Ming Zhong commited on
Commit
b6c7136
1 Parent(s): 6f3d7c7

First model version

Browse files
config.json ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "allenai/led-large-16384",
3
+ "_num_labels": 3,
4
+ "activation_dropout": 0.0,
5
+ "activation_function": "gelu",
6
+ "architectures": [
7
+ "LEDForConditionalGeneration"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "attention_window": [
11
+ 1024,
12
+ 1024,
13
+ 1024,
14
+ 1024,
15
+ 1024,
16
+ 1024,
17
+ 1024,
18
+ 1024,
19
+ 1024,
20
+ 1024,
21
+ 1024,
22
+ 1024
23
+ ],
24
+ "bos_token_id": 0,
25
+ "classif_dropout": 0.0,
26
+ "classifier_dropout": 0.0,
27
+ "d_model": 1024,
28
+ "decoder_attention_heads": 16,
29
+ "decoder_ffn_dim": 4096,
30
+ "decoder_layerdrop": 0.0,
31
+ "decoder_layers": 12,
32
+ "decoder_start_token_id": 2,
33
+ "dropout": 0.1,
34
+ "early_stopping": true,
35
+ "encoder_attention_heads": 16,
36
+ "encoder_ffn_dim": 4096,
37
+ "encoder_layerdrop": 0.0,
38
+ "encoder_layers": 12,
39
+ "eos_token_id": 2,
40
+ "gradient_checkpointing": false,
41
+ "id2label": {
42
+ "0": "LABEL_0",
43
+ "1": "LABEL_1",
44
+ "2": "LABEL_2"
45
+ },
46
+ "init_std": 0.02,
47
+ "is_encoder_decoder": true,
48
+ "label2id": {
49
+ "LABEL_0": 0,
50
+ "LABEL_1": 1,
51
+ "LABEL_2": 2
52
+ },
53
+ "max_decoder_position_embeddings": 1024,
54
+ "max_encoder_position_embeddings": 16384,
55
+ "max_length": 512,
56
+ "min_length": 100,
57
+ "model_type": "led",
58
+ "no_repeat_ngram_size": 3,
59
+ "num_beams": 6,
60
+ "num_hidden_layers": 12,
61
+ "output_past": false,
62
+ "pad_token_id": 1,
63
+ "prefix": " ",
64
+ "torch_dtype": "float32",
65
+ "transformers_version": "4.9.0.dev0",
66
+ "use_cache": true,
67
+ "vocab_size": 50265
68
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:030e327e57779cb8fc841a508f95a0faf36be77001c9420335469c9b86e08d68
3
+ size 3376450397
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d29393295cd9b6a0820ea0bb3ce16e6364d6a592af89e5aba76c0922863b1237
3
+ size 1839641975
rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a1f227facfe2afc8141bbbf0e007c1bc251fb600999ae14c368de0c260f71333
3
+ size 14593
rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2f6440254c8fc19122aaa1f143af4bd1efead550d2ee279a87e782cd47c9555
3
+ size 14593
rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca7555f50fda592bd167059addcf8df4181a0f355353217ac90bdd5920c4469e
3
+ size 14593
rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:629fe9e409112ecb2d95376d5e675725e86c553d34ac49b0eed97b4dbe1b029f
3
+ size 14593
rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94d8a5986707247d4dad0ebb0f4899a0e1fce43af8451f5412fe5b85f196b418
3
+ size 14593
rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16dd1c9912922fa8dc1b0fe3b98de5fa475ff7825af0caa39e6a0fb2b3dffb3a
3
+ size 14593
rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:527b7b0f21459bb00f20bc44bdbfe1355d903b9d73d9b65816d1eeb7b742feee
3
+ size 14593
rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a1e3f35db2ea82691debdfa5a14c5847118f2a40b9da96e6a0a8347de5beca27
3
+ size 14593
scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a05da6f1312b4ace39496d59f19dc618e85abb6b796f34fce4eca0a29a7dcf3c
3
+ size 559
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:40698bd132dea0594abce82df2ee77ed9206bd0f1b5c528ec260f5092aaef7ea
3
+ size 623
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "eos_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "unk_token": {"content": "<unk>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "sep_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "pad_token": {"content": "<pad>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "cls_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true}}
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": {"content": "<unk>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "bos_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "eos_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "add_prefix_space": false, "errors": "replace", "sep_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "cls_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "pad_token": {"content": "<pad>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "model_max_length": 16384, "special_tokens_map_file": "/home/t-zhongming/.cache/huggingface/transformers/2ad921573d53ebf0c0450d63a211e61d8e328324e84830c365abff01f2d115f1.cb2244924ab24d706b02fd7fcedaea4531566537687a539ebb94db511fd122a0", "name_or_path": "allenai/led-large-16384", "tokenizer_class": "LEDTokenizer"}
trainer_state.json ADDED
@@ -0,0 +1,2032 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 19.999910828404126,
5
+ "global_step": 168200,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.06,
12
+ "learning_rate": 2.3305588585017838e-07,
13
+ "loss": 1.5468,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.12,
18
+ "learning_rate": 4.7086801426872774e-07,
19
+ "loss": 1.1544,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.18,
24
+ "learning_rate": 7.082045184304399e-07,
25
+ "loss": 1.0372,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.24,
30
+ "learning_rate": 9.460166468489894e-07,
31
+ "loss": 0.9823,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.3,
36
+ "learning_rate": 1.1838287752675386e-06,
37
+ "loss": 0.9462,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.36,
42
+ "learning_rate": 1.4216409036860881e-06,
43
+ "loss": 0.9246,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.42,
48
+ "learning_rate": 1.6594530321046376e-06,
49
+ "loss": 0.8938,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.48,
54
+ "learning_rate": 1.8972651605231867e-06,
55
+ "loss": 0.8919,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.54,
60
+ "learning_rate": 2.135077288941736e-06,
61
+ "loss": 0.863,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.59,
66
+ "learning_rate": 2.3728894173602857e-06,
67
+ "loss": 0.8513,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.65,
72
+ "learning_rate": 2.6107015457788348e-06,
73
+ "loss": 0.8337,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.71,
78
+ "learning_rate": 2.8485136741973847e-06,
79
+ "loss": 0.8228,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.77,
84
+ "learning_rate": 3.0863258026159338e-06,
85
+ "loss": 0.787,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.83,
90
+ "learning_rate": 3.324137931034483e-06,
91
+ "loss": 0.7696,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.89,
96
+ "learning_rate": 3.5619500594530328e-06,
97
+ "loss": 0.7353,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.95,
102
+ "learning_rate": 3.799762187871582e-06,
103
+ "loss": 0.7291,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 1.01,
108
+ "learning_rate": 4.037574316290131e-06,
109
+ "loss": 0.7122,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 1.07,
114
+ "learning_rate": 4.27538644470868e-06,
115
+ "loss": 0.7098,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 1.13,
120
+ "learning_rate": 4.5131985731272295e-06,
121
+ "loss": 0.6884,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 1.19,
126
+ "learning_rate": 4.751010701545779e-06,
127
+ "loss": 0.678,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 1.25,
132
+ "learning_rate": 4.9888228299643285e-06,
133
+ "loss": 0.6688,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 1.31,
138
+ "learning_rate": 5.2266349583828776e-06,
139
+ "loss": 0.6706,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 1.37,
144
+ "learning_rate": 5.4644470868014275e-06,
145
+ "loss": 0.6705,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 1.43,
150
+ "learning_rate": 5.7022592152199766e-06,
151
+ "loss": 0.6593,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 1.49,
156
+ "learning_rate": 5.940071343638526e-06,
157
+ "loss": 0.669,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 1.55,
162
+ "learning_rate": 6.1778834720570756e-06,
163
+ "loss": 0.6495,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 1.61,
168
+ "learning_rate": 6.415695600475625e-06,
169
+ "loss": 0.6554,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 1.66,
174
+ "learning_rate": 6.653507728894174e-06,
175
+ "loss": 0.6467,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 1.72,
180
+ "learning_rate": 6.891319857312724e-06,
181
+ "loss": 0.6479,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 1.78,
186
+ "learning_rate": 7.128180737217598e-06,
187
+ "loss": 0.643,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 1.84,
192
+ "learning_rate": 7.365992865636148e-06,
193
+ "loss": 0.6369,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 1.9,
198
+ "learning_rate": 7.603804994054697e-06,
199
+ "loss": 0.6363,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 1.96,
204
+ "learning_rate": 7.841617122473245e-06,
205
+ "loss": 0.6272,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 2.02,
210
+ "learning_rate": 8.079429250891796e-06,
211
+ "loss": 0.6269,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 2.08,
216
+ "learning_rate": 8.316765755053508e-06,
217
+ "loss": 0.6116,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 2.14,
222
+ "learning_rate": 8.554577883472059e-06,
223
+ "loss": 0.6149,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 2.2,
228
+ "learning_rate": 8.792390011890606e-06,
229
+ "loss": 0.6143,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 2.26,
234
+ "learning_rate": 9.030202140309157e-06,
235
+ "loss": 0.6139,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 2.32,
240
+ "learning_rate": 9.268014268727706e-06,
241
+ "loss": 0.6048,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 2.38,
246
+ "learning_rate": 9.505826397146255e-06,
247
+ "loss": 0.6095,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 2.44,
252
+ "learning_rate": 9.743638525564804e-06,
253
+ "loss": 0.5988,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 2.5,
258
+ "learning_rate": 9.981450653983355e-06,
259
+ "loss": 0.595,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 2.56,
264
+ "learning_rate": 1.0219262782401904e-05,
265
+ "loss": 0.6005,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 2.62,
270
+ "learning_rate": 1.0457074910820453e-05,
271
+ "loss": 0.598,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 2.68,
276
+ "learning_rate": 1.0694411414982165e-05,
277
+ "loss": 0.5968,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 2.73,
282
+ "learning_rate": 1.0932223543400716e-05,
283
+ "loss": 0.5934,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 2.79,
288
+ "learning_rate": 1.1170035671819263e-05,
289
+ "loss": 0.5965,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 2.85,
294
+ "learning_rate": 1.1407847800237812e-05,
295
+ "loss": 0.5946,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 2.91,
300
+ "learning_rate": 1.1645659928656363e-05,
301
+ "loss": 0.5904,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 2.97,
306
+ "learning_rate": 1.1882996432818075e-05,
307
+ "loss": 0.5917,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 3.03,
312
+ "learning_rate": 1.2120808561236624e-05,
313
+ "loss": 0.5802,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 3.09,
318
+ "learning_rate": 1.2358620689655175e-05,
319
+ "loss": 0.5753,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 3.15,
324
+ "learning_rate": 1.2595957193816886e-05,
325
+ "loss": 0.5687,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 3.21,
330
+ "learning_rate": 1.2833769322235434e-05,
331
+ "loss": 0.5755,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 3.27,
336
+ "learning_rate": 1.3071581450653985e-05,
337
+ "loss": 0.5718,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 3.33,
342
+ "learning_rate": 1.3309393579072534e-05,
343
+ "loss": 0.5759,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 3.39,
348
+ "learning_rate": 1.3547205707491083e-05,
349
+ "loss": 0.5748,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 3.45,
354
+ "learning_rate": 1.3785017835909633e-05,
355
+ "loss": 0.571,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 3.51,
360
+ "learning_rate": 1.4022829964328183e-05,
361
+ "loss": 0.5713,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 3.57,
366
+ "learning_rate": 1.426064209274673e-05,
367
+ "loss": 0.5688,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 3.63,
372
+ "learning_rate": 1.4497978596908443e-05,
373
+ "loss": 0.5744,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 3.69,
378
+ "learning_rate": 1.4735790725326993e-05,
379
+ "loss": 0.573,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 3.75,
384
+ "learning_rate": 1.4973602853745542e-05,
385
+ "loss": 0.5711,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 3.8,
390
+ "learning_rate": 1.5211414982164092e-05,
391
+ "loss": 0.5679,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 3.86,
396
+ "learning_rate": 1.544922711058264e-05,
397
+ "loss": 0.5702,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 3.92,
402
+ "learning_rate": 1.568703923900119e-05,
403
+ "loss": 0.57,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 3.98,
408
+ "learning_rate": 1.592485136741974e-05,
409
+ "loss": 0.5659,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 4.04,
414
+ "learning_rate": 1.616266349583829e-05,
415
+ "loss": 0.5553,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 4.1,
420
+ "learning_rate": 1.6400475624256836e-05,
421
+ "loss": 0.5508,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 4.16,
426
+ "learning_rate": 1.663828775267539e-05,
427
+ "loss": 0.5466,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 4.22,
432
+ "learning_rate": 1.6876099881093938e-05,
433
+ "loss": 0.5479,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 4.28,
438
+ "learning_rate": 1.7113912009512487e-05,
439
+ "loss": 0.5472,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 4.34,
444
+ "learning_rate": 1.73512485136742e-05,
445
+ "loss": 0.5462,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 4.4,
450
+ "learning_rate": 1.758858501783591e-05,
451
+ "loss": 0.5517,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 4.46,
456
+ "learning_rate": 1.782639714625446e-05,
457
+ "loss": 0.5448,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 4.52,
462
+ "learning_rate": 1.8064209274673012e-05,
463
+ "loss": 0.5473,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 4.58,
468
+ "learning_rate": 1.8302021403091558e-05,
469
+ "loss": 0.547,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 4.64,
474
+ "learning_rate": 1.8539833531510107e-05,
475
+ "loss": 0.5481,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 4.7,
480
+ "learning_rate": 1.877764565992866e-05,
481
+ "loss": 0.5534,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 4.76,
486
+ "learning_rate": 1.901498216409037e-05,
487
+ "loss": 0.5456,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 4.82,
492
+ "learning_rate": 1.925279429250892e-05,
493
+ "loss": 0.5543,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 4.88,
498
+ "learning_rate": 1.949060642092747e-05,
499
+ "loss": 0.5462,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 4.93,
504
+ "learning_rate": 1.9728418549346018e-05,
505
+ "loss": 0.5443,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 4.99,
510
+ "learning_rate": 1.9966230677764567e-05,
511
+ "loss": 0.5429,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 5.05,
516
+ "learning_rate": 1.9977328577090766e-05,
517
+ "loss": 0.5261,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 5.11,
522
+ "learning_rate": 1.9950905007266485e-05,
523
+ "loss": 0.521,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 5.17,
528
+ "learning_rate": 1.99244814374422e-05,
529
+ "loss": 0.5263,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 5.23,
534
+ "learning_rate": 1.9898057867617915e-05,
535
+ "loss": 0.5268,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 5.29,
540
+ "learning_rate": 1.9871634297793634e-05,
541
+ "loss": 0.5232,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 5.35,
546
+ "learning_rate": 1.9845263575108998e-05,
547
+ "loss": 0.5304,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 5.41,
552
+ "learning_rate": 1.9818840005284717e-05,
553
+ "loss": 0.52,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 5.47,
558
+ "learning_rate": 1.9792416435460432e-05,
559
+ "loss": 0.5266,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 5.53,
564
+ "learning_rate": 1.976599286563615e-05,
565
+ "loss": 0.5311,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 5.59,
570
+ "learning_rate": 1.9739622142951514e-05,
571
+ "loss": 0.526,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 5.65,
576
+ "learning_rate": 1.971319857312723e-05,
577
+ "loss": 0.5272,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 5.71,
582
+ "learning_rate": 1.9686880697582245e-05,
583
+ "loss": 0.5259,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 5.77,
588
+ "learning_rate": 1.966045712775796e-05,
589
+ "loss": 0.5324,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 5.83,
594
+ "learning_rate": 1.963403355793368e-05,
595
+ "loss": 0.532,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 5.89,
600
+ "learning_rate": 1.9607609988109395e-05,
601
+ "loss": 0.525,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 5.95,
606
+ "learning_rate": 1.9581186418285113e-05,
607
+ "loss": 0.528,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 6.0,
612
+ "learning_rate": 1.955476284846083e-05,
613
+ "loss": 0.5291,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 6.06,
618
+ "learning_rate": 1.9528339278636547e-05,
619
+ "loss": 0.5051,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 6.12,
624
+ "learning_rate": 1.9501915708812262e-05,
625
+ "loss": 0.497,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 6.18,
630
+ "learning_rate": 1.9475492138987978e-05,
631
+ "loss": 0.5067,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 6.24,
636
+ "learning_rate": 1.9449068569163696e-05,
637
+ "loss": 0.5062,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 6.3,
642
+ "learning_rate": 1.942264499933941e-05,
643
+ "loss": 0.5046,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 6.36,
648
+ "learning_rate": 1.939622142951513e-05,
649
+ "loss": 0.5094,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 6.42,
654
+ "learning_rate": 1.9369797859690846e-05,
655
+ "loss": 0.5061,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 6.48,
660
+ "learning_rate": 1.9343374289866564e-05,
661
+ "loss": 0.512,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 6.54,
666
+ "learning_rate": 1.931695072004228e-05,
667
+ "loss": 0.5048,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 6.6,
672
+ "learning_rate": 1.9290579997357643e-05,
673
+ "loss": 0.5083,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 6.66,
678
+ "learning_rate": 1.9264156427533362e-05,
679
+ "loss": 0.5021,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 6.72,
684
+ "learning_rate": 1.9237732857709077e-05,
685
+ "loss": 0.5067,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 6.78,
690
+ "learning_rate": 1.9211309287884793e-05,
691
+ "loss": 0.5107,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 6.84,
696
+ "learning_rate": 1.918488571806051e-05,
697
+ "loss": 0.5121,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 6.9,
702
+ "learning_rate": 1.9158462148236226e-05,
703
+ "loss": 0.5092,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 6.96,
708
+ "learning_rate": 1.9132091425551594e-05,
709
+ "loss": 0.5106,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 7.02,
714
+ "learning_rate": 1.910566785572731e-05,
715
+ "loss": 0.5081,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 7.07,
720
+ "learning_rate": 1.9079244285903028e-05,
721
+ "loss": 0.4826,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 7.13,
726
+ "learning_rate": 1.905287356321839e-05,
727
+ "loss": 0.4854,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 7.19,
732
+ "learning_rate": 1.902644999339411e-05,
733
+ "loss": 0.48,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 7.25,
738
+ "learning_rate": 1.9000026423569825e-05,
739
+ "loss": 0.4918,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 7.31,
744
+ "learning_rate": 1.8973602853745544e-05,
745
+ "loss": 0.4867,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 7.37,
750
+ "learning_rate": 1.8947232131060908e-05,
751
+ "loss": 0.4935,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 7.43,
756
+ "learning_rate": 1.8920808561236626e-05,
757
+ "loss": 0.4855,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 7.49,
762
+ "learning_rate": 1.8894384991412342e-05,
763
+ "loss": 0.4907,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 7.55,
768
+ "learning_rate": 1.886796142158806e-05,
769
+ "loss": 0.4913,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 7.61,
774
+ "learning_rate": 1.8841537851763776e-05,
775
+ "loss": 0.4914,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 7.67,
780
+ "learning_rate": 1.881511428193949e-05,
781
+ "loss": 0.4962,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 7.73,
786
+ "learning_rate": 1.8788690712115206e-05,
787
+ "loss": 0.4904,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 7.79,
792
+ "learning_rate": 1.8762267142290925e-05,
793
+ "loss": 0.4927,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 7.85,
798
+ "learning_rate": 1.873584357246664e-05,
799
+ "loss": 0.4934,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 7.91,
804
+ "learning_rate": 1.8709472849782007e-05,
805
+ "loss": 0.4926,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 7.97,
810
+ "learning_rate": 1.8683049279957723e-05,
811
+ "loss": 0.4965,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 8.03,
816
+ "learning_rate": 1.865662571013344e-05,
817
+ "loss": 0.4783,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 8.09,
822
+ "learning_rate": 1.8630202140309157e-05,
823
+ "loss": 0.4617,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 8.15,
828
+ "learning_rate": 1.8603831417624524e-05,
829
+ "loss": 0.4649,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 8.2,
834
+ "learning_rate": 1.857740784780024e-05,
835
+ "loss": 0.4696,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 8.26,
840
+ "learning_rate": 1.8550984277975958e-05,
841
+ "loss": 0.467,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 8.32,
846
+ "learning_rate": 1.8524560708151673e-05,
847
+ "loss": 0.4689,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 8.38,
852
+ "learning_rate": 1.849813713832739e-05,
853
+ "loss": 0.4686,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 8.44,
858
+ "learning_rate": 1.8471766415642755e-05,
859
+ "loss": 0.476,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 8.5,
864
+ "learning_rate": 1.8445342845818474e-05,
865
+ "loss": 0.4721,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 8.56,
870
+ "learning_rate": 1.841891927599419e-05,
871
+ "loss": 0.483,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 8.62,
876
+ "learning_rate": 1.8392495706169905e-05,
877
+ "loss": 0.4728,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 8.68,
882
+ "learning_rate": 1.836607213634562e-05,
883
+ "loss": 0.4742,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 8.74,
888
+ "learning_rate": 1.833964856652134e-05,
889
+ "loss": 0.4774,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 8.8,
894
+ "learning_rate": 1.8313277843836702e-05,
895
+ "loss": 0.4801,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 8.86,
900
+ "learning_rate": 1.828690712115207e-05,
901
+ "loss": 0.4777,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 8.92,
906
+ "learning_rate": 1.8260483551327785e-05,
907
+ "loss": 0.4803,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 8.98,
912
+ "learning_rate": 1.8234059981503504e-05,
913
+ "loss": 0.4783,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 9.04,
918
+ "learning_rate": 1.820763641167922e-05,
919
+ "loss": 0.4596,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 9.1,
924
+ "learning_rate": 1.8181212841854937e-05,
925
+ "loss": 0.4442,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 9.16,
930
+ "learning_rate": 1.8154789272030653e-05,
931
+ "loss": 0.46,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 9.22,
936
+ "learning_rate": 1.812836570220637e-05,
937
+ "loss": 0.4597,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 9.27,
942
+ "learning_rate": 1.8101942132382087e-05,
943
+ "loss": 0.4522,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 9.33,
948
+ "learning_rate": 1.8075518562557802e-05,
949
+ "loss": 0.4592,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 9.39,
954
+ "learning_rate": 1.804914783987317e-05,
955
+ "loss": 0.4617,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 9.45,
960
+ "learning_rate": 1.8022724270048884e-05,
961
+ "loss": 0.4601,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 9.51,
966
+ "learning_rate": 1.7996300700224603e-05,
967
+ "loss": 0.4606,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 9.57,
972
+ "learning_rate": 1.7969877130400318e-05,
973
+ "loss": 0.4616,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 9.63,
978
+ "learning_rate": 1.7943506407715682e-05,
979
+ "loss": 0.4566,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 9.69,
984
+ "learning_rate": 1.79170828378914e-05,
985
+ "loss": 0.4638,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 9.75,
990
+ "learning_rate": 1.7890659268067116e-05,
991
+ "loss": 0.467,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 9.81,
996
+ "learning_rate": 1.7864235698242835e-05,
997
+ "loss": 0.4662,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 9.87,
1002
+ "learning_rate": 1.783781212841855e-05,
1003
+ "loss": 0.4636,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 9.93,
1008
+ "learning_rate": 1.781138855859427e-05,
1009
+ "loss": 0.4641,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 9.99,
1014
+ "learning_rate": 1.7784964988769984e-05,
1015
+ "loss": 0.461,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 10.05,
1020
+ "learning_rate": 1.7758594266085348e-05,
1021
+ "loss": 0.4463,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 10.11,
1026
+ "learning_rate": 1.7732223543400715e-05,
1027
+ "loss": 0.4379,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 10.17,
1032
+ "learning_rate": 1.770579997357643e-05,
1033
+ "loss": 0.4418,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 10.23,
1038
+ "learning_rate": 1.767937640375215e-05,
1039
+ "loss": 0.45,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 10.29,
1044
+ "learning_rate": 1.7652952833927864e-05,
1045
+ "loss": 0.4463,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 10.34,
1050
+ "learning_rate": 1.7626529264103583e-05,
1051
+ "loss": 0.4453,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 10.4,
1056
+ "learning_rate": 1.7600105694279298e-05,
1057
+ "loss": 0.4443,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 10.46,
1062
+ "learning_rate": 1.7573682124455017e-05,
1063
+ "loss": 0.4465,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 10.52,
1068
+ "learning_rate": 1.7547258554630732e-05,
1069
+ "loss": 0.4548,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 10.58,
1074
+ "learning_rate": 1.752083498480645e-05,
1075
+ "loss": 0.4485,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 10.64,
1080
+ "learning_rate": 1.7494411414982166e-05,
1081
+ "loss": 0.4437,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 10.7,
1086
+ "learning_rate": 1.7467987845157885e-05,
1087
+ "loss": 0.4508,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 10.76,
1092
+ "learning_rate": 1.74415642753336e-05,
1093
+ "loss": 0.4519,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 10.82,
1098
+ "learning_rate": 1.7415140705509315e-05,
1099
+ "loss": 0.4532,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 10.88,
1104
+ "learning_rate": 1.738876998282468e-05,
1105
+ "loss": 0.4536,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 10.94,
1110
+ "learning_rate": 1.7362399260140046e-05,
1111
+ "loss": 0.4571,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 11.0,
1116
+ "learning_rate": 1.733597569031576e-05,
1117
+ "loss": 0.4471,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 11.06,
1122
+ "learning_rate": 1.730955212049148e-05,
1123
+ "loss": 0.432,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 11.12,
1128
+ "learning_rate": 1.7283128550667195e-05,
1129
+ "loss": 0.4303,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 11.18,
1134
+ "learning_rate": 1.7256757827982563e-05,
1135
+ "loss": 0.4307,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 11.24,
1140
+ "learning_rate": 1.7230334258158278e-05,
1141
+ "loss": 0.4373,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 11.3,
1146
+ "learning_rate": 1.7203910688333996e-05,
1147
+ "loss": 0.4307,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 11.36,
1152
+ "learning_rate": 1.7177487118509712e-05,
1153
+ "loss": 0.4339,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 11.41,
1158
+ "learning_rate": 1.7151169242964728e-05,
1159
+ "loss": 0.4376,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 11.47,
1164
+ "learning_rate": 1.7124745673140443e-05,
1165
+ "loss": 0.4383,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 11.53,
1170
+ "learning_rate": 1.7098322103316158e-05,
1171
+ "loss": 0.4421,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 11.59,
1176
+ "learning_rate": 1.7071898533491877e-05,
1177
+ "loss": 0.4377,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 11.65,
1182
+ "learning_rate": 1.7045474963667592e-05,
1183
+ "loss": 0.4446,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 11.71,
1188
+ "learning_rate": 1.7019051393843307e-05,
1189
+ "loss": 0.4359,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 11.77,
1194
+ "learning_rate": 1.6992627824019026e-05,
1195
+ "loss": 0.4377,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 11.83,
1200
+ "learning_rate": 1.696620425419474e-05,
1201
+ "loss": 0.4412,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 11.89,
1206
+ "learning_rate": 1.693978068437046e-05,
1207
+ "loss": 0.4405,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 11.95,
1212
+ "learning_rate": 1.6913357114546175e-05,
1213
+ "loss": 0.4423,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 12.01,
1218
+ "learning_rate": 1.6886933544721894e-05,
1219
+ "loss": 0.4408,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 12.07,
1224
+ "learning_rate": 1.686050997489761e-05,
1225
+ "loss": 0.417,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 12.13,
1230
+ "learning_rate": 1.6834086405073328e-05,
1231
+ "loss": 0.4241,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 12.19,
1236
+ "learning_rate": 1.680771568238869e-05,
1237
+ "loss": 0.4251,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 12.25,
1242
+ "learning_rate": 1.678129211256441e-05,
1243
+ "loss": 0.4168,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 12.31,
1248
+ "learning_rate": 1.6754868542740125e-05,
1249
+ "loss": 0.421,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 12.37,
1254
+ "learning_rate": 1.6728444972915844e-05,
1255
+ "loss": 0.4247,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 12.43,
1260
+ "learning_rate": 1.6702074250231208e-05,
1261
+ "loss": 0.4226,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 12.49,
1266
+ "learning_rate": 1.6675650680406927e-05,
1267
+ "loss": 0.4232,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 12.54,
1272
+ "learning_rate": 1.6649227110582642e-05,
1273
+ "loss": 0.4287,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 12.6,
1278
+ "learning_rate": 1.6622803540758357e-05,
1279
+ "loss": 0.4273,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 12.66,
1284
+ "learning_rate": 1.6596379970934076e-05,
1285
+ "loss": 0.4346,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 12.72,
1290
+ "learning_rate": 1.656995640110979e-05,
1291
+ "loss": 0.4287,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 12.78,
1296
+ "learning_rate": 1.6543532831285506e-05,
1297
+ "loss": 0.429,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 12.84,
1302
+ "learning_rate": 1.6517162108600874e-05,
1303
+ "loss": 0.4254,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 12.9,
1308
+ "learning_rate": 1.649073853877659e-05,
1309
+ "loss": 0.427,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 12.96,
1314
+ "learning_rate": 1.6464314968952307e-05,
1315
+ "loss": 0.4303,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 13.02,
1320
+ "learning_rate": 1.6437891399128023e-05,
1321
+ "loss": 0.4239,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 13.08,
1326
+ "learning_rate": 1.641146782930374e-05,
1327
+ "loss": 0.401,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 13.14,
1332
+ "learning_rate": 1.6385097106619105e-05,
1333
+ "loss": 0.4092,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 13.2,
1338
+ "learning_rate": 1.6358726383934472e-05,
1339
+ "loss": 0.4076,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 13.26,
1344
+ "learning_rate": 1.6332355661249836e-05,
1345
+ "loss": 0.4118,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 13.32,
1350
+ "learning_rate": 1.6305932091425555e-05,
1351
+ "loss": 0.4145,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 13.38,
1356
+ "learning_rate": 1.627950852160127e-05,
1357
+ "loss": 0.4132,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 13.44,
1362
+ "learning_rate": 1.625308495177699e-05,
1363
+ "loss": 0.416,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 13.5,
1368
+ "learning_rate": 1.6226661381952704e-05,
1369
+ "loss": 0.416,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 13.56,
1374
+ "learning_rate": 1.620023781212842e-05,
1375
+ "loss": 0.4126,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 13.61,
1380
+ "learning_rate": 1.6173814242304135e-05,
1381
+ "loss": 0.4179,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 13.67,
1386
+ "learning_rate": 1.6147443519619502e-05,
1387
+ "loss": 0.4242,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 13.73,
1392
+ "learning_rate": 1.6121019949795217e-05,
1393
+ "loss": 0.418,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 13.79,
1398
+ "learning_rate": 1.6094596379970936e-05,
1399
+ "loss": 0.4205,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 13.85,
1404
+ "learning_rate": 1.606817281014665e-05,
1405
+ "loss": 0.422,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 13.91,
1410
+ "learning_rate": 1.604174924032237e-05,
1411
+ "loss": 0.4257,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 13.97,
1416
+ "learning_rate": 1.6015325670498085e-05,
1417
+ "loss": 0.4146,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 14.03,
1422
+ "learning_rate": 1.5988902100673804e-05,
1423
+ "loss": 0.4144,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 14.09,
1428
+ "learning_rate": 1.596247853084952e-05,
1429
+ "loss": 0.4009,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 14.15,
1434
+ "learning_rate": 1.5936054961025234e-05,
1435
+ "loss": 0.4037,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 14.21,
1440
+ "learning_rate": 1.59096842383406e-05,
1441
+ "loss": 0.4011,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 14.27,
1446
+ "learning_rate": 1.5883260668516317e-05,
1447
+ "loss": 0.4045,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 14.33,
1452
+ "learning_rate": 1.5856837098692035e-05,
1453
+ "loss": 0.4127,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 14.39,
1458
+ "learning_rate": 1.583041352886775e-05,
1459
+ "loss": 0.4051,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 14.45,
1464
+ "learning_rate": 1.580398995904347e-05,
1465
+ "loss": 0.4064,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 14.51,
1470
+ "learning_rate": 1.5777566389219184e-05,
1471
+ "loss": 0.4046,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 14.57,
1476
+ "learning_rate": 1.5751142819394903e-05,
1477
+ "loss": 0.4041,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 14.63,
1482
+ "learning_rate": 1.572471924957062e-05,
1483
+ "loss": 0.4116,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 14.68,
1488
+ "learning_rate": 1.5698295679746337e-05,
1489
+ "loss": 0.4061,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 14.74,
1494
+ "learning_rate": 1.5671872109922052e-05,
1495
+ "loss": 0.4119,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 14.8,
1500
+ "learning_rate": 1.564544854009777e-05,
1501
+ "loss": 0.4101,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 14.86,
1506
+ "learning_rate": 1.5619024970273483e-05,
1507
+ "loss": 0.412,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 14.92,
1512
+ "learning_rate": 1.55926014004492e-05,
1513
+ "loss": 0.4089,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 14.98,
1518
+ "learning_rate": 1.5566230677764565e-05,
1519
+ "loss": 0.4119,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 15.04,
1524
+ "learning_rate": 1.5539807107940284e-05,
1525
+ "loss": 0.3987,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 15.1,
1530
+ "learning_rate": 1.5513383538116e-05,
1531
+ "loss": 0.3884,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 15.16,
1536
+ "learning_rate": 1.5486959968291718e-05,
1537
+ "loss": 0.3889,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 15.22,
1542
+ "learning_rate": 1.5460589245607082e-05,
1543
+ "loss": 0.3947,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 15.28,
1548
+ "learning_rate": 1.543421852292245e-05,
1549
+ "loss": 0.3915,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 15.34,
1554
+ "learning_rate": 1.5407794953098164e-05,
1555
+ "loss": 0.3933,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 15.4,
1560
+ "learning_rate": 1.5381371383273883e-05,
1561
+ "loss": 0.3884,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 15.46,
1566
+ "learning_rate": 1.5354947813449598e-05,
1567
+ "loss": 0.3966,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 15.52,
1572
+ "learning_rate": 1.5328524243625317e-05,
1573
+ "loss": 0.4029,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 15.58,
1578
+ "learning_rate": 1.5302100673801032e-05,
1579
+ "loss": 0.394,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 15.64,
1584
+ "learning_rate": 1.527567710397675e-05,
1585
+ "loss": 0.3999,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 15.7,
1590
+ "learning_rate": 1.5249253534152466e-05,
1591
+ "loss": 0.4,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 15.75,
1596
+ "learning_rate": 1.522288281146783e-05,
1597
+ "loss": 0.4006,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 15.81,
1602
+ "learning_rate": 1.5196459241643547e-05,
1603
+ "loss": 0.4007,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 15.87,
1608
+ "learning_rate": 1.5170035671819264e-05,
1609
+ "loss": 0.4058,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 15.93,
1614
+ "learning_rate": 1.514361210199498e-05,
1615
+ "loss": 0.4032,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 15.99,
1620
+ "learning_rate": 1.5117188532170698e-05,
1621
+ "loss": 0.4027,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 16.05,
1626
+ "learning_rate": 1.5090817809486063e-05,
1627
+ "loss": 0.3827,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 16.11,
1632
+ "learning_rate": 1.5064394239661778e-05,
1633
+ "loss": 0.3889,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 16.17,
1638
+ "learning_rate": 1.5037970669837495e-05,
1639
+ "loss": 0.3876,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 16.23,
1644
+ "learning_rate": 1.5011547100013212e-05,
1645
+ "loss": 0.3825,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 16.29,
1650
+ "learning_rate": 1.498512353018893e-05,
1651
+ "loss": 0.3872,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 16.35,
1656
+ "learning_rate": 1.4958699960364646e-05,
1657
+ "loss": 0.3895,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 16.41,
1662
+ "learning_rate": 1.4932276390540363e-05,
1663
+ "loss": 0.3884,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 16.47,
1668
+ "learning_rate": 1.490585282071608e-05,
1669
+ "loss": 0.3863,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 16.53,
1674
+ "learning_rate": 1.4879482098031446e-05,
1675
+ "loss": 0.3914,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 16.59,
1680
+ "learning_rate": 1.4853111375346811e-05,
1681
+ "loss": 0.3982,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 16.65,
1686
+ "learning_rate": 1.4826687805522528e-05,
1687
+ "loss": 0.3949,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 16.71,
1692
+ "learning_rate": 1.4800264235698245e-05,
1693
+ "loss": 0.3924,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 16.77,
1698
+ "learning_rate": 1.4773893513013609e-05,
1699
+ "loss": 0.3933,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 16.83,
1704
+ "learning_rate": 1.4747469943189326e-05,
1705
+ "loss": 0.394,
1706
+ "step": 141500
1707
+ },
1708
+ {
1709
+ "epoch": 16.88,
1710
+ "learning_rate": 1.4721046373365043e-05,
1711
+ "loss": 0.3907,
1712
+ "step": 142000
1713
+ },
1714
+ {
1715
+ "epoch": 16.94,
1716
+ "learning_rate": 1.4694622803540758e-05,
1717
+ "loss": 0.3955,
1718
+ "step": 142500
1719
+ },
1720
+ {
1721
+ "epoch": 17.0,
1722
+ "learning_rate": 1.4668199233716475e-05,
1723
+ "loss": 0.3903,
1724
+ "step": 143000
1725
+ },
1726
+ {
1727
+ "epoch": 17.06,
1728
+ "learning_rate": 1.4641775663892192e-05,
1729
+ "loss": 0.3735,
1730
+ "step": 143500
1731
+ },
1732
+ {
1733
+ "epoch": 17.12,
1734
+ "learning_rate": 1.4615352094067909e-05,
1735
+ "loss": 0.379,
1736
+ "step": 144000
1737
+ },
1738
+ {
1739
+ "epoch": 17.18,
1740
+ "learning_rate": 1.4589034218522923e-05,
1741
+ "loss": 0.3799,
1742
+ "step": 144500
1743
+ },
1744
+ {
1745
+ "epoch": 17.24,
1746
+ "learning_rate": 1.456261064869864e-05,
1747
+ "loss": 0.3772,
1748
+ "step": 145000
1749
+ },
1750
+ {
1751
+ "epoch": 17.3,
1752
+ "learning_rate": 1.4536187078874357e-05,
1753
+ "loss": 0.3806,
1754
+ "step": 145500
1755
+ },
1756
+ {
1757
+ "epoch": 17.36,
1758
+ "learning_rate": 1.4509763509050074e-05,
1759
+ "loss": 0.3775,
1760
+ "step": 146000
1761
+ },
1762
+ {
1763
+ "epoch": 17.42,
1764
+ "learning_rate": 1.4483339939225791e-05,
1765
+ "loss": 0.3799,
1766
+ "step": 146500
1767
+ },
1768
+ {
1769
+ "epoch": 17.48,
1770
+ "learning_rate": 1.4456916369401508e-05,
1771
+ "loss": 0.3835,
1772
+ "step": 147000
1773
+ },
1774
+ {
1775
+ "epoch": 17.54,
1776
+ "learning_rate": 1.4430492799577225e-05,
1777
+ "loss": 0.3806,
1778
+ "step": 147500
1779
+ },
1780
+ {
1781
+ "epoch": 17.6,
1782
+ "learning_rate": 1.4404069229752942e-05,
1783
+ "loss": 0.3825,
1784
+ "step": 148000
1785
+ },
1786
+ {
1787
+ "epoch": 17.66,
1788
+ "learning_rate": 1.4377645659928659e-05,
1789
+ "loss": 0.3873,
1790
+ "step": 148500
1791
+ },
1792
+ {
1793
+ "epoch": 17.72,
1794
+ "learning_rate": 1.4351222090104376e-05,
1795
+ "loss": 0.3889,
1796
+ "step": 149000
1797
+ },
1798
+ {
1799
+ "epoch": 17.78,
1800
+ "learning_rate": 1.4324798520280093e-05,
1801
+ "loss": 0.3855,
1802
+ "step": 149500
1803
+ },
1804
+ {
1805
+ "epoch": 17.84,
1806
+ "learning_rate": 1.4298374950455806e-05,
1807
+ "loss": 0.3866,
1808
+ "step": 150000
1809
+ },
1810
+ {
1811
+ "epoch": 17.9,
1812
+ "learning_rate": 1.4271951380631523e-05,
1813
+ "loss": 0.385,
1814
+ "step": 150500
1815
+ },
1816
+ {
1817
+ "epoch": 17.95,
1818
+ "learning_rate": 1.424552781080724e-05,
1819
+ "loss": 0.391,
1820
+ "step": 151000
1821
+ },
1822
+ {
1823
+ "epoch": 18.01,
1824
+ "learning_rate": 1.4219104240982957e-05,
1825
+ "loss": 0.379,
1826
+ "step": 151500
1827
+ },
1828
+ {
1829
+ "epoch": 18.07,
1830
+ "learning_rate": 1.4192680671158674e-05,
1831
+ "loss": 0.369,
1832
+ "step": 152000
1833
+ },
1834
+ {
1835
+ "epoch": 18.13,
1836
+ "learning_rate": 1.416630994847404e-05,
1837
+ "loss": 0.3725,
1838
+ "step": 152500
1839
+ },
1840
+ {
1841
+ "epoch": 18.19,
1842
+ "learning_rate": 1.4139886378649757e-05,
1843
+ "loss": 0.3731,
1844
+ "step": 153000
1845
+ },
1846
+ {
1847
+ "epoch": 18.25,
1848
+ "learning_rate": 1.4113462808825474e-05,
1849
+ "loss": 0.3704,
1850
+ "step": 153500
1851
+ },
1852
+ {
1853
+ "epoch": 18.31,
1854
+ "learning_rate": 1.408703923900119e-05,
1855
+ "loss": 0.3696,
1856
+ "step": 154000
1857
+ },
1858
+ {
1859
+ "epoch": 18.37,
1860
+ "learning_rate": 1.4060668516316556e-05,
1861
+ "loss": 0.3682,
1862
+ "step": 154500
1863
+ },
1864
+ {
1865
+ "epoch": 18.43,
1866
+ "learning_rate": 1.4034244946492273e-05,
1867
+ "loss": 0.3772,
1868
+ "step": 155000
1869
+ },
1870
+ {
1871
+ "epoch": 18.49,
1872
+ "learning_rate": 1.400782137666799e-05,
1873
+ "loss": 0.3794,
1874
+ "step": 155500
1875
+ },
1876
+ {
1877
+ "epoch": 18.55,
1878
+ "learning_rate": 1.3981397806843705e-05,
1879
+ "loss": 0.3799,
1880
+ "step": 156000
1881
+ },
1882
+ {
1883
+ "epoch": 18.61,
1884
+ "learning_rate": 1.3954974237019422e-05,
1885
+ "loss": 0.3695,
1886
+ "step": 156500
1887
+ },
1888
+ {
1889
+ "epoch": 18.67,
1890
+ "learning_rate": 1.392855066719514e-05,
1891
+ "loss": 0.3768,
1892
+ "step": 157000
1893
+ },
1894
+ {
1895
+ "epoch": 18.73,
1896
+ "learning_rate": 1.3902179944510505e-05,
1897
+ "loss": 0.381,
1898
+ "step": 157500
1899
+ },
1900
+ {
1901
+ "epoch": 18.79,
1902
+ "learning_rate": 1.3875756374686222e-05,
1903
+ "loss": 0.375,
1904
+ "step": 158000
1905
+ },
1906
+ {
1907
+ "epoch": 18.85,
1908
+ "learning_rate": 1.3849332804861937e-05,
1909
+ "loss": 0.377,
1910
+ "step": 158500
1911
+ },
1912
+ {
1913
+ "epoch": 18.91,
1914
+ "learning_rate": 1.3822909235037654e-05,
1915
+ "loss": 0.3853,
1916
+ "step": 159000
1917
+ },
1918
+ {
1919
+ "epoch": 18.97,
1920
+ "learning_rate": 1.3796485665213371e-05,
1921
+ "loss": 0.3778,
1922
+ "step": 159500
1923
+ },
1924
+ {
1925
+ "epoch": 19.02,
1926
+ "learning_rate": 1.3770062095389088e-05,
1927
+ "loss": 0.3688,
1928
+ "step": 160000
1929
+ },
1930
+ {
1931
+ "epoch": 19.08,
1932
+ "learning_rate": 1.3743691372704453e-05,
1933
+ "loss": 0.3653,
1934
+ "step": 160500
1935
+ },
1936
+ {
1937
+ "epoch": 19.14,
1938
+ "learning_rate": 1.371726780288017e-05,
1939
+ "loss": 0.355,
1940
+ "step": 161000
1941
+ },
1942
+ {
1943
+ "epoch": 19.2,
1944
+ "learning_rate": 1.3690844233055887e-05,
1945
+ "loss": 0.3644,
1946
+ "step": 161500
1947
+ },
1948
+ {
1949
+ "epoch": 19.26,
1950
+ "learning_rate": 1.3664420663231603e-05,
1951
+ "loss": 0.3673,
1952
+ "step": 162000
1953
+ },
1954
+ {
1955
+ "epoch": 19.32,
1956
+ "learning_rate": 1.363799709340732e-05,
1957
+ "loss": 0.3691,
1958
+ "step": 162500
1959
+ },
1960
+ {
1961
+ "epoch": 19.38,
1962
+ "learning_rate": 1.3611626370722685e-05,
1963
+ "loss": 0.3544,
1964
+ "step": 163000
1965
+ },
1966
+ {
1967
+ "epoch": 19.44,
1968
+ "learning_rate": 1.3585202800898402e-05,
1969
+ "loss": 0.368,
1970
+ "step": 163500
1971
+ },
1972
+ {
1973
+ "epoch": 19.5,
1974
+ "learning_rate": 1.3558779231074119e-05,
1975
+ "loss": 0.3691,
1976
+ "step": 164000
1977
+ },
1978
+ {
1979
+ "epoch": 19.56,
1980
+ "learning_rate": 1.3532355661249836e-05,
1981
+ "loss": 0.3655,
1982
+ "step": 164500
1983
+ },
1984
+ {
1985
+ "epoch": 19.62,
1986
+ "learning_rate": 1.3505932091425553e-05,
1987
+ "loss": 0.3692,
1988
+ "step": 165000
1989
+ },
1990
+ {
1991
+ "epoch": 19.68,
1992
+ "learning_rate": 1.3479561368740918e-05,
1993
+ "loss": 0.3782,
1994
+ "step": 165500
1995
+ },
1996
+ {
1997
+ "epoch": 19.74,
1998
+ "learning_rate": 1.3453137798916635e-05,
1999
+ "loss": 0.3708,
2000
+ "step": 166000
2001
+ },
2002
+ {
2003
+ "epoch": 19.8,
2004
+ "learning_rate": 1.3426714229092352e-05,
2005
+ "loss": 0.3673,
2006
+ "step": 166500
2007
+ },
2008
+ {
2009
+ "epoch": 19.86,
2010
+ "learning_rate": 1.3400343506407716e-05,
2011
+ "loss": 0.3678,
2012
+ "step": 167000
2013
+ },
2014
+ {
2015
+ "epoch": 19.92,
2016
+ "learning_rate": 1.3373919936583433e-05,
2017
+ "loss": 0.3721,
2018
+ "step": 167500
2019
+ },
2020
+ {
2021
+ "epoch": 19.98,
2022
+ "learning_rate": 1.334749636675915e-05,
2023
+ "loss": 0.3741,
2024
+ "step": 168000
2025
+ }
2026
+ ],
2027
+ "max_steps": 420500,
2028
+ "num_train_epochs": 50,
2029
+ "total_flos": 7.76486352033908e+19,
2030
+ "trial_name": null,
2031
+ "trial_params": null
2032
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7bdd0064f3d2a01f0f02ecc012e33d97c908284e8324f5bf9c0489dd0d516a6a
3
+ size 2799
vocab.json ADDED
The diff for this file is too large to render. See raw diff