,
1114
+ "loss": 0.3703,
1115
+ "step": 158
1116
+ },
1117
+ {
1118
+ "epoch": 1.4196428571428572,
1119
+ "grad_norm": 0.21577298641204834,
1120
+ "learning_rate": 6.335913425772926e-06,
1121
+ "loss": 0.3624,
1122
+ "step": 159
1123
+ },
1124
+ {
1125
+ "epoch": 1.4285714285714286,
1126
+ "grad_norm": 0.23977944254875183,
1127
+ "learning_rate": 6.28571981484123e-06,
1128
+ "loss": 0.3663,
1129
+ "step": 160
1130
+ },
1131
+ {
1132
+ "epoch": 1.4375,
1133
+ "grad_norm": 0.2253430187702179,
1134
+ "learning_rate": 6.235387071508427e-06,
1135
+ "loss": 0.3708,
1136
+ "step": 161
1137
+ },
1138
+ {
1139
+ "epoch": 1.4464285714285714,
1140
+ "grad_norm": 0.20987780392169952,
1141
+ "learning_rate": 6.184920642463095e-06,
1142
+ "loss": 0.3703,
1143
+ "step": 162
1144
+ },
1145
+ {
1146
+ "epoch": 1.4553571428571428,
1147
+ "grad_norm": 0.23987387120723724,
1148
+ "learning_rate": 6.1343259888604335e-06,
1149
+ "loss": 0.3731,
1150
+ "step": 163
1151
+ },
1152
+ {
1153
+ "epoch": 1.4642857142857144,
1154
+ "grad_norm": 0.22025255858898163,
1155
+ "learning_rate": 6.083608585731283e-06,
1156
+ "loss": 0.3677,
1157
+ "step": 164
1158
+ },
1159
+ {
1160
+ "epoch": 1.4732142857142856,
1161
+ "grad_norm": 0.19166462123394012,
1162
+ "learning_rate": 6.032773921389655e-06,
1163
+ "loss": 0.3708,
1164
+ "step": 165
1165
+ },
1166
+ {
1167
+ "epoch": 1.4821428571428572,
1168
+ "grad_norm": 0.2034110724925995,
1169
+ "learning_rate": 5.9818274968388225e-06,
1170
+ "loss": 0.3665,
1171
+ "step": 166
1172
+ },
1173
+ {
1174
+ "epoch": 1.4910714285714286,
1175
+ "grad_norm": 0.21674256026744843,
1176
+ "learning_rate": 5.930774825176034e-06,
1177
+ "loss": 0.3625,
1178
+ "step": 167
1179
+ },
1180
+ {
1181
+ "epoch": 1.5,
1182
+ "grad_norm": 0.22111833095550537,
1183
+ "learning_rate": 5.879621430995927e-06,
1184
+ "loss": 0.3699,
1185
+ "step": 168
1186
+ },
1187
+ {
1188
+ "epoch": 1.5089285714285714,
1189
+ "grad_norm": 0.20402273535728455,
1190
+ "learning_rate": 5.8283728497926865e-06,
1191
+ "loss": 0.3678,
1192
+ "step": 169
1193
+ },
1194
+ {
1195
+ "epoch": 1.5178571428571428,
1196
+ "grad_norm": 0.2157091498374939,
1197
+ "learning_rate": 5.777034627361025e-06,
1198
+ "loss": 0.3677,
1199
+ "step": 170
1200
+ },
1201
+ {
1202
+ "epoch": 1.5267857142857144,
1203
+ "grad_norm": 0.20562361180782318,
1204
+ "learning_rate": 5.725612319196065e-06,
1205
+ "loss": 0.3697,
1206
+ "step": 171
1207
+ },
1208
+ {
1209
+ "epoch": 1.5357142857142856,
1210
+ "grad_norm": 0.1970471292734146,
1211
+ "learning_rate": 5.674111489892144e-06,
1212
+ "loss": 0.3696,
1213
+ "step": 172
1214
+ },
1215
+ {
1216
+ "epoch": 1.5446428571428572,
1217
+ "grad_norm": 0.19568419456481934,
1218
+ "learning_rate": 5.622537712540664e-06,
1219
+ "loss": 0.3744,
1220
+ "step": 173
1221
+ },
1222
+ {
1223
+ "epoch": 1.5535714285714286,
1224
+ "grad_norm": 0.20620059967041016,
1225
+ "learning_rate": 5.570896568126994e-06,
1226
+ "loss": 0.3668,
1227
+ "step": 174
1228
+ },
1229
+ {
1230
+ "epoch": 1.5625,
1231
+ "grad_norm": 0.22930361330509186,
1232
+ "learning_rate": 5.5191936449265345e-06,
1233
+ "loss": 0.3649,
1234
+ "step": 175
1235
+ },
1236
+ {
1237
+ "epoch": 1.5714285714285714,
1238
+ "grad_norm": 0.19315338134765625,
1239
+ "learning_rate": 5.4674345379e-06,
1240
+ "loss": 0.3683,
1241
+ "step": 176
1242
+ },
1243
+ {
1244
+ "epoch": 1.5803571428571428,
1245
+ "grad_norm": 0.20728451013565063,
1246
+ "learning_rate": 5.415624848087959e-06,
1247
+ "loss": 0.3658,
1248
+ "step": 177
1249
+ },
1250
+ {
1251
+ "epoch": 1.5892857142857144,
1252
+ "grad_norm": 0.1867711842060089,
1253
+ "learning_rate": 5.36377018200472e-06,
1254
+ "loss": 0.3663,
1255
+ "step": 178
1256
+ },
1257
+ {
1258
+ "epoch": 1.5982142857142856,
1259
+ "grad_norm": 0.18480746448040009,
1260
+ "learning_rate": 5.311876151031642e-06,
1261
+ "loss": 0.3618,
1262
+ "step": 179
1263
+ },
1264
+ {
1265
+ "epoch": 1.6071428571428572,
1266
+ "grad_norm": 0.20057101547718048,
1267
+ "learning_rate": 5.259948370809902e-06,
1268
+ "loss": 0.3707,
1269
+ "step": 180
1270
+ },
1271
+ {
1272
+ "epoch": 1.6160714285714286,
1273
+ "grad_norm": 0.18401624262332916,
1274
+ "learning_rate": 5.207992460632805e-06,
1275
+ "loss": 0.3647,
1276
+ "step": 181
1277
+ },
1278
+ {
1279
+ "epoch": 1.625,
1280
+ "grad_norm": 0.17545711994171143,
1281
+ "learning_rate": 5.156014042837696e-06,
1282
+ "loss": 0.3646,
1283
+ "step": 182
1284
+ },
1285
+ {
1286
+ "epoch": 1.6339285714285714,
1287
+ "grad_norm": 0.20088276267051697,
1288
+ "learning_rate": 5.104018742197557e-06,
1289
+ "loss": 0.3603,
1290
+ "step": 183
1291
+ },
1292
+ {
1293
+ "epoch": 1.6428571428571428,
1294
+ "grad_norm": 0.1916016936302185,
1295
+ "learning_rate": 5.052012185312322e-06,
1296
+ "loss": 0.3596,
1297
+ "step": 184
1298
+ },
1299
+ {
1300
+ "epoch": 1.6517857142857144,
1301
+ "grad_norm": 0.18089112639427185,
1302
+ "learning_rate": 5e-06,
1303
+ "loss": 0.3646,
1304
+ "step": 185
1305
+ },
1306
+ {
1307
+ "epoch": 1.6607142857142856,
1308
+ "grad_norm": 0.21182607114315033,
1309
+ "learning_rate": 4.94798781468768e-06,
1310
+ "loss": 0.3616,
1311
+ "step": 186
1312
+ },
1313
+ {
1314
+ "epoch": 1.6696428571428572,
1315
+ "grad_norm": 0.1948297768831253,
1316
+ "learning_rate": 4.895981257802444e-06,
1317
+ "loss": 0.366,
1318
+ "step": 187
1319
+ },
1320
+ {
1321
+ "epoch": 1.6785714285714286,
1322
+ "grad_norm": 0.18150609731674194,
1323
+ "learning_rate": 4.843985957162304e-06,
1324
+ "loss": 0.3613,
1325
+ "step": 188
1326
+ },
1327
+ {
1328
+ "epoch": 1.6875,
1329
+ "grad_norm": 0.22999927401542664,
1330
+ "learning_rate": 4.792007539367198e-06,
1331
+ "loss": 0.3614,
1332
+ "step": 189
1333
+ },
1334
+ {
1335
+ "epoch": 1.6964285714285714,
1336
+ "grad_norm": 0.18993772566318512,
1337
+ "learning_rate": 4.740051629190099e-06,
1338
+ "loss": 0.3665,
1339
+ "step": 190
1340
+ },
1341
+ {
1342
+ "epoch": 1.7053571428571428,
1343
+ "grad_norm": 0.20203571021556854,
1344
+ "learning_rate": 4.6881238489683596e-06,
1345
+ "loss": 0.3605,
1346
+ "step": 191
1347
+ },
1348
+ {
1349
+ "epoch": 1.7142857142857144,
1350
+ "grad_norm": 0.1820489913225174,
1351
+ "learning_rate": 4.636229817995281e-06,
1352
+ "loss": 0.3687,
1353
+ "step": 192
1354
+ },
1355
+ {
1356
+ "epoch": 1.7232142857142856,
1357
+ "grad_norm": 0.1792946755886078,
1358
+ "learning_rate": 4.584375151912043e-06,
1359
+ "loss": 0.37,
1360
+ "step": 193
1361
+ },
1362
+ {
1363
+ "epoch": 1.7321428571428572,
1364
+ "grad_norm": 0.1797734498977661,
1365
+ "learning_rate": 4.532565462099999e-06,
1366
+ "loss": 0.3699,
1367
+ "step": 194
1368
+ },
1369
+ {
1370
+ "epoch": 1.7410714285714286,
1371
+ "grad_norm": 0.17329907417297363,
1372
+ "learning_rate": 4.480806355073467e-06,
1373
+ "loss": 0.3634,
1374
+ "step": 195
1375
+ },
1376
+ {
1377
+ "epoch": 1.75,
1378
+ "grad_norm": 0.17865252494812012,
1379
+ "learning_rate": 4.429103431873009e-06,
1380
+ "loss": 0.3647,
1381
+ "step": 196
1382
+ },
1383
+ {
1384
+ "epoch": 1.7589285714285714,
1385
+ "grad_norm": 0.18890346586704254,
1386
+ "learning_rate": 4.377462287459338e-06,
1387
+ "loss": 0.3685,
1388
+ "step": 197
1389
+ },
1390
+ {
1391
+ "epoch": 1.7678571428571428,
1392
+ "grad_norm": 0.16928838193416595,
1393
+ "learning_rate": 4.3258885101078565e-06,
1394
+ "loss": 0.359,
1395
+ "step": 198
1396
+ },
1397
+ {
1398
+ "epoch": 1.7767857142857144,
1399
+ "grad_norm": 0.16775333881378174,
1400
+ "learning_rate": 4.274387680803936e-06,
1401
+ "loss": 0.3589,
1402
+ "step": 199
1403
+ },
1404
+ {
1405
+ "epoch": 1.7857142857142856,
1406
+ "grad_norm": 0.1853839010000229,
1407
+ "learning_rate": 4.2229653726389765e-06,
1408
+ "loss": 0.3657,
1409
+ "step": 200
1410
+ },
1411
+ {
1412
+ "epoch": 1.7946428571428572,
1413
+ "grad_norm": 0.17588171362876892,
1414
+ "learning_rate": 4.171627150207314e-06,
1415
+ "loss": 0.3637,
1416
+ "step": 201
1417
+ },
1418
+ {
1419
+ "epoch": 1.8035714285714286,
1420
+ "grad_norm": 0.1664077490568161,
1421
+ "learning_rate": 4.120378569004074e-06,
1422
+ "loss": 0.3655,
1423
+ "step": 202
1424
+ },
1425
+ {
1426
+ "epoch": 1.8125,
1427
+ "grad_norm": 0.17974969744682312,
1428
+ "learning_rate": 4.069225174823968e-06,
1429
+ "loss": 0.3662,
1430
+ "step": 203
1431
+ },
1432
+ {
1433
+ "epoch": 1.8214285714285714,
1434
+ "grad_norm": 0.17224469780921936,
1435
+ "learning_rate": 4.018172503161179e-06,
1436
+ "loss": 0.3612,
1437
+ "step": 204
1438
+ },
1439
+ {
1440
+ "epoch": 1.8303571428571428,
1441
+ "grad_norm": 0.1866443157196045,
1442
+ "learning_rate": 3.967226078610346e-06,
1443
+ "loss": 0.3594,
1444
+ "step": 205
1445
+ },
1446
+ {
1447
+ "epoch": 1.8392857142857144,
1448
+ "grad_norm": 0.17376399040222168,
1449
+ "learning_rate": 3.9163914142687185e-06,
1450
+ "loss": 0.3685,
1451
+ "step": 206
1452
+ },
1453
+ {
1454
+ "epoch": 1.8482142857142856,
1455
+ "grad_norm": 0.17465896904468536,
1456
+ "learning_rate": 3.865674011139567e-06,
1457
+ "loss": 0.3582,
1458
+ "step": 207
1459
+ },
1460
+ {
1461
+ "epoch": 1.8571428571428572,
1462
+ "grad_norm": 0.15824931859970093,
1463
+ "learning_rate": 3.815079357536907e-06,
1464
+ "loss": 0.3641,
1465
+ "step": 208
1466
+ },
1467
+ {
1468
+ "epoch": 1.8660714285714286,
1469
+ "grad_norm": 0.1589619368314743,
1470
+ "learning_rate": 3.7646129284915754e-06,
1471
+ "loss": 0.3665,
1472
+ "step": 209
1473
+ },
1474
+ {
1475
+ "epoch": 1.875,
1476
+ "grad_norm": 0.16915646195411682,
1477
+ "learning_rate": 3.714280185158771e-06,
1478
+ "loss": 0.3612,
1479
+ "step": 210
1480
+ },
1481
+ {
1482
+ "epoch": 1.8839285714285714,
1483
+ "grad_norm": 0.16680151224136353,
1484
+ "learning_rate": 3.6640865742270756e-06,
1485
+ "loss": 0.3629,
1486
+ "step": 211
1487
+ },
1488
+ {
1489
+ "epoch": 1.8928571428571428,
1490
+ "grad_norm": 0.17910495400428772,
1491
+ "learning_rate": 3.614037527329048e-06,
1492
+ "loss": 0.3646,
1493
+ "step": 212
1494
+ },
1495
+ {
1496
+ "epoch": 1.9017857142857144,
1497
+ "grad_norm": 0.14773690700531006,
1498
+ "learning_rate": 3.5641384604534503e-06,
1499
+ "loss": 0.3607,
1500
+ "step": 213
1501
+ },
1502
+ {
1503
+ "epoch": 1.9107142857142856,
1504
+ "grad_norm": 0.20091548562049866,
1505
+ "learning_rate": 3.5143947733591633e-06,
1506
+ "loss": 0.3613,
1507
+ "step": 214
1508
+ },
1509
+ {
1510
+ "epoch": 1.9196428571428572,
1511
+ "grad_norm": 0.17596937716007233,
1512
+ "learning_rate": 3.4648118489908588e-06,
1513
+ "loss": 0.365,
1514
+ "step": 215
1515
+ },
1516
+ {
1517
+ "epoch": 1.9285714285714286,
1518
+ "grad_norm": 0.18936946988105774,
1519
+ "learning_rate": 3.4153950528964867e-06,
1520
+ "loss": 0.3644,
1521
+ "step": 216
1522
+ },
1523
+ {
1524
+ "epoch": 1.9375,
1525
+ "grad_norm": 0.16055169701576233,
1526
+ "learning_rate": 3.366149732646661e-06,
1527
+ "loss": 0.3662,
1528
+ "step": 217
1529
+ },
1530
+ {
1531
+ "epoch": 1.9464285714285714,
1532
+ "grad_norm": 0.18520261347293854,
1533
+ "learning_rate": 3.3170812172559695e-06,
1534
+ "loss": 0.3598,
1535
+ "step": 218
1536
+ },
1537
+ {
1538
+ "epoch": 1.9553571428571428,
1539
+ "grad_norm": 0.16777203977108002,
1540
+ "learning_rate": 3.268194816606305e-06,
1541
+ "loss": 0.3625,
1542
+ "step": 219
1543
+ },
1544
+ {
1545
+ "epoch": 1.9642857142857144,
1546
+ "grad_norm": 0.19867220520973206,
1547
+ "learning_rate": 3.2194958208722656e-06,
1548
+ "loss": 0.364,
1549
+ "step": 220
1550
+ },
1551
+ {
1552
+ "epoch": 1.9732142857142856,
1553
+ "grad_norm": 0.15577928721904755,
1554
+ "learning_rate": 3.1709894999486828e-06,
1555
+ "loss": 0.3618,
1556
+ "step": 221
1557
+ },
1558
+ {
1559
+ "epoch": 1.9821428571428572,
1560
+ "grad_norm": 0.16875170171260834,
1561
+ "learning_rate": 3.1226811028803514e-06,
1562
+ "loss": 0.3627,
1563
+ "step": 222
1564
+ },
1565
+ {
1566
+ "epoch": 1.9910714285714286,
1567
+ "grad_norm": 0.17682914435863495,
1568
+ "learning_rate": 3.0745758572940044e-06,
1569
+ "loss": 0.3661,
1570
+ "step": 223
1571
+ },
1572
+ {
1573
+ "epoch": 2.0,
1574
+ "grad_norm": 0.17448174953460693,
1575
+ "learning_rate": 3.0266789688326187e-06,
1576
+ "loss": 0.3518,
1577
+ "step": 224
1578
+ },
1579
+ {
1580
+ "epoch": 2.0089285714285716,
1581
+ "grad_norm": 0.16910579800605774,
1582
+ "learning_rate": 2.978995620592092e-06,
1583
+ "loss": 0.3488,
1584
+ "step": 225
1585
+ },
1586
+ {
1587
+ "epoch": 2.017857142857143,
1588
+ "grad_norm": 0.1821155995130539,
1589
+ "learning_rate": 2.9315309725603596e-06,
1590
+ "loss": 0.3498,
1591
+ "step": 226
1592
+ },
1593
+ {
1594
+ "epoch": 2.0267857142857144,
1595
+ "grad_norm": 0.17017966508865356,
1596
+ "learning_rate": 2.884290161059017e-06,
1597
+ "loss": 0.3411,
1598
+ "step": 227
1599
+ },
1600
+ {
1601
+ "epoch": 2.0357142857142856,
1602
+ "grad_norm": 0.1491294950246811,
1603
+ "learning_rate": 2.8372782981874964e-06,
1604
+ "loss": 0.354,
1605
+ "step": 228
1606
+ },
1607
+ {
1608
+ "epoch": 2.044642857142857,
1609
+ "grad_norm": 0.172164186835289,
1610
+ "learning_rate": 2.7905004712698646e-06,
1611
+ "loss": 0.3454,
1612
+ "step": 229
1613
+ },
1614
+ {
1615
+ "epoch": 2.0535714285714284,
1616
+ "grad_norm": 0.180723175406456,
1617
+ "learning_rate": 2.7439617423043146e-06,
1618
+ "loss": 0.3528,
1619
+ "step": 230
1620
+ },
1621
+ {
1622
+ "epoch": 2.0625,
1623
+ "grad_norm": 0.1604950726032257,
1624
+ "learning_rate": 2.697667147415383e-06,
1625
+ "loss": 0.3489,
1626
+ "step": 231
1627
+ },
1628
+ {
1629
+ "epoch": 2.0714285714285716,
1630
+ "grad_norm": 0.16484476625919342,
1631
+ "learning_rate": 2.6516216963089698e-06,
1632
+ "loss": 0.3436,
1633
+ "step": 232
1634
+ },
1635
+ {
1636
+ "epoch": 2.080357142857143,
1637
+ "grad_norm": 0.1700424700975418,
1638
+ "learning_rate": 2.605830371730229e-06,
1639
+ "loss": 0.3432,
1640
+ "step": 233
1641
+ },
1642
+ {
1643
+ "epoch": 2.0892857142857144,
1644
+ "grad_norm": 0.16941745579242706,
1645
+ "learning_rate": 2.560298128924358e-06,
1646
+ "loss": 0.3424,
1647
+ "step": 234
1648
+ },
1649
+ {
1650
+ "epoch": 2.0982142857142856,
1651
+ "grad_norm": 0.187941312789917,
1652
+ "learning_rate": 2.5150298951003783e-06,
1653
+ "loss": 0.3421,
1654
+ "step": 235
1655
+ },
1656
+ {
1657
+ "epoch": 2.107142857142857,
1658
+ "grad_norm": 0.1790265589952469,
1659
+ "learning_rate": 2.470030568897938e-06,
1660
+ "loss": 0.3458,
1661
+ "step": 236
1662
+ },
1663
+ {
1664
+ "epoch": 2.1160714285714284,
1665
+ "grad_norm": 0.16201697289943695,
1666
+ "learning_rate": 2.425305019857222e-06,
1667
+ "loss": 0.349,
1668
+ "step": 237
1669
+ },
1670
+ {
1671
+ "epoch": 2.125,
1672
+ "grad_norm": 0.18233942985534668,
1673
+ "learning_rate": 2.3808580878919948e-06,
1674
+ "loss": 0.3507,
1675
+ "step": 238
1676
+ },
1677
+ {
1678
+ "epoch": 2.1339285714285716,
1679
+ "grad_norm": 0.16857150197029114,
1680
+ "learning_rate": 2.336694582765857e-06,
1681
+ "loss": 0.3429,
1682
+ "step": 239
1683
+ },
1684
+ {
1685
+ "epoch": 2.142857142857143,
1686
+ "grad_norm": 0.1519509106874466,
1687
+ "learning_rate": 2.2928192835717642e-06,
1688
+ "loss": 0.3459,
1689
+ "step": 240
1690
+ },
1691
+ {
1692
+ "epoch": 2.1517857142857144,
1693
+ "grad_norm": 0.15660250186920166,
1694
+ "learning_rate": 2.2492369382148634e-06,
1695
+ "loss": 0.3471,
1696
+ "step": 241
1697
+ },
1698
+ {
1699
+ "epoch": 2.1607142857142856,
1700
+ "grad_norm": 0.1472284495830536,
1701
+ "learning_rate": 2.205952262898704e-06,
1702
+ "loss": 0.3456,
1703
+ "step": 242
1704
+ },
1705
+ {
1706
+ "epoch": 2.169642857142857,
1707
+ "grad_norm": 0.15189002454280853,
1708
+ "learning_rate": 2.1629699416148832e-06,
1709
+ "loss": 0.3459,
1710
+ "step": 243
1711
+ },
1712
+ {
1713
+ "epoch": 2.1785714285714284,
1714
+ "grad_norm": 0.13611957430839539,
1715
+ "learning_rate": 2.120294625636171e-06,
1716
+ "loss": 0.3513,
1717
+ "step": 244
1718
+ },
1719
+ {
1720
+ "epoch": 2.1875,
1721
+ "grad_norm": 0.14140184223651886,
1722
+ "learning_rate": 2.077930933013182e-06,
1723
+ "loss": 0.3423,
1724
+ "step": 245
1725
+ },
1726
+ {
1727
+ "epoch": 2.1964285714285716,
1728
+ "grad_norm": 0.1375838965177536,
1729
+ "learning_rate": 2.0358834480746363e-06,
1730
+ "loss": 0.3511,
1731
+ "step": 246
1732
+ },
1733
+ {
1734
+ "epoch": 2.205357142857143,
1735
+ "grad_norm": 0.1536509245634079,
1736
+ "learning_rate": 1.9941567209312768e-06,
1737
+ "loss": 0.3412,
1738
+ "step": 247
1739
+ },
1740
+ {
1741
+ "epoch": 2.2142857142857144,
1742
+ "grad_norm": 0.15353716909885406,
1743
+ "learning_rate": 1.9527552669834797e-06,
1744
+ "loss": 0.3439,
1745
+ "step": 248
1746
+ },
1747
+ {
1748
+ "epoch": 2.2232142857142856,
1749
+ "grad_norm": 0.1261204034090042,
1750
+ "learning_rate": 1.911683566432633e-06,
1751
+ "loss": 0.3429,
1752
+ "step": 249
1753
+ },
1754
+ {
1755
+ "epoch": 2.232142857142857,
1756
+ "grad_norm": 0.1508454829454422,
1757
+ "learning_rate": 1.8709460637963123e-06,
1758
+ "loss": 0.3399,
1759
+ "step": 250
1760
+ },
1761
+ {
1762
+ "epoch": 2.2410714285714284,
1763
+ "grad_norm": 0.13777896761894226,
1764
+ "learning_rate": 1.8305471674273262e-06,
1765
+ "loss": 0.3506,
1766
+ "step": 251
1767
+ },
1768
+ {
1769
+ "epoch": 2.25,
1770
+ "grad_norm": 0.1431542932987213,
1771
+ "learning_rate": 1.7904912490366723e-06,
1772
+ "loss": 0.3421,
1773
+ "step": 252
1774
+ },
1775
+ {
1776
+ "epoch": 2.2589285714285716,
1777
+ "grad_norm": 0.15625214576721191,
1778
+ "learning_rate": 1.7507826432204571e-06,
1779
+ "loss": 0.3516,
1780
+ "step": 253
1781
+ },
1782
+ {
1783
+ "epoch": 2.267857142857143,
1784
+ "grad_norm": 0.13694243133068085,
1785
+ "learning_rate": 1.711425646990838e-06,
1786
+ "loss": 0.3476,
1787
+ "step": 254
1788
+ },
1789
+ {
1790
+ "epoch": 2.2767857142857144,
1791
+ "grad_norm": 0.13206931948661804,
1792
+ "learning_rate": 1.6724245193110178e-06,
1793
+ "loss": 0.3456,
1794
+ "step": 255
1795
+ },
1796
+ {
1797
+ "epoch": 2.2857142857142856,
1798
+ "grad_norm": 0.1381482183933258,
1799
+ "learning_rate": 1.6337834806343783e-06,
1800
+ "loss": 0.3546,
1801
+ "step": 256
1802
+ },
1803
+ {
1804
+ "epoch": 2.294642857142857,
1805
+ "grad_norm": 0.14691098034381866,
1806
+ "learning_rate": 1.595506712447768e-06,
1807
+ "loss": 0.3518,
1808
+ "step": 257
1809
+ },
1810
+ {
1811
+ "epoch": 2.3035714285714284,
1812
+ "grad_norm": 0.13349969685077667,
1813
+ "learning_rate": 1.557598356819e-06,
1814
+ "loss": 0.3488,
1815
+ "step": 258
1816
+ },
1817
+ {
1818
+ "epoch": 2.3125,
1819
+ "grad_norm": 0.14272207021713257,
1820
+ "learning_rate": 1.5200625159486322e-06,
1821
+ "loss": 0.3471,
1822
+ "step": 259
1823
+ },
1824
+ {
1825
+ "epoch": 2.3214285714285716,
1826
+ "grad_norm": 0.13418231904506683,
1827
+ "learning_rate": 1.482903251726049e-06,
1828
+ "loss": 0.3419,
1829
+ "step": 260
1830
+ },
1831
+ {
1832
+ "epoch": 2.330357142857143,
1833
+ "grad_norm": 0.13040195405483246,
1834
+ "learning_rate": 1.446124585289913e-06,
1835
+ "loss": 0.3458,
1836
+ "step": 261
1837
+ },
1838
+ {
1839
+ "epoch": 2.3392857142857144,
1840
+ "grad_norm": 0.12616047263145447,
1841
+ "learning_rate": 1.409730496593016e-06,
1842
+ "loss": 0.3493,
1843
+ "step": 262
1844
+ },
1845
+ {
1846
+ "epoch": 2.3482142857142856,
1847
+ "grad_norm": 0.20311793684959412,
1848
+ "learning_rate": 1.3737249239716043e-06,
1849
+ "loss": 0.3503,
1850
+ "step": 263
1851
+ },
1852
+ {
1853
+ "epoch": 2.357142857142857,
1854
+ "grad_norm": 0.160719633102417,
1855
+ "learning_rate": 1.3381117637191887e-06,
1856
+ "loss": 0.3493,
1857
+ "step": 264
1858
+ },
1859
+ {
1860
+ "epoch": 2.3660714285714284,
1861
+ "grad_norm": 0.1257162094116211,
1862
+ "learning_rate": 1.3028948696649164e-06,
1863
+ "loss": 0.3467,
1864
+ "step": 265
1865
+ },
1866
+ {
1867
+ "epoch": 2.375,
1868
+ "grad_norm": 0.15282198786735535,
1869
+ "learning_rate": 1.2680780527565313e-06,
1870
+ "loss": 0.3443,
1871
+ "step": 266
1872
+ },
1873
+ {
1874
+ "epoch": 2.3839285714285716,
1875
+ "grad_norm": 0.1382765769958496,
1876
+ "learning_rate": 1.2336650806479826e-06,
1877
+ "loss": 0.3461,
1878
+ "step": 267
1879
+ },
1880
+ {
1881
+ "epoch": 2.392857142857143,
1882
+ "grad_norm": 0.13694579899311066,
1883
+ "learning_rate": 1.1996596772917091e-06,
1884
+ "loss": 0.3543,
1885
+ "step": 268
1886
+ },
1887
+ {
1888
+ "epoch": 2.4017857142857144,
1889
+ "grad_norm": 0.13849100470542908,
1890
+ "learning_rate": 1.1660655225356533e-06,
1891
+ "loss": 0.3471,
1892
+ "step": 269
1893
+ },
1894
+ {
1895
+ "epoch": 2.4107142857142856,
1896
+ "grad_norm": 0.13622619211673737,
1897
+ "learning_rate": 1.132886251725061e-06,
1898
+ "loss": 0.3474,
1899
+ "step": 270
1900
+ },
1901
+ {
1902
+ "epoch": 2.419642857142857,
1903
+ "grad_norm": 0.13429802656173706,
1904
+ "learning_rate": 1.1001254553090812e-06,
1905
+ "loss": 0.3451,
1906
+ "step": 271
1907
+ },
1908
+ {
1909
+ "epoch": 2.4285714285714284,
1910
+ "grad_norm": 0.1249421089887619,
1911
+ "learning_rate": 1.0677866784522317e-06,
1912
+ "loss": 0.3474,
1913
+ "step": 272
1914
+ },
1915
+ {
1916
+ "epoch": 2.4375,
1917
+ "grad_norm": 0.1234789565205574,
1918
+ "learning_rate": 1.0358734206507642e-06,
1919
+ "loss": 0.3459,
1920
+ "step": 273
1921
+ },
1922
+ {
1923
+ "epoch": 2.4464285714285716,
1924
+ "grad_norm": 0.12073200196027756,
1925
+ "learning_rate": 1.004389135353972e-06,
1926
+ "loss": 0.3387,
1927
+ "step": 274
1928
+ },
1929
+ {
1930
+ "epoch": 2.455357142857143,
1931
+ "grad_norm": 0.1367228776216507,
1932
+ "learning_rate": 9.733372295904776e-07,
1933
+ "loss": 0.347,
1934
+ "step": 275
1935
+ },
1936
+ {
1937
+ "epoch": 2.4642857142857144,
1938
+ "grad_norm": 0.12393829226493835,
1939
+ "learning_rate": 9.427210635995482e-07,
1940
+ "loss": 0.3467,
1941
+ "step": 276
1942
+ },
1943
+ {
1944
+ "epoch": 2.4732142857142856,
1945
+ "grad_norm": 0.1407790631055832,
1946
+ "learning_rate": 9.125439504674699e-07,
1947
+ "loss": 0.3524,
1948
+ "step": 277
1949
+ },
1950
+ {
1951
+ "epoch": 2.482142857142857,
1952
+ "grad_norm": 0.1231057420372963,
1953
+ "learning_rate": 8.828091557690288e-07,
1954
+ "loss": 0.351,
1955
+ "step": 278
1956
+ },
1957
+ {
1958
+ "epoch": 2.4910714285714284,
1959
+ "grad_norm": 0.1206686794757843,
1960
+ "learning_rate": 8.535198972141295e-07,
1961
+ "loss": 0.3455,
1962
+ "step": 279
1963
+ },
1964
+ {
1965
+ "epoch": 2.5,
1966
+ "grad_norm": 0.12657472491264343,
1967
+ "learning_rate": 8.246793442995954e-07,
1968
+ "loss": 0.3441,
1969
+ "step": 280
1970
+ },
1971
+ {
1972
+ "epoch": 2.508928571428571,
1973
+ "grad_norm": 0.13261254131793976,
1974
+ "learning_rate": 7.962906179661872e-07,
1975
+ "loss": 0.338,
1976
+ "step": 281
1977
+ },
1978
+ {
1979
+ "epoch": 2.517857142857143,
1980
+ "grad_norm": 0.13191427290439606,
1981
+ "learning_rate": 7.68356790260873e-07,
1982
+ "loss": 0.3423,
1983
+ "step": 282
1984
+ },
1985
+ {
1986
+ "epoch": 2.5267857142857144,
1987
+ "grad_norm": 0.13889721035957336,
1988
+ "learning_rate": 7.408808840043913e-07,
1989
+ "loss": 0.3437,
1990
+ "step": 283
1991
+ },
1992
+ {
1993
+ "epoch": 2.5357142857142856,
1994
+ "grad_norm": 0.13105542957782745,
1995
+ "learning_rate": 7.138658724641417e-07,
1996
+ "loss": 0.3533,
1997
+ "step": 284
1998
+ },
1999
+ {
2000
+ "epoch": 2.544642857142857,
2001
+ "grad_norm": 0.12581577897071838,
2002
+ "learning_rate": 6.873146790324359e-07,
2003
+ "loss": 0.3473,
2004
+ "step": 285
2005
+ },
2006
+ {
2007
+ "epoch": 2.553571428571429,
2008
+ "grad_norm": 0.12375447899103165,
2009
+ "learning_rate": 6.612301769101464e-07,
2010
+ "loss": 0.3466,
2011
+ "step": 286
2012
+ },
2013
+ {
2014
+ "epoch": 2.5625,
2015
+ "grad_norm": 0.11879701167345047,
2016
+ "learning_rate": 6.356151887957901e-07,
2017
+ "loss": 0.3526,
2018
+ "step": 287
2019
+ },
2020
+ {
2021
+ "epoch": 2.571428571428571,
2022
+ "grad_norm": 0.1344556361436844,
2023
+ "learning_rate": 6.104724865800665e-07,
2024
+ "loss": 0.3436,
2025
+ "step": 288
2026
+ },
2027
+ {
2028
+ "epoch": 2.580357142857143,
2029
+ "grad_norm": 0.1175764948129654,
2030
+ "learning_rate": 5.858047910459109e-07,
2031
+ "loss": 0.3435,
2032
+ "step": 289
2033
+ },
2034
+ {
2035
+ "epoch": 2.5892857142857144,
2036
+ "grad_norm": 0.12031050771474838,
2037
+ "learning_rate": 5.616147715740611e-07,
2038
+ "loss": 0.3425,
2039
+ "step": 290
2040
+ },
2041
+ {
2042
+ "epoch": 2.5982142857142856,
2043
+ "grad_norm": 0.1191452294588089,
2044
+ "learning_rate": 5.379050458541995e-07,
2045
+ "loss": 0.3437,
2046
+ "step": 291
2047
+ },
2048
+ {
2049
+ "epoch": 2.607142857142857,
2050
+ "grad_norm": 0.11280887573957443,
2051
+ "learning_rate": 5.146781796016798e-07,
2052
+ "loss": 0.3477,
2053
+ "step": 292
2054
+ },
2055
+ {
2056
+ "epoch": 2.616071428571429,
2057
+ "grad_norm": 0.12038522213697433,
2058
+ "learning_rate": 4.919366862798807e-07,
2059
+ "loss": 0.3452,
2060
+ "step": 293
2061
+ },
2062
+ {
2063
+ "epoch": 2.625,
2064
+ "grad_norm": 0.1229255199432373,
2065
+ "learning_rate": 4.696830268282204e-07,
2066
+ "loss": 0.3487,
2067
+ "step": 294
2068
+ },
2069
+ {
2070
+ "epoch": 2.633928571428571,
2071
+ "grad_norm": 0.12023057788610458,
2072
+ "learning_rate": 4.479196093958421e-07,
2073
+ "loss": 0.3458,
2074
+ "step": 295
2075
+ },
2076
+ {
2077
+ "epoch": 2.642857142857143,
2078
+ "grad_norm": 0.11441769450902939,
2079
+ "learning_rate": 4.2664878908102556e-07,
2080
+ "loss": 0.3467,
2081
+ "step": 296
2082
+ },
2083
+ {
2084
+ "epoch": 2.6517857142857144,
2085
+ "grad_norm": 0.12089654058218002,
2086
+ "learning_rate": 4.058728676763313e-07,
2087
+ "loss": 0.3517,
2088
+ "step": 297
2089
+ },
2090
+ {
2091
+ "epoch": 2.6607142857142856,
2092
+ "grad_norm": 0.11661281436681747,
2093
+ "learning_rate": 3.855940934195146e-07,
2094
+ "loss": 0.3459,
2095
+ "step": 298
2096
+ },
2097
+ {
2098
+ "epoch": 2.669642857142857,
2099
+ "grad_norm": 0.11176279187202454,
2100
+ "learning_rate": 3.6581466075023443e-07,
2101
+ "loss": 0.3506,
2102
+ "step": 299
2103
+ },
2104
+ {
2105
+ "epoch": 2.678571428571429,
2106
+ "grad_norm": 0.1134781613945961,
2107
+ "learning_rate": 3.4653671007259084e-07,
2108
+ "loss": 0.3445,
2109
+ "step": 300
2110
+ },
2111
+ {
2112
+ "epoch": 2.6875,
2113
+ "grad_norm": 0.11702830344438553,
2114
+ "learning_rate": 3.2776232752349536e-07,
2115
+ "loss": 0.3417,
2116
+ "step": 301
2117
+ },
2118
+ {
2119
+ "epoch": 2.696428571428571,
2120
+ "grad_norm": 0.11138033121824265,
2121
+ "learning_rate": 3.0949354474692937e-07,
2122
+ "loss": 0.3439,
2123
+ "step": 302
2124
+ },
2125
+ {
2126
+ "epoch": 2.705357142857143,
2127
+ "grad_norm": 0.11514540761709213,
2128
+ "learning_rate": 2.9173233867409057e-07,
2129
+ "loss": 0.3439,
2130
+ "step": 303
2131
+ },
2132
+ {
2133
+ "epoch": 2.7142857142857144,
2134
+ "grad_norm": 0.11949031054973602,
2135
+ "learning_rate": 2.7448063130946224e-07,
2136
+ "loss": 0.3495,
2137
+ "step": 304
2138
+ },
2139
+ {
2140
+ "epoch": 2.7232142857142856,
2141
+ "grad_norm": 0.10531047731637955,
2142
+ "learning_rate": 2.577402895228243e-07,
2143
+ "loss": 0.3417,
2144
+ "step": 305
2145
+ },
2146
+ {
2147
+ "epoch": 2.732142857142857,
2148
+ "grad_norm": 0.11205192655324936,
2149
+ "learning_rate": 2.4151312484723465e-07,
2150
+ "loss": 0.343,
2151
+ "step": 306
2152
+ },
2153
+ {
2154
+ "epoch": 2.741071428571429,
2155
+ "grad_norm": 0.10984615981578827,
2156
+ "learning_rate": 2.2580089328299747e-07,
2157
+ "loss": 0.3462,
2158
+ "step": 307
2159
+ },
2160
+ {
2161
+ "epoch": 2.75,
2162
+ "grad_norm": 0.10972468554973602,
2163
+ "learning_rate": 2.106052951076365e-07,
2164
+ "loss": 0.3443,
2165
+ "step": 308
2166
+ },
2167
+ {
2168
+ "epoch": 2.758928571428571,
2169
+ "grad_norm": 0.14334134757518768,
2170
+ "learning_rate": 1.959279746919057e-07,
2171
+ "loss": 0.345,
2172
+ "step": 309
2173
+ },
2174
+ {
2175
+ "epoch": 2.767857142857143,
2176
+ "grad_norm": 0.11364768445491791,
2177
+ "learning_rate": 1.8177052032184285e-07,
2178
+ "loss": 0.3458,
2179
+ "step": 310
2180
+ },
2181
+ {
2182
+ "epoch": 2.7767857142857144,
2183
+ "grad_norm": 0.11008045077323914,
2184
+ "learning_rate": 1.6813446402689783e-07,
2185
+ "loss": 0.3482,
2186
+ "step": 311
2187
+ },
2188
+ {
2189
+ "epoch": 2.7857142857142856,
2190
+ "grad_norm": 0.1126861646771431,
2191
+ "learning_rate": 1.5502128141414496e-07,
2192
+ "loss": 0.3412,
2193
+ "step": 312
2194
+ },
2195
+ {
2196
+ "epoch": 2.794642857142857,
2197
+ "grad_norm": 0.11044777929782867,
2198
+ "learning_rate": 1.4243239150860123e-07,
2199
+ "loss": 0.3438,
2200
+ "step": 313
2201
+ },
2202
+ {
2203
+ "epoch": 2.803571428571429,
2204
+ "grad_norm": 0.12912730872631073,
2205
+ "learning_rate": 1.303691565996712e-07,
2206
+ "loss": 0.3497,
2207
+ "step": 314
2208
+ },
2209
+ {
2210
+ "epoch": 2.8125,
2211
+ "grad_norm": 0.11161229759454727,
2212
+ "learning_rate": 1.1883288209372512e-07,
2213
+ "loss": 0.3449,
2214
+ "step": 315
2215
+ },
2216
+ {
2217
+ "epoch": 2.821428571428571,
2218
+ "grad_norm": 0.11409316956996918,
2219
+ "learning_rate": 1.0782481637284014e-07,
2220
+ "loss": 0.3448,
2221
+ "step": 316
2222
+ },
2223
+ {
2224
+ "epoch": 2.830357142857143,
2225
+ "grad_norm": 0.11476591974496841,
2226
+ "learning_rate": 9.734615065970454e-08,
2227
+ "loss": 0.3474,
2228
+ "step": 317
2229
+ },
2230
+ {
2231
+ "epoch": 2.8392857142857144,
2232
+ "grad_norm": 0.11300153285264969,
2233
+ "learning_rate": 8.739801888871468e-08,
2234
+ "loss": 0.3484,
2235
+ "step": 318
2236
+ },
2237
+ {
2238
+ "epoch": 2.8482142857142856,
2239
+ "grad_norm": 0.10717196017503738,
2240
+ "learning_rate": 7.79814975832649e-08,
2241
+ "loss": 0.3508,
2242
+ "step": 319
2243
+ },
2244
+ {
2245
+ "epoch": 2.857142857142857,
2246
+ "grad_norm": 0.10945117473602295,
2247
+ "learning_rate": 6.909760573925561e-08,
2248
+ "loss": 0.3438,
2249
+ "step": 320
2250
+ },
2251
+ {
2252
+ "epoch": 2.866071428571429,
2253
+ "grad_norm": 0.11592280864715576,
2254
+ "learning_rate": 6.074730471482049e-08,
2255
+ "loss": 0.3466,
2256
+ "step": 321
2257
+ },
2258
+ {
2259
+ "epoch": 2.875,
2260
+ "grad_norm": 0.10622705519199371,
2261
+ "learning_rate": 5.2931498126298495e-08,
2262
+ "loss": 0.3441,
2263
+ "step": 322
2264
+ },
2265
+ {
2266
+ "epoch": 2.883928571428571,
2267
+ "grad_norm": 0.11332610249519348,
2268
+ "learning_rate": 4.565103175044883e-08,
2269
+ "loss": 0.3475,
2270
+ "step": 323
2271
+ },
2272
+ {
2273
+ "epoch": 2.892857142857143,
2274
+ "grad_norm": 0.12698130309581757,
2275
+ "learning_rate": 3.890669343292464e-08,
2276
+ "loss": 0.3478,
2277
+ "step": 324
2278
+ },
2279
+ {
2280
+ "epoch": 2.9017857142857144,
2281
+ "grad_norm": 0.11187270283699036,
2282
+ "learning_rate": 3.269921300301959e-08,
2283
+ "loss": 0.3431,
2284
+ "step": 325
2285
+ },
2286
+ {
2287
+ "epoch": 2.9107142857142856,
2288
+ "grad_norm": 0.1112656369805336,
2289
+ "learning_rate": 2.702926219468882e-08,
2290
+ "loss": 0.3441,
2291
+ "step": 326
2292
+ },
2293
+ {
2294
+ "epoch": 2.919642857142857,
2295
+ "grad_norm": 0.11196763813495636,
2296
+ "learning_rate": 2.1897454573860388e-08,
2297
+ "loss": 0.3538,
2298
+ "step": 327
2299
+ },
2300
+ {
2301
+ "epoch": 2.928571428571429,
2302
+ "grad_norm": 0.11176672577857971,
2303
+ "learning_rate": 1.7304345472035634e-08,
2304
+ "loss": 0.3431,
2305
+ "step": 328
2306
+ },
2307
+ {
2308
+ "epoch": 2.9375,
2309
+ "grad_norm": 0.11075292527675629,
2310
+ "learning_rate": 1.3250431926197793e-08,
2311
+ "loss": 0.3412,
2312
+ "step": 329
2313
+ },
2314
+ {
2315
+ "epoch": 2.946428571428571,
2316
+ "grad_norm": 0.10811440646648407,
2317
+ "learning_rate": 9.73615262502503e-09,
2318
+ "loss": 0.346,
2319
+ "step": 330
2320
+ },
2321
+ {
2322
+ "epoch": 2.955357142857143,
2323
+ "grad_norm": 0.1118391752243042,
2324
+ "learning_rate": 6.761887861417293e-09,
2325
+ "loss": 0.3491,
2326
+ "step": 331
2327
+ },
2328
+ {
2329
+ "epoch": 2.9642857142857144,
2330
+ "grad_norm": 0.10914882272481918,
2331
+ "learning_rate": 4.327959491344791e-09,
2332
+ "loss": 0.3433,
2333
+ "step": 332
2334
+ },
2335
+ {
2336
+ "epoch": 2.9732142857142856,
2337
+ "grad_norm": 0.11297401040792465,
2338
+ "learning_rate": 2.4346308990175205e-09,
2339
+ "loss": 0.3467,
2340
+ "step": 333
2341
+ },
2342
+ {
2343
+ "epoch": 2.982142857142857,
2344
+ "grad_norm": 0.1087309867143631,
2345
+ "learning_rate": 1.082106968385288e-09,
2346
+ "loss": 0.343,
2347
+ "step": 334
2348
+ },
2349
+ {
2350
+ "epoch": 2.991071428571429,
2351
+ "grad_norm": 0.1101929172873497,
2352
+ "learning_rate": 2.7053406096433366e-10,
2353
+ "loss": 0.3447,
2354
+ "step": 335
2355
+ },
2356
+ {
2357
+ "epoch": 3.0,
2358
+ "grad_norm": 0.13185350596904755,
2359
+ "learning_rate": 0.0,
2360
+ "loss": 0.3407,
2361
+ "step": 336
2362
+ },
2363
+ {
2364
+ "epoch": 3.0,
2365
+ "step": 336,
2366
+ "total_flos": 2.982866014384947e+16,
2367
+ "train_loss": 0.3923433954339652,
2368
+ "train_runtime": 48299.7361,
2369
+ "train_samples_per_second": 7.078,
2370
+ "train_steps_per_second": 0.007
2371
+ }
2372
+ ],
2373
+ "logging_steps": 1,
2374
+ "max_steps": 336,
2375
+ "num_input_tokens_seen": 0,
2376
+ "num_train_epochs": 3,
2377
+ "save_steps": 50,
2378
+ "stateful_callbacks": {
2379
+ "TrainerControl": {
2380
+ "args": {
2381
+ "should_epoch_stop": false,
2382
+ "should_evaluate": false,
2383
+ "should_log": false,
2384
+ "should_save": true,
2385
+ "should_training_stop": true
2386
+ },
2387
+ "attributes": {}
2388
+ }
2389
+ },
2390
+ "total_flos": 2.982866014384947e+16,
2391
+ "train_batch_size": 1,
2392
+ "trial_name": null,
2393
+ "trial_params": null
2394
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eb0244194182469e91b3173374e869ca007a72c3ea3f197b9369a89cb93be320
3
+ size 7608
vocab.json ADDED
The diff for this file is too large to render. See raw diff