Muennighoff commited on
Commit
fe79f60
1 Parent(s): 4915271

Upload folder using huggingface_hub

Browse files
Files changed (41) hide show
  1. checkpoint-1000/optimizer.bin +3 -0
  2. checkpoint-1000/pytorch_model.bin +3 -0
  3. checkpoint-1000/rng_state_0.pth +3 -0
  4. checkpoint-1000/rng_state_1.pth +3 -0
  5. checkpoint-1000/rng_state_2.pth +3 -0
  6. checkpoint-1000/rng_state_3.pth +3 -0
  7. checkpoint-1000/rng_state_4.pth +3 -0
  8. checkpoint-1000/rng_state_5.pth +3 -0
  9. checkpoint-1000/rng_state_6.pth +3 -0
  10. checkpoint-1000/rng_state_7.pth +3 -0
  11. checkpoint-1000/scheduler.pt +3 -0
  12. checkpoint-1000/special_tokens_map.json +24 -0
  13. checkpoint-1000/tokenizer.json +0 -0
  14. checkpoint-1000/tokenizer.model +3 -0
  15. checkpoint-1000/tokenizer_config.json +43 -0
  16. checkpoint-1000/trainer_state.json +0 -0
  17. checkpoint-1000/training_args.bin +3 -0
  18. checkpoint-500/optimizer.bin +3 -0
  19. checkpoint-500/pytorch_model.bin +3 -0
  20. checkpoint-500/rng_state_0.pth +3 -0
  21. checkpoint-500/rng_state_1.pth +3 -0
  22. checkpoint-500/rng_state_2.pth +3 -0
  23. checkpoint-500/rng_state_3.pth +3 -0
  24. checkpoint-500/rng_state_4.pth +3 -0
  25. checkpoint-500/rng_state_5.pth +3 -0
  26. checkpoint-500/rng_state_6.pth +3 -0
  27. checkpoint-500/rng_state_7.pth +3 -0
  28. checkpoint-500/scheduler.pt +3 -0
  29. checkpoint-500/special_tokens_map.json +24 -0
  30. checkpoint-500/tokenizer.json +0 -0
  31. checkpoint-500/tokenizer.model +3 -0
  32. checkpoint-500/tokenizer_config.json +43 -0
  33. checkpoint-500/trainer_state.json +3019 -0
  34. checkpoint-500/training_args.bin +3 -0
  35. config.json +31 -0
  36. pytorch_model.bin +3 -0
  37. special_tokens_map.json +24 -0
  38. tokenizer.json +0 -0
  39. tokenizer.model +3 -0
  40. tokenizer_config.json +43 -0
  41. training_args.bin +3 -0
checkpoint-1000/optimizer.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4f4f851237b7a0755e0cbdb41b0297f38e5fa2f981f8eb78180a9a32e3328443
3
+ size 28442834134
checkpoint-1000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2807191e3b5063f5f0a185ef4d36caa911e19d4eddcf54828cb5ae4153d4b279
3
+ size 14221467522
checkpoint-1000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fca36311bbd57901806c620666131a44cde8c190cbf5560d295f16934f982a60
3
+ size 15984
checkpoint-1000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6513186409e1e6081bcca47cc4ff01d87ecd1af0291a404c7e78e9524b0a0418
3
+ size 15984
checkpoint-1000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee5fcb7a571caaabb18c46317925a84fceed83d237d51d219894eec507842a52
3
+ size 15984
checkpoint-1000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:39b9eeca5ef68429a6920538de6a374f3d77603f82da23762d3bb5a0dd5e3c82
3
+ size 15984
checkpoint-1000/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f2cddacb3f050a2319baea833d0fb0821a788cce7e9c287d27e512999355648
3
+ size 15984
checkpoint-1000/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba6d68d5c751e88d19b0c12593ea9e0aa4e416ce95c720be54228e38334c24cf
3
+ size 15984
checkpoint-1000/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e5924ad176078fb57aaeca6a01dc2a6669118eec88c0b436d823c438527ab46
3
+ size 15984
checkpoint-1000/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff1590184d5fd27f49e8f0c4e2ad8d9d147db3dfe018cd0c2d1f4428f0b19d4d
3
+ size 15984
checkpoint-1000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a2895b0e3f1c28bf5c8fcf1e694d1dd1665857bd510ab8f8631b1e346e84826c
3
+ size 1064
checkpoint-1000/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "<s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
checkpoint-1000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1000/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
3
+ size 493443
checkpoint-1000/tokenizer_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "additional_special_tokens": [],
31
+ "bos_token": "<s>",
32
+ "clean_up_tokenization_spaces": false,
33
+ "eos_token": "</s>",
34
+ "legacy": true,
35
+ "model_max_length": 1000000000000000019884624838656,
36
+ "pad_token": "<s>",
37
+ "padding_side": "right",
38
+ "sp_model_kwargs": {},
39
+ "spaces_between_special_tokens": false,
40
+ "tokenizer_class": "LlamaTokenizer",
41
+ "unk_token": "<unk>",
42
+ "use_default_system_prompt": true
43
+ }
checkpoint-1000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f768ad49686ec36b5a8b570468ea76bd56614d5acbd36a18ffbbf3ba397fe3e
3
+ size 4728
checkpoint-500/optimizer.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5dc047d75f5def686fee76a6b48cf2442208e63ffde523cfd873e46d059b0be5
3
+ size 28442834134
checkpoint-500/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab470f8cc790ad6406631b14b1b0a4aa3dbc9b949a3b09ded273d3fd9a54e153
3
+ size 14221467522
checkpoint-500/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:316f8f4ba01b8af4eacf324be0e9dcd61cc461a2f357b6d03d35d0096273c67f
3
+ size 15920
checkpoint-500/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f89d809197df5975468e91a28e06dddf2b23368e1b7031b7c922bae4eb54bfa
3
+ size 15920
checkpoint-500/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b265bf483088ec1468a8dab2adf69dd0fd177cbc8d5577b6b8d0aa0796de4348
3
+ size 15920
checkpoint-500/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:abbcc0bb563141e251842f9994c1f53fd688d0858f5936ff06b69535c4a6ec8c
3
+ size 15920
checkpoint-500/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8606ebf688c0e0478e77e9538aa3265e84b6c60f6e3e7ddd56b67ead26cc8c22
3
+ size 15920
checkpoint-500/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76119ff45a6e3986f100f8ef46b98b8849ba353b26d7dc187d97188d8d57f1a6
3
+ size 15920
checkpoint-500/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ecd755eaf7f75cfc2075fbf76eca964f09d3cdcbb3caa0c3e97c3a386f660db
3
+ size 15920
checkpoint-500/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bfbd310549d8eaa959e0e6f6bbd6ef9e84bec6bc7fda2341c53b9c192d0478a2
3
+ size 15920
checkpoint-500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d340feb7bcf864481a7391a71f0217c0632c294086494acb7f8c90d216e571b3
3
+ size 1064
checkpoint-500/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "<s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
checkpoint-500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-500/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
3
+ size 493443
checkpoint-500/tokenizer_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "additional_special_tokens": [],
31
+ "bos_token": "<s>",
32
+ "clean_up_tokenization_spaces": false,
33
+ "eos_token": "</s>",
34
+ "legacy": true,
35
+ "model_max_length": 1000000000000000019884624838656,
36
+ "pad_token": "<s>",
37
+ "padding_side": "right",
38
+ "sp_model_kwargs": {},
39
+ "spaces_between_special_tokens": false,
40
+ "tokenizer_class": "LlamaTokenizer",
41
+ "unk_token": "<unk>",
42
+ "use_default_system_prompt": true
43
+ }
checkpoint-500/trainer_state.json ADDED
@@ -0,0 +1,3019 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.10434056761268781,
5
+ "eval_steps": 500,
6
+ "global_step": 500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 5.263157894736843e-07,
14
+ "loss": 5.6888,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 1.0526315789473685e-06,
20
+ "loss": 5.0454,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 1.5789473684210526e-06,
26
+ "loss": 4.8812,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.0,
31
+ "learning_rate": 2.105263157894737e-06,
32
+ "loss": 4.4258,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.0,
37
+ "learning_rate": 2.631578947368421e-06,
38
+ "loss": 3.3052,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.0,
43
+ "learning_rate": 3.157894736842105e-06,
44
+ "loss": 2.6706,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.0,
49
+ "learning_rate": 3.6842105263157896e-06,
50
+ "loss": 1.8629,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.0,
55
+ "learning_rate": 4.210526315789474e-06,
56
+ "loss": 1.3825,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.0,
61
+ "learning_rate": 4.736842105263158e-06,
62
+ "loss": 1.2787,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.0,
67
+ "learning_rate": 5.263157894736842e-06,
68
+ "loss": 0.9156,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.0,
73
+ "learning_rate": 5.789473684210527e-06,
74
+ "loss": 0.765,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.0,
79
+ "learning_rate": 6.31578947368421e-06,
80
+ "loss": 0.7249,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.0,
85
+ "learning_rate": 6.842105263157896e-06,
86
+ "loss": 0.6445,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.0,
91
+ "learning_rate": 7.368421052631579e-06,
92
+ "loss": 0.5672,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.0,
97
+ "learning_rate": 7.894736842105265e-06,
98
+ "loss": 0.6509,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.0,
103
+ "learning_rate": 8.421052631578948e-06,
104
+ "loss": 0.4831,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.0,
109
+ "learning_rate": 8.947368421052632e-06,
110
+ "loss": 0.4905,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.0,
115
+ "learning_rate": 9.473684210526315e-06,
116
+ "loss": 0.4829,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.0,
121
+ "learning_rate": 1e-05,
122
+ "loss": 0.4507,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.0,
127
+ "learning_rate": 1.0526315789473684e-05,
128
+ "loss": 0.4224,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.0,
133
+ "learning_rate": 1.105263157894737e-05,
134
+ "loss": 0.5246,
135
+ "step": 21
136
+ },
137
+ {
138
+ "epoch": 0.0,
139
+ "learning_rate": 1.1578947368421053e-05,
140
+ "loss": 0.466,
141
+ "step": 22
142
+ },
143
+ {
144
+ "epoch": 0.0,
145
+ "learning_rate": 1.2105263157894737e-05,
146
+ "loss": 0.4438,
147
+ "step": 23
148
+ },
149
+ {
150
+ "epoch": 0.01,
151
+ "learning_rate": 1.263157894736842e-05,
152
+ "loss": 0.3526,
153
+ "step": 24
154
+ },
155
+ {
156
+ "epoch": 0.01,
157
+ "learning_rate": 1.3157894736842108e-05,
158
+ "loss": 0.4781,
159
+ "step": 25
160
+ },
161
+ {
162
+ "epoch": 0.01,
163
+ "learning_rate": 1.3684210526315791e-05,
164
+ "loss": 0.352,
165
+ "step": 26
166
+ },
167
+ {
168
+ "epoch": 0.01,
169
+ "learning_rate": 1.4210526315789475e-05,
170
+ "loss": 0.4221,
171
+ "step": 27
172
+ },
173
+ {
174
+ "epoch": 0.01,
175
+ "learning_rate": 1.4736842105263159e-05,
176
+ "loss": 0.4257,
177
+ "step": 28
178
+ },
179
+ {
180
+ "epoch": 0.01,
181
+ "learning_rate": 1.5263157894736846e-05,
182
+ "loss": 0.5535,
183
+ "step": 29
184
+ },
185
+ {
186
+ "epoch": 0.01,
187
+ "learning_rate": 1.578947368421053e-05,
188
+ "loss": 0.4652,
189
+ "step": 30
190
+ },
191
+ {
192
+ "epoch": 0.01,
193
+ "learning_rate": 1.6315789473684213e-05,
194
+ "loss": 0.4701,
195
+ "step": 31
196
+ },
197
+ {
198
+ "epoch": 0.01,
199
+ "learning_rate": 1.6842105263157896e-05,
200
+ "loss": 0.5148,
201
+ "step": 32
202
+ },
203
+ {
204
+ "epoch": 0.01,
205
+ "learning_rate": 1.736842105263158e-05,
206
+ "loss": 0.4468,
207
+ "step": 33
208
+ },
209
+ {
210
+ "epoch": 0.01,
211
+ "learning_rate": 1.7894736842105264e-05,
212
+ "loss": 0.4368,
213
+ "step": 34
214
+ },
215
+ {
216
+ "epoch": 0.01,
217
+ "learning_rate": 1.8421052631578947e-05,
218
+ "loss": 0.3875,
219
+ "step": 35
220
+ },
221
+ {
222
+ "epoch": 0.01,
223
+ "learning_rate": 1.894736842105263e-05,
224
+ "loss": 0.3662,
225
+ "step": 36
226
+ },
227
+ {
228
+ "epoch": 0.01,
229
+ "learning_rate": 1.9473684210526318e-05,
230
+ "loss": 0.4751,
231
+ "step": 37
232
+ },
233
+ {
234
+ "epoch": 0.01,
235
+ "learning_rate": 2e-05,
236
+ "loss": 0.4075,
237
+ "step": 38
238
+ },
239
+ {
240
+ "epoch": 0.01,
241
+ "learning_rate": 1.998353909465021e-05,
242
+ "loss": 0.5052,
243
+ "step": 39
244
+ },
245
+ {
246
+ "epoch": 0.01,
247
+ "learning_rate": 1.9967078189300412e-05,
248
+ "loss": 0.4403,
249
+ "step": 40
250
+ },
251
+ {
252
+ "epoch": 0.01,
253
+ "learning_rate": 1.995061728395062e-05,
254
+ "loss": 0.4207,
255
+ "step": 41
256
+ },
257
+ {
258
+ "epoch": 0.01,
259
+ "learning_rate": 1.9934156378600823e-05,
260
+ "loss": 0.5641,
261
+ "step": 42
262
+ },
263
+ {
264
+ "epoch": 0.01,
265
+ "learning_rate": 1.991769547325103e-05,
266
+ "loss": 0.378,
267
+ "step": 43
268
+ },
269
+ {
270
+ "epoch": 0.01,
271
+ "learning_rate": 1.9901234567901237e-05,
272
+ "loss": 0.557,
273
+ "step": 44
274
+ },
275
+ {
276
+ "epoch": 0.01,
277
+ "learning_rate": 1.988477366255144e-05,
278
+ "loss": 0.4744,
279
+ "step": 45
280
+ },
281
+ {
282
+ "epoch": 0.01,
283
+ "learning_rate": 1.9868312757201648e-05,
284
+ "loss": 0.6027,
285
+ "step": 46
286
+ },
287
+ {
288
+ "epoch": 0.01,
289
+ "learning_rate": 1.9851851851851855e-05,
290
+ "loss": 0.4544,
291
+ "step": 47
292
+ },
293
+ {
294
+ "epoch": 0.01,
295
+ "learning_rate": 1.983539094650206e-05,
296
+ "loss": 0.4577,
297
+ "step": 48
298
+ },
299
+ {
300
+ "epoch": 0.01,
301
+ "learning_rate": 1.9818930041152265e-05,
302
+ "loss": 0.4558,
303
+ "step": 49
304
+ },
305
+ {
306
+ "epoch": 0.01,
307
+ "learning_rate": 1.9802469135802472e-05,
308
+ "loss": 0.354,
309
+ "step": 50
310
+ },
311
+ {
312
+ "epoch": 0.01,
313
+ "learning_rate": 1.9786008230452676e-05,
314
+ "loss": 0.3544,
315
+ "step": 51
316
+ },
317
+ {
318
+ "epoch": 0.01,
319
+ "learning_rate": 1.9769547325102883e-05,
320
+ "loss": 0.3501,
321
+ "step": 52
322
+ },
323
+ {
324
+ "epoch": 0.01,
325
+ "learning_rate": 1.9753086419753087e-05,
326
+ "loss": 0.4362,
327
+ "step": 53
328
+ },
329
+ {
330
+ "epoch": 0.01,
331
+ "learning_rate": 1.9736625514403294e-05,
332
+ "loss": 0.519,
333
+ "step": 54
334
+ },
335
+ {
336
+ "epoch": 0.01,
337
+ "learning_rate": 1.97201646090535e-05,
338
+ "loss": 0.4156,
339
+ "step": 55
340
+ },
341
+ {
342
+ "epoch": 0.01,
343
+ "learning_rate": 1.9703703703703704e-05,
344
+ "loss": 0.4067,
345
+ "step": 56
346
+ },
347
+ {
348
+ "epoch": 0.01,
349
+ "learning_rate": 1.968724279835391e-05,
350
+ "loss": 0.3812,
351
+ "step": 57
352
+ },
353
+ {
354
+ "epoch": 0.01,
355
+ "learning_rate": 1.9670781893004115e-05,
356
+ "loss": 0.458,
357
+ "step": 58
358
+ },
359
+ {
360
+ "epoch": 0.01,
361
+ "learning_rate": 1.9654320987654322e-05,
362
+ "loss": 0.435,
363
+ "step": 59
364
+ },
365
+ {
366
+ "epoch": 0.01,
367
+ "learning_rate": 1.963786008230453e-05,
368
+ "loss": 0.4768,
369
+ "step": 60
370
+ },
371
+ {
372
+ "epoch": 0.01,
373
+ "learning_rate": 1.9621399176954733e-05,
374
+ "loss": 0.3556,
375
+ "step": 61
376
+ },
377
+ {
378
+ "epoch": 0.01,
379
+ "learning_rate": 1.960493827160494e-05,
380
+ "loss": 0.3906,
381
+ "step": 62
382
+ },
383
+ {
384
+ "epoch": 0.01,
385
+ "learning_rate": 1.9588477366255147e-05,
386
+ "loss": 0.4025,
387
+ "step": 63
388
+ },
389
+ {
390
+ "epoch": 0.01,
391
+ "learning_rate": 1.9572016460905354e-05,
392
+ "loss": 0.4406,
393
+ "step": 64
394
+ },
395
+ {
396
+ "epoch": 0.01,
397
+ "learning_rate": 1.9555555555555557e-05,
398
+ "loss": 0.4496,
399
+ "step": 65
400
+ },
401
+ {
402
+ "epoch": 0.01,
403
+ "learning_rate": 1.9539094650205764e-05,
404
+ "loss": 0.3887,
405
+ "step": 66
406
+ },
407
+ {
408
+ "epoch": 0.01,
409
+ "learning_rate": 1.9522633744855968e-05,
410
+ "loss": 0.5111,
411
+ "step": 67
412
+ },
413
+ {
414
+ "epoch": 0.01,
415
+ "learning_rate": 1.9506172839506175e-05,
416
+ "loss": 0.4888,
417
+ "step": 68
418
+ },
419
+ {
420
+ "epoch": 0.01,
421
+ "learning_rate": 1.9489711934156382e-05,
422
+ "loss": 0.5124,
423
+ "step": 69
424
+ },
425
+ {
426
+ "epoch": 0.01,
427
+ "learning_rate": 1.9473251028806586e-05,
428
+ "loss": 0.3195,
429
+ "step": 70
430
+ },
431
+ {
432
+ "epoch": 0.01,
433
+ "learning_rate": 1.9456790123456793e-05,
434
+ "loss": 0.4139,
435
+ "step": 71
436
+ },
437
+ {
438
+ "epoch": 0.02,
439
+ "learning_rate": 1.9440329218106996e-05,
440
+ "loss": 0.4456,
441
+ "step": 72
442
+ },
443
+ {
444
+ "epoch": 0.02,
445
+ "learning_rate": 1.9423868312757203e-05,
446
+ "loss": 0.5004,
447
+ "step": 73
448
+ },
449
+ {
450
+ "epoch": 0.02,
451
+ "learning_rate": 1.9407407407407407e-05,
452
+ "loss": 0.5038,
453
+ "step": 74
454
+ },
455
+ {
456
+ "epoch": 0.02,
457
+ "learning_rate": 1.9390946502057614e-05,
458
+ "loss": 0.4608,
459
+ "step": 75
460
+ },
461
+ {
462
+ "epoch": 0.02,
463
+ "learning_rate": 1.937448559670782e-05,
464
+ "loss": 0.3851,
465
+ "step": 76
466
+ },
467
+ {
468
+ "epoch": 0.02,
469
+ "learning_rate": 1.9358024691358025e-05,
470
+ "loss": 0.5264,
471
+ "step": 77
472
+ },
473
+ {
474
+ "epoch": 0.02,
475
+ "learning_rate": 1.934156378600823e-05,
476
+ "loss": 0.5149,
477
+ "step": 78
478
+ },
479
+ {
480
+ "epoch": 0.02,
481
+ "learning_rate": 1.932510288065844e-05,
482
+ "loss": 0.5098,
483
+ "step": 79
484
+ },
485
+ {
486
+ "epoch": 0.02,
487
+ "learning_rate": 1.9308641975308646e-05,
488
+ "loss": 0.4661,
489
+ "step": 80
490
+ },
491
+ {
492
+ "epoch": 0.02,
493
+ "learning_rate": 1.929218106995885e-05,
494
+ "loss": 0.4772,
495
+ "step": 81
496
+ },
497
+ {
498
+ "epoch": 0.02,
499
+ "learning_rate": 1.9275720164609056e-05,
500
+ "loss": 0.4011,
501
+ "step": 82
502
+ },
503
+ {
504
+ "epoch": 0.02,
505
+ "learning_rate": 1.925925925925926e-05,
506
+ "loss": 0.4493,
507
+ "step": 83
508
+ },
509
+ {
510
+ "epoch": 0.02,
511
+ "learning_rate": 1.9242798353909467e-05,
512
+ "loss": 0.4032,
513
+ "step": 84
514
+ },
515
+ {
516
+ "epoch": 0.02,
517
+ "learning_rate": 1.9226337448559674e-05,
518
+ "loss": 0.3873,
519
+ "step": 85
520
+ },
521
+ {
522
+ "epoch": 0.02,
523
+ "learning_rate": 1.9209876543209878e-05,
524
+ "loss": 0.3721,
525
+ "step": 86
526
+ },
527
+ {
528
+ "epoch": 0.02,
529
+ "learning_rate": 1.9193415637860085e-05,
530
+ "loss": 0.3647,
531
+ "step": 87
532
+ },
533
+ {
534
+ "epoch": 0.02,
535
+ "learning_rate": 1.9176954732510288e-05,
536
+ "loss": 0.4311,
537
+ "step": 88
538
+ },
539
+ {
540
+ "epoch": 0.02,
541
+ "learning_rate": 1.9160493827160495e-05,
542
+ "loss": 0.543,
543
+ "step": 89
544
+ },
545
+ {
546
+ "epoch": 0.02,
547
+ "learning_rate": 1.91440329218107e-05,
548
+ "loss": 0.4659,
549
+ "step": 90
550
+ },
551
+ {
552
+ "epoch": 0.02,
553
+ "learning_rate": 1.9127572016460906e-05,
554
+ "loss": 0.3962,
555
+ "step": 91
556
+ },
557
+ {
558
+ "epoch": 0.02,
559
+ "learning_rate": 1.9111111111111113e-05,
560
+ "loss": 0.4703,
561
+ "step": 92
562
+ },
563
+ {
564
+ "epoch": 0.02,
565
+ "learning_rate": 1.9094650205761317e-05,
566
+ "loss": 0.4232,
567
+ "step": 93
568
+ },
569
+ {
570
+ "epoch": 0.02,
571
+ "learning_rate": 1.9078189300411524e-05,
572
+ "loss": 0.4152,
573
+ "step": 94
574
+ },
575
+ {
576
+ "epoch": 0.02,
577
+ "learning_rate": 1.906172839506173e-05,
578
+ "loss": 0.4784,
579
+ "step": 95
580
+ },
581
+ {
582
+ "epoch": 0.02,
583
+ "learning_rate": 1.9045267489711938e-05,
584
+ "loss": 0.364,
585
+ "step": 96
586
+ },
587
+ {
588
+ "epoch": 0.02,
589
+ "learning_rate": 1.902880658436214e-05,
590
+ "loss": 0.4181,
591
+ "step": 97
592
+ },
593
+ {
594
+ "epoch": 0.02,
595
+ "learning_rate": 1.901234567901235e-05,
596
+ "loss": 0.3675,
597
+ "step": 98
598
+ },
599
+ {
600
+ "epoch": 0.02,
601
+ "learning_rate": 1.8995884773662555e-05,
602
+ "loss": 0.5369,
603
+ "step": 99
604
+ },
605
+ {
606
+ "epoch": 0.02,
607
+ "learning_rate": 1.897942386831276e-05,
608
+ "loss": 0.3382,
609
+ "step": 100
610
+ },
611
+ {
612
+ "epoch": 0.02,
613
+ "learning_rate": 1.8962962962962966e-05,
614
+ "loss": 0.439,
615
+ "step": 101
616
+ },
617
+ {
618
+ "epoch": 0.02,
619
+ "learning_rate": 1.894650205761317e-05,
620
+ "loss": 0.3888,
621
+ "step": 102
622
+ },
623
+ {
624
+ "epoch": 0.02,
625
+ "learning_rate": 1.8930041152263377e-05,
626
+ "loss": 0.4476,
627
+ "step": 103
628
+ },
629
+ {
630
+ "epoch": 0.02,
631
+ "learning_rate": 1.891358024691358e-05,
632
+ "loss": 0.4458,
633
+ "step": 104
634
+ },
635
+ {
636
+ "epoch": 0.02,
637
+ "learning_rate": 1.8897119341563787e-05,
638
+ "loss": 0.3931,
639
+ "step": 105
640
+ },
641
+ {
642
+ "epoch": 0.02,
643
+ "learning_rate": 1.8880658436213994e-05,
644
+ "loss": 0.3731,
645
+ "step": 106
646
+ },
647
+ {
648
+ "epoch": 0.02,
649
+ "learning_rate": 1.8864197530864198e-05,
650
+ "loss": 0.4539,
651
+ "step": 107
652
+ },
653
+ {
654
+ "epoch": 0.02,
655
+ "learning_rate": 1.8847736625514405e-05,
656
+ "loss": 0.3523,
657
+ "step": 108
658
+ },
659
+ {
660
+ "epoch": 0.02,
661
+ "learning_rate": 1.883127572016461e-05,
662
+ "loss": 0.4001,
663
+ "step": 109
664
+ },
665
+ {
666
+ "epoch": 0.02,
667
+ "learning_rate": 1.8814814814814816e-05,
668
+ "loss": 0.2399,
669
+ "step": 110
670
+ },
671
+ {
672
+ "epoch": 0.02,
673
+ "learning_rate": 1.8798353909465023e-05,
674
+ "loss": 0.5446,
675
+ "step": 111
676
+ },
677
+ {
678
+ "epoch": 0.02,
679
+ "learning_rate": 1.878189300411523e-05,
680
+ "loss": 0.4942,
681
+ "step": 112
682
+ },
683
+ {
684
+ "epoch": 0.02,
685
+ "learning_rate": 1.8765432098765433e-05,
686
+ "loss": 0.3957,
687
+ "step": 113
688
+ },
689
+ {
690
+ "epoch": 0.02,
691
+ "learning_rate": 1.874897119341564e-05,
692
+ "loss": 0.4916,
693
+ "step": 114
694
+ },
695
+ {
696
+ "epoch": 0.02,
697
+ "learning_rate": 1.8732510288065847e-05,
698
+ "loss": 0.3397,
699
+ "step": 115
700
+ },
701
+ {
702
+ "epoch": 0.02,
703
+ "learning_rate": 1.871604938271605e-05,
704
+ "loss": 0.4224,
705
+ "step": 116
706
+ },
707
+ {
708
+ "epoch": 0.02,
709
+ "learning_rate": 1.8699588477366258e-05,
710
+ "loss": 0.3766,
711
+ "step": 117
712
+ },
713
+ {
714
+ "epoch": 0.02,
715
+ "learning_rate": 1.868312757201646e-05,
716
+ "loss": 0.4517,
717
+ "step": 118
718
+ },
719
+ {
720
+ "epoch": 0.02,
721
+ "learning_rate": 1.866666666666667e-05,
722
+ "loss": 0.4887,
723
+ "step": 119
724
+ },
725
+ {
726
+ "epoch": 0.03,
727
+ "learning_rate": 1.8650205761316872e-05,
728
+ "loss": 0.3778,
729
+ "step": 120
730
+ },
731
+ {
732
+ "epoch": 0.03,
733
+ "learning_rate": 1.863374485596708e-05,
734
+ "loss": 0.3754,
735
+ "step": 121
736
+ },
737
+ {
738
+ "epoch": 0.03,
739
+ "learning_rate": 1.8617283950617286e-05,
740
+ "loss": 0.3367,
741
+ "step": 122
742
+ },
743
+ {
744
+ "epoch": 0.03,
745
+ "learning_rate": 1.860082304526749e-05,
746
+ "loss": 0.4036,
747
+ "step": 123
748
+ },
749
+ {
750
+ "epoch": 0.03,
751
+ "learning_rate": 1.8584362139917697e-05,
752
+ "loss": 0.4064,
753
+ "step": 124
754
+ },
755
+ {
756
+ "epoch": 0.03,
757
+ "learning_rate": 1.85679012345679e-05,
758
+ "loss": 0.3674,
759
+ "step": 125
760
+ },
761
+ {
762
+ "epoch": 0.03,
763
+ "learning_rate": 1.8551440329218108e-05,
764
+ "loss": 0.4007,
765
+ "step": 126
766
+ },
767
+ {
768
+ "epoch": 0.03,
769
+ "learning_rate": 1.8534979423868315e-05,
770
+ "loss": 0.4,
771
+ "step": 127
772
+ },
773
+ {
774
+ "epoch": 0.03,
775
+ "learning_rate": 1.851851851851852e-05,
776
+ "loss": 0.3448,
777
+ "step": 128
778
+ },
779
+ {
780
+ "epoch": 0.03,
781
+ "learning_rate": 1.8502057613168725e-05,
782
+ "loss": 0.3736,
783
+ "step": 129
784
+ },
785
+ {
786
+ "epoch": 0.03,
787
+ "learning_rate": 1.8485596707818932e-05,
788
+ "loss": 0.4869,
789
+ "step": 130
790
+ },
791
+ {
792
+ "epoch": 0.03,
793
+ "learning_rate": 1.846913580246914e-05,
794
+ "loss": 0.4034,
795
+ "step": 131
796
+ },
797
+ {
798
+ "epoch": 0.03,
799
+ "learning_rate": 1.8452674897119343e-05,
800
+ "loss": 0.4811,
801
+ "step": 132
802
+ },
803
+ {
804
+ "epoch": 0.03,
805
+ "learning_rate": 1.843621399176955e-05,
806
+ "loss": 0.3918,
807
+ "step": 133
808
+ },
809
+ {
810
+ "epoch": 0.03,
811
+ "learning_rate": 1.8419753086419754e-05,
812
+ "loss": 0.4295,
813
+ "step": 134
814
+ },
815
+ {
816
+ "epoch": 0.03,
817
+ "learning_rate": 1.840329218106996e-05,
818
+ "loss": 0.4187,
819
+ "step": 135
820
+ },
821
+ {
822
+ "epoch": 0.03,
823
+ "learning_rate": 1.8386831275720164e-05,
824
+ "loss": 0.3605,
825
+ "step": 136
826
+ },
827
+ {
828
+ "epoch": 0.03,
829
+ "learning_rate": 1.837037037037037e-05,
830
+ "loss": 0.3696,
831
+ "step": 137
832
+ },
833
+ {
834
+ "epoch": 0.03,
835
+ "learning_rate": 1.835390946502058e-05,
836
+ "loss": 0.4477,
837
+ "step": 138
838
+ },
839
+ {
840
+ "epoch": 0.03,
841
+ "learning_rate": 1.8337448559670782e-05,
842
+ "loss": 0.5395,
843
+ "step": 139
844
+ },
845
+ {
846
+ "epoch": 0.03,
847
+ "learning_rate": 1.832098765432099e-05,
848
+ "loss": 0.4135,
849
+ "step": 140
850
+ },
851
+ {
852
+ "epoch": 0.03,
853
+ "learning_rate": 1.8304526748971193e-05,
854
+ "loss": 0.3599,
855
+ "step": 141
856
+ },
857
+ {
858
+ "epoch": 0.03,
859
+ "learning_rate": 1.82880658436214e-05,
860
+ "loss": 0.3749,
861
+ "step": 142
862
+ },
863
+ {
864
+ "epoch": 0.03,
865
+ "learning_rate": 1.8271604938271607e-05,
866
+ "loss": 0.307,
867
+ "step": 143
868
+ },
869
+ {
870
+ "epoch": 0.03,
871
+ "learning_rate": 1.8255144032921814e-05,
872
+ "loss": 0.3178,
873
+ "step": 144
874
+ },
875
+ {
876
+ "epoch": 0.03,
877
+ "learning_rate": 1.8238683127572017e-05,
878
+ "loss": 0.3468,
879
+ "step": 145
880
+ },
881
+ {
882
+ "epoch": 0.03,
883
+ "learning_rate": 1.8222222222222224e-05,
884
+ "loss": 0.3562,
885
+ "step": 146
886
+ },
887
+ {
888
+ "epoch": 0.03,
889
+ "learning_rate": 1.820576131687243e-05,
890
+ "loss": 0.3996,
891
+ "step": 147
892
+ },
893
+ {
894
+ "epoch": 0.03,
895
+ "learning_rate": 1.8189300411522635e-05,
896
+ "loss": 0.3896,
897
+ "step": 148
898
+ },
899
+ {
900
+ "epoch": 0.03,
901
+ "learning_rate": 1.8172839506172842e-05,
902
+ "loss": 0.3735,
903
+ "step": 149
904
+ },
905
+ {
906
+ "epoch": 0.03,
907
+ "learning_rate": 1.8156378600823046e-05,
908
+ "loss": 0.3152,
909
+ "step": 150
910
+ },
911
+ {
912
+ "epoch": 0.03,
913
+ "learning_rate": 1.8139917695473253e-05,
914
+ "loss": 0.4473,
915
+ "step": 151
916
+ },
917
+ {
918
+ "epoch": 0.03,
919
+ "learning_rate": 1.812345679012346e-05,
920
+ "loss": 0.4143,
921
+ "step": 152
922
+ },
923
+ {
924
+ "epoch": 0.03,
925
+ "learning_rate": 1.8106995884773663e-05,
926
+ "loss": 0.3974,
927
+ "step": 153
928
+ },
929
+ {
930
+ "epoch": 0.03,
931
+ "learning_rate": 1.809053497942387e-05,
932
+ "loss": 0.5082,
933
+ "step": 154
934
+ },
935
+ {
936
+ "epoch": 0.03,
937
+ "learning_rate": 1.8074074074074074e-05,
938
+ "loss": 0.3978,
939
+ "step": 155
940
+ },
941
+ {
942
+ "epoch": 0.03,
943
+ "learning_rate": 1.805761316872428e-05,
944
+ "loss": 0.4421,
945
+ "step": 156
946
+ },
947
+ {
948
+ "epoch": 0.03,
949
+ "learning_rate": 1.8041152263374485e-05,
950
+ "loss": 0.4486,
951
+ "step": 157
952
+ },
953
+ {
954
+ "epoch": 0.03,
955
+ "learning_rate": 1.802469135802469e-05,
956
+ "loss": 0.3171,
957
+ "step": 158
958
+ },
959
+ {
960
+ "epoch": 0.03,
961
+ "learning_rate": 1.80082304526749e-05,
962
+ "loss": 0.4101,
963
+ "step": 159
964
+ },
965
+ {
966
+ "epoch": 0.03,
967
+ "learning_rate": 1.7991769547325106e-05,
968
+ "loss": 0.3966,
969
+ "step": 160
970
+ },
971
+ {
972
+ "epoch": 0.03,
973
+ "learning_rate": 1.797530864197531e-05,
974
+ "loss": 0.4174,
975
+ "step": 161
976
+ },
977
+ {
978
+ "epoch": 0.03,
979
+ "learning_rate": 1.7958847736625516e-05,
980
+ "loss": 0.3673,
981
+ "step": 162
982
+ },
983
+ {
984
+ "epoch": 0.03,
985
+ "learning_rate": 1.7942386831275723e-05,
986
+ "loss": 0.4502,
987
+ "step": 163
988
+ },
989
+ {
990
+ "epoch": 0.03,
991
+ "learning_rate": 1.7925925925925927e-05,
992
+ "loss": 0.44,
993
+ "step": 164
994
+ },
995
+ {
996
+ "epoch": 0.03,
997
+ "learning_rate": 1.7909465020576134e-05,
998
+ "loss": 0.349,
999
+ "step": 165
1000
+ },
1001
+ {
1002
+ "epoch": 0.03,
1003
+ "learning_rate": 1.7893004115226338e-05,
1004
+ "loss": 0.3959,
1005
+ "step": 166
1006
+ },
1007
+ {
1008
+ "epoch": 0.03,
1009
+ "learning_rate": 1.7876543209876545e-05,
1010
+ "loss": 0.3136,
1011
+ "step": 167
1012
+ },
1013
+ {
1014
+ "epoch": 0.04,
1015
+ "learning_rate": 1.786008230452675e-05,
1016
+ "loss": 0.402,
1017
+ "step": 168
1018
+ },
1019
+ {
1020
+ "epoch": 0.04,
1021
+ "learning_rate": 1.7843621399176955e-05,
1022
+ "loss": 0.3123,
1023
+ "step": 169
1024
+ },
1025
+ {
1026
+ "epoch": 0.04,
1027
+ "learning_rate": 1.7827160493827162e-05,
1028
+ "loss": 0.4913,
1029
+ "step": 170
1030
+ },
1031
+ {
1032
+ "epoch": 0.04,
1033
+ "learning_rate": 1.7810699588477366e-05,
1034
+ "loss": 0.3062,
1035
+ "step": 171
1036
+ },
1037
+ {
1038
+ "epoch": 0.04,
1039
+ "learning_rate": 1.7794238683127573e-05,
1040
+ "loss": 0.3527,
1041
+ "step": 172
1042
+ },
1043
+ {
1044
+ "epoch": 0.04,
1045
+ "learning_rate": 1.7777777777777777e-05,
1046
+ "loss": 0.4381,
1047
+ "step": 173
1048
+ },
1049
+ {
1050
+ "epoch": 0.04,
1051
+ "learning_rate": 1.7761316872427984e-05,
1052
+ "loss": 0.4016,
1053
+ "step": 174
1054
+ },
1055
+ {
1056
+ "epoch": 0.04,
1057
+ "learning_rate": 1.774485596707819e-05,
1058
+ "loss": 0.4025,
1059
+ "step": 175
1060
+ },
1061
+ {
1062
+ "epoch": 0.04,
1063
+ "learning_rate": 1.7728395061728398e-05,
1064
+ "loss": 0.3111,
1065
+ "step": 176
1066
+ },
1067
+ {
1068
+ "epoch": 0.04,
1069
+ "learning_rate": 1.77119341563786e-05,
1070
+ "loss": 0.3972,
1071
+ "step": 177
1072
+ },
1073
+ {
1074
+ "epoch": 0.04,
1075
+ "learning_rate": 1.769547325102881e-05,
1076
+ "loss": 0.3924,
1077
+ "step": 178
1078
+ },
1079
+ {
1080
+ "epoch": 0.04,
1081
+ "learning_rate": 1.7679012345679015e-05,
1082
+ "loss": 0.3827,
1083
+ "step": 179
1084
+ },
1085
+ {
1086
+ "epoch": 0.04,
1087
+ "learning_rate": 1.766255144032922e-05,
1088
+ "loss": 0.3921,
1089
+ "step": 180
1090
+ },
1091
+ {
1092
+ "epoch": 0.04,
1093
+ "learning_rate": 1.7646090534979426e-05,
1094
+ "loss": 0.4678,
1095
+ "step": 181
1096
+ },
1097
+ {
1098
+ "epoch": 0.04,
1099
+ "learning_rate": 1.7629629629629633e-05,
1100
+ "loss": 0.4044,
1101
+ "step": 182
1102
+ },
1103
+ {
1104
+ "epoch": 0.04,
1105
+ "learning_rate": 1.7613168724279837e-05,
1106
+ "loss": 0.4464,
1107
+ "step": 183
1108
+ },
1109
+ {
1110
+ "epoch": 0.04,
1111
+ "learning_rate": 1.7596707818930044e-05,
1112
+ "loss": 0.4067,
1113
+ "step": 184
1114
+ },
1115
+ {
1116
+ "epoch": 0.04,
1117
+ "learning_rate": 1.7580246913580247e-05,
1118
+ "loss": 0.4357,
1119
+ "step": 185
1120
+ },
1121
+ {
1122
+ "epoch": 0.04,
1123
+ "learning_rate": 1.7563786008230454e-05,
1124
+ "loss": 0.4399,
1125
+ "step": 186
1126
+ },
1127
+ {
1128
+ "epoch": 0.04,
1129
+ "learning_rate": 1.7547325102880658e-05,
1130
+ "loss": 0.4798,
1131
+ "step": 187
1132
+ },
1133
+ {
1134
+ "epoch": 0.04,
1135
+ "learning_rate": 1.7530864197530865e-05,
1136
+ "loss": 0.4868,
1137
+ "step": 188
1138
+ },
1139
+ {
1140
+ "epoch": 0.04,
1141
+ "learning_rate": 1.7514403292181072e-05,
1142
+ "loss": 0.4026,
1143
+ "step": 189
1144
+ },
1145
+ {
1146
+ "epoch": 0.04,
1147
+ "learning_rate": 1.7497942386831276e-05,
1148
+ "loss": 0.3769,
1149
+ "step": 190
1150
+ },
1151
+ {
1152
+ "epoch": 0.04,
1153
+ "learning_rate": 1.7481481481481483e-05,
1154
+ "loss": 0.3638,
1155
+ "step": 191
1156
+ },
1157
+ {
1158
+ "epoch": 0.04,
1159
+ "learning_rate": 1.746502057613169e-05,
1160
+ "loss": 0.3285,
1161
+ "step": 192
1162
+ },
1163
+ {
1164
+ "epoch": 0.04,
1165
+ "learning_rate": 1.7448559670781893e-05,
1166
+ "loss": 0.4887,
1167
+ "step": 193
1168
+ },
1169
+ {
1170
+ "epoch": 0.04,
1171
+ "learning_rate": 1.74320987654321e-05,
1172
+ "loss": 0.3189,
1173
+ "step": 194
1174
+ },
1175
+ {
1176
+ "epoch": 0.04,
1177
+ "learning_rate": 1.7415637860082307e-05,
1178
+ "loss": 0.4173,
1179
+ "step": 195
1180
+ },
1181
+ {
1182
+ "epoch": 0.04,
1183
+ "learning_rate": 1.739917695473251e-05,
1184
+ "loss": 0.3422,
1185
+ "step": 196
1186
+ },
1187
+ {
1188
+ "epoch": 0.04,
1189
+ "learning_rate": 1.7382716049382718e-05,
1190
+ "loss": 0.3464,
1191
+ "step": 197
1192
+ },
1193
+ {
1194
+ "epoch": 0.04,
1195
+ "learning_rate": 1.7366255144032925e-05,
1196
+ "loss": 0.3964,
1197
+ "step": 198
1198
+ },
1199
+ {
1200
+ "epoch": 0.04,
1201
+ "learning_rate": 1.734979423868313e-05,
1202
+ "loss": 0.3926,
1203
+ "step": 199
1204
+ },
1205
+ {
1206
+ "epoch": 0.04,
1207
+ "learning_rate": 1.7333333333333336e-05,
1208
+ "loss": 0.4175,
1209
+ "step": 200
1210
+ },
1211
+ {
1212
+ "epoch": 0.04,
1213
+ "learning_rate": 1.731687242798354e-05,
1214
+ "loss": 0.4374,
1215
+ "step": 201
1216
+ },
1217
+ {
1218
+ "epoch": 0.04,
1219
+ "learning_rate": 1.7300411522633746e-05,
1220
+ "loss": 0.4897,
1221
+ "step": 202
1222
+ },
1223
+ {
1224
+ "epoch": 0.04,
1225
+ "learning_rate": 1.728395061728395e-05,
1226
+ "loss": 0.3859,
1227
+ "step": 203
1228
+ },
1229
+ {
1230
+ "epoch": 0.04,
1231
+ "learning_rate": 1.7267489711934157e-05,
1232
+ "loss": 0.3788,
1233
+ "step": 204
1234
+ },
1235
+ {
1236
+ "epoch": 0.04,
1237
+ "learning_rate": 1.7251028806584364e-05,
1238
+ "loss": 0.3544,
1239
+ "step": 205
1240
+ },
1241
+ {
1242
+ "epoch": 0.04,
1243
+ "learning_rate": 1.7234567901234568e-05,
1244
+ "loss": 0.3685,
1245
+ "step": 206
1246
+ },
1247
+ {
1248
+ "epoch": 0.04,
1249
+ "learning_rate": 1.7218106995884775e-05,
1250
+ "loss": 0.3924,
1251
+ "step": 207
1252
+ },
1253
+ {
1254
+ "epoch": 0.04,
1255
+ "learning_rate": 1.720164609053498e-05,
1256
+ "loss": 0.3965,
1257
+ "step": 208
1258
+ },
1259
+ {
1260
+ "epoch": 0.04,
1261
+ "learning_rate": 1.7185185185185185e-05,
1262
+ "loss": 0.3659,
1263
+ "step": 209
1264
+ },
1265
+ {
1266
+ "epoch": 0.04,
1267
+ "learning_rate": 1.7168724279835392e-05,
1268
+ "loss": 0.3092,
1269
+ "step": 210
1270
+ },
1271
+ {
1272
+ "epoch": 0.04,
1273
+ "learning_rate": 1.71522633744856e-05,
1274
+ "loss": 0.4032,
1275
+ "step": 211
1276
+ },
1277
+ {
1278
+ "epoch": 0.04,
1279
+ "learning_rate": 1.7135802469135806e-05,
1280
+ "loss": 0.4603,
1281
+ "step": 212
1282
+ },
1283
+ {
1284
+ "epoch": 0.04,
1285
+ "learning_rate": 1.711934156378601e-05,
1286
+ "loss": 0.3343,
1287
+ "step": 213
1288
+ },
1289
+ {
1290
+ "epoch": 0.04,
1291
+ "learning_rate": 1.7102880658436217e-05,
1292
+ "loss": 0.4551,
1293
+ "step": 214
1294
+ },
1295
+ {
1296
+ "epoch": 0.04,
1297
+ "learning_rate": 1.708641975308642e-05,
1298
+ "loss": 0.3632,
1299
+ "step": 215
1300
+ },
1301
+ {
1302
+ "epoch": 0.05,
1303
+ "learning_rate": 1.7069958847736628e-05,
1304
+ "loss": 0.4198,
1305
+ "step": 216
1306
+ },
1307
+ {
1308
+ "epoch": 0.05,
1309
+ "learning_rate": 1.705349794238683e-05,
1310
+ "loss": 0.3267,
1311
+ "step": 217
1312
+ },
1313
+ {
1314
+ "epoch": 0.05,
1315
+ "learning_rate": 1.7037037037037038e-05,
1316
+ "loss": 0.3981,
1317
+ "step": 218
1318
+ },
1319
+ {
1320
+ "epoch": 0.05,
1321
+ "learning_rate": 1.7020576131687242e-05,
1322
+ "loss": 0.3944,
1323
+ "step": 219
1324
+ },
1325
+ {
1326
+ "epoch": 0.05,
1327
+ "learning_rate": 1.700411522633745e-05,
1328
+ "loss": 0.4355,
1329
+ "step": 220
1330
+ },
1331
+ {
1332
+ "epoch": 0.05,
1333
+ "learning_rate": 1.6987654320987656e-05,
1334
+ "loss": 0.4652,
1335
+ "step": 221
1336
+ },
1337
+ {
1338
+ "epoch": 0.05,
1339
+ "learning_rate": 1.697119341563786e-05,
1340
+ "loss": 0.3894,
1341
+ "step": 222
1342
+ },
1343
+ {
1344
+ "epoch": 0.05,
1345
+ "learning_rate": 1.6954732510288067e-05,
1346
+ "loss": 0.4541,
1347
+ "step": 223
1348
+ },
1349
+ {
1350
+ "epoch": 0.05,
1351
+ "learning_rate": 1.6938271604938274e-05,
1352
+ "loss": 0.6168,
1353
+ "step": 224
1354
+ },
1355
+ {
1356
+ "epoch": 0.05,
1357
+ "learning_rate": 1.692181069958848e-05,
1358
+ "loss": 0.355,
1359
+ "step": 225
1360
+ },
1361
+ {
1362
+ "epoch": 0.05,
1363
+ "learning_rate": 1.6905349794238684e-05,
1364
+ "loss": 0.3723,
1365
+ "step": 226
1366
+ },
1367
+ {
1368
+ "epoch": 0.05,
1369
+ "learning_rate": 1.688888888888889e-05,
1370
+ "loss": 0.3276,
1371
+ "step": 227
1372
+ },
1373
+ {
1374
+ "epoch": 0.05,
1375
+ "learning_rate": 1.68724279835391e-05,
1376
+ "loss": 0.481,
1377
+ "step": 228
1378
+ },
1379
+ {
1380
+ "epoch": 0.05,
1381
+ "learning_rate": 1.6855967078189302e-05,
1382
+ "loss": 0.4071,
1383
+ "step": 229
1384
+ },
1385
+ {
1386
+ "epoch": 0.05,
1387
+ "learning_rate": 1.683950617283951e-05,
1388
+ "loss": 0.3346,
1389
+ "step": 230
1390
+ },
1391
+ {
1392
+ "epoch": 0.05,
1393
+ "learning_rate": 1.6823045267489713e-05,
1394
+ "loss": 0.4174,
1395
+ "step": 231
1396
+ },
1397
+ {
1398
+ "epoch": 0.05,
1399
+ "learning_rate": 1.680658436213992e-05,
1400
+ "loss": 0.4829,
1401
+ "step": 232
1402
+ },
1403
+ {
1404
+ "epoch": 0.05,
1405
+ "learning_rate": 1.6790123456790123e-05,
1406
+ "loss": 0.4351,
1407
+ "step": 233
1408
+ },
1409
+ {
1410
+ "epoch": 0.05,
1411
+ "learning_rate": 1.677366255144033e-05,
1412
+ "loss": 0.3687,
1413
+ "step": 234
1414
+ },
1415
+ {
1416
+ "epoch": 0.05,
1417
+ "learning_rate": 1.6757201646090537e-05,
1418
+ "loss": 0.4449,
1419
+ "step": 235
1420
+ },
1421
+ {
1422
+ "epoch": 0.05,
1423
+ "learning_rate": 1.674074074074074e-05,
1424
+ "loss": 0.3935,
1425
+ "step": 236
1426
+ },
1427
+ {
1428
+ "epoch": 0.05,
1429
+ "learning_rate": 1.6724279835390948e-05,
1430
+ "loss": 0.4641,
1431
+ "step": 237
1432
+ },
1433
+ {
1434
+ "epoch": 0.05,
1435
+ "learning_rate": 1.670781893004115e-05,
1436
+ "loss": 0.283,
1437
+ "step": 238
1438
+ },
1439
+ {
1440
+ "epoch": 0.05,
1441
+ "learning_rate": 1.669135802469136e-05,
1442
+ "loss": 0.3843,
1443
+ "step": 239
1444
+ },
1445
+ {
1446
+ "epoch": 0.05,
1447
+ "learning_rate": 1.6674897119341566e-05,
1448
+ "loss": 0.3586,
1449
+ "step": 240
1450
+ },
1451
+ {
1452
+ "epoch": 0.05,
1453
+ "learning_rate": 1.6658436213991773e-05,
1454
+ "loss": 0.4016,
1455
+ "step": 241
1456
+ },
1457
+ {
1458
+ "epoch": 0.05,
1459
+ "learning_rate": 1.6641975308641976e-05,
1460
+ "loss": 0.3483,
1461
+ "step": 242
1462
+ },
1463
+ {
1464
+ "epoch": 0.05,
1465
+ "learning_rate": 1.6625514403292183e-05,
1466
+ "loss": 0.3939,
1467
+ "step": 243
1468
+ },
1469
+ {
1470
+ "epoch": 0.05,
1471
+ "learning_rate": 1.660905349794239e-05,
1472
+ "loss": 0.4346,
1473
+ "step": 244
1474
+ },
1475
+ {
1476
+ "epoch": 0.05,
1477
+ "learning_rate": 1.6592592592592594e-05,
1478
+ "loss": 0.3228,
1479
+ "step": 245
1480
+ },
1481
+ {
1482
+ "epoch": 0.05,
1483
+ "learning_rate": 1.65761316872428e-05,
1484
+ "loss": 0.3403,
1485
+ "step": 246
1486
+ },
1487
+ {
1488
+ "epoch": 0.05,
1489
+ "learning_rate": 1.6559670781893005e-05,
1490
+ "loss": 0.4482,
1491
+ "step": 247
1492
+ },
1493
+ {
1494
+ "epoch": 0.05,
1495
+ "learning_rate": 1.654320987654321e-05,
1496
+ "loss": 0.3429,
1497
+ "step": 248
1498
+ },
1499
+ {
1500
+ "epoch": 0.05,
1501
+ "learning_rate": 1.6526748971193415e-05,
1502
+ "loss": 0.4401,
1503
+ "step": 249
1504
+ },
1505
+ {
1506
+ "epoch": 0.05,
1507
+ "learning_rate": 1.6510288065843622e-05,
1508
+ "loss": 0.3427,
1509
+ "step": 250
1510
+ },
1511
+ {
1512
+ "epoch": 0.05,
1513
+ "learning_rate": 1.649382716049383e-05,
1514
+ "loss": 0.3857,
1515
+ "step": 251
1516
+ },
1517
+ {
1518
+ "epoch": 0.05,
1519
+ "learning_rate": 1.6477366255144033e-05,
1520
+ "loss": 0.2981,
1521
+ "step": 252
1522
+ },
1523
+ {
1524
+ "epoch": 0.05,
1525
+ "learning_rate": 1.646090534979424e-05,
1526
+ "loss": 0.4754,
1527
+ "step": 253
1528
+ },
1529
+ {
1530
+ "epoch": 0.05,
1531
+ "learning_rate": 1.6444444444444444e-05,
1532
+ "loss": 0.3108,
1533
+ "step": 254
1534
+ },
1535
+ {
1536
+ "epoch": 0.05,
1537
+ "learning_rate": 1.642798353909465e-05,
1538
+ "loss": 0.4435,
1539
+ "step": 255
1540
+ },
1541
+ {
1542
+ "epoch": 0.05,
1543
+ "learning_rate": 1.6411522633744858e-05,
1544
+ "loss": 0.3834,
1545
+ "step": 256
1546
+ },
1547
+ {
1548
+ "epoch": 0.05,
1549
+ "learning_rate": 1.6395061728395065e-05,
1550
+ "loss": 0.4469,
1551
+ "step": 257
1552
+ },
1553
+ {
1554
+ "epoch": 0.05,
1555
+ "learning_rate": 1.6378600823045268e-05,
1556
+ "loss": 0.4308,
1557
+ "step": 258
1558
+ },
1559
+ {
1560
+ "epoch": 0.05,
1561
+ "learning_rate": 1.6362139917695475e-05,
1562
+ "loss": 0.4737,
1563
+ "step": 259
1564
+ },
1565
+ {
1566
+ "epoch": 0.05,
1567
+ "learning_rate": 1.6345679012345682e-05,
1568
+ "loss": 0.3781,
1569
+ "step": 260
1570
+ },
1571
+ {
1572
+ "epoch": 0.05,
1573
+ "learning_rate": 1.6329218106995886e-05,
1574
+ "loss": 0.3341,
1575
+ "step": 261
1576
+ },
1577
+ {
1578
+ "epoch": 0.05,
1579
+ "learning_rate": 1.6312757201646093e-05,
1580
+ "loss": 0.3724,
1581
+ "step": 262
1582
+ },
1583
+ {
1584
+ "epoch": 0.05,
1585
+ "learning_rate": 1.6296296296296297e-05,
1586
+ "loss": 0.4333,
1587
+ "step": 263
1588
+ },
1589
+ {
1590
+ "epoch": 0.06,
1591
+ "learning_rate": 1.6279835390946504e-05,
1592
+ "loss": 0.3296,
1593
+ "step": 264
1594
+ },
1595
+ {
1596
+ "epoch": 0.06,
1597
+ "learning_rate": 1.626337448559671e-05,
1598
+ "loss": 0.4457,
1599
+ "step": 265
1600
+ },
1601
+ {
1602
+ "epoch": 0.06,
1603
+ "learning_rate": 1.6246913580246914e-05,
1604
+ "loss": 0.3633,
1605
+ "step": 266
1606
+ },
1607
+ {
1608
+ "epoch": 0.06,
1609
+ "learning_rate": 1.623045267489712e-05,
1610
+ "loss": 0.3704,
1611
+ "step": 267
1612
+ },
1613
+ {
1614
+ "epoch": 0.06,
1615
+ "learning_rate": 1.6213991769547325e-05,
1616
+ "loss": 0.3162,
1617
+ "step": 268
1618
+ },
1619
+ {
1620
+ "epoch": 0.06,
1621
+ "learning_rate": 1.6197530864197532e-05,
1622
+ "loss": 0.3351,
1623
+ "step": 269
1624
+ },
1625
+ {
1626
+ "epoch": 0.06,
1627
+ "learning_rate": 1.6181069958847736e-05,
1628
+ "loss": 0.427,
1629
+ "step": 270
1630
+ },
1631
+ {
1632
+ "epoch": 0.06,
1633
+ "learning_rate": 1.6164609053497943e-05,
1634
+ "loss": 0.4071,
1635
+ "step": 271
1636
+ },
1637
+ {
1638
+ "epoch": 0.06,
1639
+ "learning_rate": 1.614814814814815e-05,
1640
+ "loss": 0.3493,
1641
+ "step": 272
1642
+ },
1643
+ {
1644
+ "epoch": 0.06,
1645
+ "learning_rate": 1.6131687242798357e-05,
1646
+ "loss": 0.3527,
1647
+ "step": 273
1648
+ },
1649
+ {
1650
+ "epoch": 0.06,
1651
+ "learning_rate": 1.611522633744856e-05,
1652
+ "loss": 0.4353,
1653
+ "step": 274
1654
+ },
1655
+ {
1656
+ "epoch": 0.06,
1657
+ "learning_rate": 1.6098765432098767e-05,
1658
+ "loss": 0.3331,
1659
+ "step": 275
1660
+ },
1661
+ {
1662
+ "epoch": 0.06,
1663
+ "learning_rate": 1.6082304526748974e-05,
1664
+ "loss": 0.4455,
1665
+ "step": 276
1666
+ },
1667
+ {
1668
+ "epoch": 0.06,
1669
+ "learning_rate": 1.6065843621399178e-05,
1670
+ "loss": 0.3604,
1671
+ "step": 277
1672
+ },
1673
+ {
1674
+ "epoch": 0.06,
1675
+ "learning_rate": 1.6049382716049385e-05,
1676
+ "loss": 0.3725,
1677
+ "step": 278
1678
+ },
1679
+ {
1680
+ "epoch": 0.06,
1681
+ "learning_rate": 1.603292181069959e-05,
1682
+ "loss": 0.3634,
1683
+ "step": 279
1684
+ },
1685
+ {
1686
+ "epoch": 0.06,
1687
+ "learning_rate": 1.6016460905349796e-05,
1688
+ "loss": 0.3471,
1689
+ "step": 280
1690
+ },
1691
+ {
1692
+ "epoch": 0.06,
1693
+ "learning_rate": 1.6000000000000003e-05,
1694
+ "loss": 0.4957,
1695
+ "step": 281
1696
+ },
1697
+ {
1698
+ "epoch": 0.06,
1699
+ "learning_rate": 1.5983539094650206e-05,
1700
+ "loss": 0.3329,
1701
+ "step": 282
1702
+ },
1703
+ {
1704
+ "epoch": 0.06,
1705
+ "learning_rate": 1.5967078189300413e-05,
1706
+ "loss": 0.3813,
1707
+ "step": 283
1708
+ },
1709
+ {
1710
+ "epoch": 0.06,
1711
+ "learning_rate": 1.5950617283950617e-05,
1712
+ "loss": 0.3336,
1713
+ "step": 284
1714
+ },
1715
+ {
1716
+ "epoch": 0.06,
1717
+ "learning_rate": 1.5934156378600824e-05,
1718
+ "loss": 0.3203,
1719
+ "step": 285
1720
+ },
1721
+ {
1722
+ "epoch": 0.06,
1723
+ "learning_rate": 1.5917695473251028e-05,
1724
+ "loss": 0.3797,
1725
+ "step": 286
1726
+ },
1727
+ {
1728
+ "epoch": 0.06,
1729
+ "learning_rate": 1.5901234567901235e-05,
1730
+ "loss": 0.3415,
1731
+ "step": 287
1732
+ },
1733
+ {
1734
+ "epoch": 0.06,
1735
+ "learning_rate": 1.588477366255144e-05,
1736
+ "loss": 0.4192,
1737
+ "step": 288
1738
+ },
1739
+ {
1740
+ "epoch": 0.06,
1741
+ "learning_rate": 1.586831275720165e-05,
1742
+ "loss": 0.4433,
1743
+ "step": 289
1744
+ },
1745
+ {
1746
+ "epoch": 0.06,
1747
+ "learning_rate": 1.5851851851851852e-05,
1748
+ "loss": 0.3177,
1749
+ "step": 290
1750
+ },
1751
+ {
1752
+ "epoch": 0.06,
1753
+ "learning_rate": 1.583539094650206e-05,
1754
+ "loss": 0.3823,
1755
+ "step": 291
1756
+ },
1757
+ {
1758
+ "epoch": 0.06,
1759
+ "learning_rate": 1.5818930041152266e-05,
1760
+ "loss": 0.4626,
1761
+ "step": 292
1762
+ },
1763
+ {
1764
+ "epoch": 0.06,
1765
+ "learning_rate": 1.580246913580247e-05,
1766
+ "loss": 0.4563,
1767
+ "step": 293
1768
+ },
1769
+ {
1770
+ "epoch": 0.06,
1771
+ "learning_rate": 1.5786008230452677e-05,
1772
+ "loss": 0.5166,
1773
+ "step": 294
1774
+ },
1775
+ {
1776
+ "epoch": 0.06,
1777
+ "learning_rate": 1.5769547325102884e-05,
1778
+ "loss": 0.3401,
1779
+ "step": 295
1780
+ },
1781
+ {
1782
+ "epoch": 0.06,
1783
+ "learning_rate": 1.5753086419753088e-05,
1784
+ "loss": 0.3565,
1785
+ "step": 296
1786
+ },
1787
+ {
1788
+ "epoch": 0.06,
1789
+ "learning_rate": 1.5736625514403295e-05,
1790
+ "loss": 0.3119,
1791
+ "step": 297
1792
+ },
1793
+ {
1794
+ "epoch": 0.06,
1795
+ "learning_rate": 1.5720164609053498e-05,
1796
+ "loss": 0.3234,
1797
+ "step": 298
1798
+ },
1799
+ {
1800
+ "epoch": 0.06,
1801
+ "learning_rate": 1.5703703703703705e-05,
1802
+ "loss": 0.4157,
1803
+ "step": 299
1804
+ },
1805
+ {
1806
+ "epoch": 0.06,
1807
+ "learning_rate": 1.568724279835391e-05,
1808
+ "loss": 0.3348,
1809
+ "step": 300
1810
+ },
1811
+ {
1812
+ "epoch": 0.06,
1813
+ "learning_rate": 1.5670781893004116e-05,
1814
+ "loss": 0.3829,
1815
+ "step": 301
1816
+ },
1817
+ {
1818
+ "epoch": 0.06,
1819
+ "learning_rate": 1.565432098765432e-05,
1820
+ "loss": 0.3549,
1821
+ "step": 302
1822
+ },
1823
+ {
1824
+ "epoch": 0.06,
1825
+ "learning_rate": 1.5637860082304527e-05,
1826
+ "loss": 0.4123,
1827
+ "step": 303
1828
+ },
1829
+ {
1830
+ "epoch": 0.06,
1831
+ "learning_rate": 1.5621399176954734e-05,
1832
+ "loss": 0.3378,
1833
+ "step": 304
1834
+ },
1835
+ {
1836
+ "epoch": 0.06,
1837
+ "learning_rate": 1.560493827160494e-05,
1838
+ "loss": 0.2703,
1839
+ "step": 305
1840
+ },
1841
+ {
1842
+ "epoch": 0.06,
1843
+ "learning_rate": 1.5588477366255144e-05,
1844
+ "loss": 0.3996,
1845
+ "step": 306
1846
+ },
1847
+ {
1848
+ "epoch": 0.06,
1849
+ "learning_rate": 1.557201646090535e-05,
1850
+ "loss": 0.2899,
1851
+ "step": 307
1852
+ },
1853
+ {
1854
+ "epoch": 0.06,
1855
+ "learning_rate": 1.555555555555556e-05,
1856
+ "loss": 0.442,
1857
+ "step": 308
1858
+ },
1859
+ {
1860
+ "epoch": 0.06,
1861
+ "learning_rate": 1.5539094650205762e-05,
1862
+ "loss": 0.3393,
1863
+ "step": 309
1864
+ },
1865
+ {
1866
+ "epoch": 0.06,
1867
+ "learning_rate": 1.552263374485597e-05,
1868
+ "loss": 0.4991,
1869
+ "step": 310
1870
+ },
1871
+ {
1872
+ "epoch": 0.06,
1873
+ "learning_rate": 1.5506172839506176e-05,
1874
+ "loss": 0.441,
1875
+ "step": 311
1876
+ },
1877
+ {
1878
+ "epoch": 0.07,
1879
+ "learning_rate": 1.548971193415638e-05,
1880
+ "loss": 0.269,
1881
+ "step": 312
1882
+ },
1883
+ {
1884
+ "epoch": 0.07,
1885
+ "learning_rate": 1.5473251028806587e-05,
1886
+ "loss": 0.3894,
1887
+ "step": 313
1888
+ },
1889
+ {
1890
+ "epoch": 0.07,
1891
+ "learning_rate": 1.545679012345679e-05,
1892
+ "loss": 0.3443,
1893
+ "step": 314
1894
+ },
1895
+ {
1896
+ "epoch": 0.07,
1897
+ "learning_rate": 1.5440329218106997e-05,
1898
+ "loss": 0.4439,
1899
+ "step": 315
1900
+ },
1901
+ {
1902
+ "epoch": 0.07,
1903
+ "learning_rate": 1.54238683127572e-05,
1904
+ "loss": 0.3945,
1905
+ "step": 316
1906
+ },
1907
+ {
1908
+ "epoch": 0.07,
1909
+ "learning_rate": 1.5407407407407408e-05,
1910
+ "loss": 0.343,
1911
+ "step": 317
1912
+ },
1913
+ {
1914
+ "epoch": 0.07,
1915
+ "learning_rate": 1.5390946502057615e-05,
1916
+ "loss": 0.37,
1917
+ "step": 318
1918
+ },
1919
+ {
1920
+ "epoch": 0.07,
1921
+ "learning_rate": 1.537448559670782e-05,
1922
+ "loss": 0.3407,
1923
+ "step": 319
1924
+ },
1925
+ {
1926
+ "epoch": 0.07,
1927
+ "learning_rate": 1.5358024691358026e-05,
1928
+ "loss": 0.4155,
1929
+ "step": 320
1930
+ },
1931
+ {
1932
+ "epoch": 0.07,
1933
+ "learning_rate": 1.5341563786008233e-05,
1934
+ "loss": 0.3395,
1935
+ "step": 321
1936
+ },
1937
+ {
1938
+ "epoch": 0.07,
1939
+ "learning_rate": 1.5325102880658436e-05,
1940
+ "loss": 0.3877,
1941
+ "step": 322
1942
+ },
1943
+ {
1944
+ "epoch": 0.07,
1945
+ "learning_rate": 1.5308641975308643e-05,
1946
+ "loss": 0.4577,
1947
+ "step": 323
1948
+ },
1949
+ {
1950
+ "epoch": 0.07,
1951
+ "learning_rate": 1.529218106995885e-05,
1952
+ "loss": 0.4159,
1953
+ "step": 324
1954
+ },
1955
+ {
1956
+ "epoch": 0.07,
1957
+ "learning_rate": 1.5275720164609057e-05,
1958
+ "loss": 0.3902,
1959
+ "step": 325
1960
+ },
1961
+ {
1962
+ "epoch": 0.07,
1963
+ "learning_rate": 1.525925925925926e-05,
1964
+ "loss": 0.379,
1965
+ "step": 326
1966
+ },
1967
+ {
1968
+ "epoch": 0.07,
1969
+ "learning_rate": 1.5242798353909466e-05,
1970
+ "loss": 0.4805,
1971
+ "step": 327
1972
+ },
1973
+ {
1974
+ "epoch": 0.07,
1975
+ "learning_rate": 1.5226337448559672e-05,
1976
+ "loss": 0.3841,
1977
+ "step": 328
1978
+ },
1979
+ {
1980
+ "epoch": 0.07,
1981
+ "learning_rate": 1.5209876543209879e-05,
1982
+ "loss": 0.3271,
1983
+ "step": 329
1984
+ },
1985
+ {
1986
+ "epoch": 0.07,
1987
+ "learning_rate": 1.5193415637860082e-05,
1988
+ "loss": 0.441,
1989
+ "step": 330
1990
+ },
1991
+ {
1992
+ "epoch": 0.07,
1993
+ "learning_rate": 1.517695473251029e-05,
1994
+ "loss": 0.4386,
1995
+ "step": 331
1996
+ },
1997
+ {
1998
+ "epoch": 0.07,
1999
+ "learning_rate": 1.5160493827160495e-05,
2000
+ "loss": 0.3656,
2001
+ "step": 332
2002
+ },
2003
+ {
2004
+ "epoch": 0.07,
2005
+ "learning_rate": 1.51440329218107e-05,
2006
+ "loss": 0.3541,
2007
+ "step": 333
2008
+ },
2009
+ {
2010
+ "epoch": 0.07,
2011
+ "learning_rate": 1.5127572016460907e-05,
2012
+ "loss": 0.3207,
2013
+ "step": 334
2014
+ },
2015
+ {
2016
+ "epoch": 0.07,
2017
+ "learning_rate": 1.5111111111111112e-05,
2018
+ "loss": 0.3754,
2019
+ "step": 335
2020
+ },
2021
+ {
2022
+ "epoch": 0.07,
2023
+ "learning_rate": 1.509465020576132e-05,
2024
+ "loss": 0.3895,
2025
+ "step": 336
2026
+ },
2027
+ {
2028
+ "epoch": 0.07,
2029
+ "learning_rate": 1.5078189300411523e-05,
2030
+ "loss": 0.3153,
2031
+ "step": 337
2032
+ },
2033
+ {
2034
+ "epoch": 0.07,
2035
+ "learning_rate": 1.506172839506173e-05,
2036
+ "loss": 0.4253,
2037
+ "step": 338
2038
+ },
2039
+ {
2040
+ "epoch": 0.07,
2041
+ "learning_rate": 1.5045267489711934e-05,
2042
+ "loss": 0.3902,
2043
+ "step": 339
2044
+ },
2045
+ {
2046
+ "epoch": 0.07,
2047
+ "learning_rate": 1.502880658436214e-05,
2048
+ "loss": 0.3469,
2049
+ "step": 340
2050
+ },
2051
+ {
2052
+ "epoch": 0.07,
2053
+ "learning_rate": 1.5012345679012348e-05,
2054
+ "loss": 0.3505,
2055
+ "step": 341
2056
+ },
2057
+ {
2058
+ "epoch": 0.07,
2059
+ "learning_rate": 1.4995884773662553e-05,
2060
+ "loss": 0.3173,
2061
+ "step": 342
2062
+ },
2063
+ {
2064
+ "epoch": 0.07,
2065
+ "learning_rate": 1.4979423868312758e-05,
2066
+ "loss": 0.4535,
2067
+ "step": 343
2068
+ },
2069
+ {
2070
+ "epoch": 0.07,
2071
+ "learning_rate": 1.4962962962962964e-05,
2072
+ "loss": 0.482,
2073
+ "step": 344
2074
+ },
2075
+ {
2076
+ "epoch": 0.07,
2077
+ "learning_rate": 1.494650205761317e-05,
2078
+ "loss": 0.3513,
2079
+ "step": 345
2080
+ },
2081
+ {
2082
+ "epoch": 0.07,
2083
+ "learning_rate": 1.4930041152263374e-05,
2084
+ "loss": 0.4741,
2085
+ "step": 346
2086
+ },
2087
+ {
2088
+ "epoch": 0.07,
2089
+ "learning_rate": 1.4913580246913581e-05,
2090
+ "loss": 0.4279,
2091
+ "step": 347
2092
+ },
2093
+ {
2094
+ "epoch": 0.07,
2095
+ "learning_rate": 1.4897119341563788e-05,
2096
+ "loss": 0.3729,
2097
+ "step": 348
2098
+ },
2099
+ {
2100
+ "epoch": 0.07,
2101
+ "learning_rate": 1.4880658436213992e-05,
2102
+ "loss": 0.3572,
2103
+ "step": 349
2104
+ },
2105
+ {
2106
+ "epoch": 0.07,
2107
+ "learning_rate": 1.4864197530864199e-05,
2108
+ "loss": 0.3935,
2109
+ "step": 350
2110
+ },
2111
+ {
2112
+ "epoch": 0.07,
2113
+ "learning_rate": 1.4847736625514404e-05,
2114
+ "loss": 0.2628,
2115
+ "step": 351
2116
+ },
2117
+ {
2118
+ "epoch": 0.07,
2119
+ "learning_rate": 1.4831275720164611e-05,
2120
+ "loss": 0.3107,
2121
+ "step": 352
2122
+ },
2123
+ {
2124
+ "epoch": 0.07,
2125
+ "learning_rate": 1.4814814814814815e-05,
2126
+ "loss": 0.4372,
2127
+ "step": 353
2128
+ },
2129
+ {
2130
+ "epoch": 0.07,
2131
+ "learning_rate": 1.4798353909465022e-05,
2132
+ "loss": 0.3705,
2133
+ "step": 354
2134
+ },
2135
+ {
2136
+ "epoch": 0.07,
2137
+ "learning_rate": 1.4781893004115229e-05,
2138
+ "loss": 0.4677,
2139
+ "step": 355
2140
+ },
2141
+ {
2142
+ "epoch": 0.07,
2143
+ "learning_rate": 1.4765432098765433e-05,
2144
+ "loss": 0.3333,
2145
+ "step": 356
2146
+ },
2147
+ {
2148
+ "epoch": 0.07,
2149
+ "learning_rate": 1.474897119341564e-05,
2150
+ "loss": 0.4412,
2151
+ "step": 357
2152
+ },
2153
+ {
2154
+ "epoch": 0.07,
2155
+ "learning_rate": 1.4732510288065845e-05,
2156
+ "loss": 0.4585,
2157
+ "step": 358
2158
+ },
2159
+ {
2160
+ "epoch": 0.07,
2161
+ "learning_rate": 1.471604938271605e-05,
2162
+ "loss": 0.4365,
2163
+ "step": 359
2164
+ },
2165
+ {
2166
+ "epoch": 0.08,
2167
+ "learning_rate": 1.4699588477366256e-05,
2168
+ "loss": 0.3039,
2169
+ "step": 360
2170
+ },
2171
+ {
2172
+ "epoch": 0.08,
2173
+ "learning_rate": 1.4683127572016463e-05,
2174
+ "loss": 0.4435,
2175
+ "step": 361
2176
+ },
2177
+ {
2178
+ "epoch": 0.08,
2179
+ "learning_rate": 1.4666666666666666e-05,
2180
+ "loss": 0.2515,
2181
+ "step": 362
2182
+ },
2183
+ {
2184
+ "epoch": 0.08,
2185
+ "learning_rate": 1.4650205761316873e-05,
2186
+ "loss": 0.3805,
2187
+ "step": 363
2188
+ },
2189
+ {
2190
+ "epoch": 0.08,
2191
+ "learning_rate": 1.463374485596708e-05,
2192
+ "loss": 0.2981,
2193
+ "step": 364
2194
+ },
2195
+ {
2196
+ "epoch": 0.08,
2197
+ "learning_rate": 1.4617283950617284e-05,
2198
+ "loss": 0.2568,
2199
+ "step": 365
2200
+ },
2201
+ {
2202
+ "epoch": 0.08,
2203
+ "learning_rate": 1.4600823045267491e-05,
2204
+ "loss": 0.3739,
2205
+ "step": 366
2206
+ },
2207
+ {
2208
+ "epoch": 0.08,
2209
+ "learning_rate": 1.4584362139917696e-05,
2210
+ "loss": 0.4102,
2211
+ "step": 367
2212
+ },
2213
+ {
2214
+ "epoch": 0.08,
2215
+ "learning_rate": 1.4567901234567903e-05,
2216
+ "loss": 0.3221,
2217
+ "step": 368
2218
+ },
2219
+ {
2220
+ "epoch": 0.08,
2221
+ "learning_rate": 1.4551440329218107e-05,
2222
+ "loss": 0.3714,
2223
+ "step": 369
2224
+ },
2225
+ {
2226
+ "epoch": 0.08,
2227
+ "learning_rate": 1.4534979423868314e-05,
2228
+ "loss": 0.4069,
2229
+ "step": 370
2230
+ },
2231
+ {
2232
+ "epoch": 0.08,
2233
+ "learning_rate": 1.4518518518518521e-05,
2234
+ "loss": 0.3146,
2235
+ "step": 371
2236
+ },
2237
+ {
2238
+ "epoch": 0.08,
2239
+ "learning_rate": 1.4502057613168725e-05,
2240
+ "loss": 0.344,
2241
+ "step": 372
2242
+ },
2243
+ {
2244
+ "epoch": 0.08,
2245
+ "learning_rate": 1.4485596707818932e-05,
2246
+ "loss": 0.4919,
2247
+ "step": 373
2248
+ },
2249
+ {
2250
+ "epoch": 0.08,
2251
+ "learning_rate": 1.4469135802469137e-05,
2252
+ "loss": 0.4057,
2253
+ "step": 374
2254
+ },
2255
+ {
2256
+ "epoch": 0.08,
2257
+ "learning_rate": 1.4452674897119342e-05,
2258
+ "loss": 0.3375,
2259
+ "step": 375
2260
+ },
2261
+ {
2262
+ "epoch": 0.08,
2263
+ "learning_rate": 1.4436213991769548e-05,
2264
+ "loss": 0.4297,
2265
+ "step": 376
2266
+ },
2267
+ {
2268
+ "epoch": 0.08,
2269
+ "learning_rate": 1.4419753086419755e-05,
2270
+ "loss": 0.4922,
2271
+ "step": 377
2272
+ },
2273
+ {
2274
+ "epoch": 0.08,
2275
+ "learning_rate": 1.4403292181069962e-05,
2276
+ "loss": 0.4206,
2277
+ "step": 378
2278
+ },
2279
+ {
2280
+ "epoch": 0.08,
2281
+ "learning_rate": 1.4386831275720165e-05,
2282
+ "loss": 0.3866,
2283
+ "step": 379
2284
+ },
2285
+ {
2286
+ "epoch": 0.08,
2287
+ "learning_rate": 1.4370370370370372e-05,
2288
+ "loss": 0.3612,
2289
+ "step": 380
2290
+ },
2291
+ {
2292
+ "epoch": 0.08,
2293
+ "learning_rate": 1.4353909465020578e-05,
2294
+ "loss": 0.2977,
2295
+ "step": 381
2296
+ },
2297
+ {
2298
+ "epoch": 0.08,
2299
+ "learning_rate": 1.4337448559670783e-05,
2300
+ "loss": 0.2621,
2301
+ "step": 382
2302
+ },
2303
+ {
2304
+ "epoch": 0.08,
2305
+ "learning_rate": 1.4320987654320988e-05,
2306
+ "loss": 0.366,
2307
+ "step": 383
2308
+ },
2309
+ {
2310
+ "epoch": 0.08,
2311
+ "learning_rate": 1.4304526748971195e-05,
2312
+ "loss": 0.2643,
2313
+ "step": 384
2314
+ },
2315
+ {
2316
+ "epoch": 0.08,
2317
+ "learning_rate": 1.4288065843621399e-05,
2318
+ "loss": 0.3707,
2319
+ "step": 385
2320
+ },
2321
+ {
2322
+ "epoch": 0.08,
2323
+ "learning_rate": 1.4271604938271606e-05,
2324
+ "loss": 0.3152,
2325
+ "step": 386
2326
+ },
2327
+ {
2328
+ "epoch": 0.08,
2329
+ "learning_rate": 1.4255144032921813e-05,
2330
+ "loss": 0.4019,
2331
+ "step": 387
2332
+ },
2333
+ {
2334
+ "epoch": 0.08,
2335
+ "learning_rate": 1.4238683127572017e-05,
2336
+ "loss": 0.5402,
2337
+ "step": 388
2338
+ },
2339
+ {
2340
+ "epoch": 0.08,
2341
+ "learning_rate": 1.4222222222222224e-05,
2342
+ "loss": 0.2943,
2343
+ "step": 389
2344
+ },
2345
+ {
2346
+ "epoch": 0.08,
2347
+ "learning_rate": 1.4205761316872429e-05,
2348
+ "loss": 0.3746,
2349
+ "step": 390
2350
+ },
2351
+ {
2352
+ "epoch": 0.08,
2353
+ "learning_rate": 1.4189300411522636e-05,
2354
+ "loss": 0.3904,
2355
+ "step": 391
2356
+ },
2357
+ {
2358
+ "epoch": 0.08,
2359
+ "learning_rate": 1.417283950617284e-05,
2360
+ "loss": 0.4219,
2361
+ "step": 392
2362
+ },
2363
+ {
2364
+ "epoch": 0.08,
2365
+ "learning_rate": 1.4156378600823047e-05,
2366
+ "loss": 0.3397,
2367
+ "step": 393
2368
+ },
2369
+ {
2370
+ "epoch": 0.08,
2371
+ "learning_rate": 1.4139917695473254e-05,
2372
+ "loss": 0.2991,
2373
+ "step": 394
2374
+ },
2375
+ {
2376
+ "epoch": 0.08,
2377
+ "learning_rate": 1.4123456790123457e-05,
2378
+ "loss": 0.3144,
2379
+ "step": 395
2380
+ },
2381
+ {
2382
+ "epoch": 0.08,
2383
+ "learning_rate": 1.4106995884773664e-05,
2384
+ "loss": 0.4191,
2385
+ "step": 396
2386
+ },
2387
+ {
2388
+ "epoch": 0.08,
2389
+ "learning_rate": 1.409053497942387e-05,
2390
+ "loss": 0.3163,
2391
+ "step": 397
2392
+ },
2393
+ {
2394
+ "epoch": 0.08,
2395
+ "learning_rate": 1.4074074074074075e-05,
2396
+ "loss": 0.4588,
2397
+ "step": 398
2398
+ },
2399
+ {
2400
+ "epoch": 0.08,
2401
+ "learning_rate": 1.405761316872428e-05,
2402
+ "loss": 0.3934,
2403
+ "step": 399
2404
+ },
2405
+ {
2406
+ "epoch": 0.08,
2407
+ "learning_rate": 1.4041152263374487e-05,
2408
+ "loss": 0.3465,
2409
+ "step": 400
2410
+ },
2411
+ {
2412
+ "epoch": 0.08,
2413
+ "learning_rate": 1.4024691358024694e-05,
2414
+ "loss": 0.3763,
2415
+ "step": 401
2416
+ },
2417
+ {
2418
+ "epoch": 0.08,
2419
+ "learning_rate": 1.4008230452674898e-05,
2420
+ "loss": 0.3422,
2421
+ "step": 402
2422
+ },
2423
+ {
2424
+ "epoch": 0.08,
2425
+ "learning_rate": 1.3991769547325105e-05,
2426
+ "loss": 0.402,
2427
+ "step": 403
2428
+ },
2429
+ {
2430
+ "epoch": 0.08,
2431
+ "learning_rate": 1.3975308641975309e-05,
2432
+ "loss": 0.3675,
2433
+ "step": 404
2434
+ },
2435
+ {
2436
+ "epoch": 0.08,
2437
+ "learning_rate": 1.3958847736625516e-05,
2438
+ "loss": 0.3448,
2439
+ "step": 405
2440
+ },
2441
+ {
2442
+ "epoch": 0.08,
2443
+ "learning_rate": 1.3942386831275721e-05,
2444
+ "loss": 0.3576,
2445
+ "step": 406
2446
+ },
2447
+ {
2448
+ "epoch": 0.08,
2449
+ "learning_rate": 1.3925925925925928e-05,
2450
+ "loss": 0.3369,
2451
+ "step": 407
2452
+ },
2453
+ {
2454
+ "epoch": 0.09,
2455
+ "learning_rate": 1.3909465020576133e-05,
2456
+ "loss": 0.2911,
2457
+ "step": 408
2458
+ },
2459
+ {
2460
+ "epoch": 0.09,
2461
+ "learning_rate": 1.3893004115226339e-05,
2462
+ "loss": 0.3524,
2463
+ "step": 409
2464
+ },
2465
+ {
2466
+ "epoch": 0.09,
2467
+ "learning_rate": 1.3876543209876546e-05,
2468
+ "loss": 0.4033,
2469
+ "step": 410
2470
+ },
2471
+ {
2472
+ "epoch": 0.09,
2473
+ "learning_rate": 1.386008230452675e-05,
2474
+ "loss": 0.3484,
2475
+ "step": 411
2476
+ },
2477
+ {
2478
+ "epoch": 0.09,
2479
+ "learning_rate": 1.3843621399176956e-05,
2480
+ "loss": 0.327,
2481
+ "step": 412
2482
+ },
2483
+ {
2484
+ "epoch": 0.09,
2485
+ "learning_rate": 1.3827160493827162e-05,
2486
+ "loss": 0.3669,
2487
+ "step": 413
2488
+ },
2489
+ {
2490
+ "epoch": 0.09,
2491
+ "learning_rate": 1.3810699588477367e-05,
2492
+ "loss": 0.3665,
2493
+ "step": 414
2494
+ },
2495
+ {
2496
+ "epoch": 0.09,
2497
+ "learning_rate": 1.3794238683127572e-05,
2498
+ "loss": 0.2736,
2499
+ "step": 415
2500
+ },
2501
+ {
2502
+ "epoch": 0.09,
2503
+ "learning_rate": 1.377777777777778e-05,
2504
+ "loss": 0.3945,
2505
+ "step": 416
2506
+ },
2507
+ {
2508
+ "epoch": 0.09,
2509
+ "learning_rate": 1.3761316872427986e-05,
2510
+ "loss": 0.3285,
2511
+ "step": 417
2512
+ },
2513
+ {
2514
+ "epoch": 0.09,
2515
+ "learning_rate": 1.374485596707819e-05,
2516
+ "loss": 0.354,
2517
+ "step": 418
2518
+ },
2519
+ {
2520
+ "epoch": 0.09,
2521
+ "learning_rate": 1.3728395061728397e-05,
2522
+ "loss": 0.3447,
2523
+ "step": 419
2524
+ },
2525
+ {
2526
+ "epoch": 0.09,
2527
+ "learning_rate": 1.37119341563786e-05,
2528
+ "loss": 0.2864,
2529
+ "step": 420
2530
+ },
2531
+ {
2532
+ "epoch": 0.09,
2533
+ "learning_rate": 1.3695473251028808e-05,
2534
+ "loss": 0.294,
2535
+ "step": 421
2536
+ },
2537
+ {
2538
+ "epoch": 0.09,
2539
+ "learning_rate": 1.3679012345679013e-05,
2540
+ "loss": 0.346,
2541
+ "step": 422
2542
+ },
2543
+ {
2544
+ "epoch": 0.09,
2545
+ "learning_rate": 1.366255144032922e-05,
2546
+ "loss": 0.3769,
2547
+ "step": 423
2548
+ },
2549
+ {
2550
+ "epoch": 0.09,
2551
+ "learning_rate": 1.3646090534979425e-05,
2552
+ "loss": 0.3316,
2553
+ "step": 424
2554
+ },
2555
+ {
2556
+ "epoch": 0.09,
2557
+ "learning_rate": 1.362962962962963e-05,
2558
+ "loss": 0.3219,
2559
+ "step": 425
2560
+ },
2561
+ {
2562
+ "epoch": 0.09,
2563
+ "learning_rate": 1.3613168724279838e-05,
2564
+ "loss": 0.3413,
2565
+ "step": 426
2566
+ },
2567
+ {
2568
+ "epoch": 0.09,
2569
+ "learning_rate": 1.3596707818930041e-05,
2570
+ "loss": 0.3495,
2571
+ "step": 427
2572
+ },
2573
+ {
2574
+ "epoch": 0.09,
2575
+ "learning_rate": 1.3580246913580248e-05,
2576
+ "loss": 0.3294,
2577
+ "step": 428
2578
+ },
2579
+ {
2580
+ "epoch": 0.09,
2581
+ "learning_rate": 1.3563786008230454e-05,
2582
+ "loss": 0.3042,
2583
+ "step": 429
2584
+ },
2585
+ {
2586
+ "epoch": 0.09,
2587
+ "learning_rate": 1.3547325102880659e-05,
2588
+ "loss": 0.3673,
2589
+ "step": 430
2590
+ },
2591
+ {
2592
+ "epoch": 0.09,
2593
+ "learning_rate": 1.3530864197530866e-05,
2594
+ "loss": 0.2992,
2595
+ "step": 431
2596
+ },
2597
+ {
2598
+ "epoch": 0.09,
2599
+ "learning_rate": 1.3514403292181071e-05,
2600
+ "loss": 0.2866,
2601
+ "step": 432
2602
+ },
2603
+ {
2604
+ "epoch": 0.09,
2605
+ "learning_rate": 1.3497942386831278e-05,
2606
+ "loss": 0.4294,
2607
+ "step": 433
2608
+ },
2609
+ {
2610
+ "epoch": 0.09,
2611
+ "learning_rate": 1.3481481481481482e-05,
2612
+ "loss": 0.381,
2613
+ "step": 434
2614
+ },
2615
+ {
2616
+ "epoch": 0.09,
2617
+ "learning_rate": 1.3465020576131689e-05,
2618
+ "loss": 0.2669,
2619
+ "step": 435
2620
+ },
2621
+ {
2622
+ "epoch": 0.09,
2623
+ "learning_rate": 1.3448559670781893e-05,
2624
+ "loss": 0.3166,
2625
+ "step": 436
2626
+ },
2627
+ {
2628
+ "epoch": 0.09,
2629
+ "learning_rate": 1.34320987654321e-05,
2630
+ "loss": 0.2913,
2631
+ "step": 437
2632
+ },
2633
+ {
2634
+ "epoch": 0.09,
2635
+ "learning_rate": 1.3415637860082307e-05,
2636
+ "loss": 0.3239,
2637
+ "step": 438
2638
+ },
2639
+ {
2640
+ "epoch": 0.09,
2641
+ "learning_rate": 1.3399176954732512e-05,
2642
+ "loss": 0.3154,
2643
+ "step": 439
2644
+ },
2645
+ {
2646
+ "epoch": 0.09,
2647
+ "learning_rate": 1.3382716049382717e-05,
2648
+ "loss": 0.4102,
2649
+ "step": 440
2650
+ },
2651
+ {
2652
+ "epoch": 0.09,
2653
+ "learning_rate": 1.3366255144032923e-05,
2654
+ "loss": 0.2735,
2655
+ "step": 441
2656
+ },
2657
+ {
2658
+ "epoch": 0.09,
2659
+ "learning_rate": 1.334979423868313e-05,
2660
+ "loss": 0.3109,
2661
+ "step": 442
2662
+ },
2663
+ {
2664
+ "epoch": 0.09,
2665
+ "learning_rate": 1.3333333333333333e-05,
2666
+ "loss": 0.41,
2667
+ "step": 443
2668
+ },
2669
+ {
2670
+ "epoch": 0.09,
2671
+ "learning_rate": 1.331687242798354e-05,
2672
+ "loss": 0.3857,
2673
+ "step": 444
2674
+ },
2675
+ {
2676
+ "epoch": 0.09,
2677
+ "learning_rate": 1.3300411522633746e-05,
2678
+ "loss": 0.3749,
2679
+ "step": 445
2680
+ },
2681
+ {
2682
+ "epoch": 0.09,
2683
+ "learning_rate": 1.3283950617283951e-05,
2684
+ "loss": 0.3313,
2685
+ "step": 446
2686
+ },
2687
+ {
2688
+ "epoch": 0.09,
2689
+ "learning_rate": 1.3267489711934158e-05,
2690
+ "loss": 0.4282,
2691
+ "step": 447
2692
+ },
2693
+ {
2694
+ "epoch": 0.09,
2695
+ "learning_rate": 1.3251028806584363e-05,
2696
+ "loss": 0.3211,
2697
+ "step": 448
2698
+ },
2699
+ {
2700
+ "epoch": 0.09,
2701
+ "learning_rate": 1.323456790123457e-05,
2702
+ "loss": 0.2531,
2703
+ "step": 449
2704
+ },
2705
+ {
2706
+ "epoch": 0.09,
2707
+ "learning_rate": 1.3218106995884774e-05,
2708
+ "loss": 0.3261,
2709
+ "step": 450
2710
+ },
2711
+ {
2712
+ "epoch": 0.09,
2713
+ "learning_rate": 1.3201646090534981e-05,
2714
+ "loss": 0.347,
2715
+ "step": 451
2716
+ },
2717
+ {
2718
+ "epoch": 0.09,
2719
+ "learning_rate": 1.3185185185185185e-05,
2720
+ "loss": 0.3412,
2721
+ "step": 452
2722
+ },
2723
+ {
2724
+ "epoch": 0.09,
2725
+ "learning_rate": 1.3168724279835392e-05,
2726
+ "loss": 0.3614,
2727
+ "step": 453
2728
+ },
2729
+ {
2730
+ "epoch": 0.09,
2731
+ "learning_rate": 1.3152263374485599e-05,
2732
+ "loss": 0.3441,
2733
+ "step": 454
2734
+ },
2735
+ {
2736
+ "epoch": 0.09,
2737
+ "learning_rate": 1.3135802469135804e-05,
2738
+ "loss": 0.3408,
2739
+ "step": 455
2740
+ },
2741
+ {
2742
+ "epoch": 0.1,
2743
+ "learning_rate": 1.311934156378601e-05,
2744
+ "loss": 0.404,
2745
+ "step": 456
2746
+ },
2747
+ {
2748
+ "epoch": 0.1,
2749
+ "learning_rate": 1.3102880658436215e-05,
2750
+ "loss": 0.3117,
2751
+ "step": 457
2752
+ },
2753
+ {
2754
+ "epoch": 0.1,
2755
+ "learning_rate": 1.3086419753086422e-05,
2756
+ "loss": 0.2961,
2757
+ "step": 458
2758
+ },
2759
+ {
2760
+ "epoch": 0.1,
2761
+ "learning_rate": 1.3069958847736625e-05,
2762
+ "loss": 0.4622,
2763
+ "step": 459
2764
+ },
2765
+ {
2766
+ "epoch": 0.1,
2767
+ "learning_rate": 1.3053497942386832e-05,
2768
+ "loss": 0.2617,
2769
+ "step": 460
2770
+ },
2771
+ {
2772
+ "epoch": 0.1,
2773
+ "learning_rate": 1.303703703703704e-05,
2774
+ "loss": 0.3371,
2775
+ "step": 461
2776
+ },
2777
+ {
2778
+ "epoch": 0.1,
2779
+ "learning_rate": 1.3020576131687243e-05,
2780
+ "loss": 0.3452,
2781
+ "step": 462
2782
+ },
2783
+ {
2784
+ "epoch": 0.1,
2785
+ "learning_rate": 1.300411522633745e-05,
2786
+ "loss": 0.3898,
2787
+ "step": 463
2788
+ },
2789
+ {
2790
+ "epoch": 0.1,
2791
+ "learning_rate": 1.2987654320987655e-05,
2792
+ "loss": 0.2803,
2793
+ "step": 464
2794
+ },
2795
+ {
2796
+ "epoch": 0.1,
2797
+ "learning_rate": 1.2971193415637862e-05,
2798
+ "loss": 0.3472,
2799
+ "step": 465
2800
+ },
2801
+ {
2802
+ "epoch": 0.1,
2803
+ "learning_rate": 1.2954732510288066e-05,
2804
+ "loss": 0.3625,
2805
+ "step": 466
2806
+ },
2807
+ {
2808
+ "epoch": 0.1,
2809
+ "learning_rate": 1.2938271604938273e-05,
2810
+ "loss": 0.3455,
2811
+ "step": 467
2812
+ },
2813
+ {
2814
+ "epoch": 0.1,
2815
+ "learning_rate": 1.2921810699588477e-05,
2816
+ "loss": 0.359,
2817
+ "step": 468
2818
+ },
2819
+ {
2820
+ "epoch": 0.1,
2821
+ "learning_rate": 1.2905349794238684e-05,
2822
+ "loss": 0.3567,
2823
+ "step": 469
2824
+ },
2825
+ {
2826
+ "epoch": 0.1,
2827
+ "learning_rate": 1.288888888888889e-05,
2828
+ "loss": 0.2884,
2829
+ "step": 470
2830
+ },
2831
+ {
2832
+ "epoch": 0.1,
2833
+ "learning_rate": 1.2872427983539096e-05,
2834
+ "loss": 0.2859,
2835
+ "step": 471
2836
+ },
2837
+ {
2838
+ "epoch": 0.1,
2839
+ "learning_rate": 1.2855967078189301e-05,
2840
+ "loss": 0.3415,
2841
+ "step": 472
2842
+ },
2843
+ {
2844
+ "epoch": 0.1,
2845
+ "learning_rate": 1.2839506172839507e-05,
2846
+ "loss": 0.3571,
2847
+ "step": 473
2848
+ },
2849
+ {
2850
+ "epoch": 0.1,
2851
+ "learning_rate": 1.2823045267489714e-05,
2852
+ "loss": 0.2856,
2853
+ "step": 474
2854
+ },
2855
+ {
2856
+ "epoch": 0.1,
2857
+ "learning_rate": 1.2806584362139917e-05,
2858
+ "loss": 0.3536,
2859
+ "step": 475
2860
+ },
2861
+ {
2862
+ "epoch": 0.1,
2863
+ "learning_rate": 1.2790123456790124e-05,
2864
+ "loss": 0.3306,
2865
+ "step": 476
2866
+ },
2867
+ {
2868
+ "epoch": 0.1,
2869
+ "learning_rate": 1.2773662551440331e-05,
2870
+ "loss": 0.444,
2871
+ "step": 477
2872
+ },
2873
+ {
2874
+ "epoch": 0.1,
2875
+ "learning_rate": 1.2757201646090535e-05,
2876
+ "loss": 0.3345,
2877
+ "step": 478
2878
+ },
2879
+ {
2880
+ "epoch": 0.1,
2881
+ "learning_rate": 1.2740740740740742e-05,
2882
+ "loss": 0.2992,
2883
+ "step": 479
2884
+ },
2885
+ {
2886
+ "epoch": 0.1,
2887
+ "learning_rate": 1.2724279835390947e-05,
2888
+ "loss": 0.3479,
2889
+ "step": 480
2890
+ },
2891
+ {
2892
+ "epoch": 0.1,
2893
+ "learning_rate": 1.2707818930041154e-05,
2894
+ "loss": 0.3584,
2895
+ "step": 481
2896
+ },
2897
+ {
2898
+ "epoch": 0.1,
2899
+ "learning_rate": 1.2691358024691358e-05,
2900
+ "loss": 0.3945,
2901
+ "step": 482
2902
+ },
2903
+ {
2904
+ "epoch": 0.1,
2905
+ "learning_rate": 1.2674897119341565e-05,
2906
+ "loss": 0.3087,
2907
+ "step": 483
2908
+ },
2909
+ {
2910
+ "epoch": 0.1,
2911
+ "learning_rate": 1.2658436213991772e-05,
2912
+ "loss": 0.4507,
2913
+ "step": 484
2914
+ },
2915
+ {
2916
+ "epoch": 0.1,
2917
+ "learning_rate": 1.2641975308641976e-05,
2918
+ "loss": 0.2884,
2919
+ "step": 485
2920
+ },
2921
+ {
2922
+ "epoch": 0.1,
2923
+ "learning_rate": 1.2625514403292183e-05,
2924
+ "loss": 0.3667,
2925
+ "step": 486
2926
+ },
2927
+ {
2928
+ "epoch": 0.1,
2929
+ "learning_rate": 1.2609053497942388e-05,
2930
+ "loss": 0.2725,
2931
+ "step": 487
2932
+ },
2933
+ {
2934
+ "epoch": 0.1,
2935
+ "learning_rate": 1.2592592592592593e-05,
2936
+ "loss": 0.3079,
2937
+ "step": 488
2938
+ },
2939
+ {
2940
+ "epoch": 0.1,
2941
+ "learning_rate": 1.2576131687242799e-05,
2942
+ "loss": 0.3374,
2943
+ "step": 489
2944
+ },
2945
+ {
2946
+ "epoch": 0.1,
2947
+ "learning_rate": 1.2559670781893006e-05,
2948
+ "loss": 0.3399,
2949
+ "step": 490
2950
+ },
2951
+ {
2952
+ "epoch": 0.1,
2953
+ "learning_rate": 1.2543209876543213e-05,
2954
+ "loss": 0.3581,
2955
+ "step": 491
2956
+ },
2957
+ {
2958
+ "epoch": 0.1,
2959
+ "learning_rate": 1.2526748971193416e-05,
2960
+ "loss": 0.3052,
2961
+ "step": 492
2962
+ },
2963
+ {
2964
+ "epoch": 0.1,
2965
+ "learning_rate": 1.2510288065843623e-05,
2966
+ "loss": 0.3658,
2967
+ "step": 493
2968
+ },
2969
+ {
2970
+ "epoch": 0.1,
2971
+ "learning_rate": 1.2493827160493827e-05,
2972
+ "loss": 0.3265,
2973
+ "step": 494
2974
+ },
2975
+ {
2976
+ "epoch": 0.1,
2977
+ "learning_rate": 1.2477366255144034e-05,
2978
+ "loss": 0.3122,
2979
+ "step": 495
2980
+ },
2981
+ {
2982
+ "epoch": 0.1,
2983
+ "learning_rate": 1.246090534979424e-05,
2984
+ "loss": 0.3255,
2985
+ "step": 496
2986
+ },
2987
+ {
2988
+ "epoch": 0.1,
2989
+ "learning_rate": 1.2444444444444446e-05,
2990
+ "loss": 0.3199,
2991
+ "step": 497
2992
+ },
2993
+ {
2994
+ "epoch": 0.1,
2995
+ "learning_rate": 1.242798353909465e-05,
2996
+ "loss": 0.2982,
2997
+ "step": 498
2998
+ },
2999
+ {
3000
+ "epoch": 0.1,
3001
+ "learning_rate": 1.2411522633744857e-05,
3002
+ "loss": 0.2721,
3003
+ "step": 499
3004
+ },
3005
+ {
3006
+ "epoch": 0.1,
3007
+ "learning_rate": 1.2395061728395064e-05,
3008
+ "loss": 0.3035,
3009
+ "step": 500
3010
+ }
3011
+ ],
3012
+ "logging_steps": 1.0,
3013
+ "max_steps": 1253,
3014
+ "num_train_epochs": 1,
3015
+ "save_steps": 500,
3016
+ "total_flos": 0.0,
3017
+ "trial_name": null,
3018
+ "trial_params": null
3019
+ }
checkpoint-500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f768ad49686ec36b5a8b570468ea76bd56614d5acbd36a18ffbbf3ba397fe3e
3
+ size 4728
config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "mistralai/Mistral-7B-v0.1",
3
+ "architectures": [
4
+ "MistralForCausalLM"
5
+ ],
6
+ "bos_token_id": 1,
7
+ "eos_token_id": 2,
8
+ "hidden_act": "silu",
9
+ "hidden_size": 4096,
10
+ "id2label": {
11
+ "0": "LABEL_0"
12
+ },
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "label2id": {
16
+ "LABEL_0": 0
17
+ },
18
+ "max_position_embeddings": 32768,
19
+ "model_type": "mistral",
20
+ "num_attention_heads": 32,
21
+ "num_hidden_layers": 32,
22
+ "num_key_value_heads": 8,
23
+ "rms_norm_eps": 1e-05,
24
+ "rope_theta": 10000.0,
25
+ "sliding_window": 4096,
26
+ "tie_word_embeddings": false,
27
+ "torch_dtype": "bfloat16",
28
+ "transformers_version": "4.36.0.dev0",
29
+ "use_cache": true,
30
+ "vocab_size": 32000
31
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c97b0457145bd7c833ba8ea4cf3b957fa4966ef2d41ea053e7e02df8209d67a1
3
+ size 14221467522
special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "<s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
3
+ size 493443
tokenizer_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "additional_special_tokens": [],
31
+ "bos_token": "<s>",
32
+ "clean_up_tokenization_spaces": false,
33
+ "eos_token": "</s>",
34
+ "legacy": true,
35
+ "model_max_length": 1000000000000000019884624838656,
36
+ "pad_token": "<s>",
37
+ "padding_side": "right",
38
+ "sp_model_kwargs": {},
39
+ "spaces_between_special_tokens": false,
40
+ "tokenizer_class": "LlamaTokenizer",
41
+ "unk_token": "<unk>",
42
+ "use_default_system_prompt": true
43
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f768ad49686ec36b5a8b570468ea76bd56614d5acbd36a18ffbbf3ba397fe3e
3
+ size 4728