Hugging Face's logo

GritLM
/
gen_m7_sq2048_ultrachat_ep1

Text Generation
Transformers
PyTorch
mistral
text-generation-inference
Inference Endpoints
Model card Files Community
Muennighoff commited on
Commit
2e37ec6
·
1 Parent(s): 293a0c6

Upload folder using huggingface_hub

Browse files
Files changed (24) hide show
  1. checkpoint-500/optimizer.bin +3 -0
  2. checkpoint-500/pytorch_model.bin +3 -0
  3. checkpoint-500/rng_state_0.pth +3 -0
  4. checkpoint-500/rng_state_1.pth +3 -0
  5. checkpoint-500/rng_state_2.pth +3 -0
  6. checkpoint-500/rng_state_3.pth +3 -0
  7. checkpoint-500/rng_state_4.pth +3 -0
  8. checkpoint-500/rng_state_5.pth +3 -0
  9. checkpoint-500/rng_state_6.pth +3 -0
  10. checkpoint-500/rng_state_7.pth +3 -0
  11. checkpoint-500/scheduler.pt +3 -0
  12. checkpoint-500/special_tokens_map.json +24 -0
  13. checkpoint-500/tokenizer.json +0 -0
  14. checkpoint-500/tokenizer.model +3 -0
  15. checkpoint-500/tokenizer_config.json +43 -0
  16. checkpoint-500/trainer_state.json +3019 -0
  17. checkpoint-500/training_args.bin +3 -0
  18. config.json +31 -0
  19. pytorch_model.bin +3 -0
  20. special_tokens_map.json +24 -0
  21. tokenizer.json +0 -0
  22. tokenizer.model +3 -0
  23. tokenizer_config.json +43 -0
  24. training_args.bin +3 -0
checkpoint-500/optimizer.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d274732c2b8bf30630f0f1fce91eb13ac266bb6f0ac2e9cc5a10cae786e3d18b
3
+ size 28967124478
checkpoint-500/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:54ecbfa032a136a8e22ed7a0a34a3faad62b6726783a7bfc645cbed45c18f389
3
+ size 14483616658
checkpoint-500/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:575119a228f98110923ffa2dedcb50e3317251b26054355d015e0b2240d566f2
3
+ size 15984
checkpoint-500/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0728b56dab7abb5ef8a0d4bae3519c5767c97467bdd886d26bf19cc8599d0312
3
+ size 15984
checkpoint-500/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4e481d4ef1546694da7337f6bb6c658b866dcb79b85deeb477da0d27ebe851e
3
+ size 15984
checkpoint-500/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:353c60be37ea56fc992fca446598ceca5d1fd002aa3bd6dbb9ad740e6f47ebb3
3
+ size 15984
checkpoint-500/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e9107fe964ba7205e354084b85210e5a5ea1c98cfd4d38adb9cd3926945dcae4
3
+ size 15984
checkpoint-500/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69d1bb1abee38b92e53f3f23549b642ce0f1edcdccf7b6129847ac61636e96d5
3
+ size 15984
checkpoint-500/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:afd5516048e20f36959601574e29e40106085a7d3cdc7bf425ce5e84633490e6
3
+ size 15984
checkpoint-500/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e2c46927fc06939b4c976a01e4b95dec1f8b98ceaea86d31a5d756fc30ff006
3
+ size 15984
checkpoint-500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9bae0606074cde62ead297fb8cda98d50aadb3979b8ed14fa52368e62ac77ddb
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": 0.6165228113440198,
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": 8.000000000000001e-07,
14
+ "loss": 0.8266,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 1.6000000000000001e-06,
20
+ "loss": 0.8352,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 2.4000000000000003e-06,
26
+ "loss": 0.8532,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.0,
31
+ "learning_rate": 3.2000000000000003e-06,
32
+ "loss": 0.845,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 4.000000000000001e-06,
38
+ "loss": 0.8296,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 4.800000000000001e-06,
44
+ "loss": 0.8202,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.01,
49
+ "learning_rate": 5.600000000000001e-06,
50
+ "loss": 0.8037,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.01,
55
+ "learning_rate": 6.4000000000000006e-06,
56
+ "loss": 0.8097,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.01,
61
+ "learning_rate": 7.2000000000000005e-06,
62
+ "loss": 0.8035,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.01,
67
+ "learning_rate": 8.000000000000001e-06,
68
+ "loss": 0.779,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.01,
73
+ "learning_rate": 8.8e-06,
74
+ "loss": 0.8298,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.01,
79
+ "learning_rate": 9.600000000000001e-06,
80
+ "loss": 0.8062,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.02,
85
+ "learning_rate": 1.04e-05,
86
+ "loss": 0.8089,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.02,
91
+ "learning_rate": 1.1200000000000001e-05,
92
+ "loss": 0.8245,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.02,
97
+ "learning_rate": 1.2e-05,
98
+ "loss": 0.8143,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.02,
103
+ "learning_rate": 1.2800000000000001e-05,
104
+ "loss": 0.8057,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.02,
109
+ "learning_rate": 1.3600000000000002e-05,
110
+ "loss": 0.7979,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.02,
115
+ "learning_rate": 1.4400000000000001e-05,
116
+ "loss": 0.7662,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.02,
121
+ "learning_rate": 1.5200000000000002e-05,
122
+ "loss": 0.8049,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.02,
127
+ "learning_rate": 1.6000000000000003e-05,
128
+ "loss": 0.7981,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.03,
133
+ "learning_rate": 1.6800000000000002e-05,
134
+ "loss": 0.8098,
135
+ "step": 21
136
+ },
137
+ {
138
+ "epoch": 0.03,
139
+ "learning_rate": 1.76e-05,
140
+ "loss": 0.7649,
141
+ "step": 22
142
+ },
143
+ {
144
+ "epoch": 0.03,
145
+ "learning_rate": 1.8400000000000003e-05,
146
+ "loss": 0.8131,
147
+ "step": 23
148
+ },
149
+ {
150
+ "epoch": 0.03,
151
+ "learning_rate": 1.9200000000000003e-05,
152
+ "loss": 0.7954,
153
+ "step": 24
154
+ },
155
+ {
156
+ "epoch": 0.03,
157
+ "learning_rate": 2e-05,
158
+ "loss": 0.8081,
159
+ "step": 25
160
+ },
161
+ {
162
+ "epoch": 0.03,
163
+ "learning_rate": 1.9974554707379136e-05,
164
+ "loss": 0.8338,
165
+ "step": 26
166
+ },
167
+ {
168
+ "epoch": 0.03,
169
+ "learning_rate": 1.994910941475827e-05,
170
+ "loss": 0.8243,
171
+ "step": 27
172
+ },
173
+ {
174
+ "epoch": 0.03,
175
+ "learning_rate": 1.9923664122137406e-05,
176
+ "loss": 0.8006,
177
+ "step": 28
178
+ },
179
+ {
180
+ "epoch": 0.04,
181
+ "learning_rate": 1.989821882951654e-05,
182
+ "loss": 0.778,
183
+ "step": 29
184
+ },
185
+ {
186
+ "epoch": 0.04,
187
+ "learning_rate": 1.9872773536895676e-05,
188
+ "loss": 0.7981,
189
+ "step": 30
190
+ },
191
+ {
192
+ "epoch": 0.04,
193
+ "learning_rate": 1.984732824427481e-05,
194
+ "loss": 0.8286,
195
+ "step": 31
196
+ },
197
+ {
198
+ "epoch": 0.04,
199
+ "learning_rate": 1.9821882951653945e-05,
200
+ "loss": 0.7917,
201
+ "step": 32
202
+ },
203
+ {
204
+ "epoch": 0.04,
205
+ "learning_rate": 1.979643765903308e-05,
206
+ "loss": 0.8026,
207
+ "step": 33
208
+ },
209
+ {
210
+ "epoch": 0.04,
211
+ "learning_rate": 1.9770992366412215e-05,
212
+ "loss": 0.8098,
213
+ "step": 34
214
+ },
215
+ {
216
+ "epoch": 0.04,
217
+ "learning_rate": 1.974554707379135e-05,
218
+ "loss": 0.7908,
219
+ "step": 35
220
+ },
221
+ {
222
+ "epoch": 0.04,
223
+ "learning_rate": 1.9720101781170485e-05,
224
+ "loss": 0.8318,
225
+ "step": 36
226
+ },
227
+ {
228
+ "epoch": 0.05,
229
+ "learning_rate": 1.969465648854962e-05,
230
+ "loss": 0.7805,
231
+ "step": 37
232
+ },
233
+ {
234
+ "epoch": 0.05,
235
+ "learning_rate": 1.9669211195928754e-05,
236
+ "loss": 0.841,
237
+ "step": 38
238
+ },
239
+ {
240
+ "epoch": 0.05,
241
+ "learning_rate": 1.964376590330789e-05,
242
+ "loss": 0.8339,
243
+ "step": 39
244
+ },
245
+ {
246
+ "epoch": 0.05,
247
+ "learning_rate": 1.9618320610687024e-05,
248
+ "loss": 0.8446,
249
+ "step": 40
250
+ },
251
+ {
252
+ "epoch": 0.05,
253
+ "learning_rate": 1.959287531806616e-05,
254
+ "loss": 0.8089,
255
+ "step": 41
256
+ },
257
+ {
258
+ "epoch": 0.05,
259
+ "learning_rate": 1.9567430025445294e-05,
260
+ "loss": 0.8422,
261
+ "step": 42
262
+ },
263
+ {
264
+ "epoch": 0.05,
265
+ "learning_rate": 1.954198473282443e-05,
266
+ "loss": 0.8238,
267
+ "step": 43
268
+ },
269
+ {
270
+ "epoch": 0.05,
271
+ "learning_rate": 1.9516539440203563e-05,
272
+ "loss": 0.8063,
273
+ "step": 44
274
+ },
275
+ {
276
+ "epoch": 0.06,
277
+ "learning_rate": 1.9491094147582698e-05,
278
+ "loss": 0.8135,
279
+ "step": 45
280
+ },
281
+ {
282
+ "epoch": 0.06,
283
+ "learning_rate": 1.9465648854961833e-05,
284
+ "loss": 0.8519,
285
+ "step": 46
286
+ },
287
+ {
288
+ "epoch": 0.06,
289
+ "learning_rate": 1.9440203562340968e-05,
290
+ "loss": 0.8441,
291
+ "step": 47
292
+ },
293
+ {
294
+ "epoch": 0.06,
295
+ "learning_rate": 1.9414758269720103e-05,
296
+ "loss": 0.8496,
297
+ "step": 48
298
+ },
299
+ {
300
+ "epoch": 0.06,
301
+ "learning_rate": 1.9389312977099238e-05,
302
+ "loss": 0.8291,
303
+ "step": 49
304
+ },
305
+ {
306
+ "epoch": 0.06,
307
+ "learning_rate": 1.9363867684478372e-05,
308
+ "loss": 0.7992,
309
+ "step": 50
310
+ },
311
+ {
312
+ "epoch": 0.06,
313
+ "learning_rate": 1.9338422391857507e-05,
314
+ "loss": 0.8375,
315
+ "step": 51
316
+ },
317
+ {
318
+ "epoch": 0.06,
319
+ "learning_rate": 1.9312977099236642e-05,
320
+ "loss": 0.795,
321
+ "step": 52
322
+ },
323
+ {
324
+ "epoch": 0.07,
325
+ "learning_rate": 1.9287531806615777e-05,
326
+ "loss": 0.8077,
327
+ "step": 53
328
+ },
329
+ {
330
+ "epoch": 0.07,
331
+ "learning_rate": 1.926208651399491e-05,
332
+ "loss": 0.811,
333
+ "step": 54
334
+ },
335
+ {
336
+ "epoch": 0.07,
337
+ "learning_rate": 1.923664122137405e-05,
338
+ "loss": 0.8112,
339
+ "step": 55
340
+ },
341
+ {
342
+ "epoch": 0.07,
343
+ "learning_rate": 1.921119592875318e-05,
344
+ "loss": 0.7858,
345
+ "step": 56
346
+ },
347
+ {
348
+ "epoch": 0.07,
349
+ "learning_rate": 1.9185750636132316e-05,
350
+ "loss": 0.837,
351
+ "step": 57
352
+ },
353
+ {
354
+ "epoch": 0.07,
355
+ "learning_rate": 1.916030534351145e-05,
356
+ "loss": 0.809,
357
+ "step": 58
358
+ },
359
+ {
360
+ "epoch": 0.07,
361
+ "learning_rate": 1.9134860050890586e-05,
362
+ "loss": 0.8575,
363
+ "step": 59
364
+ },
365
+ {
366
+ "epoch": 0.07,
367
+ "learning_rate": 1.910941475826972e-05,
368
+ "loss": 0.8098,
369
+ "step": 60
370
+ },
371
+ {
372
+ "epoch": 0.08,
373
+ "learning_rate": 1.908396946564886e-05,
374
+ "loss": 0.8357,
375
+ "step": 61
376
+ },
377
+ {
378
+ "epoch": 0.08,
379
+ "learning_rate": 1.9058524173027994e-05,
380
+ "loss": 0.8313,
381
+ "step": 62
382
+ },
383
+ {
384
+ "epoch": 0.08,
385
+ "learning_rate": 1.9033078880407125e-05,
386
+ "loss": 0.8082,
387
+ "step": 63
388
+ },
389
+ {
390
+ "epoch": 0.08,
391
+ "learning_rate": 1.900763358778626e-05,
392
+ "loss": 0.8316,
393
+ "step": 64
394
+ },
395
+ {
396
+ "epoch": 0.08,
397
+ "learning_rate": 1.8982188295165395e-05,
398
+ "loss": 0.7987,
399
+ "step": 65
400
+ },
401
+ {
402
+ "epoch": 0.08,
403
+ "learning_rate": 1.895674300254453e-05,
404
+ "loss": 0.8333,
405
+ "step": 66
406
+ },
407
+ {
408
+ "epoch": 0.08,
409
+ "learning_rate": 1.8931297709923668e-05,
410
+ "loss": 0.823,
411
+ "step": 67
412
+ },
413
+ {
414
+ "epoch": 0.08,
415
+ "learning_rate": 1.8905852417302803e-05,
416
+ "loss": 0.8245,
417
+ "step": 68
418
+ },
419
+ {
420
+ "epoch": 0.09,
421
+ "learning_rate": 1.8880407124681937e-05,
422
+ "loss": 0.825,
423
+ "step": 69
424
+ },
425
+ {
426
+ "epoch": 0.09,
427
+ "learning_rate": 1.885496183206107e-05,
428
+ "loss": 0.8213,
429
+ "step": 70
430
+ },
431
+ {
432
+ "epoch": 0.09,
433
+ "learning_rate": 1.8829516539440204e-05,
434
+ "loss": 0.8425,
435
+ "step": 71
436
+ },
437
+ {
438
+ "epoch": 0.09,
439
+ "learning_rate": 1.880407124681934e-05,
440
+ "loss": 0.824,
441
+ "step": 72
442
+ },
443
+ {
444
+ "epoch": 0.09,
445
+ "learning_rate": 1.8778625954198473e-05,
446
+ "loss": 0.8289,
447
+ "step": 73
448
+ },
449
+ {
450
+ "epoch": 0.09,
451
+ "learning_rate": 1.875318066157761e-05,
452
+ "loss": 0.8098,
453
+ "step": 74
454
+ },
455
+ {
456
+ "epoch": 0.09,
457
+ "learning_rate": 1.8727735368956746e-05,
458
+ "loss": 0.7983,
459
+ "step": 75
460
+ },
461
+ {
462
+ "epoch": 0.09,
463
+ "learning_rate": 1.8702290076335878e-05,
464
+ "loss": 0.7866,
465
+ "step": 76
466
+ },
467
+ {
468
+ "epoch": 0.09,
469
+ "learning_rate": 1.8676844783715013e-05,
470
+ "loss": 0.8254,
471
+ "step": 77
472
+ },
473
+ {
474
+ "epoch": 0.1,
475
+ "learning_rate": 1.8651399491094148e-05,
476
+ "loss": 0.7949,
477
+ "step": 78
478
+ },
479
+ {
480
+ "epoch": 0.1,
481
+ "learning_rate": 1.8625954198473282e-05,
482
+ "loss": 0.8324,
483
+ "step": 79
484
+ },
485
+ {
486
+ "epoch": 0.1,
487
+ "learning_rate": 1.860050890585242e-05,
488
+ "loss": 0.8328,
489
+ "step": 80
490
+ },
491
+ {
492
+ "epoch": 0.1,
493
+ "learning_rate": 1.8575063613231555e-05,
494
+ "loss": 0.8056,
495
+ "step": 81
496
+ },
497
+ {
498
+ "epoch": 0.1,
499
+ "learning_rate": 1.854961832061069e-05,
500
+ "loss": 0.8349,
501
+ "step": 82
502
+ },
503
+ {
504
+ "epoch": 0.1,
505
+ "learning_rate": 1.852417302798982e-05,
506
+ "loss": 0.8192,
507
+ "step": 83
508
+ },
509
+ {
510
+ "epoch": 0.1,
511
+ "learning_rate": 1.8498727735368956e-05,
512
+ "loss": 0.8216,
513
+ "step": 84
514
+ },
515
+ {
516
+ "epoch": 0.1,
517
+ "learning_rate": 1.847328244274809e-05,
518
+ "loss": 0.8042,
519
+ "step": 85
520
+ },
521
+ {
522
+ "epoch": 0.11,
523
+ "learning_rate": 1.844783715012723e-05,
524
+ "loss": 0.8278,
525
+ "step": 86
526
+ },
527
+ {
528
+ "epoch": 0.11,
529
+ "learning_rate": 1.8422391857506364e-05,
530
+ "loss": 0.7775,
531
+ "step": 87
532
+ },
533
+ {
534
+ "epoch": 0.11,
535
+ "learning_rate": 1.83969465648855e-05,
536
+ "loss": 0.8462,
537
+ "step": 88
538
+ },
539
+ {
540
+ "epoch": 0.11,
541
+ "learning_rate": 1.8371501272264634e-05,
542
+ "loss": 0.7989,
543
+ "step": 89
544
+ },
545
+ {
546
+ "epoch": 0.11,
547
+ "learning_rate": 1.8346055979643765e-05,
548
+ "loss": 0.8196,
549
+ "step": 90
550
+ },
551
+ {
552
+ "epoch": 0.11,
553
+ "learning_rate": 1.83206106870229e-05,
554
+ "loss": 0.8412,
555
+ "step": 91
556
+ },
557
+ {
558
+ "epoch": 0.11,
559
+ "learning_rate": 1.829516539440204e-05,
560
+ "loss": 0.8466,
561
+ "step": 92
562
+ },
563
+ {
564
+ "epoch": 0.11,
565
+ "learning_rate": 1.8269720101781173e-05,
566
+ "loss": 0.8204,
567
+ "step": 93
568
+ },
569
+ {
570
+ "epoch": 0.12,
571
+ "learning_rate": 1.8244274809160308e-05,
572
+ "loss": 0.8152,
573
+ "step": 94
574
+ },
575
+ {
576
+ "epoch": 0.12,
577
+ "learning_rate": 1.8218829516539443e-05,
578
+ "loss": 0.8614,
579
+ "step": 95
580
+ },
581
+ {
582
+ "epoch": 0.12,
583
+ "learning_rate": 1.8193384223918578e-05,
584
+ "loss": 0.8579,
585
+ "step": 96
586
+ },
587
+ {
588
+ "epoch": 0.12,
589
+ "learning_rate": 1.816793893129771e-05,
590
+ "loss": 0.8276,
591
+ "step": 97
592
+ },
593
+ {
594
+ "epoch": 0.12,
595
+ "learning_rate": 1.8142493638676847e-05,
596
+ "loss": 0.8455,
597
+ "step": 98
598
+ },
599
+ {
600
+ "epoch": 0.12,
601
+ "learning_rate": 1.8117048346055982e-05,
602
+ "loss": 0.8032,
603
+ "step": 99
604
+ },
605
+ {
606
+ "epoch": 0.12,
607
+ "learning_rate": 1.8091603053435117e-05,
608
+ "loss": 0.7993,
609
+ "step": 100
610
+ },
611
+ {
612
+ "epoch": 0.12,
613
+ "learning_rate": 1.8066157760814252e-05,
614
+ "loss": 0.846,
615
+ "step": 101
616
+ },
617
+ {
618
+ "epoch": 0.13,
619
+ "learning_rate": 1.8040712468193387e-05,
620
+ "loss": 0.8601,
621
+ "step": 102
622
+ },
623
+ {
624
+ "epoch": 0.13,
625
+ "learning_rate": 1.8015267175572518e-05,
626
+ "loss": 0.813,
627
+ "step": 103
628
+ },
629
+ {
630
+ "epoch": 0.13,
631
+ "learning_rate": 1.7989821882951653e-05,
632
+ "loss": 0.8183,
633
+ "step": 104
634
+ },
635
+ {
636
+ "epoch": 0.13,
637
+ "learning_rate": 1.796437659033079e-05,
638
+ "loss": 0.8244,
639
+ "step": 105
640
+ },
641
+ {
642
+ "epoch": 0.13,
643
+ "learning_rate": 1.7938931297709926e-05,
644
+ "loss": 0.8334,
645
+ "step": 106
646
+ },
647
+ {
648
+ "epoch": 0.13,
649
+ "learning_rate": 1.791348600508906e-05,
650
+ "loss": 0.8105,
651
+ "step": 107
652
+ },
653
+ {
654
+ "epoch": 0.13,
655
+ "learning_rate": 1.7888040712468196e-05,
656
+ "loss": 0.8148,
657
+ "step": 108
658
+ },
659
+ {
660
+ "epoch": 0.13,
661
+ "learning_rate": 1.786259541984733e-05,
662
+ "loss": 0.7892,
663
+ "step": 109
664
+ },
665
+ {
666
+ "epoch": 0.14,
667
+ "learning_rate": 1.7837150127226462e-05,
668
+ "loss": 0.8514,
669
+ "step": 110
670
+ },
671
+ {
672
+ "epoch": 0.14,
673
+ "learning_rate": 1.78117048346056e-05,
674
+ "loss": 0.8151,
675
+ "step": 111
676
+ },
677
+ {
678
+ "epoch": 0.14,
679
+ "learning_rate": 1.7786259541984735e-05,
680
+ "loss": 0.8408,
681
+ "step": 112
682
+ },
683
+ {
684
+ "epoch": 0.14,
685
+ "learning_rate": 1.776081424936387e-05,
686
+ "loss": 0.8087,
687
+ "step": 113
688
+ },
689
+ {
690
+ "epoch": 0.14,
691
+ "learning_rate": 1.7735368956743005e-05,
692
+ "loss": 0.7902,
693
+ "step": 114
694
+ },
695
+ {
696
+ "epoch": 0.14,
697
+ "learning_rate": 1.770992366412214e-05,
698
+ "loss": 0.8691,
699
+ "step": 115
700
+ },
701
+ {
702
+ "epoch": 0.14,
703
+ "learning_rate": 1.7684478371501274e-05,
704
+ "loss": 0.8045,
705
+ "step": 116
706
+ },
707
+ {
708
+ "epoch": 0.14,
709
+ "learning_rate": 1.765903307888041e-05,
710
+ "loss": 0.8286,
711
+ "step": 117
712
+ },
713
+ {
714
+ "epoch": 0.15,
715
+ "learning_rate": 1.7633587786259544e-05,
716
+ "loss": 0.8271,
717
+ "step": 118
718
+ },
719
+ {
720
+ "epoch": 0.15,
721
+ "learning_rate": 1.760814249363868e-05,
722
+ "loss": 0.8203,
723
+ "step": 119
724
+ },
725
+ {
726
+ "epoch": 0.15,
727
+ "learning_rate": 1.7582697201017814e-05,
728
+ "loss": 0.8212,
729
+ "step": 120
730
+ },
731
+ {
732
+ "epoch": 0.15,
733
+ "learning_rate": 1.755725190839695e-05,
734
+ "loss": 0.8217,
735
+ "step": 121
736
+ },
737
+ {
738
+ "epoch": 0.15,
739
+ "learning_rate": 1.7531806615776083e-05,
740
+ "loss": 0.828,
741
+ "step": 122
742
+ },
743
+ {
744
+ "epoch": 0.15,
745
+ "learning_rate": 1.7506361323155218e-05,
746
+ "loss": 0.8448,
747
+ "step": 123
748
+ },
749
+ {
750
+ "epoch": 0.15,
751
+ "learning_rate": 1.7480916030534353e-05,
752
+ "loss": 0.8328,
753
+ "step": 124
754
+ },
755
+ {
756
+ "epoch": 0.15,
757
+ "learning_rate": 1.7455470737913488e-05,
758
+ "loss": 0.7982,
759
+ "step": 125
760
+ },
761
+ {
762
+ "epoch": 0.16,
763
+ "learning_rate": 1.7430025445292623e-05,
764
+ "loss": 0.8134,
765
+ "step": 126
766
+ },
767
+ {
768
+ "epoch": 0.16,
769
+ "learning_rate": 1.7404580152671757e-05,
770
+ "loss": 0.8511,
771
+ "step": 127
772
+ },
773
+ {
774
+ "epoch": 0.16,
775
+ "learning_rate": 1.7379134860050892e-05,
776
+ "loss": 0.8345,
777
+ "step": 128
778
+ },
779
+ {
780
+ "epoch": 0.16,
781
+ "learning_rate": 1.7353689567430027e-05,
782
+ "loss": 0.8349,
783
+ "step": 129
784
+ },
785
+ {
786
+ "epoch": 0.16,
787
+ "learning_rate": 1.7328244274809162e-05,
788
+ "loss": 0.7941,
789
+ "step": 130
790
+ },
791
+ {
792
+ "epoch": 0.16,
793
+ "learning_rate": 1.7302798982188297e-05,
794
+ "loss": 0.8267,
795
+ "step": 131
796
+ },
797
+ {
798
+ "epoch": 0.16,
799
+ "learning_rate": 1.727735368956743e-05,
800
+ "loss": 0.846,
801
+ "step": 132
802
+ },
803
+ {
804
+ "epoch": 0.16,
805
+ "learning_rate": 1.7251908396946566e-05,
806
+ "loss": 0.8577,
807
+ "step": 133
808
+ },
809
+ {
810
+ "epoch": 0.17,
811
+ "learning_rate": 1.72264631043257e-05,
812
+ "loss": 0.819,
813
+ "step": 134
814
+ },
815
+ {
816
+ "epoch": 0.17,
817
+ "learning_rate": 1.7201017811704836e-05,
818
+ "loss": 0.8394,
819
+ "step": 135
820
+ },
821
+ {
822
+ "epoch": 0.17,
823
+ "learning_rate": 1.717557251908397e-05,
824
+ "loss": 0.7877,
825
+ "step": 136
826
+ },
827
+ {
828
+ "epoch": 0.17,
829
+ "learning_rate": 1.7150127226463106e-05,
830
+ "loss": 0.7775,
831
+ "step": 137
832
+ },
833
+ {
834
+ "epoch": 0.17,
835
+ "learning_rate": 1.712468193384224e-05,
836
+ "loss": 0.7958,
837
+ "step": 138
838
+ },
839
+ {
840
+ "epoch": 0.17,
841
+ "learning_rate": 1.7099236641221375e-05,
842
+ "loss": 0.8083,
843
+ "step": 139
844
+ },
845
+ {
846
+ "epoch": 0.17,
847
+ "learning_rate": 1.707379134860051e-05,
848
+ "loss": 0.7965,
849
+ "step": 140
850
+ },
851
+ {
852
+ "epoch": 0.17,
853
+ "learning_rate": 1.7048346055979645e-05,
854
+ "loss": 0.8244,
855
+ "step": 141
856
+ },
857
+ {
858
+ "epoch": 0.18,
859
+ "learning_rate": 1.702290076335878e-05,
860
+ "loss": 0.8298,
861
+ "step": 142
862
+ },
863
+ {
864
+ "epoch": 0.18,
865
+ "learning_rate": 1.6997455470737915e-05,
866
+ "loss": 0.7986,
867
+ "step": 143
868
+ },
869
+ {
870
+ "epoch": 0.18,
871
+ "learning_rate": 1.697201017811705e-05,
872
+ "loss": 0.8166,
873
+ "step": 144
874
+ },
875
+ {
876
+ "epoch": 0.18,
877
+ "learning_rate": 1.6946564885496184e-05,
878
+ "loss": 0.8519,
879
+ "step": 145
880
+ },
881
+ {
882
+ "epoch": 0.18,
883
+ "learning_rate": 1.692111959287532e-05,
884
+ "loss": 0.8205,
885
+ "step": 146
886
+ },
887
+ {
888
+ "epoch": 0.18,
889
+ "learning_rate": 1.6895674300254454e-05,
890
+ "loss": 0.8231,
891
+ "step": 147
892
+ },
893
+ {
894
+ "epoch": 0.18,
895
+ "learning_rate": 1.687022900763359e-05,
896
+ "loss": 0.8123,
897
+ "step": 148
898
+ },
899
+ {
900
+ "epoch": 0.18,
901
+ "learning_rate": 1.6844783715012724e-05,
902
+ "loss": 0.8189,
903
+ "step": 149
904
+ },
905
+ {
906
+ "epoch": 0.18,
907
+ "learning_rate": 1.681933842239186e-05,
908
+ "loss": 0.8057,
909
+ "step": 150
910
+ },
911
+ {
912
+ "epoch": 0.19,
913
+ "learning_rate": 1.6793893129770993e-05,
914
+ "loss": 0.8088,
915
+ "step": 151
916
+ },
917
+ {
918
+ "epoch": 0.19,
919
+ "learning_rate": 1.6768447837150128e-05,
920
+ "loss": 0.8504,
921
+ "step": 152
922
+ },
923
+ {
924
+ "epoch": 0.19,
925
+ "learning_rate": 1.6743002544529263e-05,
926
+ "loss": 0.8424,
927
+ "step": 153
928
+ },
929
+ {
930
+ "epoch": 0.19,
931
+ "learning_rate": 1.6717557251908398e-05,
932
+ "loss": 0.8135,
933
+ "step": 154
934
+ },
935
+ {
936
+ "epoch": 0.19,
937
+ "learning_rate": 1.6692111959287533e-05,
938
+ "loss": 0.8179,
939
+ "step": 155
940
+ },
941
+ {
942
+ "epoch": 0.19,
943
+ "learning_rate": 1.6666666666666667e-05,
944
+ "loss": 0.8367,
945
+ "step": 156
946
+ },
947
+ {
948
+ "epoch": 0.19,
949
+ "learning_rate": 1.6641221374045802e-05,
950
+ "loss": 0.8113,
951
+ "step": 157
952
+ },
953
+ {
954
+ "epoch": 0.19,
955
+ "learning_rate": 1.6615776081424937e-05,
956
+ "loss": 0.8083,
957
+ "step": 158
958
+ },
959
+ {
960
+ "epoch": 0.2,
961
+ "learning_rate": 1.6590330788804072e-05,
962
+ "loss": 0.83,
963
+ "step": 159
964
+ },
965
+ {
966
+ "epoch": 0.2,
967
+ "learning_rate": 1.6564885496183207e-05,
968
+ "loss": 0.8066,
969
+ "step": 160
970
+ },
971
+ {
972
+ "epoch": 0.2,
973
+ "learning_rate": 1.653944020356234e-05,
974
+ "loss": 0.7994,
975
+ "step": 161
976
+ },
977
+ {
978
+ "epoch": 0.2,
979
+ "learning_rate": 1.6513994910941476e-05,
980
+ "loss": 0.8223,
981
+ "step": 162
982
+ },
983
+ {
984
+ "epoch": 0.2,
985
+ "learning_rate": 1.648854961832061e-05,
986
+ "loss": 0.8133,
987
+ "step": 163
988
+ },
989
+ {
990
+ "epoch": 0.2,
991
+ "learning_rate": 1.6463104325699746e-05,
992
+ "loss": 0.8491,
993
+ "step": 164
994
+ },
995
+ {
996
+ "epoch": 0.2,
997
+ "learning_rate": 1.643765903307888e-05,
998
+ "loss": 0.82,
999
+ "step": 165
1000
+ },
1001
+ {
1002
+ "epoch": 0.2,
1003
+ "learning_rate": 1.6412213740458016e-05,
1004
+ "loss": 0.8242,
1005
+ "step": 166
1006
+ },
1007
+ {
1008
+ "epoch": 0.21,
1009
+ "learning_rate": 1.638676844783715e-05,
1010
+ "loss": 0.8221,
1011
+ "step": 167
1012
+ },
1013
+ {
1014
+ "epoch": 0.21,
1015
+ "learning_rate": 1.6361323155216285e-05,
1016
+ "loss": 0.8349,
1017
+ "step": 168
1018
+ },
1019
+ {
1020
+ "epoch": 0.21,
1021
+ "learning_rate": 1.633587786259542e-05,
1022
+ "loss": 0.8067,
1023
+ "step": 169
1024
+ },
1025
+ {
1026
+ "epoch": 0.21,
1027
+ "learning_rate": 1.631043256997456e-05,
1028
+ "loss": 0.807,
1029
+ "step": 170
1030
+ },
1031
+ {
1032
+ "epoch": 0.21,
1033
+ "learning_rate": 1.628498727735369e-05,
1034
+ "loss": 0.8141,
1035
+ "step": 171
1036
+ },
1037
+ {
1038
+ "epoch": 0.21,
1039
+ "learning_rate": 1.6259541984732825e-05,
1040
+ "loss": 0.8158,
1041
+ "step": 172
1042
+ },
1043
+ {
1044
+ "epoch": 0.21,
1045
+ "learning_rate": 1.623409669211196e-05,
1046
+ "loss": 0.8053,
1047
+ "step": 173
1048
+ },
1049
+ {
1050
+ "epoch": 0.21,
1051
+ "learning_rate": 1.6208651399491094e-05,
1052
+ "loss": 0.8267,
1053
+ "step": 174
1054
+ },
1055
+ {
1056
+ "epoch": 0.22,
1057
+ "learning_rate": 1.618320610687023e-05,
1058
+ "loss": 0.7809,
1059
+ "step": 175
1060
+ },
1061
+ {
1062
+ "epoch": 0.22,
1063
+ "learning_rate": 1.6157760814249367e-05,
1064
+ "loss": 0.8495,
1065
+ "step": 176
1066
+ },
1067
+ {
1068
+ "epoch": 0.22,
1069
+ "learning_rate": 1.6132315521628502e-05,
1070
+ "loss": 0.8149,
1071
+ "step": 177
1072
+ },
1073
+ {
1074
+ "epoch": 0.22,
1075
+ "learning_rate": 1.6106870229007634e-05,
1076
+ "loss": 0.8217,
1077
+ "step": 178
1078
+ },
1079
+ {
1080
+ "epoch": 0.22,
1081
+ "learning_rate": 1.608142493638677e-05,
1082
+ "loss": 0.821,
1083
+ "step": 179
1084
+ },
1085
+ {
1086
+ "epoch": 0.22,
1087
+ "learning_rate": 1.6055979643765903e-05,
1088
+ "loss": 0.7847,
1089
+ "step": 180
1090
+ },
1091
+ {
1092
+ "epoch": 0.22,
1093
+ "learning_rate": 1.6030534351145038e-05,
1094
+ "loss": 0.8436,
1095
+ "step": 181
1096
+ },
1097
+ {
1098
+ "epoch": 0.22,
1099
+ "learning_rate": 1.6005089058524176e-05,
1100
+ "loss": 0.8318,
1101
+ "step": 182
1102
+ },
1103
+ {
1104
+ "epoch": 0.23,
1105
+ "learning_rate": 1.597964376590331e-05,
1106
+ "loss": 0.8021,
1107
+ "step": 183
1108
+ },
1109
+ {
1110
+ "epoch": 0.23,
1111
+ "learning_rate": 1.5954198473282446e-05,
1112
+ "loss": 0.8232,
1113
+ "step": 184
1114
+ },
1115
+ {
1116
+ "epoch": 0.23,
1117
+ "learning_rate": 1.5928753180661577e-05,
1118
+ "loss": 0.8305,
1119
+ "step": 185
1120
+ },
1121
+ {
1122
+ "epoch": 0.23,
1123
+ "learning_rate": 1.5903307888040712e-05,
1124
+ "loss": 0.7975,
1125
+ "step": 186
1126
+ },
1127
+ {
1128
+ "epoch": 0.23,
1129
+ "learning_rate": 1.5877862595419847e-05,
1130
+ "loss": 0.8288,
1131
+ "step": 187
1132
+ },
1133
+ {
1134
+ "epoch": 0.23,
1135
+ "learning_rate": 1.5852417302798982e-05,
1136
+ "loss": 0.8123,
1137
+ "step": 188
1138
+ },
1139
+ {
1140
+ "epoch": 0.23,
1141
+ "learning_rate": 1.582697201017812e-05,
1142
+ "loss": 0.8017,
1143
+ "step": 189
1144
+ },
1145
+ {
1146
+ "epoch": 0.23,
1147
+ "learning_rate": 1.5801526717557255e-05,
1148
+ "loss": 0.8329,
1149
+ "step": 190
1150
+ },
1151
+ {
1152
+ "epoch": 0.24,
1153
+ "learning_rate": 1.577608142493639e-05,
1154
+ "loss": 0.8103,
1155
+ "step": 191
1156
+ },
1157
+ {
1158
+ "epoch": 0.24,
1159
+ "learning_rate": 1.575063613231552e-05,
1160
+ "loss": 0.8243,
1161
+ "step": 192
1162
+ },
1163
+ {
1164
+ "epoch": 0.24,
1165
+ "learning_rate": 1.5725190839694656e-05,
1166
+ "loss": 0.8264,
1167
+ "step": 193
1168
+ },
1169
+ {
1170
+ "epoch": 0.24,
1171
+ "learning_rate": 1.569974554707379e-05,
1172
+ "loss": 0.8223,
1173
+ "step": 194
1174
+ },
1175
+ {
1176
+ "epoch": 0.24,
1177
+ "learning_rate": 1.567430025445293e-05,
1178
+ "loss": 0.8236,
1179
+ "step": 195
1180
+ },
1181
+ {
1182
+ "epoch": 0.24,
1183
+ "learning_rate": 1.5648854961832064e-05,
1184
+ "loss": 0.7997,
1185
+ "step": 196
1186
+ },
1187
+ {
1188
+ "epoch": 0.24,
1189
+ "learning_rate": 1.56234096692112e-05,
1190
+ "loss": 0.8069,
1191
+ "step": 197
1192
+ },
1193
+ {
1194
+ "epoch": 0.24,
1195
+ "learning_rate": 1.559796437659033e-05,
1196
+ "loss": 0.8146,
1197
+ "step": 198
1198
+ },
1199
+ {
1200
+ "epoch": 0.25,
1201
+ "learning_rate": 1.5572519083969465e-05,
1202
+ "loss": 0.8507,
1203
+ "step": 199
1204
+ },
1205
+ {
1206
+ "epoch": 0.25,
1207
+ "learning_rate": 1.55470737913486e-05,
1208
+ "loss": 0.8013,
1209
+ "step": 200
1210
+ },
1211
+ {
1212
+ "epoch": 0.25,
1213
+ "learning_rate": 1.5521628498727738e-05,
1214
+ "loss": 0.7905,
1215
+ "step": 201
1216
+ },
1217
+ {
1218
+ "epoch": 0.25,
1219
+ "learning_rate": 1.5496183206106873e-05,
1220
+ "loss": 0.7836,
1221
+ "step": 202
1222
+ },
1223
+ {
1224
+ "epoch": 0.25,
1225
+ "learning_rate": 1.5470737913486008e-05,
1226
+ "loss": 0.8132,
1227
+ "step": 203
1228
+ },
1229
+ {
1230
+ "epoch": 0.25,
1231
+ "learning_rate": 1.5445292620865143e-05,
1232
+ "loss": 0.8581,
1233
+ "step": 204
1234
+ },
1235
+ {
1236
+ "epoch": 0.25,
1237
+ "learning_rate": 1.5419847328244274e-05,
1238
+ "loss": 0.8297,
1239
+ "step": 205
1240
+ },
1241
+ {
1242
+ "epoch": 0.25,
1243
+ "learning_rate": 1.539440203562341e-05,
1244
+ "loss": 0.8093,
1245
+ "step": 206
1246
+ },
1247
+ {
1248
+ "epoch": 0.26,
1249
+ "learning_rate": 1.5368956743002547e-05,
1250
+ "loss": 0.7955,
1251
+ "step": 207
1252
+ },
1253
+ {
1254
+ "epoch": 0.26,
1255
+ "learning_rate": 1.5343511450381682e-05,
1256
+ "loss": 0.8216,
1257
+ "step": 208
1258
+ },
1259
+ {
1260
+ "epoch": 0.26,
1261
+ "learning_rate": 1.5318066157760817e-05,
1262
+ "loss": 0.8334,
1263
+ "step": 209
1264
+ },
1265
+ {
1266
+ "epoch": 0.26,
1267
+ "learning_rate": 1.529262086513995e-05,
1268
+ "loss": 0.8088,
1269
+ "step": 210
1270
+ },
1271
+ {
1272
+ "epoch": 0.26,
1273
+ "learning_rate": 1.5267175572519086e-05,
1274
+ "loss": 0.8165,
1275
+ "step": 211
1276
+ },
1277
+ {
1278
+ "epoch": 0.26,
1279
+ "learning_rate": 1.524173027989822e-05,
1280
+ "loss": 0.7831,
1281
+ "step": 212
1282
+ },
1283
+ {
1284
+ "epoch": 0.26,
1285
+ "learning_rate": 1.5216284987277354e-05,
1286
+ "loss": 0.8178,
1287
+ "step": 213
1288
+ },
1289
+ {
1290
+ "epoch": 0.26,
1291
+ "learning_rate": 1.519083969465649e-05,
1292
+ "loss": 0.8033,
1293
+ "step": 214
1294
+ },
1295
+ {
1296
+ "epoch": 0.27,
1297
+ "learning_rate": 1.5165394402035624e-05,
1298
+ "loss": 0.8089,
1299
+ "step": 215
1300
+ },
1301
+ {
1302
+ "epoch": 0.27,
1303
+ "learning_rate": 1.513994910941476e-05,
1304
+ "loss": 0.7977,
1305
+ "step": 216
1306
+ },
1307
+ {
1308
+ "epoch": 0.27,
1309
+ "learning_rate": 1.5114503816793895e-05,
1310
+ "loss": 0.8236,
1311
+ "step": 217
1312
+ },
1313
+ {
1314
+ "epoch": 0.27,
1315
+ "learning_rate": 1.508905852417303e-05,
1316
+ "loss": 0.7973,
1317
+ "step": 218
1318
+ },
1319
+ {
1320
+ "epoch": 0.27,
1321
+ "learning_rate": 1.5063613231552163e-05,
1322
+ "loss": 0.8159,
1323
+ "step": 219
1324
+ },
1325
+ {
1326
+ "epoch": 0.27,
1327
+ "learning_rate": 1.5038167938931298e-05,
1328
+ "loss": 0.8309,
1329
+ "step": 220
1330
+ },
1331
+ {
1332
+ "epoch": 0.27,
1333
+ "learning_rate": 1.5012722646310433e-05,
1334
+ "loss": 0.8062,
1335
+ "step": 221
1336
+ },
1337
+ {
1338
+ "epoch": 0.27,
1339
+ "learning_rate": 1.498727735368957e-05,
1340
+ "loss": 0.8043,
1341
+ "step": 222
1342
+ },
1343
+ {
1344
+ "epoch": 0.27,
1345
+ "learning_rate": 1.4961832061068704e-05,
1346
+ "loss": 0.8255,
1347
+ "step": 223
1348
+ },
1349
+ {
1350
+ "epoch": 0.28,
1351
+ "learning_rate": 1.4936386768447839e-05,
1352
+ "loss": 0.8489,
1353
+ "step": 224
1354
+ },
1355
+ {
1356
+ "epoch": 0.28,
1357
+ "learning_rate": 1.4910941475826972e-05,
1358
+ "loss": 0.8045,
1359
+ "step": 225
1360
+ },
1361
+ {
1362
+ "epoch": 0.28,
1363
+ "learning_rate": 1.4885496183206107e-05,
1364
+ "loss": 0.8173,
1365
+ "step": 226
1366
+ },
1367
+ {
1368
+ "epoch": 0.28,
1369
+ "learning_rate": 1.4860050890585242e-05,
1370
+ "loss": 0.8537,
1371
+ "step": 227
1372
+ },
1373
+ {
1374
+ "epoch": 0.28,
1375
+ "learning_rate": 1.4834605597964378e-05,
1376
+ "loss": 0.84,
1377
+ "step": 228
1378
+ },
1379
+ {
1380
+ "epoch": 0.28,
1381
+ "learning_rate": 1.4809160305343513e-05,
1382
+ "loss": 0.8321,
1383
+ "step": 229
1384
+ },
1385
+ {
1386
+ "epoch": 0.28,
1387
+ "learning_rate": 1.4783715012722648e-05,
1388
+ "loss": 0.8304,
1389
+ "step": 230
1390
+ },
1391
+ {
1392
+ "epoch": 0.28,
1393
+ "learning_rate": 1.4758269720101783e-05,
1394
+ "loss": 0.8076,
1395
+ "step": 231
1396
+ },
1397
+ {
1398
+ "epoch": 0.29,
1399
+ "learning_rate": 1.4732824427480916e-05,
1400
+ "loss": 0.8124,
1401
+ "step": 232
1402
+ },
1403
+ {
1404
+ "epoch": 0.29,
1405
+ "learning_rate": 1.4707379134860051e-05,
1406
+ "loss": 0.8221,
1407
+ "step": 233
1408
+ },
1409
+ {
1410
+ "epoch": 0.29,
1411
+ "learning_rate": 1.4681933842239187e-05,
1412
+ "loss": 0.7866,
1413
+ "step": 234
1414
+ },
1415
+ {
1416
+ "epoch": 0.29,
1417
+ "learning_rate": 1.4656488549618322e-05,
1418
+ "loss": 0.8212,
1419
+ "step": 235
1420
+ },
1421
+ {
1422
+ "epoch": 0.29,
1423
+ "learning_rate": 1.4631043256997457e-05,
1424
+ "loss": 0.8377,
1425
+ "step": 236
1426
+ },
1427
+ {
1428
+ "epoch": 0.29,
1429
+ "learning_rate": 1.4605597964376592e-05,
1430
+ "loss": 0.8037,
1431
+ "step": 237
1432
+ },
1433
+ {
1434
+ "epoch": 0.29,
1435
+ "learning_rate": 1.4580152671755727e-05,
1436
+ "loss": 0.805,
1437
+ "step": 238
1438
+ },
1439
+ {
1440
+ "epoch": 0.29,
1441
+ "learning_rate": 1.455470737913486e-05,
1442
+ "loss": 0.7626,
1443
+ "step": 239
1444
+ },
1445
+ {
1446
+ "epoch": 0.3,
1447
+ "learning_rate": 1.4529262086513995e-05,
1448
+ "loss": 0.8299,
1449
+ "step": 240
1450
+ },
1451
+ {
1452
+ "epoch": 0.3,
1453
+ "learning_rate": 1.4503816793893131e-05,
1454
+ "loss": 0.7801,
1455
+ "step": 241
1456
+ },
1457
+ {
1458
+ "epoch": 0.3,
1459
+ "learning_rate": 1.4478371501272266e-05,
1460
+ "loss": 0.8103,
1461
+ "step": 242
1462
+ },
1463
+ {
1464
+ "epoch": 0.3,
1465
+ "learning_rate": 1.4452926208651401e-05,
1466
+ "loss": 0.7977,
1467
+ "step": 243
1468
+ },
1469
+ {
1470
+ "epoch": 0.3,
1471
+ "learning_rate": 1.4427480916030536e-05,
1472
+ "loss": 0.8012,
1473
+ "step": 244
1474
+ },
1475
+ {
1476
+ "epoch": 0.3,
1477
+ "learning_rate": 1.4402035623409672e-05,
1478
+ "loss": 0.8185,
1479
+ "step": 245
1480
+ },
1481
+ {
1482
+ "epoch": 0.3,
1483
+ "learning_rate": 1.4376590330788804e-05,
1484
+ "loss": 0.8048,
1485
+ "step": 246
1486
+ },
1487
+ {
1488
+ "epoch": 0.3,
1489
+ "learning_rate": 1.435114503816794e-05,
1490
+ "loss": 0.8045,
1491
+ "step": 247
1492
+ },
1493
+ {
1494
+ "epoch": 0.31,
1495
+ "learning_rate": 1.4325699745547075e-05,
1496
+ "loss": 0.8213,
1497
+ "step": 248
1498
+ },
1499
+ {
1500
+ "epoch": 0.31,
1501
+ "learning_rate": 1.430025445292621e-05,
1502
+ "loss": 0.8134,
1503
+ "step": 249
1504
+ },
1505
+ {
1506
+ "epoch": 0.31,
1507
+ "learning_rate": 1.4274809160305345e-05,
1508
+ "loss": 0.8056,
1509
+ "step": 250
1510
+ },
1511
+ {
1512
+ "epoch": 0.31,
1513
+ "learning_rate": 1.424936386768448e-05,
1514
+ "loss": 0.8212,
1515
+ "step": 251
1516
+ },
1517
+ {
1518
+ "epoch": 0.31,
1519
+ "learning_rate": 1.4223918575063613e-05,
1520
+ "loss": 0.7922,
1521
+ "step": 252
1522
+ },
1523
+ {
1524
+ "epoch": 0.31,
1525
+ "learning_rate": 1.4198473282442749e-05,
1526
+ "loss": 0.8234,
1527
+ "step": 253
1528
+ },
1529
+ {
1530
+ "epoch": 0.31,
1531
+ "learning_rate": 1.4173027989821884e-05,
1532
+ "loss": 0.7652,
1533
+ "step": 254
1534
+ },
1535
+ {
1536
+ "epoch": 0.31,
1537
+ "learning_rate": 1.4147582697201019e-05,
1538
+ "loss": 0.7972,
1539
+ "step": 255
1540
+ },
1541
+ {
1542
+ "epoch": 0.32,
1543
+ "learning_rate": 1.4122137404580154e-05,
1544
+ "loss": 0.813,
1545
+ "step": 256
1546
+ },
1547
+ {
1548
+ "epoch": 0.32,
1549
+ "learning_rate": 1.4096692111959288e-05,
1550
+ "loss": 0.8071,
1551
+ "step": 257
1552
+ },
1553
+ {
1554
+ "epoch": 0.32,
1555
+ "learning_rate": 1.4071246819338425e-05,
1556
+ "loss": 0.7848,
1557
+ "step": 258
1558
+ },
1559
+ {
1560
+ "epoch": 0.32,
1561
+ "learning_rate": 1.4045801526717558e-05,
1562
+ "loss": 0.8106,
1563
+ "step": 259
1564
+ },
1565
+ {
1566
+ "epoch": 0.32,
1567
+ "learning_rate": 1.4020356234096693e-05,
1568
+ "loss": 0.7902,
1569
+ "step": 260
1570
+ },
1571
+ {
1572
+ "epoch": 0.32,
1573
+ "learning_rate": 1.3994910941475828e-05,
1574
+ "loss": 0.8135,
1575
+ "step": 261
1576
+ },
1577
+ {
1578
+ "epoch": 0.32,
1579
+ "learning_rate": 1.3969465648854963e-05,
1580
+ "loss": 0.7981,
1581
+ "step": 262
1582
+ },
1583
+ {
1584
+ "epoch": 0.32,
1585
+ "learning_rate": 1.3944020356234097e-05,
1586
+ "loss": 0.827,
1587
+ "step": 263
1588
+ },
1589
+ {
1590
+ "epoch": 0.33,
1591
+ "learning_rate": 1.3918575063613234e-05,
1592
+ "loss": 0.7908,
1593
+ "step": 264
1594
+ },
1595
+ {
1596
+ "epoch": 0.33,
1597
+ "learning_rate": 1.3893129770992369e-05,
1598
+ "loss": 0.825,
1599
+ "step": 265
1600
+ },
1601
+ {
1602
+ "epoch": 0.33,
1603
+ "learning_rate": 1.3867684478371502e-05,
1604
+ "loss": 0.8236,
1605
+ "step": 266
1606
+ },
1607
+ {
1608
+ "epoch": 0.33,
1609
+ "learning_rate": 1.3842239185750637e-05,
1610
+ "loss": 0.8169,
1611
+ "step": 267
1612
+ },
1613
+ {
1614
+ "epoch": 0.33,
1615
+ "learning_rate": 1.3816793893129772e-05,
1616
+ "loss": 0.7962,
1617
+ "step": 268
1618
+ },
1619
+ {
1620
+ "epoch": 0.33,
1621
+ "learning_rate": 1.3791348600508906e-05,
1622
+ "loss": 0.7938,
1623
+ "step": 269
1624
+ },
1625
+ {
1626
+ "epoch": 0.33,
1627
+ "learning_rate": 1.3765903307888043e-05,
1628
+ "loss": 0.7996,
1629
+ "step": 270
1630
+ },
1631
+ {
1632
+ "epoch": 0.33,
1633
+ "learning_rate": 1.3740458015267178e-05,
1634
+ "loss": 0.8133,
1635
+ "step": 271
1636
+ },
1637
+ {
1638
+ "epoch": 0.34,
1639
+ "learning_rate": 1.3715012722646313e-05,
1640
+ "loss": 0.7995,
1641
+ "step": 272
1642
+ },
1643
+ {
1644
+ "epoch": 0.34,
1645
+ "learning_rate": 1.3689567430025446e-05,
1646
+ "loss": 0.8223,
1647
+ "step": 273
1648
+ },
1649
+ {
1650
+ "epoch": 0.34,
1651
+ "learning_rate": 1.366412213740458e-05,
1652
+ "loss": 0.7924,
1653
+ "step": 274
1654
+ },
1655
+ {
1656
+ "epoch": 0.34,
1657
+ "learning_rate": 1.3638676844783715e-05,
1658
+ "loss": 0.8331,
1659
+ "step": 275
1660
+ },
1661
+ {
1662
+ "epoch": 0.34,
1663
+ "learning_rate": 1.3613231552162852e-05,
1664
+ "loss": 0.8134,
1665
+ "step": 276
1666
+ },
1667
+ {
1668
+ "epoch": 0.34,
1669
+ "learning_rate": 1.3587786259541987e-05,
1670
+ "loss": 0.8018,
1671
+ "step": 277
1672
+ },
1673
+ {
1674
+ "epoch": 0.34,
1675
+ "learning_rate": 1.3562340966921122e-05,
1676
+ "loss": 0.789,
1677
+ "step": 278
1678
+ },
1679
+ {
1680
+ "epoch": 0.34,
1681
+ "learning_rate": 1.3536895674300256e-05,
1682
+ "loss": 0.816,
1683
+ "step": 279
1684
+ },
1685
+ {
1686
+ "epoch": 0.35,
1687
+ "learning_rate": 1.351145038167939e-05,
1688
+ "loss": 0.8142,
1689
+ "step": 280
1690
+ },
1691
+ {
1692
+ "epoch": 0.35,
1693
+ "learning_rate": 1.3486005089058524e-05,
1694
+ "loss": 0.833,
1695
+ "step": 281
1696
+ },
1697
+ {
1698
+ "epoch": 0.35,
1699
+ "learning_rate": 1.346055979643766e-05,
1700
+ "loss": 0.8419,
1701
+ "step": 282
1702
+ },
1703
+ {
1704
+ "epoch": 0.35,
1705
+ "learning_rate": 1.3435114503816796e-05,
1706
+ "loss": 0.7914,
1707
+ "step": 283
1708
+ },
1709
+ {
1710
+ "epoch": 0.35,
1711
+ "learning_rate": 1.340966921119593e-05,
1712
+ "loss": 0.7716,
1713
+ "step": 284
1714
+ },
1715
+ {
1716
+ "epoch": 0.35,
1717
+ "learning_rate": 1.3384223918575065e-05,
1718
+ "loss": 0.797,
1719
+ "step": 285
1720
+ },
1721
+ {
1722
+ "epoch": 0.35,
1723
+ "learning_rate": 1.3358778625954198e-05,
1724
+ "loss": 0.8074,
1725
+ "step": 286
1726
+ },
1727
+ {
1728
+ "epoch": 0.35,
1729
+ "learning_rate": 1.3333333333333333e-05,
1730
+ "loss": 0.817,
1731
+ "step": 287
1732
+ },
1733
+ {
1734
+ "epoch": 0.36,
1735
+ "learning_rate": 1.3307888040712468e-05,
1736
+ "loss": 0.8007,
1737
+ "step": 288
1738
+ },
1739
+ {
1740
+ "epoch": 0.36,
1741
+ "learning_rate": 1.3282442748091605e-05,
1742
+ "loss": 0.8075,
1743
+ "step": 289
1744
+ },
1745
+ {
1746
+ "epoch": 0.36,
1747
+ "learning_rate": 1.325699745547074e-05,
1748
+ "loss": 0.7962,
1749
+ "step": 290
1750
+ },
1751
+ {
1752
+ "epoch": 0.36,
1753
+ "learning_rate": 1.3231552162849874e-05,
1754
+ "loss": 0.7909,
1755
+ "step": 291
1756
+ },
1757
+ {
1758
+ "epoch": 0.36,
1759
+ "learning_rate": 1.3206106870229009e-05,
1760
+ "loss": 0.8124,
1761
+ "step": 292
1762
+ },
1763
+ {
1764
+ "epoch": 0.36,
1765
+ "learning_rate": 1.3180661577608142e-05,
1766
+ "loss": 0.8064,
1767
+ "step": 293
1768
+ },
1769
+ {
1770
+ "epoch": 0.36,
1771
+ "learning_rate": 1.3155216284987277e-05,
1772
+ "loss": 0.8419,
1773
+ "step": 294
1774
+ },
1775
+ {
1776
+ "epoch": 0.36,
1777
+ "learning_rate": 1.3129770992366414e-05,
1778
+ "loss": 0.7884,
1779
+ "step": 295
1780
+ },
1781
+ {
1782
+ "epoch": 0.36,
1783
+ "learning_rate": 1.3104325699745548e-05,
1784
+ "loss": 0.8232,
1785
+ "step": 296
1786
+ },
1787
+ {
1788
+ "epoch": 0.37,
1789
+ "learning_rate": 1.3078880407124683e-05,
1790
+ "loss": 0.8327,
1791
+ "step": 297
1792
+ },
1793
+ {
1794
+ "epoch": 0.37,
1795
+ "learning_rate": 1.3053435114503818e-05,
1796
+ "loss": 0.8124,
1797
+ "step": 298
1798
+ },
1799
+ {
1800
+ "epoch": 0.37,
1801
+ "learning_rate": 1.3027989821882953e-05,
1802
+ "loss": 0.8181,
1803
+ "step": 299
1804
+ },
1805
+ {
1806
+ "epoch": 0.37,
1807
+ "learning_rate": 1.3002544529262086e-05,
1808
+ "loss": 0.7472,
1809
+ "step": 300
1810
+ },
1811
+ {
1812
+ "epoch": 0.37,
1813
+ "learning_rate": 1.2977099236641223e-05,
1814
+ "loss": 0.7523,
1815
+ "step": 301
1816
+ },
1817
+ {
1818
+ "epoch": 0.37,
1819
+ "learning_rate": 1.2951653944020357e-05,
1820
+ "loss": 0.777,
1821
+ "step": 302
1822
+ },
1823
+ {
1824
+ "epoch": 0.37,
1825
+ "learning_rate": 1.2926208651399492e-05,
1826
+ "loss": 0.805,
1827
+ "step": 303
1828
+ },
1829
+ {
1830
+ "epoch": 0.37,
1831
+ "learning_rate": 1.2900763358778627e-05,
1832
+ "loss": 0.7844,
1833
+ "step": 304
1834
+ },
1835
+ {
1836
+ "epoch": 0.38,
1837
+ "learning_rate": 1.2875318066157762e-05,
1838
+ "loss": 0.7474,
1839
+ "step": 305
1840
+ },
1841
+ {
1842
+ "epoch": 0.38,
1843
+ "learning_rate": 1.2849872773536898e-05,
1844
+ "loss": 0.7819,
1845
+ "step": 306
1846
+ },
1847
+ {
1848
+ "epoch": 0.38,
1849
+ "learning_rate": 1.2824427480916032e-05,
1850
+ "loss": 0.8223,
1851
+ "step": 307
1852
+ },
1853
+ {
1854
+ "epoch": 0.38,
1855
+ "learning_rate": 1.2798982188295166e-05,
1856
+ "loss": 0.7577,
1857
+ "step": 308
1858
+ },
1859
+ {
1860
+ "epoch": 0.38,
1861
+ "learning_rate": 1.2773536895674301e-05,
1862
+ "loss": 0.7975,
1863
+ "step": 309
1864
+ },
1865
+ {
1866
+ "epoch": 0.38,
1867
+ "learning_rate": 1.2748091603053436e-05,
1868
+ "loss": 0.8095,
1869
+ "step": 310
1870
+ },
1871
+ {
1872
+ "epoch": 0.38,
1873
+ "learning_rate": 1.2722646310432571e-05,
1874
+ "loss": 0.7697,
1875
+ "step": 311
1876
+ },
1877
+ {
1878
+ "epoch": 0.38,
1879
+ "learning_rate": 1.2697201017811707e-05,
1880
+ "loss": 0.8154,
1881
+ "step": 312
1882
+ },
1883
+ {
1884
+ "epoch": 0.39,
1885
+ "learning_rate": 1.2671755725190839e-05,
1886
+ "loss": 0.8007,
1887
+ "step": 313
1888
+ },
1889
+ {
1890
+ "epoch": 0.39,
1891
+ "learning_rate": 1.2646310432569975e-05,
1892
+ "loss": 0.81,
1893
+ "step": 314
1894
+ },
1895
+ {
1896
+ "epoch": 0.39,
1897
+ "learning_rate": 1.262086513994911e-05,
1898
+ "loss": 0.7897,
1899
+ "step": 315
1900
+ },
1901
+ {
1902
+ "epoch": 0.39,
1903
+ "learning_rate": 1.2595419847328245e-05,
1904
+ "loss": 0.8132,
1905
+ "step": 316
1906
+ },
1907
+ {
1908
+ "epoch": 0.39,
1909
+ "learning_rate": 1.256997455470738e-05,
1910
+ "loss": 0.7748,
1911
+ "step": 317
1912
+ },
1913
+ {
1914
+ "epoch": 0.39,
1915
+ "learning_rate": 1.2544529262086516e-05,
1916
+ "loss": 0.8145,
1917
+ "step": 318
1918
+ },
1919
+ {
1920
+ "epoch": 0.39,
1921
+ "learning_rate": 1.2519083969465651e-05,
1922
+ "loss": 0.7718,
1923
+ "step": 319
1924
+ },
1925
+ {
1926
+ "epoch": 0.39,
1927
+ "learning_rate": 1.2493638676844784e-05,
1928
+ "loss": 0.8176,
1929
+ "step": 320
1930
+ },
1931
+ {
1932
+ "epoch": 0.4,
1933
+ "learning_rate": 1.2468193384223919e-05,
1934
+ "loss": 0.772,
1935
+ "step": 321
1936
+ },
1937
+ {
1938
+ "epoch": 0.4,
1939
+ "learning_rate": 1.2442748091603054e-05,
1940
+ "loss": 0.8236,
1941
+ "step": 322
1942
+ },
1943
+ {
1944
+ "epoch": 0.4,
1945
+ "learning_rate": 1.2417302798982189e-05,
1946
+ "loss": 0.8173,
1947
+ "step": 323
1948
+ },
1949
+ {
1950
+ "epoch": 0.4,
1951
+ "learning_rate": 1.2391857506361324e-05,
1952
+ "loss": 0.802,
1953
+ "step": 324
1954
+ },
1955
+ {
1956
+ "epoch": 0.4,
1957
+ "learning_rate": 1.236641221374046e-05,
1958
+ "loss": 0.8048,
1959
+ "step": 325
1960
+ },
1961
+ {
1962
+ "epoch": 0.4,
1963
+ "learning_rate": 1.2340966921119595e-05,
1964
+ "loss": 0.7729,
1965
+ "step": 326
1966
+ },
1967
+ {
1968
+ "epoch": 0.4,
1969
+ "learning_rate": 1.2315521628498728e-05,
1970
+ "loss": 0.8161,
1971
+ "step": 327
1972
+ },
1973
+ {
1974
+ "epoch": 0.4,
1975
+ "learning_rate": 1.2290076335877863e-05,
1976
+ "loss": 0.7838,
1977
+ "step": 328
1978
+ },
1979
+ {
1980
+ "epoch": 0.41,
1981
+ "learning_rate": 1.2264631043256998e-05,
1982
+ "loss": 0.793,
1983
+ "step": 329
1984
+ },
1985
+ {
1986
+ "epoch": 0.41,
1987
+ "learning_rate": 1.2239185750636133e-05,
1988
+ "loss": 0.7653,
1989
+ "step": 330
1990
+ },
1991
+ {
1992
+ "epoch": 0.41,
1993
+ "learning_rate": 1.2213740458015269e-05,
1994
+ "loss": 0.8032,
1995
+ "step": 331
1996
+ },
1997
+ {
1998
+ "epoch": 0.41,
1999
+ "learning_rate": 1.2188295165394404e-05,
2000
+ "loss": 0.8163,
2001
+ "step": 332
2002
+ },
2003
+ {
2004
+ "epoch": 0.41,
2005
+ "learning_rate": 1.2162849872773539e-05,
2006
+ "loss": 0.7815,
2007
+ "step": 333
2008
+ },
2009
+ {
2010
+ "epoch": 0.41,
2011
+ "learning_rate": 1.2137404580152672e-05,
2012
+ "loss": 0.7783,
2013
+ "step": 334
2014
+ },
2015
+ {
2016
+ "epoch": 0.41,
2017
+ "learning_rate": 1.2111959287531807e-05,
2018
+ "loss": 0.7745,
2019
+ "step": 335
2020
+ },
2021
+ {
2022
+ "epoch": 0.41,
2023
+ "learning_rate": 1.2086513994910942e-05,
2024
+ "loss": 0.7743,
2025
+ "step": 336
2026
+ },
2027
+ {
2028
+ "epoch": 0.42,
2029
+ "learning_rate": 1.2061068702290078e-05,
2030
+ "loss": 0.7846,
2031
+ "step": 337
2032
+ },
2033
+ {
2034
+ "epoch": 0.42,
2035
+ "learning_rate": 1.2035623409669213e-05,
2036
+ "loss": 0.7933,
2037
+ "step": 338
2038
+ },
2039
+ {
2040
+ "epoch": 0.42,
2041
+ "learning_rate": 1.2010178117048348e-05,
2042
+ "loss": 0.7788,
2043
+ "step": 339
2044
+ },
2045
+ {
2046
+ "epoch": 0.42,
2047
+ "learning_rate": 1.1984732824427483e-05,
2048
+ "loss": 0.8189,
2049
+ "step": 340
2050
+ },
2051
+ {
2052
+ "epoch": 0.42,
2053
+ "learning_rate": 1.1959287531806616e-05,
2054
+ "loss": 0.7676,
2055
+ "step": 341
2056
+ },
2057
+ {
2058
+ "epoch": 0.42,
2059
+ "learning_rate": 1.193384223918575e-05,
2060
+ "loss": 0.8016,
2061
+ "step": 342
2062
+ },
2063
+ {
2064
+ "epoch": 0.42,
2065
+ "learning_rate": 1.1908396946564887e-05,
2066
+ "loss": 0.798,
2067
+ "step": 343
2068
+ },
2069
+ {
2070
+ "epoch": 0.42,
2071
+ "learning_rate": 1.1882951653944022e-05,
2072
+ "loss": 0.7961,
2073
+ "step": 344
2074
+ },
2075
+ {
2076
+ "epoch": 0.43,
2077
+ "learning_rate": 1.1857506361323157e-05,
2078
+ "loss": 0.7811,
2079
+ "step": 345
2080
+ },
2081
+ {
2082
+ "epoch": 0.43,
2083
+ "learning_rate": 1.1832061068702292e-05,
2084
+ "loss": 0.7938,
2085
+ "step": 346
2086
+ },
2087
+ {
2088
+ "epoch": 0.43,
2089
+ "learning_rate": 1.1806615776081425e-05,
2090
+ "loss": 0.8402,
2091
+ "step": 347
2092
+ },
2093
+ {
2094
+ "epoch": 0.43,
2095
+ "learning_rate": 1.178117048346056e-05,
2096
+ "loss": 0.7691,
2097
+ "step": 348
2098
+ },
2099
+ {
2100
+ "epoch": 0.43,
2101
+ "learning_rate": 1.1755725190839696e-05,
2102
+ "loss": 0.7851,
2103
+ "step": 349
2104
+ },
2105
+ {
2106
+ "epoch": 0.43,
2107
+ "learning_rate": 1.173027989821883e-05,
2108
+ "loss": 0.8005,
2109
+ "step": 350
2110
+ },
2111
+ {
2112
+ "epoch": 0.43,
2113
+ "learning_rate": 1.1704834605597966e-05,
2114
+ "loss": 0.8049,
2115
+ "step": 351
2116
+ },
2117
+ {
2118
+ "epoch": 0.43,
2119
+ "learning_rate": 1.16793893129771e-05,
2120
+ "loss": 0.8211,
2121
+ "step": 352
2122
+ },
2123
+ {
2124
+ "epoch": 0.44,
2125
+ "learning_rate": 1.1653944020356235e-05,
2126
+ "loss": 0.7973,
2127
+ "step": 353
2128
+ },
2129
+ {
2130
+ "epoch": 0.44,
2131
+ "learning_rate": 1.1628498727735368e-05,
2132
+ "loss": 0.8232,
2133
+ "step": 354
2134
+ },
2135
+ {
2136
+ "epoch": 0.44,
2137
+ "learning_rate": 1.1603053435114503e-05,
2138
+ "loss": 0.7613,
2139
+ "step": 355
2140
+ },
2141
+ {
2142
+ "epoch": 0.44,
2143
+ "learning_rate": 1.157760814249364e-05,
2144
+ "loss": 0.7813,
2145
+ "step": 356
2146
+ },
2147
+ {
2148
+ "epoch": 0.44,
2149
+ "learning_rate": 1.1552162849872775e-05,
2150
+ "loss": 0.7843,
2151
+ "step": 357
2152
+ },
2153
+ {
2154
+ "epoch": 0.44,
2155
+ "learning_rate": 1.152671755725191e-05,
2156
+ "loss": 0.7724,
2157
+ "step": 358
2158
+ },
2159
+ {
2160
+ "epoch": 0.44,
2161
+ "learning_rate": 1.1501272264631044e-05,
2162
+ "loss": 0.7733,
2163
+ "step": 359
2164
+ },
2165
+ {
2166
+ "epoch": 0.44,
2167
+ "learning_rate": 1.147582697201018e-05,
2168
+ "loss": 0.7517,
2169
+ "step": 360
2170
+ },
2171
+ {
2172
+ "epoch": 0.45,
2173
+ "learning_rate": 1.1450381679389312e-05,
2174
+ "loss": 0.7999,
2175
+ "step": 361
2176
+ },
2177
+ {
2178
+ "epoch": 0.45,
2179
+ "learning_rate": 1.1424936386768449e-05,
2180
+ "loss": 0.7879,
2181
+ "step": 362
2182
+ },
2183
+ {
2184
+ "epoch": 0.45,
2185
+ "learning_rate": 1.1399491094147584e-05,
2186
+ "loss": 0.8242,
2187
+ "step": 363
2188
+ },
2189
+ {
2190
+ "epoch": 0.45,
2191
+ "learning_rate": 1.1374045801526718e-05,
2192
+ "loss": 0.7993,
2193
+ "step": 364
2194
+ },
2195
+ {
2196
+ "epoch": 0.45,
2197
+ "learning_rate": 1.1348600508905853e-05,
2198
+ "loss": 0.7994,
2199
+ "step": 365
2200
+ },
2201
+ {
2202
+ "epoch": 0.45,
2203
+ "learning_rate": 1.132315521628499e-05,
2204
+ "loss": 0.7804,
2205
+ "step": 366
2206
+ },
2207
+ {
2208
+ "epoch": 0.45,
2209
+ "learning_rate": 1.1297709923664125e-05,
2210
+ "loss": 0.8016,
2211
+ "step": 367
2212
+ },
2213
+ {
2214
+ "epoch": 0.45,
2215
+ "learning_rate": 1.1272264631043258e-05,
2216
+ "loss": 0.7949,
2217
+ "step": 368
2218
+ },
2219
+ {
2220
+ "epoch": 0.45,
2221
+ "learning_rate": 1.1246819338422393e-05,
2222
+ "loss": 0.7758,
2223
+ "step": 369
2224
+ },
2225
+ {
2226
+ "epoch": 0.46,
2227
+ "learning_rate": 1.1221374045801527e-05,
2228
+ "loss": 0.7772,
2229
+ "step": 370
2230
+ },
2231
+ {
2232
+ "epoch": 0.46,
2233
+ "learning_rate": 1.1195928753180662e-05,
2234
+ "loss": 0.7905,
2235
+ "step": 371
2236
+ },
2237
+ {
2238
+ "epoch": 0.46,
2239
+ "learning_rate": 1.1170483460559797e-05,
2240
+ "loss": 0.7626,
2241
+ "step": 372
2242
+ },
2243
+ {
2244
+ "epoch": 0.46,
2245
+ "learning_rate": 1.1145038167938934e-05,
2246
+ "loss": 0.7646,
2247
+ "step": 373
2248
+ },
2249
+ {
2250
+ "epoch": 0.46,
2251
+ "learning_rate": 1.1119592875318067e-05,
2252
+ "loss": 0.7925,
2253
+ "step": 374
2254
+ },
2255
+ {
2256
+ "epoch": 0.46,
2257
+ "learning_rate": 1.1094147582697202e-05,
2258
+ "loss": 0.8305,
2259
+ "step": 375
2260
+ },
2261
+ {
2262
+ "epoch": 0.46,
2263
+ "learning_rate": 1.1068702290076336e-05,
2264
+ "loss": 0.7818,
2265
+ "step": 376
2266
+ },
2267
+ {
2268
+ "epoch": 0.46,
2269
+ "learning_rate": 1.1043256997455471e-05,
2270
+ "loss": 0.7926,
2271
+ "step": 377
2272
+ },
2273
+ {
2274
+ "epoch": 0.47,
2275
+ "learning_rate": 1.1017811704834606e-05,
2276
+ "loss": 0.8238,
2277
+ "step": 378
2278
+ },
2279
+ {
2280
+ "epoch": 0.47,
2281
+ "learning_rate": 1.0992366412213743e-05,
2282
+ "loss": 0.7864,
2283
+ "step": 379
2284
+ },
2285
+ {
2286
+ "epoch": 0.47,
2287
+ "learning_rate": 1.0966921119592877e-05,
2288
+ "loss": 0.7909,
2289
+ "step": 380
2290
+ },
2291
+ {
2292
+ "epoch": 0.47,
2293
+ "learning_rate": 1.094147582697201e-05,
2294
+ "loss": 0.7571,
2295
+ "step": 381
2296
+ },
2297
+ {
2298
+ "epoch": 0.47,
2299
+ "learning_rate": 1.0916030534351145e-05,
2300
+ "loss": 0.7977,
2301
+ "step": 382
2302
+ },
2303
+ {
2304
+ "epoch": 0.47,
2305
+ "learning_rate": 1.089058524173028e-05,
2306
+ "loss": 0.7998,
2307
+ "step": 383
2308
+ },
2309
+ {
2310
+ "epoch": 0.47,
2311
+ "learning_rate": 1.0865139949109415e-05,
2312
+ "loss": 0.7794,
2313
+ "step": 384
2314
+ },
2315
+ {
2316
+ "epoch": 0.47,
2317
+ "learning_rate": 1.0839694656488552e-05,
2318
+ "loss": 0.816,
2319
+ "step": 385
2320
+ },
2321
+ {
2322
+ "epoch": 0.48,
2323
+ "learning_rate": 1.0814249363867686e-05,
2324
+ "loss": 0.8044,
2325
+ "step": 386
2326
+ },
2327
+ {
2328
+ "epoch": 0.48,
2329
+ "learning_rate": 1.0788804071246821e-05,
2330
+ "loss": 0.7634,
2331
+ "step": 387
2332
+ },
2333
+ {
2334
+ "epoch": 0.48,
2335
+ "learning_rate": 1.0763358778625954e-05,
2336
+ "loss": 0.7763,
2337
+ "step": 388
2338
+ },
2339
+ {
2340
+ "epoch": 0.48,
2341
+ "learning_rate": 1.0737913486005089e-05,
2342
+ "loss": 0.779,
2343
+ "step": 389
2344
+ },
2345
+ {
2346
+ "epoch": 0.48,
2347
+ "learning_rate": 1.0712468193384224e-05,
2348
+ "loss": 0.7838,
2349
+ "step": 390
2350
+ },
2351
+ {
2352
+ "epoch": 0.48,
2353
+ "learning_rate": 1.068702290076336e-05,
2354
+ "loss": 0.7542,
2355
+ "step": 391
2356
+ },
2357
+ {
2358
+ "epoch": 0.48,
2359
+ "learning_rate": 1.0661577608142495e-05,
2360
+ "loss": 0.7927,
2361
+ "step": 392
2362
+ },
2363
+ {
2364
+ "epoch": 0.48,
2365
+ "learning_rate": 1.063613231552163e-05,
2366
+ "loss": 0.7867,
2367
+ "step": 393
2368
+ },
2369
+ {
2370
+ "epoch": 0.49,
2371
+ "learning_rate": 1.0610687022900765e-05,
2372
+ "loss": 0.7875,
2373
+ "step": 394
2374
+ },
2375
+ {
2376
+ "epoch": 0.49,
2377
+ "learning_rate": 1.0585241730279898e-05,
2378
+ "loss": 0.7999,
2379
+ "step": 395
2380
+ },
2381
+ {
2382
+ "epoch": 0.49,
2383
+ "learning_rate": 1.0559796437659033e-05,
2384
+ "loss": 0.7645,
2385
+ "step": 396
2386
+ },
2387
+ {
2388
+ "epoch": 0.49,
2389
+ "learning_rate": 1.0534351145038168e-05,
2390
+ "loss": 0.8013,
2391
+ "step": 397
2392
+ },
2393
+ {
2394
+ "epoch": 0.49,
2395
+ "learning_rate": 1.0508905852417304e-05,
2396
+ "loss": 0.7937,
2397
+ "step": 398
2398
+ },
2399
+ {
2400
+ "epoch": 0.49,
2401
+ "learning_rate": 1.0483460559796439e-05,
2402
+ "loss": 0.7774,
2403
+ "step": 399
2404
+ },
2405
+ {
2406
+ "epoch": 0.49,
2407
+ "learning_rate": 1.0458015267175574e-05,
2408
+ "loss": 0.774,
2409
+ "step": 400
2410
+ },
2411
+ {
2412
+ "epoch": 0.49,
2413
+ "learning_rate": 1.0432569974554709e-05,
2414
+ "loss": 0.8111,
2415
+ "step": 401
2416
+ },
2417
+ {
2418
+ "epoch": 0.5,
2419
+ "learning_rate": 1.0407124681933842e-05,
2420
+ "loss": 0.7667,
2421
+ "step": 402
2422
+ },
2423
+ {
2424
+ "epoch": 0.5,
2425
+ "learning_rate": 1.0381679389312977e-05,
2426
+ "loss": 0.7813,
2427
+ "step": 403
2428
+ },
2429
+ {
2430
+ "epoch": 0.5,
2431
+ "learning_rate": 1.0356234096692113e-05,
2432
+ "loss": 0.7822,
2433
+ "step": 404
2434
+ },
2435
+ {
2436
+ "epoch": 0.5,
2437
+ "learning_rate": 1.0330788804071248e-05,
2438
+ "loss": 0.7816,
2439
+ "step": 405
2440
+ },
2441
+ {
2442
+ "epoch": 0.5,
2443
+ "learning_rate": 1.0305343511450383e-05,
2444
+ "loss": 0.7896,
2445
+ "step": 406
2446
+ },
2447
+ {
2448
+ "epoch": 0.5,
2449
+ "learning_rate": 1.0279898218829518e-05,
2450
+ "loss": 0.7864,
2451
+ "step": 407
2452
+ },
2453
+ {
2454
+ "epoch": 0.5,
2455
+ "learning_rate": 1.025445292620865e-05,
2456
+ "loss": 0.7823,
2457
+ "step": 408
2458
+ },
2459
+ {
2460
+ "epoch": 0.5,
2461
+ "learning_rate": 1.0229007633587786e-05,
2462
+ "loss": 0.7632,
2463
+ "step": 409
2464
+ },
2465
+ {
2466
+ "epoch": 0.51,
2467
+ "learning_rate": 1.0203562340966922e-05,
2468
+ "loss": 0.7953,
2469
+ "step": 410
2470
+ },
2471
+ {
2472
+ "epoch": 0.51,
2473
+ "learning_rate": 1.0178117048346057e-05,
2474
+ "loss": 0.7478,
2475
+ "step": 411
2476
+ },
2477
+ {
2478
+ "epoch": 0.51,
2479
+ "learning_rate": 1.0152671755725192e-05,
2480
+ "loss": 0.8003,
2481
+ "step": 412
2482
+ },
2483
+ {
2484
+ "epoch": 0.51,
2485
+ "learning_rate": 1.0127226463104327e-05,
2486
+ "loss": 0.7848,
2487
+ "step": 413
2488
+ },
2489
+ {
2490
+ "epoch": 0.51,
2491
+ "learning_rate": 1.0101781170483462e-05,
2492
+ "loss": 0.7936,
2493
+ "step": 414
2494
+ },
2495
+ {
2496
+ "epoch": 0.51,
2497
+ "learning_rate": 1.0076335877862595e-05,
2498
+ "loss": 0.8239,
2499
+ "step": 415
2500
+ },
2501
+ {
2502
+ "epoch": 0.51,
2503
+ "learning_rate": 1.0050890585241731e-05,
2504
+ "loss": 0.7778,
2505
+ "step": 416
2506
+ },
2507
+ {
2508
+ "epoch": 0.51,
2509
+ "learning_rate": 1.0025445292620866e-05,
2510
+ "loss": 0.7765,
2511
+ "step": 417
2512
+ },
2513
+ {
2514
+ "epoch": 0.52,
2515
+ "learning_rate": 1e-05,
2516
+ "loss": 0.7456,
2517
+ "step": 418
2518
+ },
2519
+ {
2520
+ "epoch": 0.52,
2521
+ "learning_rate": 9.974554707379136e-06,
2522
+ "loss": 0.7621,
2523
+ "step": 419
2524
+ },
2525
+ {
2526
+ "epoch": 0.52,
2527
+ "learning_rate": 9.94910941475827e-06,
2528
+ "loss": 0.8068,
2529
+ "step": 420
2530
+ },
2531
+ {
2532
+ "epoch": 0.52,
2533
+ "learning_rate": 9.923664122137405e-06,
2534
+ "loss": 0.7886,
2535
+ "step": 421
2536
+ },
2537
+ {
2538
+ "epoch": 0.52,
2539
+ "learning_rate": 9.89821882951654e-06,
2540
+ "loss": 0.7915,
2541
+ "step": 422
2542
+ },
2543
+ {
2544
+ "epoch": 0.52,
2545
+ "learning_rate": 9.872773536895675e-06,
2546
+ "loss": 0.7942,
2547
+ "step": 423
2548
+ },
2549
+ {
2550
+ "epoch": 0.52,
2551
+ "learning_rate": 9.84732824427481e-06,
2552
+ "loss": 0.7917,
2553
+ "step": 424
2554
+ },
2555
+ {
2556
+ "epoch": 0.52,
2557
+ "learning_rate": 9.821882951653945e-06,
2558
+ "loss": 0.7602,
2559
+ "step": 425
2560
+ },
2561
+ {
2562
+ "epoch": 0.53,
2563
+ "learning_rate": 9.79643765903308e-06,
2564
+ "loss": 0.7677,
2565
+ "step": 426
2566
+ },
2567
+ {
2568
+ "epoch": 0.53,
2569
+ "learning_rate": 9.770992366412214e-06,
2570
+ "loss": 0.7593,
2571
+ "step": 427
2572
+ },
2573
+ {
2574
+ "epoch": 0.53,
2575
+ "learning_rate": 9.745547073791349e-06,
2576
+ "loss": 0.7761,
2577
+ "step": 428
2578
+ },
2579
+ {
2580
+ "epoch": 0.53,
2581
+ "learning_rate": 9.720101781170484e-06,
2582
+ "loss": 0.7617,
2583
+ "step": 429
2584
+ },
2585
+ {
2586
+ "epoch": 0.53,
2587
+ "learning_rate": 9.694656488549619e-06,
2588
+ "loss": 0.797,
2589
+ "step": 430
2590
+ },
2591
+ {
2592
+ "epoch": 0.53,
2593
+ "learning_rate": 9.669211195928754e-06,
2594
+ "loss": 0.7706,
2595
+ "step": 431
2596
+ },
2597
+ {
2598
+ "epoch": 0.53,
2599
+ "learning_rate": 9.643765903307888e-06,
2600
+ "loss": 0.773,
2601
+ "step": 432
2602
+ },
2603
+ {
2604
+ "epoch": 0.53,
2605
+ "learning_rate": 9.618320610687025e-06,
2606
+ "loss": 0.7392,
2607
+ "step": 433
2608
+ },
2609
+ {
2610
+ "epoch": 0.54,
2611
+ "learning_rate": 9.592875318066158e-06,
2612
+ "loss": 0.8147,
2613
+ "step": 434
2614
+ },
2615
+ {
2616
+ "epoch": 0.54,
2617
+ "learning_rate": 9.567430025445293e-06,
2618
+ "loss": 0.7338,
2619
+ "step": 435
2620
+ },
2621
+ {
2622
+ "epoch": 0.54,
2623
+ "learning_rate": 9.54198473282443e-06,
2624
+ "loss": 0.7699,
2625
+ "step": 436
2626
+ },
2627
+ {
2628
+ "epoch": 0.54,
2629
+ "learning_rate": 9.516539440203563e-06,
2630
+ "loss": 0.7778,
2631
+ "step": 437
2632
+ },
2633
+ {
2634
+ "epoch": 0.54,
2635
+ "learning_rate": 9.491094147582697e-06,
2636
+ "loss": 0.7772,
2637
+ "step": 438
2638
+ },
2639
+ {
2640
+ "epoch": 0.54,
2641
+ "learning_rate": 9.465648854961834e-06,
2642
+ "loss": 0.7823,
2643
+ "step": 439
2644
+ },
2645
+ {
2646
+ "epoch": 0.54,
2647
+ "learning_rate": 9.440203562340969e-06,
2648
+ "loss": 0.7403,
2649
+ "step": 440
2650
+ },
2651
+ {
2652
+ "epoch": 0.54,
2653
+ "learning_rate": 9.414758269720102e-06,
2654
+ "loss": 0.7894,
2655
+ "step": 441
2656
+ },
2657
+ {
2658
+ "epoch": 0.55,
2659
+ "learning_rate": 9.389312977099237e-06,
2660
+ "loss": 0.7673,
2661
+ "step": 442
2662
+ },
2663
+ {
2664
+ "epoch": 0.55,
2665
+ "learning_rate": 9.363867684478373e-06,
2666
+ "loss": 0.7633,
2667
+ "step": 443
2668
+ },
2669
+ {
2670
+ "epoch": 0.55,
2671
+ "learning_rate": 9.338422391857506e-06,
2672
+ "loss": 0.7629,
2673
+ "step": 444
2674
+ },
2675
+ {
2676
+ "epoch": 0.55,
2677
+ "learning_rate": 9.312977099236641e-06,
2678
+ "loss": 0.7795,
2679
+ "step": 445
2680
+ },
2681
+ {
2682
+ "epoch": 0.55,
2683
+ "learning_rate": 9.287531806615778e-06,
2684
+ "loss": 0.7861,
2685
+ "step": 446
2686
+ },
2687
+ {
2688
+ "epoch": 0.55,
2689
+ "learning_rate": 9.26208651399491e-06,
2690
+ "loss": 0.7865,
2691
+ "step": 447
2692
+ },
2693
+ {
2694
+ "epoch": 0.55,
2695
+ "learning_rate": 9.236641221374046e-06,
2696
+ "loss": 0.7722,
2697
+ "step": 448
2698
+ },
2699
+ {
2700
+ "epoch": 0.55,
2701
+ "learning_rate": 9.211195928753182e-06,
2702
+ "loss": 0.7991,
2703
+ "step": 449
2704
+ },
2705
+ {
2706
+ "epoch": 0.55,
2707
+ "learning_rate": 9.185750636132317e-06,
2708
+ "loss": 0.7732,
2709
+ "step": 450
2710
+ },
2711
+ {
2712
+ "epoch": 0.56,
2713
+ "learning_rate": 9.16030534351145e-06,
2714
+ "loss": 0.7687,
2715
+ "step": 451
2716
+ },
2717
+ {
2718
+ "epoch": 0.56,
2719
+ "learning_rate": 9.134860050890587e-06,
2720
+ "loss": 0.7663,
2721
+ "step": 452
2722
+ },
2723
+ {
2724
+ "epoch": 0.56,
2725
+ "learning_rate": 9.109414758269721e-06,
2726
+ "loss": 0.7762,
2727
+ "step": 453
2728
+ },
2729
+ {
2730
+ "epoch": 0.56,
2731
+ "learning_rate": 9.083969465648855e-06,
2732
+ "loss": 0.7643,
2733
+ "step": 454
2734
+ },
2735
+ {
2736
+ "epoch": 0.56,
2737
+ "learning_rate": 9.058524173027991e-06,
2738
+ "loss": 0.7694,
2739
+ "step": 455
2740
+ },
2741
+ {
2742
+ "epoch": 0.56,
2743
+ "learning_rate": 9.033078880407126e-06,
2744
+ "loss": 0.7918,
2745
+ "step": 456
2746
+ },
2747
+ {
2748
+ "epoch": 0.56,
2749
+ "learning_rate": 9.007633587786259e-06,
2750
+ "loss": 0.7893,
2751
+ "step": 457
2752
+ },
2753
+ {
2754
+ "epoch": 0.56,
2755
+ "learning_rate": 8.982188295165396e-06,
2756
+ "loss": 0.7898,
2757
+ "step": 458
2758
+ },
2759
+ {
2760
+ "epoch": 0.57,
2761
+ "learning_rate": 8.95674300254453e-06,
2762
+ "loss": 0.7376,
2763
+ "step": 459
2764
+ },
2765
+ {
2766
+ "epoch": 0.57,
2767
+ "learning_rate": 8.931297709923665e-06,
2768
+ "loss": 0.793,
2769
+ "step": 460
2770
+ },
2771
+ {
2772
+ "epoch": 0.57,
2773
+ "learning_rate": 8.9058524173028e-06,
2774
+ "loss": 0.8003,
2775
+ "step": 461
2776
+ },
2777
+ {
2778
+ "epoch": 0.57,
2779
+ "learning_rate": 8.880407124681935e-06,
2780
+ "loss": 0.7373,
2781
+ "step": 462
2782
+ },
2783
+ {
2784
+ "epoch": 0.57,
2785
+ "learning_rate": 8.85496183206107e-06,
2786
+ "loss": 0.749,
2787
+ "step": 463
2788
+ },
2789
+ {
2790
+ "epoch": 0.57,
2791
+ "learning_rate": 8.829516539440205e-06,
2792
+ "loss": 0.7714,
2793
+ "step": 464
2794
+ },
2795
+ {
2796
+ "epoch": 0.57,
2797
+ "learning_rate": 8.80407124681934e-06,
2798
+ "loss": 0.8103,
2799
+ "step": 465
2800
+ },
2801
+ {
2802
+ "epoch": 0.57,
2803
+ "learning_rate": 8.778625954198474e-06,
2804
+ "loss": 0.7754,
2805
+ "step": 466
2806
+ },
2807
+ {
2808
+ "epoch": 0.58,
2809
+ "learning_rate": 8.753180661577609e-06,
2810
+ "loss": 0.8023,
2811
+ "step": 467
2812
+ },
2813
+ {
2814
+ "epoch": 0.58,
2815
+ "learning_rate": 8.727735368956744e-06,
2816
+ "loss": 0.7566,
2817
+ "step": 468
2818
+ },
2819
+ {
2820
+ "epoch": 0.58,
2821
+ "learning_rate": 8.702290076335879e-06,
2822
+ "loss": 0.7562,
2823
+ "step": 469
2824
+ },
2825
+ {
2826
+ "epoch": 0.58,
2827
+ "learning_rate": 8.676844783715014e-06,
2828
+ "loss": 0.7638,
2829
+ "step": 470
2830
+ },
2831
+ {
2832
+ "epoch": 0.58,
2833
+ "learning_rate": 8.651399491094148e-06,
2834
+ "loss": 0.7629,
2835
+ "step": 471
2836
+ },
2837
+ {
2838
+ "epoch": 0.58,
2839
+ "learning_rate": 8.625954198473283e-06,
2840
+ "loss": 0.7708,
2841
+ "step": 472
2842
+ },
2843
+ {
2844
+ "epoch": 0.58,
2845
+ "learning_rate": 8.600508905852418e-06,
2846
+ "loss": 0.7399,
2847
+ "step": 473
2848
+ },
2849
+ {
2850
+ "epoch": 0.58,
2851
+ "learning_rate": 8.575063613231553e-06,
2852
+ "loss": 0.7733,
2853
+ "step": 474
2854
+ },
2855
+ {
2856
+ "epoch": 0.59,
2857
+ "learning_rate": 8.549618320610688e-06,
2858
+ "loss": 0.7958,
2859
+ "step": 475
2860
+ },
2861
+ {
2862
+ "epoch": 0.59,
2863
+ "learning_rate": 8.524173027989823e-06,
2864
+ "loss": 0.771,
2865
+ "step": 476
2866
+ },
2867
+ {
2868
+ "epoch": 0.59,
2869
+ "learning_rate": 8.498727735368957e-06,
2870
+ "loss": 0.7736,
2871
+ "step": 477
2872
+ },
2873
+ {
2874
+ "epoch": 0.59,
2875
+ "learning_rate": 8.473282442748092e-06,
2876
+ "loss": 0.7686,
2877
+ "step": 478
2878
+ },
2879
+ {
2880
+ "epoch": 0.59,
2881
+ "learning_rate": 8.447837150127227e-06,
2882
+ "loss": 0.7704,
2883
+ "step": 479
2884
+ },
2885
+ {
2886
+ "epoch": 0.59,
2887
+ "learning_rate": 8.422391857506362e-06,
2888
+ "loss": 0.7755,
2889
+ "step": 480
2890
+ },
2891
+ {
2892
+ "epoch": 0.59,
2893
+ "learning_rate": 8.396946564885497e-06,
2894
+ "loss": 0.7795,
2895
+ "step": 481
2896
+ },
2897
+ {
2898
+ "epoch": 0.59,
2899
+ "learning_rate": 8.371501272264631e-06,
2900
+ "loss": 0.7885,
2901
+ "step": 482
2902
+ },
2903
+ {
2904
+ "epoch": 0.6,
2905
+ "learning_rate": 8.346055979643766e-06,
2906
+ "loss": 0.7562,
2907
+ "step": 483
2908
+ },
2909
+ {
2910
+ "epoch": 0.6,
2911
+ "learning_rate": 8.320610687022901e-06,
2912
+ "loss": 0.769,
2913
+ "step": 484
2914
+ },
2915
+ {
2916
+ "epoch": 0.6,
2917
+ "learning_rate": 8.295165394402036e-06,
2918
+ "loss": 0.7976,
2919
+ "step": 485
2920
+ },
2921
+ {
2922
+ "epoch": 0.6,
2923
+ "learning_rate": 8.26972010178117e-06,
2924
+ "loss": 0.7762,
2925
+ "step": 486
2926
+ },
2927
+ {
2928
+ "epoch": 0.6,
2929
+ "learning_rate": 8.244274809160306e-06,
2930
+ "loss": 0.7615,
2931
+ "step": 487
2932
+ },
2933
+ {
2934
+ "epoch": 0.6,
2935
+ "learning_rate": 8.21882951653944e-06,
2936
+ "loss": 0.7555,
2937
+ "step": 488
2938
+ },
2939
+ {
2940
+ "epoch": 0.6,
2941
+ "learning_rate": 8.193384223918575e-06,
2942
+ "loss": 0.7615,
2943
+ "step": 489
2944
+ },
2945
+ {
2946
+ "epoch": 0.6,
2947
+ "learning_rate": 8.16793893129771e-06,
2948
+ "loss": 0.7778,
2949
+ "step": 490
2950
+ },
2951
+ {
2952
+ "epoch": 0.61,
2953
+ "learning_rate": 8.142493638676845e-06,
2954
+ "loss": 0.8184,
2955
+ "step": 491
2956
+ },
2957
+ {
2958
+ "epoch": 0.61,
2959
+ "learning_rate": 8.11704834605598e-06,
2960
+ "loss": 0.7501,
2961
+ "step": 492
2962
+ },
2963
+ {
2964
+ "epoch": 0.61,
2965
+ "learning_rate": 8.091603053435115e-06,
2966
+ "loss": 0.7728,
2967
+ "step": 493
2968
+ },
2969
+ {
2970
+ "epoch": 0.61,
2971
+ "learning_rate": 8.066157760814251e-06,
2972
+ "loss": 0.8017,
2973
+ "step": 494
2974
+ },
2975
+ {
2976
+ "epoch": 0.61,
2977
+ "learning_rate": 8.040712468193384e-06,
2978
+ "loss": 0.8018,
2979
+ "step": 495
2980
+ },
2981
+ {
2982
+ "epoch": 0.61,
2983
+ "learning_rate": 8.015267175572519e-06,
2984
+ "loss": 0.7942,
2985
+ "step": 496
2986
+ },
2987
+ {
2988
+ "epoch": 0.61,
2989
+ "learning_rate": 7.989821882951656e-06,
2990
+ "loss": 0.7882,
2991
+ "step": 497
2992
+ },
2993
+ {
2994
+ "epoch": 0.61,
2995
+ "learning_rate": 7.964376590330789e-06,
2996
+ "loss": 0.7822,
2997
+ "step": 498
2998
+ },
2999
+ {
3000
+ "epoch": 0.62,
3001
+ "learning_rate": 7.938931297709924e-06,
3002
+ "loss": 0.7923,
3003
+ "step": 499
3004
+ },
3005
+ {
3006
+ "epoch": 0.62,
3007
+ "learning_rate": 7.91348600508906e-06,
3008
+ "loss": 0.7733,
3009
+ "step": 500
3010
+ }
3011
+ ],
3012
+ "logging_steps": 1.0,
3013
+ "max_steps": 811,
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 sha256:9ce16922e39dce9c6f5794bc20d08e280ea7976c4c32fb1cad8a78b8b615f56b
3
+ size 4792
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 sha256:38a8636ac4d2c67615b92668093722e128fa806f1f3fd210d79979bba2143101
3
+ size 14483567250
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 sha256:9ce16922e39dce9c6f5794bc20d08e280ea7976c4c32fb1cad8a78b8b615f56b
3
+ size 4792