abhinavp commited on
Commit
e8780dd
1 Parent(s): be3f7f8

Training in progress, epoch 0, checkpoint

Browse files
checkpoint-228960/config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_remove_final_layer_norm": false,
3
+ "activation_function": "relu",
4
+ "architectures": [
5
+ "OPTForCausalLM"
6
+ ],
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 2,
9
+ "do_layer_norm_before": true,
10
+ "dropout": 0.1,
11
+ "enable_bias": true,
12
+ "eos_token_id": 2,
13
+ "ffn_dim": 768,
14
+ "hidden_size": 768,
15
+ "init_std": 0.02,
16
+ "layer_norm_elementwise_affine": true,
17
+ "layerdrop": 0.0,
18
+ "max_position_embeddings": 512,
19
+ "model_type": "opt",
20
+ "num_attention_heads": 8,
21
+ "num_hidden_layers": 8,
22
+ "pad_token_id": 1,
23
+ "torch_dtype": "float32",
24
+ "transformers_version": "4.33.3",
25
+ "use_cache": true,
26
+ "vocab_size": 50002,
27
+ "word_embed_proj_dim": 768
28
+ }
checkpoint-228960/generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 2,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 1,
6
+ "transformers_version": "4.33.3"
7
+ }
checkpoint-228960/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f552267d95d0e04f4bd5578d964e29dde3acdd70c62d95dce759d42ee31922f0
3
+ size 537476677
checkpoint-228960/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c111bb5fff53c2ee3b24630f622ca5e10723ae84f4d7da9e86578f648325ca6
3
+ size 268727709
checkpoint-228960/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:522dd2d16a56d20dd16a1b0aae89cd72fcadba82a22f56fa4489445935f27a7f
3
+ size 14575
checkpoint-228960/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f2ab14e1a74215e058c54917b0be8a2b400ec5f5259761532b6b9c8e235953e
3
+ size 627
checkpoint-228960/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "eos_token": "<eos>",
3
+ "pad_token": "<pad>",
4
+ "unk_token": "<unk>"
5
+ }
checkpoint-228960/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-228960/tokenizer_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "clean_up_tokenization_spaces": true,
3
+ "eos_token": "<eos>",
4
+ "max_length": null,
5
+ "model_max_length": 512,
6
+ "pad_to_multiple_of": null,
7
+ "pad_token": "<pad>",
8
+ "pad_token_type_id": 0,
9
+ "padding_side": "right",
10
+ "tokenizer_class": "PreTrainedTokenizerFast",
11
+ "unk_token": "<unk>"
12
+ }
checkpoint-228960/trainer_state.json ADDED
@@ -0,0 +1,2731 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 3.957947015762329,
3
+ "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/passive/transformer/1/checkpoints/checkpoint-228960",
4
+ "epoch": 0.025000606015738065,
5
+ "eval_steps": 10,
6
+ "global_step": 228960,
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": 4.999998362119627e-05,
14
+ "loss": 10.9323,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 4.999161405248948e-05,
20
+ "loss": 6.829,
21
+ "step": 512
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 4.998322810497896e-05,
26
+ "loss": 6.1955,
27
+ "step": 1024
28
+ },
29
+ {
30
+ "epoch": 0.0,
31
+ "learning_rate": 4.997484215746844e-05,
32
+ "loss": 5.9876,
33
+ "step": 1536
34
+ },
35
+ {
36
+ "epoch": 0.0,
37
+ "learning_rate": 4.996645620995792e-05,
38
+ "loss": 5.8238,
39
+ "step": 2048
40
+ },
41
+ {
42
+ "epoch": 0.0,
43
+ "learning_rate": 4.99580702624474e-05,
44
+ "loss": 5.7039,
45
+ "step": 2560
46
+ },
47
+ {
48
+ "epoch": 0.0,
49
+ "learning_rate": 4.994968431493688e-05,
50
+ "loss": 5.6145,
51
+ "step": 3072
52
+ },
53
+ {
54
+ "epoch": 0.0,
55
+ "learning_rate": 4.994129836742636e-05,
56
+ "loss": 5.5373,
57
+ "step": 3584
58
+ },
59
+ {
60
+ "epoch": 0.0,
61
+ "learning_rate": 4.993291241991584e-05,
62
+ "loss": 5.4552,
63
+ "step": 4096
64
+ },
65
+ {
66
+ "epoch": 0.0,
67
+ "learning_rate": 4.992452647240532e-05,
68
+ "loss": 5.4069,
69
+ "step": 4608
70
+ },
71
+ {
72
+ "epoch": 0.0,
73
+ "learning_rate": 4.99161405248948e-05,
74
+ "loss": 5.3539,
75
+ "step": 5120
76
+ },
77
+ {
78
+ "epoch": 0.0,
79
+ "learning_rate": 4.990775457738428e-05,
80
+ "loss": 5.3134,
81
+ "step": 5632
82
+ },
83
+ {
84
+ "epoch": 0.0,
85
+ "learning_rate": 4.989936862987376e-05,
86
+ "loss": 5.2658,
87
+ "step": 6144
88
+ },
89
+ {
90
+ "epoch": 0.0,
91
+ "learning_rate": 4.989099906116697e-05,
92
+ "loss": 5.2176,
93
+ "step": 6656
94
+ },
95
+ {
96
+ "epoch": 0.0,
97
+ "learning_rate": 4.988261311365645e-05,
98
+ "loss": 5.178,
99
+ "step": 7168
100
+ },
101
+ {
102
+ "epoch": 0.0,
103
+ "learning_rate": 4.987422716614593e-05,
104
+ "loss": 5.1336,
105
+ "step": 7680
106
+ },
107
+ {
108
+ "epoch": 0.0,
109
+ "learning_rate": 4.986584121863541e-05,
110
+ "loss": 5.115,
111
+ "step": 8192
112
+ },
113
+ {
114
+ "epoch": 0.0,
115
+ "learning_rate": 4.985745527112489e-05,
116
+ "loss": 5.0825,
117
+ "step": 8704
118
+ },
119
+ {
120
+ "epoch": 0.0,
121
+ "learning_rate": 4.984906932361437e-05,
122
+ "loss": 5.0539,
123
+ "step": 9216
124
+ },
125
+ {
126
+ "epoch": 0.0,
127
+ "learning_rate": 4.984068337610385e-05,
128
+ "loss": 5.0235,
129
+ "step": 9728
130
+ },
131
+ {
132
+ "epoch": 0.0,
133
+ "learning_rate": 4.983229742859333e-05,
134
+ "loss": 5.015,
135
+ "step": 10240
136
+ },
137
+ {
138
+ "epoch": 0.0,
139
+ "learning_rate": 4.9823927859886547e-05,
140
+ "loss": 4.9821,
141
+ "step": 10752
142
+ },
143
+ {
144
+ "epoch": 0.0,
145
+ "learning_rate": 4.9815541912376026e-05,
146
+ "loss": 4.9593,
147
+ "step": 11264
148
+ },
149
+ {
150
+ "epoch": 0.0,
151
+ "learning_rate": 4.9807155964865506e-05,
152
+ "loss": 4.9296,
153
+ "step": 11776
154
+ },
155
+ {
156
+ "epoch": 0.0,
157
+ "learning_rate": 4.9798770017354986e-05,
158
+ "loss": 4.9278,
159
+ "step": 12288
160
+ },
161
+ {
162
+ "epoch": 0.0,
163
+ "learning_rate": 4.9790400448648195e-05,
164
+ "loss": 4.8883,
165
+ "step": 12800
166
+ },
167
+ {
168
+ "epoch": 0.0,
169
+ "learning_rate": 4.9782014501137675e-05,
170
+ "loss": 4.8781,
171
+ "step": 13312
172
+ },
173
+ {
174
+ "epoch": 0.0,
175
+ "learning_rate": 4.9773628553627155e-05,
176
+ "loss": 4.8482,
177
+ "step": 13824
178
+ },
179
+ {
180
+ "epoch": 0.0,
181
+ "learning_rate": 4.9765242606116635e-05,
182
+ "loss": 4.8377,
183
+ "step": 14336
184
+ },
185
+ {
186
+ "epoch": 0.0,
187
+ "learning_rate": 4.9756856658606115e-05,
188
+ "loss": 4.8239,
189
+ "step": 14848
190
+ },
191
+ {
192
+ "epoch": 0.01,
193
+ "learning_rate": 4.9748470711095595e-05,
194
+ "loss": 4.8065,
195
+ "step": 15360
196
+ },
197
+ {
198
+ "epoch": 0.01,
199
+ "learning_rate": 4.9740117521192533e-05,
200
+ "loss": 4.7973,
201
+ "step": 15872
202
+ },
203
+ {
204
+ "epoch": 0.01,
205
+ "learning_rate": 4.9731731573682013e-05,
206
+ "loss": 4.784,
207
+ "step": 16384
208
+ },
209
+ {
210
+ "epoch": 0.01,
211
+ "learning_rate": 4.97233456261715e-05,
212
+ "loss": 4.7687,
213
+ "step": 16896
214
+ },
215
+ {
216
+ "epoch": 0.01,
217
+ "learning_rate": 4.971495967866098e-05,
218
+ "loss": 4.7659,
219
+ "step": 17408
220
+ },
221
+ {
222
+ "epoch": 0.01,
223
+ "learning_rate": 4.970657373115046e-05,
224
+ "loss": 4.746,
225
+ "step": 17920
226
+ },
227
+ {
228
+ "epoch": 0.01,
229
+ "learning_rate": 4.969820416244367e-05,
230
+ "loss": 4.7399,
231
+ "step": 18432
232
+ },
233
+ {
234
+ "epoch": 0.01,
235
+ "learning_rate": 4.968981821493315e-05,
236
+ "loss": 4.6986,
237
+ "step": 18944
238
+ },
239
+ {
240
+ "epoch": 0.01,
241
+ "learning_rate": 4.968143226742263e-05,
242
+ "loss": 4.7107,
243
+ "step": 19456
244
+ },
245
+ {
246
+ "epoch": 0.01,
247
+ "learning_rate": 4.967304631991211e-05,
248
+ "loss": 4.6746,
249
+ "step": 19968
250
+ },
251
+ {
252
+ "epoch": 0.01,
253
+ "learning_rate": 4.966466037240159e-05,
254
+ "loss": 4.6858,
255
+ "step": 20480
256
+ },
257
+ {
258
+ "epoch": 0.01,
259
+ "learning_rate": 4.965627442489107e-05,
260
+ "loss": 4.6678,
261
+ "step": 20992
262
+ },
263
+ {
264
+ "epoch": 0.01,
265
+ "learning_rate": 4.964788847738054e-05,
266
+ "loss": 4.6666,
267
+ "step": 21504
268
+ },
269
+ {
270
+ "epoch": 0.01,
271
+ "learning_rate": 4.963950252987002e-05,
272
+ "loss": 4.6418,
273
+ "step": 22016
274
+ },
275
+ {
276
+ "epoch": 0.01,
277
+ "learning_rate": 4.963113296116324e-05,
278
+ "loss": 4.653,
279
+ "step": 22528
280
+ },
281
+ {
282
+ "epoch": 0.01,
283
+ "learning_rate": 4.962274701365272e-05,
284
+ "loss": 4.6397,
285
+ "step": 23040
286
+ },
287
+ {
288
+ "epoch": 0.01,
289
+ "learning_rate": 4.96143610661422e-05,
290
+ "loss": 4.6337,
291
+ "step": 23552
292
+ },
293
+ {
294
+ "epoch": 0.01,
295
+ "learning_rate": 4.960597511863168e-05,
296
+ "loss": 4.6287,
297
+ "step": 24064
298
+ },
299
+ {
300
+ "epoch": 0.01,
301
+ "learning_rate": 4.9597605549924894e-05,
302
+ "loss": 4.5955,
303
+ "step": 24576
304
+ },
305
+ {
306
+ "epoch": 0.01,
307
+ "learning_rate": 4.9589219602414374e-05,
308
+ "loss": 4.5991,
309
+ "step": 25088
310
+ },
311
+ {
312
+ "epoch": 0.01,
313
+ "learning_rate": 4.958083365490385e-05,
314
+ "loss": 4.6011,
315
+ "step": 25600
316
+ },
317
+ {
318
+ "epoch": 0.01,
319
+ "learning_rate": 4.957244770739333e-05,
320
+ "loss": 4.5891,
321
+ "step": 26112
322
+ },
323
+ {
324
+ "epoch": 0.01,
325
+ "learning_rate": 4.956407813868654e-05,
326
+ "loss": 4.5802,
327
+ "step": 26624
328
+ },
329
+ {
330
+ "epoch": 0.01,
331
+ "learning_rate": 4.9555692191176016e-05,
332
+ "loss": 4.5608,
333
+ "step": 27136
334
+ },
335
+ {
336
+ "epoch": 0.01,
337
+ "learning_rate": 4.9547306243665496e-05,
338
+ "loss": 4.5591,
339
+ "step": 27648
340
+ },
341
+ {
342
+ "epoch": 0.01,
343
+ "learning_rate": 4.9538920296154976e-05,
344
+ "loss": 4.5413,
345
+ "step": 28160
346
+ },
347
+ {
348
+ "epoch": 0.01,
349
+ "learning_rate": 4.9530534348644456e-05,
350
+ "loss": 4.5632,
351
+ "step": 28672
352
+ },
353
+ {
354
+ "epoch": 0.01,
355
+ "learning_rate": 4.9522148401133936e-05,
356
+ "loss": 4.5192,
357
+ "step": 29184
358
+ },
359
+ {
360
+ "epoch": 0.01,
361
+ "learning_rate": 4.951377883242715e-05,
362
+ "loss": 4.5418,
363
+ "step": 29696
364
+ },
365
+ {
366
+ "epoch": 0.01,
367
+ "learning_rate": 4.950539288491663e-05,
368
+ "loss": 4.5339,
369
+ "step": 30208
370
+ },
371
+ {
372
+ "epoch": 0.01,
373
+ "learning_rate": 4.949700693740611e-05,
374
+ "loss": 4.5138,
375
+ "step": 30720
376
+ },
377
+ {
378
+ "epoch": 0.01,
379
+ "learning_rate": 4.948862098989559e-05,
380
+ "loss": 4.5131,
381
+ "step": 31232
382
+ },
383
+ {
384
+ "epoch": 0.01,
385
+ "learning_rate": 4.94802514211888e-05,
386
+ "loss": 4.504,
387
+ "step": 31744
388
+ },
389
+ {
390
+ "epoch": 0.01,
391
+ "learning_rate": 4.947186547367828e-05,
392
+ "loss": 4.4957,
393
+ "step": 32256
394
+ },
395
+ {
396
+ "epoch": 0.01,
397
+ "learning_rate": 4.946347952616776e-05,
398
+ "loss": 4.4877,
399
+ "step": 32768
400
+ },
401
+ {
402
+ "epoch": 0.01,
403
+ "learning_rate": 4.945509357865724e-05,
404
+ "loss": 4.4991,
405
+ "step": 33280
406
+ },
407
+ {
408
+ "epoch": 0.01,
409
+ "learning_rate": 4.944672400995045e-05,
410
+ "loss": 4.4766,
411
+ "step": 33792
412
+ },
413
+ {
414
+ "epoch": 0.01,
415
+ "learning_rate": 4.943833806243993e-05,
416
+ "loss": 4.4741,
417
+ "step": 34304
418
+ },
419
+ {
420
+ "epoch": 0.01,
421
+ "learning_rate": 4.942995211492941e-05,
422
+ "loss": 4.4664,
423
+ "step": 34816
424
+ },
425
+ {
426
+ "epoch": 0.01,
427
+ "learning_rate": 4.9421582546222625e-05,
428
+ "loss": 4.4654,
429
+ "step": 35328
430
+ },
431
+ {
432
+ "epoch": 0.01,
433
+ "learning_rate": 4.9413196598712105e-05,
434
+ "loss": 4.4755,
435
+ "step": 35840
436
+ },
437
+ {
438
+ "epoch": 0.01,
439
+ "learning_rate": 4.9404810651201585e-05,
440
+ "loss": 4.4707,
441
+ "step": 36352
442
+ },
443
+ {
444
+ "epoch": 0.01,
445
+ "learning_rate": 4.9396424703691065e-05,
446
+ "loss": 4.4465,
447
+ "step": 36864
448
+ },
449
+ {
450
+ "epoch": 0.01,
451
+ "learning_rate": 4.9388038756180545e-05,
452
+ "loss": 4.4613,
453
+ "step": 37376
454
+ },
455
+ {
456
+ "epoch": 0.01,
457
+ "learning_rate": 4.9379652808670025e-05,
458
+ "loss": 4.4609,
459
+ "step": 37888
460
+ },
461
+ {
462
+ "epoch": 0.01,
463
+ "learning_rate": 4.9371266861159505e-05,
464
+ "loss": 4.4338,
465
+ "step": 38400
466
+ },
467
+ {
468
+ "epoch": 0.01,
469
+ "learning_rate": 4.9362897292452714e-05,
470
+ "loss": 4.4412,
471
+ "step": 38912
472
+ },
473
+ {
474
+ "epoch": 0.01,
475
+ "learning_rate": 4.9354511344942194e-05,
476
+ "loss": 4.4307,
477
+ "step": 39424
478
+ },
479
+ {
480
+ "epoch": 0.01,
481
+ "learning_rate": 4.9346125397431674e-05,
482
+ "loss": 4.4251,
483
+ "step": 39936
484
+ },
485
+ {
486
+ "epoch": 0.01,
487
+ "learning_rate": 4.9337739449921154e-05,
488
+ "loss": 4.4148,
489
+ "step": 40448
490
+ },
491
+ {
492
+ "epoch": 0.01,
493
+ "learning_rate": 4.9329353502410634e-05,
494
+ "loss": 4.4198,
495
+ "step": 40960
496
+ },
497
+ {
498
+ "epoch": 0.01,
499
+ "learning_rate": 4.9320967554900114e-05,
500
+ "loss": 4.4241,
501
+ "step": 41472
502
+ },
503
+ {
504
+ "epoch": 0.01,
505
+ "learning_rate": 4.9312581607389594e-05,
506
+ "loss": 4.4273,
507
+ "step": 41984
508
+ },
509
+ {
510
+ "epoch": 0.01,
511
+ "learning_rate": 4.930421203868281e-05,
512
+ "loss": 4.4024,
513
+ "step": 42496
514
+ },
515
+ {
516
+ "epoch": 0.01,
517
+ "learning_rate": 4.929582609117229e-05,
518
+ "loss": 4.3811,
519
+ "step": 43008
520
+ },
521
+ {
522
+ "epoch": 0.01,
523
+ "learning_rate": 4.928744014366177e-05,
524
+ "loss": 4.3965,
525
+ "step": 43520
526
+ },
527
+ {
528
+ "epoch": 0.01,
529
+ "learning_rate": 4.927905419615125e-05,
530
+ "loss": 4.3952,
531
+ "step": 44032
532
+ },
533
+ {
534
+ "epoch": 0.01,
535
+ "learning_rate": 4.927066824864073e-05,
536
+ "loss": 4.3948,
537
+ "step": 44544
538
+ },
539
+ {
540
+ "epoch": 0.01,
541
+ "learning_rate": 4.926229867993394e-05,
542
+ "loss": 4.3824,
543
+ "step": 45056
544
+ },
545
+ {
546
+ "epoch": 0.01,
547
+ "learning_rate": 4.925392911122715e-05,
548
+ "loss": 4.3843,
549
+ "step": 45568
550
+ },
551
+ {
552
+ "epoch": 0.02,
553
+ "learning_rate": 4.924554316371663e-05,
554
+ "loss": 4.3804,
555
+ "step": 46080
556
+ },
557
+ {
558
+ "epoch": 0.02,
559
+ "learning_rate": 4.923715721620611e-05,
560
+ "loss": 4.377,
561
+ "step": 46592
562
+ },
563
+ {
564
+ "epoch": 0.02,
565
+ "learning_rate": 4.922877126869559e-05,
566
+ "loss": 4.3679,
567
+ "step": 47104
568
+ },
569
+ {
570
+ "epoch": 0.02,
571
+ "learning_rate": 4.922038532118507e-05,
572
+ "loss": 4.3705,
573
+ "step": 47616
574
+ },
575
+ {
576
+ "epoch": 0.02,
577
+ "learning_rate": 4.921201575247828e-05,
578
+ "loss": 4.3589,
579
+ "step": 48128
580
+ },
581
+ {
582
+ "epoch": 0.02,
583
+ "learning_rate": 4.920362980496776e-05,
584
+ "loss": 4.3724,
585
+ "step": 48640
586
+ },
587
+ {
588
+ "epoch": 0.02,
589
+ "learning_rate": 4.919524385745724e-05,
590
+ "loss": 4.3578,
591
+ "step": 49152
592
+ },
593
+ {
594
+ "epoch": 0.02,
595
+ "learning_rate": 4.918685790994672e-05,
596
+ "loss": 4.3395,
597
+ "step": 49664
598
+ },
599
+ {
600
+ "epoch": 0.02,
601
+ "learning_rate": 4.91784719624362e-05,
602
+ "loss": 4.3522,
603
+ "step": 50176
604
+ },
605
+ {
606
+ "epoch": 0.02,
607
+ "learning_rate": 4.9170086014925676e-05,
608
+ "loss": 4.3529,
609
+ "step": 50688
610
+ },
611
+ {
612
+ "epoch": 0.02,
613
+ "learning_rate": 4.9161700067415156e-05,
614
+ "loss": 4.356,
615
+ "step": 51200
616
+ },
617
+ {
618
+ "epoch": 0.02,
619
+ "learning_rate": 4.9153314119904636e-05,
620
+ "loss": 4.3399,
621
+ "step": 51712
622
+ },
623
+ {
624
+ "epoch": 0.02,
625
+ "learning_rate": 4.914494455119785e-05,
626
+ "loss": 4.3391,
627
+ "step": 52224
628
+ },
629
+ {
630
+ "epoch": 0.02,
631
+ "learning_rate": 4.9136558603687325e-05,
632
+ "loss": 4.3242,
633
+ "step": 52736
634
+ },
635
+ {
636
+ "epoch": 0.02,
637
+ "learning_rate": 4.9128172656176805e-05,
638
+ "loss": 4.3343,
639
+ "step": 53248
640
+ },
641
+ {
642
+ "epoch": 0.02,
643
+ "learning_rate": 4.9119786708666285e-05,
644
+ "loss": 4.3156,
645
+ "step": 53760
646
+ },
647
+ {
648
+ "epoch": 0.02,
649
+ "learning_rate": 4.911143351876323e-05,
650
+ "loss": 4.3236,
651
+ "step": 54272
652
+ },
653
+ {
654
+ "epoch": 0.02,
655
+ "learning_rate": 4.910304757125272e-05,
656
+ "loss": 4.321,
657
+ "step": 54784
658
+ },
659
+ {
660
+ "epoch": 0.02,
661
+ "learning_rate": 4.90946616237422e-05,
662
+ "loss": 4.3199,
663
+ "step": 55296
664
+ },
665
+ {
666
+ "epoch": 0.02,
667
+ "learning_rate": 4.908627567623168e-05,
668
+ "loss": 4.3109,
669
+ "step": 55808
670
+ },
671
+ {
672
+ "epoch": 0.02,
673
+ "learning_rate": 4.907788972872115e-05,
674
+ "loss": 4.3137,
675
+ "step": 56320
676
+ },
677
+ {
678
+ "epoch": 0.02,
679
+ "learning_rate": 4.906950378121063e-05,
680
+ "loss": 4.3049,
681
+ "step": 56832
682
+ },
683
+ {
684
+ "epoch": 0.02,
685
+ "learning_rate": 4.906111783370011e-05,
686
+ "loss": 4.3124,
687
+ "step": 57344
688
+ },
689
+ {
690
+ "epoch": 0.02,
691
+ "learning_rate": 4.9052748264993326e-05,
692
+ "loss": 4.3029,
693
+ "step": 57856
694
+ },
695
+ {
696
+ "epoch": 0.02,
697
+ "learning_rate": 4.90443623174828e-05,
698
+ "loss": 4.2991,
699
+ "step": 58368
700
+ },
701
+ {
702
+ "epoch": 0.02,
703
+ "learning_rate": 4.903597636997228e-05,
704
+ "loss": 4.3147,
705
+ "step": 58880
706
+ },
707
+ {
708
+ "epoch": 0.02,
709
+ "learning_rate": 4.902759042246176e-05,
710
+ "loss": 4.3027,
711
+ "step": 59392
712
+ },
713
+ {
714
+ "epoch": 0.02,
715
+ "learning_rate": 4.9019220853754975e-05,
716
+ "loss": 4.2995,
717
+ "step": 59904
718
+ },
719
+ {
720
+ "epoch": 0.02,
721
+ "learning_rate": 4.9010834906244455e-05,
722
+ "loss": 4.2886,
723
+ "step": 60416
724
+ },
725
+ {
726
+ "epoch": 0.02,
727
+ "learning_rate": 4.9002448958733935e-05,
728
+ "loss": 4.2941,
729
+ "step": 60928
730
+ },
731
+ {
732
+ "epoch": 0.02,
733
+ "learning_rate": 4.8994063011223415e-05,
734
+ "loss": 4.2865,
735
+ "step": 61440
736
+ },
737
+ {
738
+ "epoch": 0.02,
739
+ "learning_rate": 4.8985677063712895e-05,
740
+ "loss": 4.2855,
741
+ "step": 61952
742
+ },
743
+ {
744
+ "epoch": 0.02,
745
+ "learning_rate": 4.8977291116202375e-05,
746
+ "loss": 4.2943,
747
+ "step": 62464
748
+ },
749
+ {
750
+ "epoch": 0.02,
751
+ "learning_rate": 4.8968921547495584e-05,
752
+ "loss": 4.2783,
753
+ "step": 62976
754
+ },
755
+ {
756
+ "epoch": 0.02,
757
+ "learning_rate": 4.8960535599985064e-05,
758
+ "loss": 4.2839,
759
+ "step": 63488
760
+ },
761
+ {
762
+ "epoch": 0.02,
763
+ "learning_rate": 4.8952149652474544e-05,
764
+ "loss": 4.2651,
765
+ "step": 64000
766
+ },
767
+ {
768
+ "epoch": 0.02,
769
+ "learning_rate": 4.8943763704964024e-05,
770
+ "loss": 4.2693,
771
+ "step": 64512
772
+ },
773
+ {
774
+ "epoch": 0.02,
775
+ "learning_rate": 4.8935377757453504e-05,
776
+ "loss": 4.2748,
777
+ "step": 65024
778
+ },
779
+ {
780
+ "epoch": 0.02,
781
+ "learning_rate": 4.8926991809942984e-05,
782
+ "loss": 4.2663,
783
+ "step": 65536
784
+ },
785
+ {
786
+ "epoch": 0.02,
787
+ "learning_rate": 4.891862224123619e-05,
788
+ "loss": 4.2585,
789
+ "step": 66048
790
+ },
791
+ {
792
+ "epoch": 0.02,
793
+ "learning_rate": 4.891023629372567e-05,
794
+ "loss": 4.2845,
795
+ "step": 66560
796
+ },
797
+ {
798
+ "epoch": 0.02,
799
+ "learning_rate": 4.890185034621515e-05,
800
+ "loss": 4.2653,
801
+ "step": 67072
802
+ },
803
+ {
804
+ "epoch": 0.02,
805
+ "learning_rate": 4.889346439870464e-05,
806
+ "loss": 4.2586,
807
+ "step": 67584
808
+ },
809
+ {
810
+ "epoch": 0.02,
811
+ "learning_rate": 4.888507845119412e-05,
812
+ "loss": 4.2576,
813
+ "step": 68096
814
+ },
815
+ {
816
+ "epoch": 0.02,
817
+ "learning_rate": 4.887670888248733e-05,
818
+ "loss": 4.2578,
819
+ "step": 68608
820
+ },
821
+ {
822
+ "epoch": 0.02,
823
+ "learning_rate": 4.886832293497681e-05,
824
+ "loss": 4.2585,
825
+ "step": 69120
826
+ },
827
+ {
828
+ "epoch": 0.02,
829
+ "learning_rate": 4.885993698746629e-05,
830
+ "loss": 4.2567,
831
+ "step": 69632
832
+ },
833
+ {
834
+ "epoch": 0.02,
835
+ "learning_rate": 4.885155103995577e-05,
836
+ "loss": 4.2509,
837
+ "step": 70144
838
+ },
839
+ {
840
+ "epoch": 0.02,
841
+ "learning_rate": 4.884318147124898e-05,
842
+ "loss": 4.2584,
843
+ "step": 70656
844
+ },
845
+ {
846
+ "epoch": 0.02,
847
+ "learning_rate": 4.883479552373846e-05,
848
+ "loss": 4.2557,
849
+ "step": 71168
850
+ },
851
+ {
852
+ "epoch": 0.02,
853
+ "learning_rate": 4.882640957622794e-05,
854
+ "loss": 4.2495,
855
+ "step": 71680
856
+ },
857
+ {
858
+ "epoch": 0.02,
859
+ "learning_rate": 4.881802362871742e-05,
860
+ "loss": 4.2407,
861
+ "step": 72192
862
+ },
863
+ {
864
+ "epoch": 0.02,
865
+ "learning_rate": 4.88096376812069e-05,
866
+ "loss": 4.246,
867
+ "step": 72704
868
+ },
869
+ {
870
+ "epoch": 0.02,
871
+ "learning_rate": 4.880125173369638e-05,
872
+ "loss": 4.2374,
873
+ "step": 73216
874
+ },
875
+ {
876
+ "epoch": 0.02,
877
+ "learning_rate": 4.879286578618586e-05,
878
+ "loss": 4.2389,
879
+ "step": 73728
880
+ },
881
+ {
882
+ "epoch": 0.02,
883
+ "learning_rate": 4.878447983867534e-05,
884
+ "loss": 4.2352,
885
+ "step": 74240
886
+ },
887
+ {
888
+ "epoch": 0.02,
889
+ "learning_rate": 4.877611026996855e-05,
890
+ "loss": 4.2332,
891
+ "step": 74752
892
+ },
893
+ {
894
+ "epoch": 0.02,
895
+ "learning_rate": 4.876772432245803e-05,
896
+ "loss": 4.239,
897
+ "step": 75264
898
+ },
899
+ {
900
+ "epoch": 0.02,
901
+ "learning_rate": 4.875933837494751e-05,
902
+ "loss": 4.2262,
903
+ "step": 75776
904
+ },
905
+ {
906
+ "epoch": 0.02,
907
+ "learning_rate": 4.8750952427436986e-05,
908
+ "loss": 4.2233,
909
+ "step": 76288
910
+ },
911
+ {
912
+ "epoch": 0.03,
913
+ "eval_loss": 4.200882911682129,
914
+ "eval_runtime": 302.3694,
915
+ "eval_samples_per_second": 1262.003,
916
+ "eval_steps_per_second": 39.439,
917
+ "step": 76320
918
+ },
919
+ {
920
+ "epoch": 1.0,
921
+ "learning_rate": 4.87425828587302e-05,
922
+ "loss": 4.2084,
923
+ "step": 76800
924
+ },
925
+ {
926
+ "epoch": 1.0,
927
+ "learning_rate": 4.873419691121968e-05,
928
+ "loss": 4.21,
929
+ "step": 77312
930
+ },
931
+ {
932
+ "epoch": 1.0,
933
+ "learning_rate": 4.872582734251289e-05,
934
+ "loss": 4.2365,
935
+ "step": 77824
936
+ },
937
+ {
938
+ "epoch": 1.0,
939
+ "learning_rate": 4.871744139500237e-05,
940
+ "loss": 4.2209,
941
+ "step": 78336
942
+ },
943
+ {
944
+ "epoch": 1.0,
945
+ "learning_rate": 4.870905544749185e-05,
946
+ "loss": 4.2144,
947
+ "step": 78848
948
+ },
949
+ {
950
+ "epoch": 1.0,
951
+ "learning_rate": 4.870066949998133e-05,
952
+ "loss": 4.2116,
953
+ "step": 79360
954
+ },
955
+ {
956
+ "epoch": 1.0,
957
+ "learning_rate": 4.869228355247081e-05,
958
+ "loss": 4.2083,
959
+ "step": 79872
960
+ },
961
+ {
962
+ "epoch": 1.0,
963
+ "learning_rate": 4.868389760496029e-05,
964
+ "loss": 4.1963,
965
+ "step": 80384
966
+ },
967
+ {
968
+ "epoch": 1.0,
969
+ "learning_rate": 4.8675528036253507e-05,
970
+ "loss": 4.2087,
971
+ "step": 80896
972
+ },
973
+ {
974
+ "epoch": 1.0,
975
+ "learning_rate": 4.8667142088742986e-05,
976
+ "loss": 4.2032,
977
+ "step": 81408
978
+ },
979
+ {
980
+ "epoch": 1.0,
981
+ "learning_rate": 4.865875614123246e-05,
982
+ "loss": 4.206,
983
+ "step": 81920
984
+ },
985
+ {
986
+ "epoch": 1.0,
987
+ "learning_rate": 4.865037019372194e-05,
988
+ "loss": 4.2077,
989
+ "step": 82432
990
+ },
991
+ {
992
+ "epoch": 1.0,
993
+ "learning_rate": 4.8642000625015155e-05,
994
+ "loss": 4.1922,
995
+ "step": 82944
996
+ },
997
+ {
998
+ "epoch": 1.0,
999
+ "learning_rate": 4.8633614677504635e-05,
1000
+ "loss": 4.1882,
1001
+ "step": 83456
1002
+ },
1003
+ {
1004
+ "epoch": 1.0,
1005
+ "learning_rate": 4.862522872999411e-05,
1006
+ "loss": 4.1812,
1007
+ "step": 83968
1008
+ },
1009
+ {
1010
+ "epoch": 1.0,
1011
+ "learning_rate": 4.861684278248359e-05,
1012
+ "loss": 4.1794,
1013
+ "step": 84480
1014
+ },
1015
+ {
1016
+ "epoch": 1.0,
1017
+ "learning_rate": 4.860845683497307e-05,
1018
+ "loss": 4.1899,
1019
+ "step": 84992
1020
+ },
1021
+ {
1022
+ "epoch": 1.0,
1023
+ "learning_rate": 4.860007088746255e-05,
1024
+ "loss": 4.1797,
1025
+ "step": 85504
1026
+ },
1027
+ {
1028
+ "epoch": 1.0,
1029
+ "learning_rate": 4.859168493995203e-05,
1030
+ "loss": 4.1838,
1031
+ "step": 86016
1032
+ },
1033
+ {
1034
+ "epoch": 1.0,
1035
+ "learning_rate": 4.858329899244151e-05,
1036
+ "loss": 4.2022,
1037
+ "step": 86528
1038
+ },
1039
+ {
1040
+ "epoch": 1.0,
1041
+ "learning_rate": 4.8574913044930995e-05,
1042
+ "loss": 4.1809,
1043
+ "step": 87040
1044
+ },
1045
+ {
1046
+ "epoch": 1.0,
1047
+ "learning_rate": 4.8566543476224204e-05,
1048
+ "loss": 4.1847,
1049
+ "step": 87552
1050
+ },
1051
+ {
1052
+ "epoch": 1.0,
1053
+ "learning_rate": 4.8558157528713684e-05,
1054
+ "loss": 4.172,
1055
+ "step": 88064
1056
+ },
1057
+ {
1058
+ "epoch": 1.0,
1059
+ "learning_rate": 4.8549771581203164e-05,
1060
+ "loss": 4.1883,
1061
+ "step": 88576
1062
+ },
1063
+ {
1064
+ "epoch": 1.0,
1065
+ "learning_rate": 4.8541385633692644e-05,
1066
+ "loss": 4.1681,
1067
+ "step": 89088
1068
+ },
1069
+ {
1070
+ "epoch": 1.0,
1071
+ "learning_rate": 4.853301606498585e-05,
1072
+ "loss": 4.1691,
1073
+ "step": 89600
1074
+ },
1075
+ {
1076
+ "epoch": 1.0,
1077
+ "learning_rate": 4.852463011747533e-05,
1078
+ "loss": 4.1632,
1079
+ "step": 90112
1080
+ },
1081
+ {
1082
+ "epoch": 1.0,
1083
+ "learning_rate": 4.851624416996481e-05,
1084
+ "loss": 4.1672,
1085
+ "step": 90624
1086
+ },
1087
+ {
1088
+ "epoch": 1.0,
1089
+ "learning_rate": 4.850785822245429e-05,
1090
+ "loss": 4.1594,
1091
+ "step": 91136
1092
+ },
1093
+ {
1094
+ "epoch": 1.01,
1095
+ "learning_rate": 4.849947227494377e-05,
1096
+ "loss": 4.1576,
1097
+ "step": 91648
1098
+ },
1099
+ {
1100
+ "epoch": 1.01,
1101
+ "learning_rate": 4.849110270623698e-05,
1102
+ "loss": 4.1613,
1103
+ "step": 92160
1104
+ },
1105
+ {
1106
+ "epoch": 1.01,
1107
+ "learning_rate": 4.848271675872646e-05,
1108
+ "loss": 4.1675,
1109
+ "step": 92672
1110
+ },
1111
+ {
1112
+ "epoch": 1.01,
1113
+ "learning_rate": 4.847433081121595e-05,
1114
+ "loss": 4.1589,
1115
+ "step": 93184
1116
+ },
1117
+ {
1118
+ "epoch": 1.01,
1119
+ "learning_rate": 4.846594486370543e-05,
1120
+ "loss": 4.1619,
1121
+ "step": 93696
1122
+ },
1123
+ {
1124
+ "epoch": 1.01,
1125
+ "learning_rate": 4.845755891619491e-05,
1126
+ "loss": 4.1548,
1127
+ "step": 94208
1128
+ },
1129
+ {
1130
+ "epoch": 1.01,
1131
+ "learning_rate": 4.844918934748812e-05,
1132
+ "loss": 4.1643,
1133
+ "step": 94720
1134
+ },
1135
+ {
1136
+ "epoch": 1.01,
1137
+ "learning_rate": 4.84408033999776e-05,
1138
+ "loss": 4.1372,
1139
+ "step": 95232
1140
+ },
1141
+ {
1142
+ "epoch": 1.01,
1143
+ "learning_rate": 4.843241745246708e-05,
1144
+ "loss": 4.1529,
1145
+ "step": 95744
1146
+ },
1147
+ {
1148
+ "epoch": 1.01,
1149
+ "learning_rate": 4.842403150495656e-05,
1150
+ "loss": 4.1308,
1151
+ "step": 96256
1152
+ },
1153
+ {
1154
+ "epoch": 1.01,
1155
+ "learning_rate": 4.841566193624977e-05,
1156
+ "loss": 4.1516,
1157
+ "step": 96768
1158
+ },
1159
+ {
1160
+ "epoch": 1.01,
1161
+ "learning_rate": 4.840727598873925e-05,
1162
+ "loss": 4.1364,
1163
+ "step": 97280
1164
+ },
1165
+ {
1166
+ "epoch": 1.01,
1167
+ "learning_rate": 4.839889004122873e-05,
1168
+ "loss": 4.1497,
1169
+ "step": 97792
1170
+ },
1171
+ {
1172
+ "epoch": 1.01,
1173
+ "learning_rate": 4.839050409371821e-05,
1174
+ "loss": 4.1317,
1175
+ "step": 98304
1176
+ },
1177
+ {
1178
+ "epoch": 1.01,
1179
+ "learning_rate": 4.8382134525011416e-05,
1180
+ "loss": 4.1458,
1181
+ "step": 98816
1182
+ },
1183
+ {
1184
+ "epoch": 1.01,
1185
+ "learning_rate": 4.83737485775009e-05,
1186
+ "loss": 4.144,
1187
+ "step": 99328
1188
+ },
1189
+ {
1190
+ "epoch": 1.01,
1191
+ "learning_rate": 4.836536262999038e-05,
1192
+ "loss": 4.1423,
1193
+ "step": 99840
1194
+ },
1195
+ {
1196
+ "epoch": 1.01,
1197
+ "learning_rate": 4.835697668247986e-05,
1198
+ "loss": 4.1457,
1199
+ "step": 100352
1200
+ },
1201
+ {
1202
+ "epoch": 1.01,
1203
+ "learning_rate": 4.834860711377307e-05,
1204
+ "loss": 4.1179,
1205
+ "step": 100864
1206
+ },
1207
+ {
1208
+ "epoch": 1.01,
1209
+ "learning_rate": 4.834022116626255e-05,
1210
+ "loss": 4.1238,
1211
+ "step": 101376
1212
+ },
1213
+ {
1214
+ "epoch": 1.01,
1215
+ "learning_rate": 4.833183521875203e-05,
1216
+ "loss": 4.1366,
1217
+ "step": 101888
1218
+ },
1219
+ {
1220
+ "epoch": 1.01,
1221
+ "learning_rate": 4.832344927124151e-05,
1222
+ "loss": 4.1342,
1223
+ "step": 102400
1224
+ },
1225
+ {
1226
+ "epoch": 1.01,
1227
+ "learning_rate": 4.831507970253472e-05,
1228
+ "loss": 4.1264,
1229
+ "step": 102912
1230
+ },
1231
+ {
1232
+ "epoch": 1.01,
1233
+ "learning_rate": 4.83066937550242e-05,
1234
+ "loss": 4.1215,
1235
+ "step": 103424
1236
+ },
1237
+ {
1238
+ "epoch": 1.01,
1239
+ "learning_rate": 4.829830780751368e-05,
1240
+ "loss": 4.1168,
1241
+ "step": 103936
1242
+ },
1243
+ {
1244
+ "epoch": 1.01,
1245
+ "learning_rate": 4.828992186000316e-05,
1246
+ "loss": 4.1084,
1247
+ "step": 104448
1248
+ },
1249
+ {
1250
+ "epoch": 1.01,
1251
+ "learning_rate": 4.828155229129637e-05,
1252
+ "loss": 4.1234,
1253
+ "step": 104960
1254
+ },
1255
+ {
1256
+ "epoch": 1.01,
1257
+ "learning_rate": 4.8273166343785856e-05,
1258
+ "loss": 4.105,
1259
+ "step": 105472
1260
+ },
1261
+ {
1262
+ "epoch": 1.01,
1263
+ "learning_rate": 4.8264780396275336e-05,
1264
+ "loss": 4.1149,
1265
+ "step": 105984
1266
+ },
1267
+ {
1268
+ "epoch": 1.01,
1269
+ "learning_rate": 4.8256394448764816e-05,
1270
+ "loss": 4.1243,
1271
+ "step": 106496
1272
+ },
1273
+ {
1274
+ "epoch": 1.01,
1275
+ "learning_rate": 4.8248024880058025e-05,
1276
+ "loss": 4.1045,
1277
+ "step": 107008
1278
+ },
1279
+ {
1280
+ "epoch": 1.01,
1281
+ "learning_rate": 4.8239638932547505e-05,
1282
+ "loss": 4.1043,
1283
+ "step": 107520
1284
+ },
1285
+ {
1286
+ "epoch": 1.01,
1287
+ "learning_rate": 4.8231252985036985e-05,
1288
+ "loss": 4.1042,
1289
+ "step": 108032
1290
+ },
1291
+ {
1292
+ "epoch": 1.01,
1293
+ "learning_rate": 4.8222867037526465e-05,
1294
+ "loss": 4.0964,
1295
+ "step": 108544
1296
+ },
1297
+ {
1298
+ "epoch": 1.01,
1299
+ "learning_rate": 4.8214497468819674e-05,
1300
+ "loss": 4.0984,
1301
+ "step": 109056
1302
+ },
1303
+ {
1304
+ "epoch": 1.01,
1305
+ "learning_rate": 4.8206111521309154e-05,
1306
+ "loss": 4.1156,
1307
+ "step": 109568
1308
+ },
1309
+ {
1310
+ "epoch": 1.01,
1311
+ "learning_rate": 4.8197725573798634e-05,
1312
+ "loss": 4.0981,
1313
+ "step": 110080
1314
+ },
1315
+ {
1316
+ "epoch": 1.01,
1317
+ "learning_rate": 4.8189339626288114e-05,
1318
+ "loss": 4.0959,
1319
+ "step": 110592
1320
+ },
1321
+ {
1322
+ "epoch": 1.01,
1323
+ "learning_rate": 4.818097005758132e-05,
1324
+ "loss": 4.0935,
1325
+ "step": 111104
1326
+ },
1327
+ {
1328
+ "epoch": 1.01,
1329
+ "learning_rate": 4.817258411007081e-05,
1330
+ "loss": 4.0933,
1331
+ "step": 111616
1332
+ },
1333
+ {
1334
+ "epoch": 1.01,
1335
+ "learning_rate": 4.816419816256029e-05,
1336
+ "loss": 4.1088,
1337
+ "step": 112128
1338
+ },
1339
+ {
1340
+ "epoch": 1.01,
1341
+ "learning_rate": 4.815581221504977e-05,
1342
+ "loss": 4.1027,
1343
+ "step": 112640
1344
+ },
1345
+ {
1346
+ "epoch": 1.01,
1347
+ "learning_rate": 4.814742626753924e-05,
1348
+ "loss": 4.0923,
1349
+ "step": 113152
1350
+ },
1351
+ {
1352
+ "epoch": 1.01,
1353
+ "learning_rate": 4.813905669883246e-05,
1354
+ "loss": 4.104,
1355
+ "step": 113664
1356
+ },
1357
+ {
1358
+ "epoch": 1.01,
1359
+ "learning_rate": 4.813067075132194e-05,
1360
+ "loss": 4.1052,
1361
+ "step": 114176
1362
+ },
1363
+ {
1364
+ "epoch": 1.01,
1365
+ "learning_rate": 4.812228480381141e-05,
1366
+ "loss": 4.0876,
1367
+ "step": 114688
1368
+ },
1369
+ {
1370
+ "epoch": 1.01,
1371
+ "learning_rate": 4.811391523510463e-05,
1372
+ "loss": 4.0936,
1373
+ "step": 115200
1374
+ },
1375
+ {
1376
+ "epoch": 1.01,
1377
+ "learning_rate": 4.810552928759411e-05,
1378
+ "loss": 4.0909,
1379
+ "step": 115712
1380
+ },
1381
+ {
1382
+ "epoch": 1.01,
1383
+ "learning_rate": 4.809714334008359e-05,
1384
+ "loss": 4.0888,
1385
+ "step": 116224
1386
+ },
1387
+ {
1388
+ "epoch": 1.01,
1389
+ "learning_rate": 4.808875739257306e-05,
1390
+ "loss": 4.0813,
1391
+ "step": 116736
1392
+ },
1393
+ {
1394
+ "epoch": 1.01,
1395
+ "learning_rate": 4.808037144506255e-05,
1396
+ "loss": 4.0844,
1397
+ "step": 117248
1398
+ },
1399
+ {
1400
+ "epoch": 1.01,
1401
+ "learning_rate": 4.807200187635576e-05,
1402
+ "loss": 4.0962,
1403
+ "step": 117760
1404
+ },
1405
+ {
1406
+ "epoch": 1.01,
1407
+ "learning_rate": 4.8063615928845243e-05,
1408
+ "loss": 4.0938,
1409
+ "step": 118272
1410
+ },
1411
+ {
1412
+ "epoch": 1.01,
1413
+ "learning_rate": 4.8055229981334717e-05,
1414
+ "loss": 4.0827,
1415
+ "step": 118784
1416
+ },
1417
+ {
1418
+ "epoch": 1.01,
1419
+ "learning_rate": 4.804686041262793e-05,
1420
+ "loss": 4.0602,
1421
+ "step": 119296
1422
+ },
1423
+ {
1424
+ "epoch": 1.01,
1425
+ "learning_rate": 4.803847446511741e-05,
1426
+ "loss": 4.0764,
1427
+ "step": 119808
1428
+ },
1429
+ {
1430
+ "epoch": 1.01,
1431
+ "learning_rate": 4.8030088517606886e-05,
1432
+ "loss": 4.0819,
1433
+ "step": 120320
1434
+ },
1435
+ {
1436
+ "epoch": 1.01,
1437
+ "learning_rate": 4.8021702570096366e-05,
1438
+ "loss": 4.0806,
1439
+ "step": 120832
1440
+ },
1441
+ {
1442
+ "epoch": 1.01,
1443
+ "learning_rate": 4.801333300138958e-05,
1444
+ "loss": 4.0763,
1445
+ "step": 121344
1446
+ },
1447
+ {
1448
+ "epoch": 1.01,
1449
+ "learning_rate": 4.800494705387906e-05,
1450
+ "loss": 4.0722,
1451
+ "step": 121856
1452
+ },
1453
+ {
1454
+ "epoch": 1.02,
1455
+ "learning_rate": 4.799657748517227e-05,
1456
+ "loss": 4.0791,
1457
+ "step": 122368
1458
+ },
1459
+ {
1460
+ "epoch": 1.02,
1461
+ "learning_rate": 4.798819153766175e-05,
1462
+ "loss": 4.0707,
1463
+ "step": 122880
1464
+ },
1465
+ {
1466
+ "epoch": 1.02,
1467
+ "learning_rate": 4.797980559015123e-05,
1468
+ "loss": 4.0663,
1469
+ "step": 123392
1470
+ },
1471
+ {
1472
+ "epoch": 1.02,
1473
+ "learning_rate": 4.797141964264071e-05,
1474
+ "loss": 4.079,
1475
+ "step": 123904
1476
+ },
1477
+ {
1478
+ "epoch": 1.02,
1479
+ "learning_rate": 4.796303369513019e-05,
1480
+ "loss": 4.0618,
1481
+ "step": 124416
1482
+ },
1483
+ {
1484
+ "epoch": 1.02,
1485
+ "learning_rate": 4.7954664126423406e-05,
1486
+ "loss": 4.0825,
1487
+ "step": 124928
1488
+ },
1489
+ {
1490
+ "epoch": 1.02,
1491
+ "learning_rate": 4.7946278178912886e-05,
1492
+ "loss": 4.0719,
1493
+ "step": 125440
1494
+ },
1495
+ {
1496
+ "epoch": 1.02,
1497
+ "learning_rate": 4.793789223140236e-05,
1498
+ "loss": 4.0517,
1499
+ "step": 125952
1500
+ },
1501
+ {
1502
+ "epoch": 1.02,
1503
+ "learning_rate": 4.792950628389184e-05,
1504
+ "loss": 4.0694,
1505
+ "step": 126464
1506
+ },
1507
+ {
1508
+ "epoch": 1.02,
1509
+ "learning_rate": 4.792112033638132e-05,
1510
+ "loss": 4.0647,
1511
+ "step": 126976
1512
+ },
1513
+ {
1514
+ "epoch": 1.02,
1515
+ "learning_rate": 4.79127343888708e-05,
1516
+ "loss": 4.075,
1517
+ "step": 127488
1518
+ },
1519
+ {
1520
+ "epoch": 1.02,
1521
+ "learning_rate": 4.790434844136028e-05,
1522
+ "loss": 4.0654,
1523
+ "step": 128000
1524
+ },
1525
+ {
1526
+ "epoch": 1.02,
1527
+ "learning_rate": 4.789596249384976e-05,
1528
+ "loss": 4.0621,
1529
+ "step": 128512
1530
+ },
1531
+ {
1532
+ "epoch": 1.02,
1533
+ "learning_rate": 4.788759292514297e-05,
1534
+ "loss": 4.0499,
1535
+ "step": 129024
1536
+ },
1537
+ {
1538
+ "epoch": 1.02,
1539
+ "learning_rate": 4.787920697763245e-05,
1540
+ "loss": 4.0635,
1541
+ "step": 129536
1542
+ },
1543
+ {
1544
+ "epoch": 1.02,
1545
+ "learning_rate": 4.7870821030121935e-05,
1546
+ "loss": 4.0452,
1547
+ "step": 130048
1548
+ },
1549
+ {
1550
+ "epoch": 1.02,
1551
+ "learning_rate": 4.7862451461415144e-05,
1552
+ "loss": 4.0532,
1553
+ "step": 130560
1554
+ },
1555
+ {
1556
+ "epoch": 1.02,
1557
+ "learning_rate": 4.7854065513904624e-05,
1558
+ "loss": 4.0588,
1559
+ "step": 131072
1560
+ },
1561
+ {
1562
+ "epoch": 1.02,
1563
+ "learning_rate": 4.7845679566394104e-05,
1564
+ "loss": 4.0558,
1565
+ "step": 131584
1566
+ },
1567
+ {
1568
+ "epoch": 1.02,
1569
+ "learning_rate": 4.7837293618883584e-05,
1570
+ "loss": 4.0446,
1571
+ "step": 132096
1572
+ },
1573
+ {
1574
+ "epoch": 1.02,
1575
+ "learning_rate": 4.7828907671373064e-05,
1576
+ "loss": 4.0554,
1577
+ "step": 132608
1578
+ },
1579
+ {
1580
+ "epoch": 1.02,
1581
+ "learning_rate": 4.7820521723862544e-05,
1582
+ "loss": 4.049,
1583
+ "step": 133120
1584
+ },
1585
+ {
1586
+ "epoch": 1.02,
1587
+ "learning_rate": 4.7812135776352024e-05,
1588
+ "loss": 4.0541,
1589
+ "step": 133632
1590
+ },
1591
+ {
1592
+ "epoch": 1.02,
1593
+ "learning_rate": 4.7803749828841504e-05,
1594
+ "loss": 4.0507,
1595
+ "step": 134144
1596
+ },
1597
+ {
1598
+ "epoch": 1.02,
1599
+ "learning_rate": 4.779538026013471e-05,
1600
+ "loss": 4.0429,
1601
+ "step": 134656
1602
+ },
1603
+ {
1604
+ "epoch": 1.02,
1605
+ "learning_rate": 4.778699431262419e-05,
1606
+ "loss": 4.0674,
1607
+ "step": 135168
1608
+ },
1609
+ {
1610
+ "epoch": 1.02,
1611
+ "learning_rate": 4.777860836511367e-05,
1612
+ "loss": 4.0555,
1613
+ "step": 135680
1614
+ },
1615
+ {
1616
+ "epoch": 1.02,
1617
+ "learning_rate": 4.777022241760315e-05,
1618
+ "loss": 4.0525,
1619
+ "step": 136192
1620
+ },
1621
+ {
1622
+ "epoch": 1.02,
1623
+ "learning_rate": 4.776185284889637e-05,
1624
+ "loss": 4.0391,
1625
+ "step": 136704
1626
+ },
1627
+ {
1628
+ "epoch": 1.02,
1629
+ "learning_rate": 4.775346690138585e-05,
1630
+ "loss": 4.0466,
1631
+ "step": 137216
1632
+ },
1633
+ {
1634
+ "epoch": 1.02,
1635
+ "learning_rate": 4.774508095387533e-05,
1636
+ "loss": 4.0452,
1637
+ "step": 137728
1638
+ },
1639
+ {
1640
+ "epoch": 1.02,
1641
+ "learning_rate": 4.773669500636481e-05,
1642
+ "loss": 4.0392,
1643
+ "step": 138240
1644
+ },
1645
+ {
1646
+ "epoch": 1.02,
1647
+ "learning_rate": 4.772832543765802e-05,
1648
+ "loss": 4.0547,
1649
+ "step": 138752
1650
+ },
1651
+ {
1652
+ "epoch": 1.02,
1653
+ "learning_rate": 4.77199394901475e-05,
1654
+ "loss": 4.0425,
1655
+ "step": 139264
1656
+ },
1657
+ {
1658
+ "epoch": 1.02,
1659
+ "learning_rate": 4.771155354263698e-05,
1660
+ "loss": 4.0413,
1661
+ "step": 139776
1662
+ },
1663
+ {
1664
+ "epoch": 1.02,
1665
+ "learning_rate": 4.770316759512646e-05,
1666
+ "loss": 4.035,
1667
+ "step": 140288
1668
+ },
1669
+ {
1670
+ "epoch": 1.02,
1671
+ "learning_rate": 4.7694798026419666e-05,
1672
+ "loss": 4.0369,
1673
+ "step": 140800
1674
+ },
1675
+ {
1676
+ "epoch": 1.02,
1677
+ "learning_rate": 4.7686412078909146e-05,
1678
+ "loss": 4.038,
1679
+ "step": 141312
1680
+ },
1681
+ {
1682
+ "epoch": 1.02,
1683
+ "learning_rate": 4.7678026131398626e-05,
1684
+ "loss": 4.037,
1685
+ "step": 141824
1686
+ },
1687
+ {
1688
+ "epoch": 1.02,
1689
+ "learning_rate": 4.766965656269184e-05,
1690
+ "loss": 4.0266,
1691
+ "step": 142336
1692
+ },
1693
+ {
1694
+ "epoch": 1.02,
1695
+ "learning_rate": 4.766127061518132e-05,
1696
+ "loss": 4.0544,
1697
+ "step": 142848
1698
+ },
1699
+ {
1700
+ "epoch": 1.02,
1701
+ "learning_rate": 4.76528846676708e-05,
1702
+ "loss": 4.0437,
1703
+ "step": 143360
1704
+ },
1705
+ {
1706
+ "epoch": 1.02,
1707
+ "learning_rate": 4.764449872016028e-05,
1708
+ "loss": 4.0336,
1709
+ "step": 143872
1710
+ },
1711
+ {
1712
+ "epoch": 1.02,
1713
+ "learning_rate": 4.763611277264976e-05,
1714
+ "loss": 4.0305,
1715
+ "step": 144384
1716
+ },
1717
+ {
1718
+ "epoch": 1.02,
1719
+ "learning_rate": 4.762772682513924e-05,
1720
+ "loss": 4.0299,
1721
+ "step": 144896
1722
+ },
1723
+ {
1724
+ "epoch": 1.02,
1725
+ "learning_rate": 4.761934087762872e-05,
1726
+ "loss": 4.0397,
1727
+ "step": 145408
1728
+ },
1729
+ {
1730
+ "epoch": 1.02,
1731
+ "learning_rate": 4.7610954930118195e-05,
1732
+ "loss": 4.0356,
1733
+ "step": 145920
1734
+ },
1735
+ {
1736
+ "epoch": 1.02,
1737
+ "learning_rate": 4.760258536141141e-05,
1738
+ "loss": 4.0309,
1739
+ "step": 146432
1740
+ },
1741
+ {
1742
+ "epoch": 1.02,
1743
+ "learning_rate": 4.759419941390089e-05,
1744
+ "loss": 4.0415,
1745
+ "step": 146944
1746
+ },
1747
+ {
1748
+ "epoch": 1.02,
1749
+ "learning_rate": 4.758581346639037e-05,
1750
+ "loss": 4.0358,
1751
+ "step": 147456
1752
+ },
1753
+ {
1754
+ "epoch": 1.02,
1755
+ "learning_rate": 4.7577427518879844e-05,
1756
+ "loss": 4.0378,
1757
+ "step": 147968
1758
+ },
1759
+ {
1760
+ "epoch": 1.02,
1761
+ "learning_rate": 4.756905795017306e-05,
1762
+ "loss": 4.0258,
1763
+ "step": 148480
1764
+ },
1765
+ {
1766
+ "epoch": 1.02,
1767
+ "learning_rate": 4.756067200266254e-05,
1768
+ "loss": 4.0328,
1769
+ "step": 148992
1770
+ },
1771
+ {
1772
+ "epoch": 1.02,
1773
+ "learning_rate": 4.755228605515202e-05,
1774
+ "loss": 4.028,
1775
+ "step": 149504
1776
+ },
1777
+ {
1778
+ "epoch": 1.02,
1779
+ "learning_rate": 4.75439001076415e-05,
1780
+ "loss": 4.0261,
1781
+ "step": 150016
1782
+ },
1783
+ {
1784
+ "epoch": 1.02,
1785
+ "learning_rate": 4.7535530538934716e-05,
1786
+ "loss": 4.0267,
1787
+ "step": 150528
1788
+ },
1789
+ {
1790
+ "epoch": 1.02,
1791
+ "learning_rate": 4.7527144591424196e-05,
1792
+ "loss": 4.0249,
1793
+ "step": 151040
1794
+ },
1795
+ {
1796
+ "epoch": 1.02,
1797
+ "learning_rate": 4.751875864391367e-05,
1798
+ "loss": 4.0292,
1799
+ "step": 151552
1800
+ },
1801
+ {
1802
+ "epoch": 1.02,
1803
+ "learning_rate": 4.751037269640315e-05,
1804
+ "loss": 4.0176,
1805
+ "step": 152064
1806
+ },
1807
+ {
1808
+ "epoch": 1.02,
1809
+ "learning_rate": 4.7502003127696365e-05,
1810
+ "loss": 4.0186,
1811
+ "step": 152576
1812
+ },
1813
+ {
1814
+ "epoch": 1.03,
1815
+ "eval_loss": 4.032747268676758,
1816
+ "eval_runtime": 302.674,
1817
+ "eval_samples_per_second": 1260.733,
1818
+ "eval_steps_per_second": 39.399,
1819
+ "step": 152640
1820
+ },
1821
+ {
1822
+ "epoch": 0.0,
1823
+ "learning_rate": 4.7493617180185845e-05,
1824
+ "loss": 4.0048,
1825
+ "step": 153088
1826
+ },
1827
+ {
1828
+ "epoch": 0.0,
1829
+ "learning_rate": 4.748523123267532e-05,
1830
+ "loss": 4.007,
1831
+ "step": 153600
1832
+ },
1833
+ {
1834
+ "epoch": 0.0,
1835
+ "learning_rate": 4.74768452851648e-05,
1836
+ "loss": 4.0337,
1837
+ "step": 154112
1838
+ },
1839
+ {
1840
+ "epoch": 0.0,
1841
+ "learning_rate": 4.746845933765428e-05,
1842
+ "loss": 4.0255,
1843
+ "step": 154624
1844
+ },
1845
+ {
1846
+ "epoch": 0.0,
1847
+ "learning_rate": 4.7460073390143764e-05,
1848
+ "loss": 4.0172,
1849
+ "step": 155136
1850
+ },
1851
+ {
1852
+ "epoch": 0.0,
1853
+ "learning_rate": 4.7451687442633244e-05,
1854
+ "loss": 4.0119,
1855
+ "step": 155648
1856
+ },
1857
+ {
1858
+ "epoch": 0.0,
1859
+ "learning_rate": 4.7443301495122724e-05,
1860
+ "loss": 4.015,
1861
+ "step": 156160
1862
+ },
1863
+ {
1864
+ "epoch": 0.0,
1865
+ "learning_rate": 4.7434915547612204e-05,
1866
+ "loss": 4.0007,
1867
+ "step": 156672
1868
+ },
1869
+ {
1870
+ "epoch": 0.0,
1871
+ "learning_rate": 4.7426529600101684e-05,
1872
+ "loss": 4.0185,
1873
+ "step": 157184
1874
+ },
1875
+ {
1876
+ "epoch": 0.0,
1877
+ "learning_rate": 4.7418143652591164e-05,
1878
+ "loss": 4.0119,
1879
+ "step": 157696
1880
+ },
1881
+ {
1882
+ "epoch": 0.0,
1883
+ "learning_rate": 4.7409757705080644e-05,
1884
+ "loss": 4.0115,
1885
+ "step": 158208
1886
+ },
1887
+ {
1888
+ "epoch": 0.0,
1889
+ "learning_rate": 4.740138813637385e-05,
1890
+ "loss": 4.0153,
1891
+ "step": 158720
1892
+ },
1893
+ {
1894
+ "epoch": 0.0,
1895
+ "learning_rate": 4.739303494647079e-05,
1896
+ "loss": 4.002,
1897
+ "step": 159232
1898
+ },
1899
+ {
1900
+ "epoch": 0.0,
1901
+ "learning_rate": 4.738464899896027e-05,
1902
+ "loss": 4.0025,
1903
+ "step": 159744
1904
+ },
1905
+ {
1906
+ "epoch": 0.0,
1907
+ "learning_rate": 4.737626305144975e-05,
1908
+ "loss": 3.9976,
1909
+ "step": 160256
1910
+ },
1911
+ {
1912
+ "epoch": 0.0,
1913
+ "learning_rate": 4.736787710393923e-05,
1914
+ "loss": 3.9954,
1915
+ "step": 160768
1916
+ },
1917
+ {
1918
+ "epoch": 0.0,
1919
+ "learning_rate": 4.735949115642871e-05,
1920
+ "loss": 3.9998,
1921
+ "step": 161280
1922
+ },
1923
+ {
1924
+ "epoch": 0.0,
1925
+ "learning_rate": 4.73511052089182e-05,
1926
+ "loss": 3.994,
1927
+ "step": 161792
1928
+ },
1929
+ {
1930
+ "epoch": 0.0,
1931
+ "learning_rate": 4.734271926140768e-05,
1932
+ "loss": 4.0013,
1933
+ "step": 162304
1934
+ },
1935
+ {
1936
+ "epoch": 0.0,
1937
+ "learning_rate": 4.733433331389716e-05,
1938
+ "loss": 4.0239,
1939
+ "step": 162816
1940
+ },
1941
+ {
1942
+ "epoch": 0.0,
1943
+ "learning_rate": 4.732594736638664e-05,
1944
+ "loss": 3.9984,
1945
+ "step": 163328
1946
+ },
1947
+ {
1948
+ "epoch": 0.0,
1949
+ "learning_rate": 4.731756141887612e-05,
1950
+ "loss": 4.0038,
1951
+ "step": 163840
1952
+ },
1953
+ {
1954
+ "epoch": 0.0,
1955
+ "learning_rate": 4.73091754713656e-05,
1956
+ "loss": 3.9984,
1957
+ "step": 164352
1958
+ },
1959
+ {
1960
+ "epoch": 0.0,
1961
+ "learning_rate": 4.730080590265881e-05,
1962
+ "loss": 4.0074,
1963
+ "step": 164864
1964
+ },
1965
+ {
1966
+ "epoch": 0.0,
1967
+ "learning_rate": 4.729241995514829e-05,
1968
+ "loss": 3.9885,
1969
+ "step": 165376
1970
+ },
1971
+ {
1972
+ "epoch": 0.0,
1973
+ "learning_rate": 4.728403400763777e-05,
1974
+ "loss": 3.9984,
1975
+ "step": 165888
1976
+ },
1977
+ {
1978
+ "epoch": 0.0,
1979
+ "learning_rate": 4.727564806012725e-05,
1980
+ "loss": 3.9898,
1981
+ "step": 166400
1982
+ },
1983
+ {
1984
+ "epoch": 0.0,
1985
+ "learning_rate": 4.7267278491420456e-05,
1986
+ "loss": 3.9903,
1987
+ "step": 166912
1988
+ },
1989
+ {
1990
+ "epoch": 0.0,
1991
+ "learning_rate": 4.7258892543909936e-05,
1992
+ "loss": 3.9847,
1993
+ "step": 167424
1994
+ },
1995
+ {
1996
+ "epoch": 0.01,
1997
+ "learning_rate": 4.7250506596399416e-05,
1998
+ "loss": 3.985,
1999
+ "step": 167936
2000
+ },
2001
+ {
2002
+ "epoch": 0.01,
2003
+ "learning_rate": 4.7242120648888896e-05,
2004
+ "loss": 3.9899,
2005
+ "step": 168448
2006
+ },
2007
+ {
2008
+ "epoch": 0.01,
2009
+ "learning_rate": 4.723375108018211e-05,
2010
+ "loss": 4.0005,
2011
+ "step": 168960
2012
+ },
2013
+ {
2014
+ "epoch": 0.01,
2015
+ "learning_rate": 4.722536513267159e-05,
2016
+ "loss": 3.9893,
2017
+ "step": 169472
2018
+ },
2019
+ {
2020
+ "epoch": 0.01,
2021
+ "learning_rate": 4.721697918516107e-05,
2022
+ "loss": 3.9968,
2023
+ "step": 169984
2024
+ },
2025
+ {
2026
+ "epoch": 0.01,
2027
+ "learning_rate": 4.720859323765055e-05,
2028
+ "loss": 3.9822,
2029
+ "step": 170496
2030
+ },
2031
+ {
2032
+ "epoch": 0.01,
2033
+ "learning_rate": 4.720020729014003e-05,
2034
+ "loss": 4.0003,
2035
+ "step": 171008
2036
+ },
2037
+ {
2038
+ "epoch": 0.01,
2039
+ "learning_rate": 4.7191821342629505e-05,
2040
+ "loss": 3.97,
2041
+ "step": 171520
2042
+ },
2043
+ {
2044
+ "epoch": 0.01,
2045
+ "learning_rate": 4.7183435395118985e-05,
2046
+ "loss": 3.9902,
2047
+ "step": 172032
2048
+ },
2049
+ {
2050
+ "epoch": 0.01,
2051
+ "learning_rate": 4.71750658264122e-05,
2052
+ "loss": 3.9657,
2053
+ "step": 172544
2054
+ },
2055
+ {
2056
+ "epoch": 0.01,
2057
+ "learning_rate": 4.716667987890168e-05,
2058
+ "loss": 3.9867,
2059
+ "step": 173056
2060
+ },
2061
+ {
2062
+ "epoch": 0.01,
2063
+ "learning_rate": 4.7158293931391154e-05,
2064
+ "loss": 3.9753,
2065
+ "step": 173568
2066
+ },
2067
+ {
2068
+ "epoch": 0.01,
2069
+ "learning_rate": 4.7149907983880634e-05,
2070
+ "loss": 3.9899,
2071
+ "step": 174080
2072
+ },
2073
+ {
2074
+ "epoch": 0.01,
2075
+ "learning_rate": 4.714153841517385e-05,
2076
+ "loss": 3.9752,
2077
+ "step": 174592
2078
+ },
2079
+ {
2080
+ "epoch": 0.01,
2081
+ "learning_rate": 4.713315246766333e-05,
2082
+ "loss": 3.9808,
2083
+ "step": 175104
2084
+ },
2085
+ {
2086
+ "epoch": 0.01,
2087
+ "learning_rate": 4.712476652015281e-05,
2088
+ "loss": 3.9836,
2089
+ "step": 175616
2090
+ },
2091
+ {
2092
+ "epoch": 0.01,
2093
+ "learning_rate": 4.711638057264229e-05,
2094
+ "loss": 3.9864,
2095
+ "step": 176128
2096
+ },
2097
+ {
2098
+ "epoch": 0.01,
2099
+ "learning_rate": 4.7108011003935505e-05,
2100
+ "loss": 3.9837,
2101
+ "step": 176640
2102
+ },
2103
+ {
2104
+ "epoch": 0.01,
2105
+ "learning_rate": 4.709962505642498e-05,
2106
+ "loss": 3.9659,
2107
+ "step": 177152
2108
+ },
2109
+ {
2110
+ "epoch": 0.01,
2111
+ "learning_rate": 4.709123910891446e-05,
2112
+ "loss": 3.9652,
2113
+ "step": 177664
2114
+ },
2115
+ {
2116
+ "epoch": 0.01,
2117
+ "learning_rate": 4.708285316140394e-05,
2118
+ "loss": 3.9821,
2119
+ "step": 178176
2120
+ },
2121
+ {
2122
+ "epoch": 0.01,
2123
+ "learning_rate": 4.7074483592697154e-05,
2124
+ "loss": 3.98,
2125
+ "step": 178688
2126
+ },
2127
+ {
2128
+ "epoch": 0.01,
2129
+ "learning_rate": 4.706611402399036e-05,
2130
+ "loss": 3.9719,
2131
+ "step": 179200
2132
+ },
2133
+ {
2134
+ "epoch": 0.01,
2135
+ "learning_rate": 4.705772807647984e-05,
2136
+ "loss": 3.9691,
2137
+ "step": 179712
2138
+ },
2139
+ {
2140
+ "epoch": 0.01,
2141
+ "learning_rate": 4.704934212896932e-05,
2142
+ "loss": 3.9675,
2143
+ "step": 180224
2144
+ },
2145
+ {
2146
+ "epoch": 0.01,
2147
+ "learning_rate": 4.70409561814588e-05,
2148
+ "loss": 3.9536,
2149
+ "step": 180736
2150
+ },
2151
+ {
2152
+ "epoch": 0.01,
2153
+ "learning_rate": 4.703257023394828e-05,
2154
+ "loss": 3.9688,
2155
+ "step": 181248
2156
+ },
2157
+ {
2158
+ "epoch": 0.01,
2159
+ "learning_rate": 4.702418428643776e-05,
2160
+ "loss": 3.9559,
2161
+ "step": 181760
2162
+ },
2163
+ {
2164
+ "epoch": 0.01,
2165
+ "learning_rate": 4.701579833892724e-05,
2166
+ "loss": 3.9672,
2167
+ "step": 182272
2168
+ },
2169
+ {
2170
+ "epoch": 0.01,
2171
+ "learning_rate": 4.700742877022045e-05,
2172
+ "loss": 3.9763,
2173
+ "step": 182784
2174
+ },
2175
+ {
2176
+ "epoch": 0.01,
2177
+ "learning_rate": 4.699904282270993e-05,
2178
+ "loss": 3.9564,
2179
+ "step": 183296
2180
+ },
2181
+ {
2182
+ "epoch": 0.01,
2183
+ "learning_rate": 4.699065687519941e-05,
2184
+ "loss": 3.9572,
2185
+ "step": 183808
2186
+ },
2187
+ {
2188
+ "epoch": 0.01,
2189
+ "learning_rate": 4.698227092768889e-05,
2190
+ "loss": 3.9613,
2191
+ "step": 184320
2192
+ },
2193
+ {
2194
+ "epoch": 0.01,
2195
+ "learning_rate": 4.69739013589821e-05,
2196
+ "loss": 3.9492,
2197
+ "step": 184832
2198
+ },
2199
+ {
2200
+ "epoch": 0.01,
2201
+ "learning_rate": 4.696551541147158e-05,
2202
+ "loss": 3.9492,
2203
+ "step": 185344
2204
+ },
2205
+ {
2206
+ "epoch": 0.01,
2207
+ "learning_rate": 4.695712946396106e-05,
2208
+ "loss": 3.97,
2209
+ "step": 185856
2210
+ },
2211
+ {
2212
+ "epoch": 0.01,
2213
+ "learning_rate": 4.694874351645054e-05,
2214
+ "loss": 3.9539,
2215
+ "step": 186368
2216
+ },
2217
+ {
2218
+ "epoch": 0.01,
2219
+ "learning_rate": 4.694037394774376e-05,
2220
+ "loss": 3.9525,
2221
+ "step": 186880
2222
+ },
2223
+ {
2224
+ "epoch": 0.01,
2225
+ "learning_rate": 4.693198800023324e-05,
2226
+ "loss": 3.9506,
2227
+ "step": 187392
2228
+ },
2229
+ {
2230
+ "epoch": 0.01,
2231
+ "learning_rate": 4.692361843152645e-05,
2232
+ "loss": 3.9528,
2233
+ "step": 187904
2234
+ },
2235
+ {
2236
+ "epoch": 0.01,
2237
+ "learning_rate": 4.6915232484015926e-05,
2238
+ "loss": 3.9649,
2239
+ "step": 188416
2240
+ },
2241
+ {
2242
+ "epoch": 0.01,
2243
+ "learning_rate": 4.6906846536505406e-05,
2244
+ "loss": 3.9618,
2245
+ "step": 188928
2246
+ },
2247
+ {
2248
+ "epoch": 0.01,
2249
+ "learning_rate": 4.6898460588994886e-05,
2250
+ "loss": 3.9499,
2251
+ "step": 189440
2252
+ },
2253
+ {
2254
+ "epoch": 0.01,
2255
+ "learning_rate": 4.6890074641484366e-05,
2256
+ "loss": 3.9648,
2257
+ "step": 189952
2258
+ },
2259
+ {
2260
+ "epoch": 0.01,
2261
+ "learning_rate": 4.6881688693973846e-05,
2262
+ "loss": 3.961,
2263
+ "step": 190464
2264
+ },
2265
+ {
2266
+ "epoch": 0.01,
2267
+ "learning_rate": 4.6873302746463326e-05,
2268
+ "loss": 3.9504,
2269
+ "step": 190976
2270
+ },
2271
+ {
2272
+ "epoch": 0.01,
2273
+ "learning_rate": 4.6864916798952806e-05,
2274
+ "loss": 3.9549,
2275
+ "step": 191488
2276
+ },
2277
+ {
2278
+ "epoch": 0.01,
2279
+ "learning_rate": 4.6856547230246015e-05,
2280
+ "loss": 3.9563,
2281
+ "step": 192000
2282
+ },
2283
+ {
2284
+ "epoch": 0.01,
2285
+ "learning_rate": 4.6848161282735495e-05,
2286
+ "loss": 3.9511,
2287
+ "step": 192512
2288
+ },
2289
+ {
2290
+ "epoch": 0.01,
2291
+ "learning_rate": 4.683979171402871e-05,
2292
+ "loss": 3.948,
2293
+ "step": 193024
2294
+ },
2295
+ {
2296
+ "epoch": 0.01,
2297
+ "learning_rate": 4.683140576651819e-05,
2298
+ "loss": 3.9472,
2299
+ "step": 193536
2300
+ },
2301
+ {
2302
+ "epoch": 0.01,
2303
+ "learning_rate": 4.682301981900767e-05,
2304
+ "loss": 3.96,
2305
+ "step": 194048
2306
+ },
2307
+ {
2308
+ "epoch": 0.01,
2309
+ "learning_rate": 4.681463387149715e-05,
2310
+ "loss": 3.9609,
2311
+ "step": 194560
2312
+ },
2313
+ {
2314
+ "epoch": 0.01,
2315
+ "learning_rate": 4.680624792398663e-05,
2316
+ "loss": 3.9449,
2317
+ "step": 195072
2318
+ },
2319
+ {
2320
+ "epoch": 0.01,
2321
+ "learning_rate": 4.679786197647611e-05,
2322
+ "loss": 3.9335,
2323
+ "step": 195584
2324
+ },
2325
+ {
2326
+ "epoch": 0.01,
2327
+ "learning_rate": 4.678947602896559e-05,
2328
+ "loss": 3.9396,
2329
+ "step": 196096
2330
+ },
2331
+ {
2332
+ "epoch": 0.01,
2333
+ "learning_rate": 4.678109008145507e-05,
2334
+ "loss": 3.9483,
2335
+ "step": 196608
2336
+ },
2337
+ {
2338
+ "epoch": 0.01,
2339
+ "learning_rate": 4.677272051274828e-05,
2340
+ "loss": 3.9506,
2341
+ "step": 197120
2342
+ },
2343
+ {
2344
+ "epoch": 0.01,
2345
+ "learning_rate": 4.676435094404149e-05,
2346
+ "loss": 3.9422,
2347
+ "step": 197632
2348
+ },
2349
+ {
2350
+ "epoch": 0.01,
2351
+ "learning_rate": 4.675596499653097e-05,
2352
+ "loss": 3.9392,
2353
+ "step": 198144
2354
+ },
2355
+ {
2356
+ "epoch": 0.02,
2357
+ "learning_rate": 4.674757904902045e-05,
2358
+ "loss": 3.9536,
2359
+ "step": 198656
2360
+ },
2361
+ {
2362
+ "epoch": 0.02,
2363
+ "learning_rate": 4.6739193101509935e-05,
2364
+ "loss": 3.9364,
2365
+ "step": 199168
2366
+ },
2367
+ {
2368
+ "epoch": 0.02,
2369
+ "learning_rate": 4.6730807153999415e-05,
2370
+ "loss": 3.939,
2371
+ "step": 199680
2372
+ },
2373
+ {
2374
+ "epoch": 0.02,
2375
+ "learning_rate": 4.6722421206488895e-05,
2376
+ "loss": 3.9469,
2377
+ "step": 200192
2378
+ },
2379
+ {
2380
+ "epoch": 0.02,
2381
+ "learning_rate": 4.6714051637782104e-05,
2382
+ "loss": 3.9354,
2383
+ "step": 200704
2384
+ },
2385
+ {
2386
+ "epoch": 0.02,
2387
+ "learning_rate": 4.6705665690271584e-05,
2388
+ "loss": 3.9547,
2389
+ "step": 201216
2390
+ },
2391
+ {
2392
+ "epoch": 0.02,
2393
+ "learning_rate": 4.6697279742761064e-05,
2394
+ "loss": 3.9413,
2395
+ "step": 201728
2396
+ },
2397
+ {
2398
+ "epoch": 0.02,
2399
+ "learning_rate": 4.6688893795250544e-05,
2400
+ "loss": 3.9252,
2401
+ "step": 202240
2402
+ },
2403
+ {
2404
+ "epoch": 0.02,
2405
+ "learning_rate": 4.6680507847740024e-05,
2406
+ "loss": 3.942,
2407
+ "step": 202752
2408
+ },
2409
+ {
2410
+ "epoch": 0.02,
2411
+ "learning_rate": 4.6672121900229504e-05,
2412
+ "loss": 3.9358,
2413
+ "step": 203264
2414
+ },
2415
+ {
2416
+ "epoch": 0.02,
2417
+ "learning_rate": 4.6663735952718984e-05,
2418
+ "loss": 3.952,
2419
+ "step": 203776
2420
+ },
2421
+ {
2422
+ "epoch": 0.02,
2423
+ "learning_rate": 4.665535000520846e-05,
2424
+ "loss": 3.942,
2425
+ "step": 204288
2426
+ },
2427
+ {
2428
+ "epoch": 0.02,
2429
+ "learning_rate": 4.66469968153054e-05,
2430
+ "loss": 3.9341,
2431
+ "step": 204800
2432
+ },
2433
+ {
2434
+ "epoch": 0.02,
2435
+ "learning_rate": 4.663861086779488e-05,
2436
+ "loss": 3.922,
2437
+ "step": 205312
2438
+ },
2439
+ {
2440
+ "epoch": 0.02,
2441
+ "learning_rate": 4.663022492028437e-05,
2442
+ "loss": 3.94,
2443
+ "step": 205824
2444
+ },
2445
+ {
2446
+ "epoch": 0.02,
2447
+ "learning_rate": 4.662183897277385e-05,
2448
+ "loss": 3.9226,
2449
+ "step": 206336
2450
+ },
2451
+ {
2452
+ "epoch": 0.02,
2453
+ "learning_rate": 4.661345302526333e-05,
2454
+ "loss": 3.9328,
2455
+ "step": 206848
2456
+ },
2457
+ {
2458
+ "epoch": 0.02,
2459
+ "learning_rate": 4.660508345655654e-05,
2460
+ "loss": 3.9358,
2461
+ "step": 207360
2462
+ },
2463
+ {
2464
+ "epoch": 0.02,
2465
+ "learning_rate": 4.659669750904602e-05,
2466
+ "loss": 3.9322,
2467
+ "step": 207872
2468
+ },
2469
+ {
2470
+ "epoch": 0.02,
2471
+ "learning_rate": 4.65883115615355e-05,
2472
+ "loss": 3.9218,
2473
+ "step": 208384
2474
+ },
2475
+ {
2476
+ "epoch": 0.02,
2477
+ "learning_rate": 4.657992561402498e-05,
2478
+ "loss": 3.9407,
2479
+ "step": 208896
2480
+ },
2481
+ {
2482
+ "epoch": 0.02,
2483
+ "learning_rate": 4.657153966651446e-05,
2484
+ "loss": 3.9249,
2485
+ "step": 209408
2486
+ },
2487
+ {
2488
+ "epoch": 0.02,
2489
+ "learning_rate": 4.656315371900393e-05,
2490
+ "loss": 3.9292,
2491
+ "step": 209920
2492
+ },
2493
+ {
2494
+ "epoch": 0.02,
2495
+ "learning_rate": 4.6554784150297146e-05,
2496
+ "loss": 3.9343,
2497
+ "step": 210432
2498
+ },
2499
+ {
2500
+ "epoch": 0.02,
2501
+ "learning_rate": 4.6546398202786626e-05,
2502
+ "loss": 3.9238,
2503
+ "step": 210944
2504
+ },
2505
+ {
2506
+ "epoch": 0.02,
2507
+ "learning_rate": 4.6538012255276106e-05,
2508
+ "loss": 3.9455,
2509
+ "step": 211456
2510
+ },
2511
+ {
2512
+ "epoch": 0.02,
2513
+ "learning_rate": 4.6529626307765586e-05,
2514
+ "loss": 3.934,
2515
+ "step": 211968
2516
+ },
2517
+ {
2518
+ "epoch": 0.02,
2519
+ "learning_rate": 4.65212567390588e-05,
2520
+ "loss": 3.9351,
2521
+ "step": 212480
2522
+ },
2523
+ {
2524
+ "epoch": 0.02,
2525
+ "learning_rate": 4.651287079154828e-05,
2526
+ "loss": 3.9219,
2527
+ "step": 212992
2528
+ },
2529
+ {
2530
+ "epoch": 0.02,
2531
+ "learning_rate": 4.650448484403776e-05,
2532
+ "loss": 3.9289,
2533
+ "step": 213504
2534
+ },
2535
+ {
2536
+ "epoch": 0.02,
2537
+ "learning_rate": 4.6496098896527235e-05,
2538
+ "loss": 3.9306,
2539
+ "step": 214016
2540
+ },
2541
+ {
2542
+ "epoch": 0.02,
2543
+ "learning_rate": 4.6487712949016715e-05,
2544
+ "loss": 3.9174,
2545
+ "step": 214528
2546
+ },
2547
+ {
2548
+ "epoch": 0.02,
2549
+ "learning_rate": 4.6479327001506195e-05,
2550
+ "loss": 3.9337,
2551
+ "step": 215040
2552
+ },
2553
+ {
2554
+ "epoch": 0.02,
2555
+ "learning_rate": 4.6470941053995675e-05,
2556
+ "loss": 3.9266,
2557
+ "step": 215552
2558
+ },
2559
+ {
2560
+ "epoch": 0.02,
2561
+ "learning_rate": 4.6462555106485155e-05,
2562
+ "loss": 3.9249,
2563
+ "step": 216064
2564
+ },
2565
+ {
2566
+ "epoch": 0.02,
2567
+ "learning_rate": 4.6454185537778364e-05,
2568
+ "loss": 3.9247,
2569
+ "step": 216576
2570
+ },
2571
+ {
2572
+ "epoch": 0.02,
2573
+ "learning_rate": 4.6445799590267844e-05,
2574
+ "loss": 3.9238,
2575
+ "step": 217088
2576
+ },
2577
+ {
2578
+ "epoch": 0.02,
2579
+ "learning_rate": 4.6437413642757324e-05,
2580
+ "loss": 3.9162,
2581
+ "step": 217600
2582
+ },
2583
+ {
2584
+ "epoch": 0.02,
2585
+ "learning_rate": 4.6429027695246804e-05,
2586
+ "loss": 3.9262,
2587
+ "step": 218112
2588
+ },
2589
+ {
2590
+ "epoch": 0.02,
2591
+ "learning_rate": 4.642065812654002e-05,
2592
+ "loss": 3.9156,
2593
+ "step": 218624
2594
+ },
2595
+ {
2596
+ "epoch": 0.02,
2597
+ "learning_rate": 4.64122721790295e-05,
2598
+ "loss": 3.9364,
2599
+ "step": 219136
2600
+ },
2601
+ {
2602
+ "epoch": 0.02,
2603
+ "learning_rate": 4.640388623151898e-05,
2604
+ "loss": 3.9335,
2605
+ "step": 219648
2606
+ },
2607
+ {
2608
+ "epoch": 0.02,
2609
+ "learning_rate": 4.639550028400846e-05,
2610
+ "loss": 3.9185,
2611
+ "step": 220160
2612
+ },
2613
+ {
2614
+ "epoch": 0.02,
2615
+ "learning_rate": 4.638711433649794e-05,
2616
+ "loss": 3.9168,
2617
+ "step": 220672
2618
+ },
2619
+ {
2620
+ "epoch": 0.02,
2621
+ "learning_rate": 4.637872838898742e-05,
2622
+ "loss": 3.918,
2623
+ "step": 221184
2624
+ },
2625
+ {
2626
+ "epoch": 0.02,
2627
+ "learning_rate": 4.63703424414769e-05,
2628
+ "loss": 3.9255,
2629
+ "step": 221696
2630
+ },
2631
+ {
2632
+ "epoch": 0.02,
2633
+ "learning_rate": 4.636195649396638e-05,
2634
+ "loss": 3.9251,
2635
+ "step": 222208
2636
+ },
2637
+ {
2638
+ "epoch": 0.02,
2639
+ "learning_rate": 4.635360330406332e-05,
2640
+ "loss": 3.9197,
2641
+ "step": 222720
2642
+ },
2643
+ {
2644
+ "epoch": 0.02,
2645
+ "learning_rate": 4.63452173565528e-05,
2646
+ "loss": 3.9264,
2647
+ "step": 223232
2648
+ },
2649
+ {
2650
+ "epoch": 0.02,
2651
+ "learning_rate": 4.633683140904228e-05,
2652
+ "loss": 3.9247,
2653
+ "step": 223744
2654
+ },
2655
+ {
2656
+ "epoch": 0.02,
2657
+ "learning_rate": 4.632844546153176e-05,
2658
+ "loss": 3.9284,
2659
+ "step": 224256
2660
+ },
2661
+ {
2662
+ "epoch": 0.02,
2663
+ "learning_rate": 4.6320059514021244e-05,
2664
+ "loss": 3.9127,
2665
+ "step": 224768
2666
+ },
2667
+ {
2668
+ "epoch": 0.02,
2669
+ "learning_rate": 4.6311689945314454e-05,
2670
+ "loss": 3.9265,
2671
+ "step": 225280
2672
+ },
2673
+ {
2674
+ "epoch": 0.02,
2675
+ "learning_rate": 4.6303303997803934e-05,
2676
+ "loss": 3.9211,
2677
+ "step": 225792
2678
+ },
2679
+ {
2680
+ "epoch": 0.02,
2681
+ "learning_rate": 4.6294918050293413e-05,
2682
+ "loss": 3.9114,
2683
+ "step": 226304
2684
+ },
2685
+ {
2686
+ "epoch": 0.02,
2687
+ "learning_rate": 4.6286532102782893e-05,
2688
+ "loss": 3.9212,
2689
+ "step": 226816
2690
+ },
2691
+ {
2692
+ "epoch": 0.02,
2693
+ "learning_rate": 4.6278146155272373e-05,
2694
+ "loss": 3.9196,
2695
+ "step": 227328
2696
+ },
2697
+ {
2698
+ "epoch": 0.02,
2699
+ "learning_rate": 4.626976020776185e-05,
2700
+ "loss": 3.9171,
2701
+ "step": 227840
2702
+ },
2703
+ {
2704
+ "epoch": 0.02,
2705
+ "learning_rate": 4.626137426025133e-05,
2706
+ "loss": 3.9079,
2707
+ "step": 228352
2708
+ },
2709
+ {
2710
+ "epoch": 0.02,
2711
+ "learning_rate": 4.625300469154454e-05,
2712
+ "loss": 3.9093,
2713
+ "step": 228864
2714
+ },
2715
+ {
2716
+ "epoch": 0.03,
2717
+ "eval_loss": 3.957947015762329,
2718
+ "eval_runtime": 310.3378,
2719
+ "eval_samples_per_second": 1229.599,
2720
+ "eval_steps_per_second": 38.426,
2721
+ "step": 228960
2722
+ }
2723
+ ],
2724
+ "logging_steps": 512,
2725
+ "max_steps": 3052726,
2726
+ "num_train_epochs": 9223372036854775807,
2727
+ "save_steps": 10,
2728
+ "total_flos": 1.5937334177716224e+17,
2729
+ "trial_name": null,
2730
+ "trial_params": null
2731
+ }
checkpoint-228960/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc007bd407836b97ec90388042036c3b95e075537017857e7d3d9b0e9a5ee3d5
3
+ size 4219