dcfidalgo commited on
Commit
b37798c
1 Parent(s): cb7b17b

Add first version 62000 checkpoint (except optimizer.pt/scheduler.pt)

Browse files
config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "dccuchile/bert-base-spanish-wwm-cased",
3
+ "architectures": [
4
+ "BertForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "LABEL_0",
13
+ "1": "LABEL_1",
14
+ "2": "LABEL_2"
15
+ },
16
+ "initializer_range": 0.02,
17
+ "intermediate_size": 3072,
18
+ "label2id": {
19
+ "LABEL_0": 0,
20
+ "LABEL_1": 1,
21
+ "LABEL_2": 2
22
+ },
23
+ "layer_norm_eps": 1e-12,
24
+ "max_position_embeddings": 512,
25
+ "model_type": "bert",
26
+ "num_attention_heads": 12,
27
+ "num_hidden_layers": 12,
28
+ "output_past": true,
29
+ "pad_token_id": 1,
30
+ "position_embedding_type": "absolute",
31
+ "transformers_version": "4.3.3",
32
+ "type_vocab_size": 2,
33
+ "use_cache": true,
34
+ "vocab_size": 31002
35
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8f7f0e947cf0426a7e01a66c7bdf058f86b82ed1c8c4cccf3dea046a6b56008
3
+ size 439502089
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"do_lower_case": false, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "special_tokens_map_file": "/home/ec2-user/.cache/huggingface/transformers/9848a00af462c42dfb4ec88ef438fbab5256330f7f6f50badc48d277f9367d49.dd8bd9bfd3664b530ea4e645105f557769387b3da9f79bdb55ed556bdd80611d", "name_or_path": "dccuchile/bert-base-spanish-wwm-cased", "do_basic_tokenize": true, "never_split": null}
trainer_state.json ADDED
@@ -0,0 +1,2496 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.8164658634538152,
3
+ "best_model_checkpoint": "./results/checkpoint-62000",
4
+ "epoch": 2.526075619295958,
5
+ "global_step": 62000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 2.5e-05,
13
+ "loss": 1.0225,
14
+ "step": 250
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 5e-05,
19
+ "loss": 0.8271,
20
+ "step": 500
21
+ },
22
+ {
23
+ "epoch": 0.02,
24
+ "eval_accuracy": 0.661847389558233,
25
+ "eval_loss": 0.8168992400169373,
26
+ "eval_runtime": 9.3368,
27
+ "eval_samples_per_second": 266.686,
28
+ "step": 500
29
+ },
30
+ {
31
+ "epoch": 0.03,
32
+ "learning_rate": 4.98290761909971e-05,
33
+ "loss": 0.7821,
34
+ "step": 750
35
+ },
36
+ {
37
+ "epoch": 0.04,
38
+ "learning_rate": 4.965815238199421e-05,
39
+ "loss": 0.7463,
40
+ "step": 1000
41
+ },
42
+ {
43
+ "epoch": 0.04,
44
+ "eval_accuracy": 0.708433734939759,
45
+ "eval_loss": 0.7267000675201416,
46
+ "eval_runtime": 9.0373,
47
+ "eval_samples_per_second": 275.525,
48
+ "step": 1000
49
+ },
50
+ {
51
+ "epoch": 0.05,
52
+ "learning_rate": 4.9487228572991307e-05,
53
+ "loss": 0.7266,
54
+ "step": 1250
55
+ },
56
+ {
57
+ "epoch": 0.06,
58
+ "learning_rate": 4.9316304763988406e-05,
59
+ "loss": 0.7156,
60
+ "step": 1500
61
+ },
62
+ {
63
+ "epoch": 0.06,
64
+ "eval_accuracy": 0.7180722891566265,
65
+ "eval_loss": 0.6909603476524353,
66
+ "eval_runtime": 9.3259,
67
+ "eval_samples_per_second": 266.998,
68
+ "step": 1500
69
+ },
70
+ {
71
+ "epoch": 0.07,
72
+ "learning_rate": 4.914538095498551e-05,
73
+ "loss": 0.7193,
74
+ "step": 1750
75
+ },
76
+ {
77
+ "epoch": 0.08,
78
+ "learning_rate": 4.897445714598261e-05,
79
+ "loss": 0.7073,
80
+ "step": 2000
81
+ },
82
+ {
83
+ "epoch": 0.08,
84
+ "eval_accuracy": 0.7176706827309237,
85
+ "eval_loss": 0.6720558404922485,
86
+ "eval_runtime": 9.0667,
87
+ "eval_samples_per_second": 274.632,
88
+ "step": 2000
89
+ },
90
+ {
91
+ "epoch": 0.09,
92
+ "learning_rate": 4.880353333697971e-05,
93
+ "loss": 0.6817,
94
+ "step": 2250
95
+ },
96
+ {
97
+ "epoch": 0.1,
98
+ "learning_rate": 4.8632609527976816e-05,
99
+ "loss": 0.6825,
100
+ "step": 2500
101
+ },
102
+ {
103
+ "epoch": 0.1,
104
+ "eval_accuracy": 0.7373493975903614,
105
+ "eval_loss": 0.6640393137931824,
106
+ "eval_runtime": 9.298,
107
+ "eval_samples_per_second": 267.799,
108
+ "step": 2500
109
+ },
110
+ {
111
+ "epoch": 0.11,
112
+ "learning_rate": 4.8461685718973915e-05,
113
+ "loss": 0.6759,
114
+ "step": 2750
115
+ },
116
+ {
117
+ "epoch": 0.12,
118
+ "learning_rate": 4.8290761909971014e-05,
119
+ "loss": 0.6886,
120
+ "step": 3000
121
+ },
122
+ {
123
+ "epoch": 0.12,
124
+ "eval_accuracy": 0.7578313253012048,
125
+ "eval_loss": 0.6005652546882629,
126
+ "eval_runtime": 9.0815,
127
+ "eval_samples_per_second": 274.184,
128
+ "step": 3000
129
+ },
130
+ {
131
+ "epoch": 0.13,
132
+ "learning_rate": 4.811983810096811e-05,
133
+ "loss": 0.6971,
134
+ "step": 3250
135
+ },
136
+ {
137
+ "epoch": 0.14,
138
+ "learning_rate": 4.794891429196521e-05,
139
+ "loss": 0.6642,
140
+ "step": 3500
141
+ },
142
+ {
143
+ "epoch": 0.14,
144
+ "eval_accuracy": 0.7582329317269076,
145
+ "eval_loss": 0.5993230938911438,
146
+ "eval_runtime": 8.992,
147
+ "eval_samples_per_second": 276.912,
148
+ "step": 3500
149
+ },
150
+ {
151
+ "epoch": 0.15,
152
+ "learning_rate": 4.777799048296232e-05,
153
+ "loss": 0.664,
154
+ "step": 3750
155
+ },
156
+ {
157
+ "epoch": 0.16,
158
+ "learning_rate": 4.760706667395942e-05,
159
+ "loss": 0.671,
160
+ "step": 4000
161
+ },
162
+ {
163
+ "epoch": 0.16,
164
+ "eval_accuracy": 0.7610441767068273,
165
+ "eval_loss": 0.5954618453979492,
166
+ "eval_runtime": 8.999,
167
+ "eval_samples_per_second": 276.698,
168
+ "step": 4000
169
+ },
170
+ {
171
+ "epoch": 0.17,
172
+ "learning_rate": 4.7436142864956516e-05,
173
+ "loss": 0.6421,
174
+ "step": 4250
175
+ },
176
+ {
177
+ "epoch": 0.18,
178
+ "learning_rate": 4.726521905595362e-05,
179
+ "loss": 0.6768,
180
+ "step": 4500
181
+ },
182
+ {
183
+ "epoch": 0.18,
184
+ "eval_accuracy": 0.7200803212851405,
185
+ "eval_loss": 0.6515903472900391,
186
+ "eval_runtime": 9.1435,
187
+ "eval_samples_per_second": 272.325,
188
+ "step": 4500
189
+ },
190
+ {
191
+ "epoch": 0.19,
192
+ "learning_rate": 4.709429524695072e-05,
193
+ "loss": 0.6606,
194
+ "step": 4750
195
+ },
196
+ {
197
+ "epoch": 0.2,
198
+ "learning_rate": 4.692337143794782e-05,
199
+ "loss": 0.6511,
200
+ "step": 5000
201
+ },
202
+ {
203
+ "epoch": 0.2,
204
+ "eval_accuracy": 0.7518072289156627,
205
+ "eval_loss": 0.6293097734451294,
206
+ "eval_runtime": 9.1699,
207
+ "eval_samples_per_second": 271.54,
208
+ "step": 5000
209
+ },
210
+ {
211
+ "epoch": 0.21,
212
+ "learning_rate": 4.6752447628944926e-05,
213
+ "loss": 0.651,
214
+ "step": 5250
215
+ },
216
+ {
217
+ "epoch": 0.22,
218
+ "learning_rate": 4.6581523819942026e-05,
219
+ "loss": 0.6729,
220
+ "step": 5500
221
+ },
222
+ {
223
+ "epoch": 0.22,
224
+ "eval_accuracy": 0.748995983935743,
225
+ "eval_loss": 0.6285788416862488,
226
+ "eval_runtime": 9.1397,
227
+ "eval_samples_per_second": 272.439,
228
+ "step": 5500
229
+ },
230
+ {
231
+ "epoch": 0.23,
232
+ "learning_rate": 4.6410600010939125e-05,
233
+ "loss": 0.6573,
234
+ "step": 5750
235
+ },
236
+ {
237
+ "epoch": 0.24,
238
+ "learning_rate": 4.623967620193623e-05,
239
+ "loss": 0.6335,
240
+ "step": 6000
241
+ },
242
+ {
243
+ "epoch": 0.24,
244
+ "eval_accuracy": 0.7485943775100402,
245
+ "eval_loss": 0.6102361679077148,
246
+ "eval_runtime": 9.2187,
247
+ "eval_samples_per_second": 270.105,
248
+ "step": 6000
249
+ },
250
+ {
251
+ "epoch": 0.25,
252
+ "learning_rate": 4.606875239293333e-05,
253
+ "loss": 0.6171,
254
+ "step": 6250
255
+ },
256
+ {
257
+ "epoch": 0.26,
258
+ "learning_rate": 4.589782858393043e-05,
259
+ "loss": 0.6445,
260
+ "step": 6500
261
+ },
262
+ {
263
+ "epoch": 0.26,
264
+ "eval_accuracy": 0.7578313253012048,
265
+ "eval_loss": 0.5952242612838745,
266
+ "eval_runtime": 9.1091,
267
+ "eval_samples_per_second": 273.352,
268
+ "step": 6500
269
+ },
270
+ {
271
+ "epoch": 0.28,
272
+ "learning_rate": 4.5726904774927535e-05,
273
+ "loss": 0.6371,
274
+ "step": 6750
275
+ },
276
+ {
277
+ "epoch": 0.29,
278
+ "learning_rate": 4.5555980965924634e-05,
279
+ "loss": 0.6285,
280
+ "step": 7000
281
+ },
282
+ {
283
+ "epoch": 0.29,
284
+ "eval_accuracy": 0.7550200803212851,
285
+ "eval_loss": 0.5844168663024902,
286
+ "eval_runtime": 9.3454,
287
+ "eval_samples_per_second": 266.442,
288
+ "step": 7000
289
+ },
290
+ {
291
+ "epoch": 0.3,
292
+ "learning_rate": 4.538505715692173e-05,
293
+ "loss": 0.645,
294
+ "step": 7250
295
+ },
296
+ {
297
+ "epoch": 0.31,
298
+ "learning_rate": 4.521413334791884e-05,
299
+ "loss": 0.6614,
300
+ "step": 7500
301
+ },
302
+ {
303
+ "epoch": 0.31,
304
+ "eval_accuracy": 0.7706827309236948,
305
+ "eval_loss": 0.5955267548561096,
306
+ "eval_runtime": 9.1185,
307
+ "eval_samples_per_second": 273.072,
308
+ "step": 7500
309
+ },
310
+ {
311
+ "epoch": 0.32,
312
+ "learning_rate": 4.504320953891594e-05,
313
+ "loss": 0.6042,
314
+ "step": 7750
315
+ },
316
+ {
317
+ "epoch": 0.33,
318
+ "learning_rate": 4.487228572991304e-05,
319
+ "loss": 0.6497,
320
+ "step": 8000
321
+ },
322
+ {
323
+ "epoch": 0.33,
324
+ "eval_accuracy": 0.7803212851405622,
325
+ "eval_loss": 0.5732316970825195,
326
+ "eval_runtime": 9.5547,
327
+ "eval_samples_per_second": 260.605,
328
+ "step": 8000
329
+ },
330
+ {
331
+ "epoch": 0.34,
332
+ "learning_rate": 4.470136192091014e-05,
333
+ "loss": 0.6394,
334
+ "step": 8250
335
+ },
336
+ {
337
+ "epoch": 0.35,
338
+ "learning_rate": 4.453043811190724e-05,
339
+ "loss": 0.6109,
340
+ "step": 8500
341
+ },
342
+ {
343
+ "epoch": 0.35,
344
+ "eval_accuracy": 0.7594377510040161,
345
+ "eval_loss": 0.5940960049629211,
346
+ "eval_runtime": 9.1183,
347
+ "eval_samples_per_second": 273.077,
348
+ "step": 8500
349
+ },
350
+ {
351
+ "epoch": 0.36,
352
+ "learning_rate": 4.4359514302904335e-05,
353
+ "loss": 0.6324,
354
+ "step": 8750
355
+ },
356
+ {
357
+ "epoch": 0.37,
358
+ "learning_rate": 4.418859049390144e-05,
359
+ "loss": 0.6266,
360
+ "step": 9000
361
+ },
362
+ {
363
+ "epoch": 0.37,
364
+ "eval_accuracy": 0.7730923694779116,
365
+ "eval_loss": 0.5806799530982971,
366
+ "eval_runtime": 9.6937,
367
+ "eval_samples_per_second": 256.869,
368
+ "step": 9000
369
+ },
370
+ {
371
+ "epoch": 0.38,
372
+ "learning_rate": 4.401766668489854e-05,
373
+ "loss": 0.622,
374
+ "step": 9250
375
+ },
376
+ {
377
+ "epoch": 0.39,
378
+ "learning_rate": 4.384674287589564e-05,
379
+ "loss": 0.629,
380
+ "step": 9500
381
+ },
382
+ {
383
+ "epoch": 0.39,
384
+ "eval_accuracy": 0.770281124497992,
385
+ "eval_loss": 0.5808362364768982,
386
+ "eval_runtime": 9.089,
387
+ "eval_samples_per_second": 273.959,
388
+ "step": 9500
389
+ },
390
+ {
391
+ "epoch": 0.4,
392
+ "learning_rate": 4.3675819066892745e-05,
393
+ "loss": 0.6107,
394
+ "step": 9750
395
+ },
396
+ {
397
+ "epoch": 0.41,
398
+ "learning_rate": 4.3504895257889844e-05,
399
+ "loss": 0.6347,
400
+ "step": 10000
401
+ },
402
+ {
403
+ "epoch": 0.41,
404
+ "eval_accuracy": 0.7650602409638554,
405
+ "eval_loss": 0.5932815074920654,
406
+ "eval_runtime": 9.5702,
407
+ "eval_samples_per_second": 260.184,
408
+ "step": 10000
409
+ },
410
+ {
411
+ "epoch": 0.42,
412
+ "learning_rate": 4.333397144888694e-05,
413
+ "loss": 0.6441,
414
+ "step": 10250
415
+ },
416
+ {
417
+ "epoch": 0.43,
418
+ "learning_rate": 4.316304763988405e-05,
419
+ "loss": 0.6181,
420
+ "step": 10500
421
+ },
422
+ {
423
+ "epoch": 0.43,
424
+ "eval_accuracy": 0.7815261044176707,
425
+ "eval_loss": 0.5563480257987976,
426
+ "eval_runtime": 9.0503,
427
+ "eval_samples_per_second": 275.128,
428
+ "step": 10500
429
+ },
430
+ {
431
+ "epoch": 0.44,
432
+ "learning_rate": 4.299212383088115e-05,
433
+ "loss": 0.6133,
434
+ "step": 10750
435
+ },
436
+ {
437
+ "epoch": 0.45,
438
+ "learning_rate": 4.282120002187825e-05,
439
+ "loss": 0.6173,
440
+ "step": 11000
441
+ },
442
+ {
443
+ "epoch": 0.45,
444
+ "eval_accuracy": 0.7775100401606426,
445
+ "eval_loss": 0.5630077719688416,
446
+ "eval_runtime": 9.4887,
447
+ "eval_samples_per_second": 262.417,
448
+ "step": 11000
449
+ },
450
+ {
451
+ "epoch": 0.46,
452
+ "learning_rate": 4.265027621287535e-05,
453
+ "loss": 0.5979,
454
+ "step": 11250
455
+ },
456
+ {
457
+ "epoch": 0.47,
458
+ "learning_rate": 4.247935240387245e-05,
459
+ "loss": 0.6114,
460
+ "step": 11500
461
+ },
462
+ {
463
+ "epoch": 0.47,
464
+ "eval_accuracy": 0.7666666666666667,
465
+ "eval_loss": 0.5745715498924255,
466
+ "eval_runtime": 9.0458,
467
+ "eval_samples_per_second": 275.267,
468
+ "step": 11500
469
+ },
470
+ {
471
+ "epoch": 0.48,
472
+ "learning_rate": 4.230842859486955e-05,
473
+ "loss": 0.6168,
474
+ "step": 11750
475
+ },
476
+ {
477
+ "epoch": 0.49,
478
+ "learning_rate": 4.213750478586666e-05,
479
+ "loss": 0.602,
480
+ "step": 12000
481
+ },
482
+ {
483
+ "epoch": 0.49,
484
+ "eval_accuracy": 0.7614457831325301,
485
+ "eval_loss": 0.5948830246925354,
486
+ "eval_runtime": 9.3904,
487
+ "eval_samples_per_second": 265.165,
488
+ "step": 12000
489
+ },
490
+ {
491
+ "epoch": 0.5,
492
+ "learning_rate": 4.1966580976863756e-05,
493
+ "loss": 0.6268,
494
+ "step": 12250
495
+ },
496
+ {
497
+ "epoch": 0.51,
498
+ "learning_rate": 4.1795657167860855e-05,
499
+ "loss": 0.6041,
500
+ "step": 12500
501
+ },
502
+ {
503
+ "epoch": 0.51,
504
+ "eval_accuracy": 0.7755020080321285,
505
+ "eval_loss": 0.5688386559486389,
506
+ "eval_runtime": 9.0467,
507
+ "eval_samples_per_second": 275.238,
508
+ "step": 12500
509
+ },
510
+ {
511
+ "epoch": 0.52,
512
+ "learning_rate": 4.162473335885796e-05,
513
+ "loss": 0.6284,
514
+ "step": 12750
515
+ },
516
+ {
517
+ "epoch": 0.53,
518
+ "learning_rate": 4.145380954985506e-05,
519
+ "loss": 0.6193,
520
+ "step": 13000
521
+ },
522
+ {
523
+ "epoch": 0.53,
524
+ "eval_accuracy": 0.7590361445783133,
525
+ "eval_loss": 0.583430290222168,
526
+ "eval_runtime": 9.3111,
527
+ "eval_samples_per_second": 267.423,
528
+ "step": 13000
529
+ },
530
+ {
531
+ "epoch": 0.54,
532
+ "learning_rate": 4.128288574085216e-05,
533
+ "loss": 0.5877,
534
+ "step": 13250
535
+ },
536
+ {
537
+ "epoch": 0.55,
538
+ "learning_rate": 4.1111961931849265e-05,
539
+ "loss": 0.5842,
540
+ "step": 13500
541
+ },
542
+ {
543
+ "epoch": 0.55,
544
+ "eval_accuracy": 0.7614457831325301,
545
+ "eval_loss": 0.5992956161499023,
546
+ "eval_runtime": 9.0426,
547
+ "eval_samples_per_second": 275.365,
548
+ "step": 13500
549
+ },
550
+ {
551
+ "epoch": 0.56,
552
+ "learning_rate": 4.0941038122846365e-05,
553
+ "loss": 0.592,
554
+ "step": 13750
555
+ },
556
+ {
557
+ "epoch": 0.57,
558
+ "learning_rate": 4.0770114313843464e-05,
559
+ "loss": 0.6068,
560
+ "step": 14000
561
+ },
562
+ {
563
+ "epoch": 0.57,
564
+ "eval_accuracy": 0.7666666666666667,
565
+ "eval_loss": 0.582902193069458,
566
+ "eval_runtime": 9.2607,
567
+ "eval_samples_per_second": 268.878,
568
+ "step": 14000
569
+ },
570
+ {
571
+ "epoch": 0.58,
572
+ "learning_rate": 4.059919050484056e-05,
573
+ "loss": 0.6125,
574
+ "step": 14250
575
+ },
576
+ {
577
+ "epoch": 0.59,
578
+ "learning_rate": 4.042826669583766e-05,
579
+ "loss": 0.6012,
580
+ "step": 14500
581
+ },
582
+ {
583
+ "epoch": 0.59,
584
+ "eval_accuracy": 0.7799196787148595,
585
+ "eval_loss": 0.565805196762085,
586
+ "eval_runtime": 9.0222,
587
+ "eval_samples_per_second": 275.985,
588
+ "step": 14500
589
+ },
590
+ {
591
+ "epoch": 0.6,
592
+ "learning_rate": 4.025734288683477e-05,
593
+ "loss": 0.5845,
594
+ "step": 14750
595
+ },
596
+ {
597
+ "epoch": 0.61,
598
+ "learning_rate": 4.008641907783187e-05,
599
+ "loss": 0.6159,
600
+ "step": 15000
601
+ },
602
+ {
603
+ "epoch": 0.61,
604
+ "eval_accuracy": 0.7538152610441767,
605
+ "eval_loss": 0.6279062628746033,
606
+ "eval_runtime": 9.2223,
607
+ "eval_samples_per_second": 269.997,
608
+ "step": 15000
609
+ },
610
+ {
611
+ "epoch": 0.62,
612
+ "learning_rate": 3.9915495268828966e-05,
613
+ "loss": 0.5897,
614
+ "step": 15250
615
+ },
616
+ {
617
+ "epoch": 0.63,
618
+ "learning_rate": 3.974457145982607e-05,
619
+ "loss": 0.5996,
620
+ "step": 15500
621
+ },
622
+ {
623
+ "epoch": 0.63,
624
+ "eval_accuracy": 0.7726907630522089,
625
+ "eval_loss": 0.5615552067756653,
626
+ "eval_runtime": 9.0848,
627
+ "eval_samples_per_second": 274.085,
628
+ "step": 15500
629
+ },
630
+ {
631
+ "epoch": 0.64,
632
+ "learning_rate": 3.957364765082317e-05,
633
+ "loss": 0.5899,
634
+ "step": 15750
635
+ },
636
+ {
637
+ "epoch": 0.65,
638
+ "learning_rate": 3.940272384182027e-05,
639
+ "loss": 0.6014,
640
+ "step": 16000
641
+ },
642
+ {
643
+ "epoch": 0.65,
644
+ "eval_accuracy": 0.7883534136546185,
645
+ "eval_loss": 0.5467997193336487,
646
+ "eval_runtime": 9.0085,
647
+ "eval_samples_per_second": 276.404,
648
+ "step": 16000
649
+ },
650
+ {
651
+ "epoch": 0.66,
652
+ "learning_rate": 3.9231800032817376e-05,
653
+ "loss": 0.6057,
654
+ "step": 16250
655
+ },
656
+ {
657
+ "epoch": 0.67,
658
+ "learning_rate": 3.9060876223814475e-05,
659
+ "loss": 0.6118,
660
+ "step": 16500
661
+ },
662
+ {
663
+ "epoch": 0.67,
664
+ "eval_accuracy": 0.7626506024096386,
665
+ "eval_loss": 0.572376012802124,
666
+ "eval_runtime": 8.9948,
667
+ "eval_samples_per_second": 276.827,
668
+ "step": 16500
669
+ },
670
+ {
671
+ "epoch": 0.68,
672
+ "learning_rate": 3.8889952414811574e-05,
673
+ "loss": 0.6071,
674
+ "step": 16750
675
+ },
676
+ {
677
+ "epoch": 0.69,
678
+ "learning_rate": 3.871902860580868e-05,
679
+ "loss": 0.5986,
680
+ "step": 17000
681
+ },
682
+ {
683
+ "epoch": 0.69,
684
+ "eval_accuracy": 0.7718875502008032,
685
+ "eval_loss": 0.5707191228866577,
686
+ "eval_runtime": 9.1769,
687
+ "eval_samples_per_second": 271.333,
688
+ "step": 17000
689
+ },
690
+ {
691
+ "epoch": 0.7,
692
+ "learning_rate": 3.854810479680578e-05,
693
+ "loss": 0.5837,
694
+ "step": 17250
695
+ },
696
+ {
697
+ "epoch": 0.71,
698
+ "learning_rate": 3.837718098780288e-05,
699
+ "loss": 0.5946,
700
+ "step": 17500
701
+ },
702
+ {
703
+ "epoch": 0.71,
704
+ "eval_accuracy": 0.7726907630522089,
705
+ "eval_loss": 0.5725845098495483,
706
+ "eval_runtime": 9.1347,
707
+ "eval_samples_per_second": 272.587,
708
+ "step": 17500
709
+ },
710
+ {
711
+ "epoch": 0.72,
712
+ "learning_rate": 3.8206257178799984e-05,
713
+ "loss": 0.5858,
714
+ "step": 17750
715
+ },
716
+ {
717
+ "epoch": 0.73,
718
+ "learning_rate": 3.8035333369797084e-05,
719
+ "loss": 0.5768,
720
+ "step": 18000
721
+ },
722
+ {
723
+ "epoch": 0.73,
724
+ "eval_accuracy": 0.763855421686747,
725
+ "eval_loss": 0.5683776140213013,
726
+ "eval_runtime": 9.114,
727
+ "eval_samples_per_second": 273.207,
728
+ "step": 18000
729
+ },
730
+ {
731
+ "epoch": 0.74,
732
+ "learning_rate": 3.786440956079418e-05,
733
+ "loss": 0.6339,
734
+ "step": 18250
735
+ },
736
+ {
737
+ "epoch": 0.75,
738
+ "learning_rate": 3.769348575179129e-05,
739
+ "loss": 0.577,
740
+ "step": 18500
741
+ },
742
+ {
743
+ "epoch": 0.75,
744
+ "eval_accuracy": 0.7803212851405622,
745
+ "eval_loss": 0.5440804958343506,
746
+ "eval_runtime": 9.2459,
747
+ "eval_samples_per_second": 269.31,
748
+ "step": 18500
749
+ },
750
+ {
751
+ "epoch": 0.76,
752
+ "learning_rate": 3.752256194278839e-05,
753
+ "loss": 0.6009,
754
+ "step": 18750
755
+ },
756
+ {
757
+ "epoch": 0.77,
758
+ "learning_rate": 3.735163813378549e-05,
759
+ "loss": 0.5862,
760
+ "step": 19000
761
+ },
762
+ {
763
+ "epoch": 0.77,
764
+ "eval_accuracy": 0.7863453815261044,
765
+ "eval_loss": 0.5475257635116577,
766
+ "eval_runtime": 9.1052,
767
+ "eval_samples_per_second": 273.471,
768
+ "step": 19000
769
+ },
770
+ {
771
+ "epoch": 0.78,
772
+ "learning_rate": 3.7180714324782586e-05,
773
+ "loss": 0.591,
774
+ "step": 19250
775
+ },
776
+ {
777
+ "epoch": 0.79,
778
+ "learning_rate": 3.7009790515779685e-05,
779
+ "loss": 0.6009,
780
+ "step": 19500
781
+ },
782
+ {
783
+ "epoch": 0.79,
784
+ "eval_accuracy": 0.7730923694779116,
785
+ "eval_loss": 0.5383695960044861,
786
+ "eval_runtime": 9.4316,
787
+ "eval_samples_per_second": 264.007,
788
+ "step": 19500
789
+ },
790
+ {
791
+ "epoch": 0.8,
792
+ "learning_rate": 3.6838866706776784e-05,
793
+ "loss": 0.5783,
794
+ "step": 19750
795
+ },
796
+ {
797
+ "epoch": 0.81,
798
+ "learning_rate": 3.666794289777389e-05,
799
+ "loss": 0.5803,
800
+ "step": 20000
801
+ },
802
+ {
803
+ "epoch": 0.81,
804
+ "eval_accuracy": 0.7799196787148595,
805
+ "eval_loss": 0.5557690858840942,
806
+ "eval_runtime": 9.0883,
807
+ "eval_samples_per_second": 273.978,
808
+ "step": 20000
809
+ },
810
+ {
811
+ "epoch": 0.83,
812
+ "learning_rate": 3.649701908877099e-05,
813
+ "loss": 0.5826,
814
+ "step": 20250
815
+ },
816
+ {
817
+ "epoch": 0.84,
818
+ "learning_rate": 3.632609527976809e-05,
819
+ "loss": 0.5811,
820
+ "step": 20500
821
+ },
822
+ {
823
+ "epoch": 0.84,
824
+ "eval_accuracy": 0.7650602409638554,
825
+ "eval_loss": 0.5788130760192871,
826
+ "eval_runtime": 9.68,
827
+ "eval_samples_per_second": 257.231,
828
+ "step": 20500
829
+ },
830
+ {
831
+ "epoch": 0.85,
832
+ "learning_rate": 3.6155171470765194e-05,
833
+ "loss": 0.5937,
834
+ "step": 20750
835
+ },
836
+ {
837
+ "epoch": 0.86,
838
+ "learning_rate": 3.598424766176229e-05,
839
+ "loss": 0.597,
840
+ "step": 21000
841
+ },
842
+ {
843
+ "epoch": 0.86,
844
+ "eval_accuracy": 0.791566265060241,
845
+ "eval_loss": 0.528762698173523,
846
+ "eval_runtime": 9.0827,
847
+ "eval_samples_per_second": 274.147,
848
+ "step": 21000
849
+ },
850
+ {
851
+ "epoch": 0.87,
852
+ "learning_rate": 3.581332385275939e-05,
853
+ "loss": 0.5732,
854
+ "step": 21250
855
+ },
856
+ {
857
+ "epoch": 0.88,
858
+ "learning_rate": 3.56424000437565e-05,
859
+ "loss": 0.6004,
860
+ "step": 21500
861
+ },
862
+ {
863
+ "epoch": 0.88,
864
+ "eval_accuracy": 0.7819277108433735,
865
+ "eval_loss": 0.5339077711105347,
866
+ "eval_runtime": 9.6053,
867
+ "eval_samples_per_second": 259.233,
868
+ "step": 21500
869
+ },
870
+ {
871
+ "epoch": 0.89,
872
+ "learning_rate": 3.54714762347536e-05,
873
+ "loss": 0.5548,
874
+ "step": 21750
875
+ },
876
+ {
877
+ "epoch": 0.9,
878
+ "learning_rate": 3.53005524257507e-05,
879
+ "loss": 0.5718,
880
+ "step": 22000
881
+ },
882
+ {
883
+ "epoch": 0.9,
884
+ "eval_accuracy": 0.7759036144578313,
885
+ "eval_loss": 0.5448992848396301,
886
+ "eval_runtime": 8.9793,
887
+ "eval_samples_per_second": 277.305,
888
+ "step": 22000
889
+ },
890
+ {
891
+ "epoch": 0.91,
892
+ "learning_rate": 3.51296286167478e-05,
893
+ "loss": 0.5747,
894
+ "step": 22250
895
+ },
896
+ {
897
+ "epoch": 0.92,
898
+ "learning_rate": 3.49587048077449e-05,
899
+ "loss": 0.5922,
900
+ "step": 22500
901
+ },
902
+ {
903
+ "epoch": 0.92,
904
+ "eval_accuracy": 0.7827309236947791,
905
+ "eval_loss": 0.534243643283844,
906
+ "eval_runtime": 9.0048,
907
+ "eval_samples_per_second": 276.52,
908
+ "step": 22500
909
+ },
910
+ {
911
+ "epoch": 0.93,
912
+ "learning_rate": 3.4787780998742e-05,
913
+ "loss": 0.5665,
914
+ "step": 22750
915
+ },
916
+ {
917
+ "epoch": 0.94,
918
+ "learning_rate": 3.461685718973911e-05,
919
+ "loss": 0.5834,
920
+ "step": 23000
921
+ },
922
+ {
923
+ "epoch": 0.94,
924
+ "eval_accuracy": 0.8012048192771084,
925
+ "eval_loss": 0.517599880695343,
926
+ "eval_runtime": 9.1146,
927
+ "eval_samples_per_second": 273.189,
928
+ "step": 23000
929
+ },
930
+ {
931
+ "epoch": 0.95,
932
+ "learning_rate": 3.4445933380736206e-05,
933
+ "loss": 0.5759,
934
+ "step": 23250
935
+ },
936
+ {
937
+ "epoch": 0.96,
938
+ "learning_rate": 3.4275009571733305e-05,
939
+ "loss": 0.5791,
940
+ "step": 23500
941
+ },
942
+ {
943
+ "epoch": 0.96,
944
+ "eval_accuracy": 0.7879518072289157,
945
+ "eval_loss": 0.5282337665557861,
946
+ "eval_runtime": 9.3392,
947
+ "eval_samples_per_second": 266.619,
948
+ "step": 23500
949
+ },
950
+ {
951
+ "epoch": 0.97,
952
+ "learning_rate": 3.410408576273041e-05,
953
+ "loss": 0.5892,
954
+ "step": 23750
955
+ },
956
+ {
957
+ "epoch": 0.98,
958
+ "learning_rate": 3.393316195372751e-05,
959
+ "loss": 0.5598,
960
+ "step": 24000
961
+ },
962
+ {
963
+ "epoch": 0.98,
964
+ "eval_accuracy": 0.7855421686746988,
965
+ "eval_loss": 0.5417413711547852,
966
+ "eval_runtime": 9.0517,
967
+ "eval_samples_per_second": 275.086,
968
+ "step": 24000
969
+ },
970
+ {
971
+ "epoch": 0.99,
972
+ "learning_rate": 3.376223814472461e-05,
973
+ "loss": 0.5695,
974
+ "step": 24250
975
+ },
976
+ {
977
+ "epoch": 1.0,
978
+ "learning_rate": 3.3591314335721715e-05,
979
+ "loss": 0.5663,
980
+ "step": 24500
981
+ },
982
+ {
983
+ "epoch": 1.0,
984
+ "eval_accuracy": 0.7811244979919679,
985
+ "eval_loss": 0.5498207807540894,
986
+ "eval_runtime": 9.2845,
987
+ "eval_samples_per_second": 268.188,
988
+ "step": 24500
989
+ },
990
+ {
991
+ "epoch": 1.01,
992
+ "learning_rate": 3.342039052671881e-05,
993
+ "loss": 0.4729,
994
+ "step": 24750
995
+ },
996
+ {
997
+ "epoch": 1.02,
998
+ "learning_rate": 3.324946671771591e-05,
999
+ "loss": 0.4511,
1000
+ "step": 25000
1001
+ },
1002
+ {
1003
+ "epoch": 1.02,
1004
+ "eval_accuracy": 0.7779116465863454,
1005
+ "eval_loss": 0.5648518204689026,
1006
+ "eval_runtime": 9.0553,
1007
+ "eval_samples_per_second": 274.977,
1008
+ "step": 25000
1009
+ },
1010
+ {
1011
+ "epoch": 1.03,
1012
+ "learning_rate": 3.307854290871301e-05,
1013
+ "loss": 0.4426,
1014
+ "step": 25250
1015
+ },
1016
+ {
1017
+ "epoch": 1.04,
1018
+ "learning_rate": 3.290761909971011e-05,
1019
+ "loss": 0.4361,
1020
+ "step": 25500
1021
+ },
1022
+ {
1023
+ "epoch": 1.04,
1024
+ "eval_accuracy": 0.7959839357429719,
1025
+ "eval_loss": 0.5434826016426086,
1026
+ "eval_runtime": 9.2435,
1027
+ "eval_samples_per_second": 269.378,
1028
+ "step": 25500
1029
+ },
1030
+ {
1031
+ "epoch": 1.05,
1032
+ "learning_rate": 3.273669529070722e-05,
1033
+ "loss": 0.4643,
1034
+ "step": 25750
1035
+ },
1036
+ {
1037
+ "epoch": 1.06,
1038
+ "learning_rate": 3.2565771481704317e-05,
1039
+ "loss": 0.4686,
1040
+ "step": 26000
1041
+ },
1042
+ {
1043
+ "epoch": 1.06,
1044
+ "eval_accuracy": 0.7947791164658634,
1045
+ "eval_loss": 0.5344994068145752,
1046
+ "eval_runtime": 9.0719,
1047
+ "eval_samples_per_second": 274.473,
1048
+ "step": 26000
1049
+ },
1050
+ {
1051
+ "epoch": 1.07,
1052
+ "learning_rate": 3.2394847672701416e-05,
1053
+ "loss": 0.4505,
1054
+ "step": 26250
1055
+ },
1056
+ {
1057
+ "epoch": 1.08,
1058
+ "learning_rate": 3.222392386369852e-05,
1059
+ "loss": 0.4483,
1060
+ "step": 26500
1061
+ },
1062
+ {
1063
+ "epoch": 1.08,
1064
+ "eval_accuracy": 0.7983935742971887,
1065
+ "eval_loss": 0.5487410426139832,
1066
+ "eval_runtime": 9.2215,
1067
+ "eval_samples_per_second": 270.021,
1068
+ "step": 26500
1069
+ },
1070
+ {
1071
+ "epoch": 1.09,
1072
+ "learning_rate": 3.205300005469562e-05,
1073
+ "loss": 0.4497,
1074
+ "step": 26750
1075
+ },
1076
+ {
1077
+ "epoch": 1.1,
1078
+ "learning_rate": 3.188207624569272e-05,
1079
+ "loss": 0.4608,
1080
+ "step": 27000
1081
+ },
1082
+ {
1083
+ "epoch": 1.1,
1084
+ "eval_accuracy": 0.7879518072289157,
1085
+ "eval_loss": 0.5529894232749939,
1086
+ "eval_runtime": 9.098,
1087
+ "eval_samples_per_second": 273.687,
1088
+ "step": 27000
1089
+ },
1090
+ {
1091
+ "epoch": 1.11,
1092
+ "learning_rate": 3.1711152436689826e-05,
1093
+ "loss": 0.4549,
1094
+ "step": 27250
1095
+ },
1096
+ {
1097
+ "epoch": 1.12,
1098
+ "learning_rate": 3.1540228627686925e-05,
1099
+ "loss": 0.4636,
1100
+ "step": 27500
1101
+ },
1102
+ {
1103
+ "epoch": 1.12,
1104
+ "eval_accuracy": 0.7859437751004016,
1105
+ "eval_loss": 0.5583757162094116,
1106
+ "eval_runtime": 9.1565,
1107
+ "eval_samples_per_second": 271.938,
1108
+ "step": 27500
1109
+ },
1110
+ {
1111
+ "epoch": 1.13,
1112
+ "learning_rate": 3.1369304818684024e-05,
1113
+ "loss": 0.4537,
1114
+ "step": 27750
1115
+ },
1116
+ {
1117
+ "epoch": 1.14,
1118
+ "learning_rate": 3.119838100968113e-05,
1119
+ "loss": 0.4628,
1120
+ "step": 28000
1121
+ },
1122
+ {
1123
+ "epoch": 1.14,
1124
+ "eval_accuracy": 0.795582329317269,
1125
+ "eval_loss": 0.5474898815155029,
1126
+ "eval_runtime": 9.1471,
1127
+ "eval_samples_per_second": 272.217,
1128
+ "step": 28000
1129
+ },
1130
+ {
1131
+ "epoch": 1.15,
1132
+ "learning_rate": 3.102745720067823e-05,
1133
+ "loss": 0.4648,
1134
+ "step": 28250
1135
+ },
1136
+ {
1137
+ "epoch": 1.16,
1138
+ "learning_rate": 3.085653339167533e-05,
1139
+ "loss": 0.476,
1140
+ "step": 28500
1141
+ },
1142
+ {
1143
+ "epoch": 1.16,
1144
+ "eval_accuracy": 0.7911646586345381,
1145
+ "eval_loss": 0.5516586899757385,
1146
+ "eval_runtime": 8.9921,
1147
+ "eval_samples_per_second": 276.91,
1148
+ "step": 28500
1149
+ },
1150
+ {
1151
+ "epoch": 1.17,
1152
+ "learning_rate": 3.0685609582672434e-05,
1153
+ "loss": 0.4794,
1154
+ "step": 28750
1155
+ },
1156
+ {
1157
+ "epoch": 1.18,
1158
+ "learning_rate": 3.0514685773669533e-05,
1159
+ "loss": 0.4537,
1160
+ "step": 29000
1161
+ },
1162
+ {
1163
+ "epoch": 1.18,
1164
+ "eval_accuracy": 0.804417670682731,
1165
+ "eval_loss": 0.5304700136184692,
1166
+ "eval_runtime": 9.0138,
1167
+ "eval_samples_per_second": 276.242,
1168
+ "step": 29000
1169
+ },
1170
+ {
1171
+ "epoch": 1.19,
1172
+ "learning_rate": 3.0343761964666632e-05,
1173
+ "loss": 0.481,
1174
+ "step": 29250
1175
+ },
1176
+ {
1177
+ "epoch": 1.2,
1178
+ "learning_rate": 3.0172838155663735e-05,
1179
+ "loss": 0.4715,
1180
+ "step": 29500
1181
+ },
1182
+ {
1183
+ "epoch": 1.2,
1184
+ "eval_accuracy": 0.7939759036144578,
1185
+ "eval_loss": 0.5461502075195312,
1186
+ "eval_runtime": 9.0986,
1187
+ "eval_samples_per_second": 273.667,
1188
+ "step": 29500
1189
+ },
1190
+ {
1191
+ "epoch": 1.21,
1192
+ "learning_rate": 3.0001914346660837e-05,
1193
+ "loss": 0.4385,
1194
+ "step": 29750
1195
+ },
1196
+ {
1197
+ "epoch": 1.22,
1198
+ "learning_rate": 2.9830990537657933e-05,
1199
+ "loss": 0.4633,
1200
+ "step": 30000
1201
+ },
1202
+ {
1203
+ "epoch": 1.22,
1204
+ "eval_accuracy": 0.793574297188755,
1205
+ "eval_loss": 0.5552789568901062,
1206
+ "eval_runtime": 9.3994,
1207
+ "eval_samples_per_second": 264.911,
1208
+ "step": 30000
1209
+ },
1210
+ {
1211
+ "epoch": 1.23,
1212
+ "learning_rate": 2.9660066728655032e-05,
1213
+ "loss": 0.4748,
1214
+ "step": 30250
1215
+ },
1216
+ {
1217
+ "epoch": 1.24,
1218
+ "learning_rate": 2.9489142919652135e-05,
1219
+ "loss": 0.466,
1220
+ "step": 30500
1221
+ },
1222
+ {
1223
+ "epoch": 1.24,
1224
+ "eval_accuracy": 0.7907630522088354,
1225
+ "eval_loss": 0.5751305222511292,
1226
+ "eval_runtime": 9.0876,
1227
+ "eval_samples_per_second": 274.0,
1228
+ "step": 30500
1229
+ },
1230
+ {
1231
+ "epoch": 1.25,
1232
+ "learning_rate": 2.9318219110649237e-05,
1233
+ "loss": 0.4705,
1234
+ "step": 30750
1235
+ },
1236
+ {
1237
+ "epoch": 1.26,
1238
+ "learning_rate": 2.9147295301646336e-05,
1239
+ "loss": 0.4815,
1240
+ "step": 31000
1241
+ },
1242
+ {
1243
+ "epoch": 1.26,
1244
+ "eval_accuracy": 0.8008032128514057,
1245
+ "eval_loss": 0.5165457725524902,
1246
+ "eval_runtime": 9.6998,
1247
+ "eval_samples_per_second": 256.707,
1248
+ "step": 31000
1249
+ },
1250
+ {
1251
+ "epoch": 1.27,
1252
+ "learning_rate": 2.897637149264344e-05,
1253
+ "loss": 0.4638,
1254
+ "step": 31250
1255
+ },
1256
+ {
1257
+ "epoch": 1.28,
1258
+ "learning_rate": 2.880544768364054e-05,
1259
+ "loss": 0.4544,
1260
+ "step": 31500
1261
+ },
1262
+ {
1263
+ "epoch": 1.28,
1264
+ "eval_accuracy": 0.7847389558232932,
1265
+ "eval_loss": 0.5818995833396912,
1266
+ "eval_runtime": 9.0916,
1267
+ "eval_samples_per_second": 273.879,
1268
+ "step": 31500
1269
+ },
1270
+ {
1271
+ "epoch": 1.29,
1272
+ "learning_rate": 2.863452387463764e-05,
1273
+ "loss": 0.4639,
1274
+ "step": 31750
1275
+ },
1276
+ {
1277
+ "epoch": 1.3,
1278
+ "learning_rate": 2.8463600065634743e-05,
1279
+ "loss": 0.4626,
1280
+ "step": 32000
1281
+ },
1282
+ {
1283
+ "epoch": 1.3,
1284
+ "eval_accuracy": 0.8012048192771084,
1285
+ "eval_loss": 0.5314044952392578,
1286
+ "eval_runtime": 9.5865,
1287
+ "eval_samples_per_second": 259.741,
1288
+ "step": 32000
1289
+ },
1290
+ {
1291
+ "epoch": 1.31,
1292
+ "learning_rate": 2.8292676256631846e-05,
1293
+ "loss": 0.4502,
1294
+ "step": 32250
1295
+ },
1296
+ {
1297
+ "epoch": 1.32,
1298
+ "learning_rate": 2.8121752447628945e-05,
1299
+ "loss": 0.4742,
1300
+ "step": 32500
1301
+ },
1302
+ {
1303
+ "epoch": 1.32,
1304
+ "eval_accuracy": 0.8036144578313253,
1305
+ "eval_loss": 0.5086196064949036,
1306
+ "eval_runtime": 9.0534,
1307
+ "eval_samples_per_second": 275.036,
1308
+ "step": 32500
1309
+ },
1310
+ {
1311
+ "epoch": 1.33,
1312
+ "learning_rate": 2.7950828638626047e-05,
1313
+ "loss": 0.4555,
1314
+ "step": 32750
1315
+ },
1316
+ {
1317
+ "epoch": 1.34,
1318
+ "learning_rate": 2.777990482962315e-05,
1319
+ "loss": 0.4643,
1320
+ "step": 33000
1321
+ },
1322
+ {
1323
+ "epoch": 1.34,
1324
+ "eval_accuracy": 0.8012048192771084,
1325
+ "eval_loss": 0.5454714894294739,
1326
+ "eval_runtime": 9.453,
1327
+ "eval_samples_per_second": 263.409,
1328
+ "step": 33000
1329
+ },
1330
+ {
1331
+ "epoch": 1.35,
1332
+ "learning_rate": 2.760898102062025e-05,
1333
+ "loss": 0.4597,
1334
+ "step": 33250
1335
+ },
1336
+ {
1337
+ "epoch": 1.36,
1338
+ "learning_rate": 2.743805721161735e-05,
1339
+ "loss": 0.4727,
1340
+ "step": 33500
1341
+ },
1342
+ {
1343
+ "epoch": 1.36,
1344
+ "eval_accuracy": 0.7943775100401607,
1345
+ "eval_loss": 0.5575410723686218,
1346
+ "eval_runtime": 8.9953,
1347
+ "eval_samples_per_second": 276.812,
1348
+ "step": 33500
1349
+ },
1350
+ {
1351
+ "epoch": 1.38,
1352
+ "learning_rate": 2.7267133402614454e-05,
1353
+ "loss": 0.4612,
1354
+ "step": 33750
1355
+ },
1356
+ {
1357
+ "epoch": 1.39,
1358
+ "learning_rate": 2.7096209593611553e-05,
1359
+ "loss": 0.4646,
1360
+ "step": 34000
1361
+ },
1362
+ {
1363
+ "epoch": 1.39,
1364
+ "eval_accuracy": 0.7987951807228916,
1365
+ "eval_loss": 0.5220562219619751,
1366
+ "eval_runtime": 9.4179,
1367
+ "eval_samples_per_second": 264.391,
1368
+ "step": 34000
1369
+ },
1370
+ {
1371
+ "epoch": 1.4,
1372
+ "learning_rate": 2.6925285784608655e-05,
1373
+ "loss": 0.46,
1374
+ "step": 34250
1375
+ },
1376
+ {
1377
+ "epoch": 1.41,
1378
+ "learning_rate": 2.6754361975605758e-05,
1379
+ "loss": 0.4573,
1380
+ "step": 34500
1381
+ },
1382
+ {
1383
+ "epoch": 1.41,
1384
+ "eval_accuracy": 0.7939759036144578,
1385
+ "eval_loss": 0.5481248497962952,
1386
+ "eval_runtime": 9.0056,
1387
+ "eval_samples_per_second": 276.493,
1388
+ "step": 34500
1389
+ },
1390
+ {
1391
+ "epoch": 1.42,
1392
+ "learning_rate": 2.6583438166602857e-05,
1393
+ "loss": 0.4633,
1394
+ "step": 34750
1395
+ },
1396
+ {
1397
+ "epoch": 1.43,
1398
+ "learning_rate": 2.641251435759996e-05,
1399
+ "loss": 0.4532,
1400
+ "step": 35000
1401
+ },
1402
+ {
1403
+ "epoch": 1.43,
1404
+ "eval_accuracy": 0.793574297188755,
1405
+ "eval_loss": 0.5569635629653931,
1406
+ "eval_runtime": 9.3588,
1407
+ "eval_samples_per_second": 266.061,
1408
+ "step": 35000
1409
+ },
1410
+ {
1411
+ "epoch": 1.44,
1412
+ "learning_rate": 2.6241590548597055e-05,
1413
+ "loss": 0.4431,
1414
+ "step": 35250
1415
+ },
1416
+ {
1417
+ "epoch": 1.45,
1418
+ "learning_rate": 2.6070666739594158e-05,
1419
+ "loss": 0.4637,
1420
+ "step": 35500
1421
+ },
1422
+ {
1423
+ "epoch": 1.45,
1424
+ "eval_accuracy": 0.7971887550200804,
1425
+ "eval_loss": 0.5141083002090454,
1426
+ "eval_runtime": 9.013,
1427
+ "eval_samples_per_second": 276.266,
1428
+ "step": 35500
1429
+ },
1430
+ {
1431
+ "epoch": 1.46,
1432
+ "learning_rate": 2.5899742930591257e-05,
1433
+ "loss": 0.4337,
1434
+ "step": 35750
1435
+ },
1436
+ {
1437
+ "epoch": 1.47,
1438
+ "learning_rate": 2.572881912158836e-05,
1439
+ "loss": 0.4572,
1440
+ "step": 36000
1441
+ },
1442
+ {
1443
+ "epoch": 1.47,
1444
+ "eval_accuracy": 0.7887550200803213,
1445
+ "eval_loss": 0.5640882253646851,
1446
+ "eval_runtime": 9.316,
1447
+ "eval_samples_per_second": 267.283,
1448
+ "step": 36000
1449
+ },
1450
+ {
1451
+ "epoch": 1.48,
1452
+ "learning_rate": 2.5557895312585462e-05,
1453
+ "loss": 0.4485,
1454
+ "step": 36250
1455
+ },
1456
+ {
1457
+ "epoch": 1.49,
1458
+ "learning_rate": 2.538697150358256e-05,
1459
+ "loss": 0.4613,
1460
+ "step": 36500
1461
+ },
1462
+ {
1463
+ "epoch": 1.49,
1464
+ "eval_accuracy": 0.8004016064257028,
1465
+ "eval_loss": 0.5442628860473633,
1466
+ "eval_runtime": 9.0407,
1467
+ "eval_samples_per_second": 275.422,
1468
+ "step": 36500
1469
+ },
1470
+ {
1471
+ "epoch": 1.5,
1472
+ "learning_rate": 2.5216047694579664e-05,
1473
+ "loss": 0.4547,
1474
+ "step": 36750
1475
+ },
1476
+ {
1477
+ "epoch": 1.51,
1478
+ "learning_rate": 2.5045123885576766e-05,
1479
+ "loss": 0.4614,
1480
+ "step": 37000
1481
+ },
1482
+ {
1483
+ "epoch": 1.51,
1484
+ "eval_accuracy": 0.7987951807228916,
1485
+ "eval_loss": 0.5075089931488037,
1486
+ "eval_runtime": 9.22,
1487
+ "eval_samples_per_second": 270.065,
1488
+ "step": 37000
1489
+ },
1490
+ {
1491
+ "epoch": 1.52,
1492
+ "learning_rate": 2.4874200076573865e-05,
1493
+ "loss": 0.4545,
1494
+ "step": 37250
1495
+ },
1496
+ {
1497
+ "epoch": 1.53,
1498
+ "learning_rate": 2.4703276267570968e-05,
1499
+ "loss": 0.4724,
1500
+ "step": 37500
1501
+ },
1502
+ {
1503
+ "epoch": 1.53,
1504
+ "eval_accuracy": 0.8092369477911646,
1505
+ "eval_loss": 0.5035138726234436,
1506
+ "eval_runtime": 9.0797,
1507
+ "eval_samples_per_second": 274.237,
1508
+ "step": 37500
1509
+ },
1510
+ {
1511
+ "epoch": 1.54,
1512
+ "learning_rate": 2.453235245856807e-05,
1513
+ "loss": 0.4451,
1514
+ "step": 37750
1515
+ },
1516
+ {
1517
+ "epoch": 1.55,
1518
+ "learning_rate": 2.436142864956517e-05,
1519
+ "loss": 0.4396,
1520
+ "step": 38000
1521
+ },
1522
+ {
1523
+ "epoch": 1.55,
1524
+ "eval_accuracy": 0.7971887550200804,
1525
+ "eval_loss": 0.5381002426147461,
1526
+ "eval_runtime": 9.1884,
1527
+ "eval_samples_per_second": 270.995,
1528
+ "step": 38000
1529
+ },
1530
+ {
1531
+ "epoch": 1.56,
1532
+ "learning_rate": 2.4190504840562272e-05,
1533
+ "loss": 0.459,
1534
+ "step": 38250
1535
+ },
1536
+ {
1537
+ "epoch": 1.57,
1538
+ "learning_rate": 2.4019581031559374e-05,
1539
+ "loss": 0.4561,
1540
+ "step": 38500
1541
+ },
1542
+ {
1543
+ "epoch": 1.57,
1544
+ "eval_accuracy": 0.8100401606425702,
1545
+ "eval_loss": 0.4913768470287323,
1546
+ "eval_runtime": 9.0787,
1547
+ "eval_samples_per_second": 274.268,
1548
+ "step": 38500
1549
+ },
1550
+ {
1551
+ "epoch": 1.58,
1552
+ "learning_rate": 2.3848657222556474e-05,
1553
+ "loss": 0.4503,
1554
+ "step": 38750
1555
+ },
1556
+ {
1557
+ "epoch": 1.59,
1558
+ "learning_rate": 2.3677733413553576e-05,
1559
+ "loss": 0.4643,
1560
+ "step": 39000
1561
+ },
1562
+ {
1563
+ "epoch": 1.59,
1564
+ "eval_accuracy": 0.8092369477911646,
1565
+ "eval_loss": 0.5171190500259399,
1566
+ "eval_runtime": 8.9997,
1567
+ "eval_samples_per_second": 276.677,
1568
+ "step": 39000
1569
+ },
1570
+ {
1571
+ "epoch": 1.6,
1572
+ "learning_rate": 2.3506809604550675e-05,
1573
+ "loss": 0.4624,
1574
+ "step": 39250
1575
+ },
1576
+ {
1577
+ "epoch": 1.61,
1578
+ "learning_rate": 2.3335885795547778e-05,
1579
+ "loss": 0.443,
1580
+ "step": 39500
1581
+ },
1582
+ {
1583
+ "epoch": 1.61,
1584
+ "eval_accuracy": 0.7951807228915663,
1585
+ "eval_loss": 0.5365191102027893,
1586
+ "eval_runtime": 9.0082,
1587
+ "eval_samples_per_second": 276.415,
1588
+ "step": 39500
1589
+ },
1590
+ {
1591
+ "epoch": 1.62,
1592
+ "learning_rate": 2.3164961986544877e-05,
1593
+ "loss": 0.4467,
1594
+ "step": 39750
1595
+ },
1596
+ {
1597
+ "epoch": 1.63,
1598
+ "learning_rate": 2.299403817754198e-05,
1599
+ "loss": 0.4525,
1600
+ "step": 40000
1601
+ },
1602
+ {
1603
+ "epoch": 1.63,
1604
+ "eval_accuracy": 0.8036144578313253,
1605
+ "eval_loss": 0.5161953568458557,
1606
+ "eval_runtime": 9.0761,
1607
+ "eval_samples_per_second": 274.346,
1608
+ "step": 40000
1609
+ },
1610
+ {
1611
+ "epoch": 1.64,
1612
+ "learning_rate": 2.2823114368539082e-05,
1613
+ "loss": 0.4591,
1614
+ "step": 40250
1615
+ },
1616
+ {
1617
+ "epoch": 1.65,
1618
+ "learning_rate": 2.265219055953618e-05,
1619
+ "loss": 0.4618,
1620
+ "step": 40500
1621
+ },
1622
+ {
1623
+ "epoch": 1.65,
1624
+ "eval_accuracy": 0.8080321285140563,
1625
+ "eval_loss": 0.49769964814186096,
1626
+ "eval_runtime": 9.2939,
1627
+ "eval_samples_per_second": 267.916,
1628
+ "step": 40500
1629
+ },
1630
+ {
1631
+ "epoch": 1.66,
1632
+ "learning_rate": 2.2481266750533284e-05,
1633
+ "loss": 0.4434,
1634
+ "step": 40750
1635
+ },
1636
+ {
1637
+ "epoch": 1.67,
1638
+ "learning_rate": 2.2310342941530386e-05,
1639
+ "loss": 0.4561,
1640
+ "step": 41000
1641
+ },
1642
+ {
1643
+ "epoch": 1.67,
1644
+ "eval_accuracy": 0.8092369477911646,
1645
+ "eval_loss": 0.4977361559867859,
1646
+ "eval_runtime": 8.9616,
1647
+ "eval_samples_per_second": 277.852,
1648
+ "step": 41000
1649
+ },
1650
+ {
1651
+ "epoch": 1.68,
1652
+ "learning_rate": 2.2139419132527485e-05,
1653
+ "loss": 0.4483,
1654
+ "step": 41250
1655
+ },
1656
+ {
1657
+ "epoch": 1.69,
1658
+ "learning_rate": 2.1968495323524588e-05,
1659
+ "loss": 0.4512,
1660
+ "step": 41500
1661
+ },
1662
+ {
1663
+ "epoch": 1.69,
1664
+ "eval_accuracy": 0.802008032128514,
1665
+ "eval_loss": 0.4957820475101471,
1666
+ "eval_runtime": 9.0164,
1667
+ "eval_samples_per_second": 276.164,
1668
+ "step": 41500
1669
+ },
1670
+ {
1671
+ "epoch": 1.7,
1672
+ "learning_rate": 2.179757151452169e-05,
1673
+ "loss": 0.4473,
1674
+ "step": 41750
1675
+ },
1676
+ {
1677
+ "epoch": 1.71,
1678
+ "learning_rate": 2.162664770551879e-05,
1679
+ "loss": 0.4564,
1680
+ "step": 42000
1681
+ },
1682
+ {
1683
+ "epoch": 1.71,
1684
+ "eval_accuracy": 0.7987951807228916,
1685
+ "eval_loss": 0.5506803393363953,
1686
+ "eval_runtime": 9.1173,
1687
+ "eval_samples_per_second": 273.106,
1688
+ "step": 42000
1689
+ },
1690
+ {
1691
+ "epoch": 1.72,
1692
+ "learning_rate": 2.145572389651589e-05,
1693
+ "loss": 0.4491,
1694
+ "step": 42250
1695
+ },
1696
+ {
1697
+ "epoch": 1.73,
1698
+ "learning_rate": 2.128480008751299e-05,
1699
+ "loss": 0.4331,
1700
+ "step": 42500
1701
+ },
1702
+ {
1703
+ "epoch": 1.73,
1704
+ "eval_accuracy": 0.8140562248995984,
1705
+ "eval_loss": 0.5114361047744751,
1706
+ "eval_runtime": 9.554,
1707
+ "eval_samples_per_second": 260.625,
1708
+ "step": 42500
1709
+ },
1710
+ {
1711
+ "epoch": 1.74,
1712
+ "learning_rate": 2.1113876278510094e-05,
1713
+ "loss": 0.4631,
1714
+ "step": 42750
1715
+ },
1716
+ {
1717
+ "epoch": 1.75,
1718
+ "learning_rate": 2.0942952469507193e-05,
1719
+ "loss": 0.4598,
1720
+ "step": 43000
1721
+ },
1722
+ {
1723
+ "epoch": 1.75,
1724
+ "eval_accuracy": 0.8032128514056225,
1725
+ "eval_loss": 0.5167751312255859,
1726
+ "eval_runtime": 9.065,
1727
+ "eval_samples_per_second": 274.682,
1728
+ "step": 43000
1729
+ },
1730
+ {
1731
+ "epoch": 1.76,
1732
+ "learning_rate": 2.0772028660504295e-05,
1733
+ "loss": 0.433,
1734
+ "step": 43250
1735
+ },
1736
+ {
1737
+ "epoch": 1.77,
1738
+ "learning_rate": 2.0601104851501398e-05,
1739
+ "loss": 0.4711,
1740
+ "step": 43500
1741
+ },
1742
+ {
1743
+ "epoch": 1.77,
1744
+ "eval_accuracy": 0.8100401606425702,
1745
+ "eval_loss": 0.49471431970596313,
1746
+ "eval_runtime": 9.4303,
1747
+ "eval_samples_per_second": 264.043,
1748
+ "step": 43500
1749
+ },
1750
+ {
1751
+ "epoch": 1.78,
1752
+ "learning_rate": 2.0430181042498497e-05,
1753
+ "loss": 0.4638,
1754
+ "step": 43750
1755
+ },
1756
+ {
1757
+ "epoch": 1.79,
1758
+ "learning_rate": 2.02592572334956e-05,
1759
+ "loss": 0.4133,
1760
+ "step": 44000
1761
+ },
1762
+ {
1763
+ "epoch": 1.79,
1764
+ "eval_accuracy": 0.8144578313253013,
1765
+ "eval_loss": 0.5469810962677002,
1766
+ "eval_runtime": 9.0719,
1767
+ "eval_samples_per_second": 274.474,
1768
+ "step": 44000
1769
+ },
1770
+ {
1771
+ "epoch": 1.8,
1772
+ "learning_rate": 2.0088333424492702e-05,
1773
+ "loss": 0.4558,
1774
+ "step": 44250
1775
+ },
1776
+ {
1777
+ "epoch": 1.81,
1778
+ "learning_rate": 1.99174096154898e-05,
1779
+ "loss": 0.4442,
1780
+ "step": 44500
1781
+ },
1782
+ {
1783
+ "epoch": 1.81,
1784
+ "eval_accuracy": 0.8024096385542169,
1785
+ "eval_loss": 0.5169267654418945,
1786
+ "eval_runtime": 9.3219,
1787
+ "eval_samples_per_second": 267.114,
1788
+ "step": 44500
1789
+ },
1790
+ {
1791
+ "epoch": 1.82,
1792
+ "learning_rate": 1.97464858064869e-05,
1793
+ "loss": 0.4412,
1794
+ "step": 44750
1795
+ },
1796
+ {
1797
+ "epoch": 1.83,
1798
+ "learning_rate": 1.9575561997484003e-05,
1799
+ "loss": 0.4425,
1800
+ "step": 45000
1801
+ },
1802
+ {
1803
+ "epoch": 1.83,
1804
+ "eval_accuracy": 0.8076305220883534,
1805
+ "eval_loss": 0.5110898613929749,
1806
+ "eval_runtime": 9.0459,
1807
+ "eval_samples_per_second": 275.262,
1808
+ "step": 45000
1809
+ },
1810
+ {
1811
+ "epoch": 1.84,
1812
+ "learning_rate": 1.9404638188481102e-05,
1813
+ "loss": 0.4443,
1814
+ "step": 45250
1815
+ },
1816
+ {
1817
+ "epoch": 1.85,
1818
+ "learning_rate": 1.9233714379478204e-05,
1819
+ "loss": 0.4522,
1820
+ "step": 45500
1821
+ },
1822
+ {
1823
+ "epoch": 1.85,
1824
+ "eval_accuracy": 0.8032128514056225,
1825
+ "eval_loss": 0.527312159538269,
1826
+ "eval_runtime": 8.9896,
1827
+ "eval_samples_per_second": 276.986,
1828
+ "step": 45500
1829
+ },
1830
+ {
1831
+ "epoch": 1.86,
1832
+ "learning_rate": 1.9062790570475307e-05,
1833
+ "loss": 0.45,
1834
+ "step": 45750
1835
+ },
1836
+ {
1837
+ "epoch": 1.87,
1838
+ "learning_rate": 1.8891866761472406e-05,
1839
+ "loss": 0.4451,
1840
+ "step": 46000
1841
+ },
1842
+ {
1843
+ "epoch": 1.87,
1844
+ "eval_accuracy": 0.8132530120481928,
1845
+ "eval_loss": 0.4912641644477844,
1846
+ "eval_runtime": 9.0608,
1847
+ "eval_samples_per_second": 274.811,
1848
+ "step": 46000
1849
+ },
1850
+ {
1851
+ "epoch": 1.88,
1852
+ "learning_rate": 1.872094295246951e-05,
1853
+ "loss": 0.4321,
1854
+ "step": 46250
1855
+ },
1856
+ {
1857
+ "epoch": 1.89,
1858
+ "learning_rate": 1.855001914346661e-05,
1859
+ "loss": 0.463,
1860
+ "step": 46500
1861
+ },
1862
+ {
1863
+ "epoch": 1.89,
1864
+ "eval_accuracy": 0.8040160642570281,
1865
+ "eval_loss": 0.5067523717880249,
1866
+ "eval_runtime": 8.9793,
1867
+ "eval_samples_per_second": 277.305,
1868
+ "step": 46500
1869
+ },
1870
+ {
1871
+ "epoch": 1.9,
1872
+ "learning_rate": 1.837909533446371e-05,
1873
+ "loss": 0.44,
1874
+ "step": 46750
1875
+ },
1876
+ {
1877
+ "epoch": 1.91,
1878
+ "learning_rate": 1.8208171525460813e-05,
1879
+ "loss": 0.4173,
1880
+ "step": 47000
1881
+ },
1882
+ {
1883
+ "epoch": 1.91,
1884
+ "eval_accuracy": 0.8008032128514057,
1885
+ "eval_loss": 0.5309551954269409,
1886
+ "eval_runtime": 9.0236,
1887
+ "eval_samples_per_second": 275.944,
1888
+ "step": 47000
1889
+ },
1890
+ {
1891
+ "epoch": 1.93,
1892
+ "learning_rate": 1.803724771645791e-05,
1893
+ "loss": 0.4377,
1894
+ "step": 47250
1895
+ },
1896
+ {
1897
+ "epoch": 1.94,
1898
+ "learning_rate": 1.7866323907455014e-05,
1899
+ "loss": 0.4336,
1900
+ "step": 47500
1901
+ },
1902
+ {
1903
+ "epoch": 1.94,
1904
+ "eval_accuracy": 0.8036144578313253,
1905
+ "eval_loss": 0.5289146900177002,
1906
+ "eval_runtime": 9.1489,
1907
+ "eval_samples_per_second": 272.164,
1908
+ "step": 47500
1909
+ },
1910
+ {
1911
+ "epoch": 1.95,
1912
+ "learning_rate": 1.7695400098452113e-05,
1913
+ "loss": 0.4424,
1914
+ "step": 47750
1915
+ },
1916
+ {
1917
+ "epoch": 1.96,
1918
+ "learning_rate": 1.7524476289449216e-05,
1919
+ "loss": 0.4266,
1920
+ "step": 48000
1921
+ },
1922
+ {
1923
+ "epoch": 1.96,
1924
+ "eval_accuracy": 0.8048192771084337,
1925
+ "eval_loss": 0.516535758972168,
1926
+ "eval_runtime": 9.1038,
1927
+ "eval_samples_per_second": 273.513,
1928
+ "step": 48000
1929
+ },
1930
+ {
1931
+ "epoch": 1.97,
1932
+ "learning_rate": 1.735355248044632e-05,
1933
+ "loss": 0.4276,
1934
+ "step": 48250
1935
+ },
1936
+ {
1937
+ "epoch": 1.98,
1938
+ "learning_rate": 1.7182628671443417e-05,
1939
+ "loss": 0.4336,
1940
+ "step": 48500
1941
+ },
1942
+ {
1943
+ "epoch": 1.98,
1944
+ "eval_accuracy": 0.8104417670682731,
1945
+ "eval_loss": 0.5314404368400574,
1946
+ "eval_runtime": 9.1174,
1947
+ "eval_samples_per_second": 273.103,
1948
+ "step": 48500
1949
+ },
1950
+ {
1951
+ "epoch": 1.99,
1952
+ "learning_rate": 1.701170486244052e-05,
1953
+ "loss": 0.4286,
1954
+ "step": 48750
1955
+ },
1956
+ {
1957
+ "epoch": 2.0,
1958
+ "learning_rate": 1.6840781053437622e-05,
1959
+ "loss": 0.4342,
1960
+ "step": 49000
1961
+ },
1962
+ {
1963
+ "epoch": 2.0,
1964
+ "eval_accuracy": 0.814859437751004,
1965
+ "eval_loss": 0.4977148473262787,
1966
+ "eval_runtime": 9.12,
1967
+ "eval_samples_per_second": 273.026,
1968
+ "step": 49000
1969
+ },
1970
+ {
1971
+ "epoch": 2.01,
1972
+ "learning_rate": 1.666985724443472e-05,
1973
+ "loss": 0.3449,
1974
+ "step": 49250
1975
+ },
1976
+ {
1977
+ "epoch": 2.02,
1978
+ "learning_rate": 1.6498933435431824e-05,
1979
+ "loss": 0.3156,
1980
+ "step": 49500
1981
+ },
1982
+ {
1983
+ "epoch": 2.02,
1984
+ "eval_accuracy": 0.8128514056224899,
1985
+ "eval_loss": 0.5999274849891663,
1986
+ "eval_runtime": 9.1186,
1987
+ "eval_samples_per_second": 273.069,
1988
+ "step": 49500
1989
+ },
1990
+ {
1991
+ "epoch": 2.03,
1992
+ "learning_rate": 1.6328009626428927e-05,
1993
+ "loss": 0.2989,
1994
+ "step": 49750
1995
+ },
1996
+ {
1997
+ "epoch": 2.04,
1998
+ "learning_rate": 1.6157085817426022e-05,
1999
+ "loss": 0.3013,
2000
+ "step": 50000
2001
+ },
2002
+ {
2003
+ "epoch": 2.04,
2004
+ "eval_accuracy": 0.8028112449799196,
2005
+ "eval_loss": 0.6256367564201355,
2006
+ "eval_runtime": 9.2774,
2007
+ "eval_samples_per_second": 268.395,
2008
+ "step": 50000
2009
+ },
2010
+ {
2011
+ "epoch": 2.05,
2012
+ "learning_rate": 1.5986162008423125e-05,
2013
+ "loss": 0.2947,
2014
+ "step": 50250
2015
+ },
2016
+ {
2017
+ "epoch": 2.06,
2018
+ "learning_rate": 1.5815238199420227e-05,
2019
+ "loss": 0.2849,
2020
+ "step": 50500
2021
+ },
2022
+ {
2023
+ "epoch": 2.06,
2024
+ "eval_accuracy": 0.8116465863453816,
2025
+ "eval_loss": 0.6139395236968994,
2026
+ "eval_runtime": 9.009,
2027
+ "eval_samples_per_second": 276.389,
2028
+ "step": 50500
2029
+ },
2030
+ {
2031
+ "epoch": 2.07,
2032
+ "learning_rate": 1.5644314390417327e-05,
2033
+ "loss": 0.2929,
2034
+ "step": 50750
2035
+ },
2036
+ {
2037
+ "epoch": 2.08,
2038
+ "learning_rate": 1.547339058141443e-05,
2039
+ "loss": 0.2927,
2040
+ "step": 51000
2041
+ },
2042
+ {
2043
+ "epoch": 2.08,
2044
+ "eval_accuracy": 0.808433734939759,
2045
+ "eval_loss": 0.5880476236343384,
2046
+ "eval_runtime": 8.9456,
2047
+ "eval_samples_per_second": 278.35,
2048
+ "step": 51000
2049
+ },
2050
+ {
2051
+ "epoch": 2.09,
2052
+ "learning_rate": 1.530246677241153e-05,
2053
+ "loss": 0.2725,
2054
+ "step": 51250
2055
+ },
2056
+ {
2057
+ "epoch": 2.1,
2058
+ "learning_rate": 1.5131542963408632e-05,
2059
+ "loss": 0.2944,
2060
+ "step": 51500
2061
+ },
2062
+ {
2063
+ "epoch": 2.1,
2064
+ "eval_accuracy": 0.8064257028112449,
2065
+ "eval_loss": 0.6175798177719116,
2066
+ "eval_runtime": 9.0878,
2067
+ "eval_samples_per_second": 273.995,
2068
+ "step": 51500
2069
+ },
2070
+ {
2071
+ "epoch": 2.11,
2072
+ "learning_rate": 1.4960619154405733e-05,
2073
+ "loss": 0.2812,
2074
+ "step": 51750
2075
+ },
2076
+ {
2077
+ "epoch": 2.12,
2078
+ "learning_rate": 1.4789695345402834e-05,
2079
+ "loss": 0.2869,
2080
+ "step": 52000
2081
+ },
2082
+ {
2083
+ "epoch": 2.12,
2084
+ "eval_accuracy": 0.8100401606425702,
2085
+ "eval_loss": 0.5959600806236267,
2086
+ "eval_runtime": 9.5023,
2087
+ "eval_samples_per_second": 262.042,
2088
+ "step": 52000
2089
+ },
2090
+ {
2091
+ "epoch": 2.13,
2092
+ "learning_rate": 1.4618771536399937e-05,
2093
+ "loss": 0.2726,
2094
+ "step": 52250
2095
+ },
2096
+ {
2097
+ "epoch": 2.14,
2098
+ "learning_rate": 1.4447847727397037e-05,
2099
+ "loss": 0.3034,
2100
+ "step": 52500
2101
+ },
2102
+ {
2103
+ "epoch": 2.14,
2104
+ "eval_accuracy": 0.7907630522088354,
2105
+ "eval_loss": 0.6271839141845703,
2106
+ "eval_runtime": 9.0834,
2107
+ "eval_samples_per_second": 274.126,
2108
+ "step": 52500
2109
+ },
2110
+ {
2111
+ "epoch": 2.15,
2112
+ "learning_rate": 1.4276923918394136e-05,
2113
+ "loss": 0.3019,
2114
+ "step": 52750
2115
+ },
2116
+ {
2117
+ "epoch": 2.16,
2118
+ "learning_rate": 1.4106000109391237e-05,
2119
+ "loss": 0.279,
2120
+ "step": 53000
2121
+ },
2122
+ {
2123
+ "epoch": 2.16,
2124
+ "eval_accuracy": 0.8096385542168675,
2125
+ "eval_loss": 0.6030941009521484,
2126
+ "eval_runtime": 9.4494,
2127
+ "eval_samples_per_second": 263.51,
2128
+ "step": 53000
2129
+ },
2130
+ {
2131
+ "epoch": 2.17,
2132
+ "learning_rate": 1.3935076300388338e-05,
2133
+ "loss": 0.3079,
2134
+ "step": 53250
2135
+ },
2136
+ {
2137
+ "epoch": 2.18,
2138
+ "learning_rate": 1.376415249138544e-05,
2139
+ "loss": 0.2896,
2140
+ "step": 53500
2141
+ },
2142
+ {
2143
+ "epoch": 2.18,
2144
+ "eval_accuracy": 0.8068273092369478,
2145
+ "eval_loss": 0.6132158637046814,
2146
+ "eval_runtime": 9.0141,
2147
+ "eval_samples_per_second": 276.233,
2148
+ "step": 53500
2149
+ },
2150
+ {
2151
+ "epoch": 2.19,
2152
+ "learning_rate": 1.3593228682382541e-05,
2153
+ "loss": 0.2823,
2154
+ "step": 53750
2155
+ },
2156
+ {
2157
+ "epoch": 2.2,
2158
+ "learning_rate": 1.3422304873379642e-05,
2159
+ "loss": 0.2952,
2160
+ "step": 54000
2161
+ },
2162
+ {
2163
+ "epoch": 2.2,
2164
+ "eval_accuracy": 0.8064257028112449,
2165
+ "eval_loss": 0.6195886135101318,
2166
+ "eval_runtime": 8.9618,
2167
+ "eval_samples_per_second": 277.847,
2168
+ "step": 54000
2169
+ },
2170
+ {
2171
+ "epoch": 2.21,
2172
+ "learning_rate": 1.3251381064376745e-05,
2173
+ "loss": 0.3049,
2174
+ "step": 54250
2175
+ },
2176
+ {
2177
+ "epoch": 2.22,
2178
+ "learning_rate": 1.3080457255373846e-05,
2179
+ "loss": 0.2921,
2180
+ "step": 54500
2181
+ },
2182
+ {
2183
+ "epoch": 2.22,
2184
+ "eval_accuracy": 0.8076305220883534,
2185
+ "eval_loss": 0.6113378405570984,
2186
+ "eval_runtime": 9.0032,
2187
+ "eval_samples_per_second": 276.569,
2188
+ "step": 54500
2189
+ },
2190
+ {
2191
+ "epoch": 2.23,
2192
+ "learning_rate": 1.2909533446370946e-05,
2193
+ "loss": 0.2765,
2194
+ "step": 54750
2195
+ },
2196
+ {
2197
+ "epoch": 2.24,
2198
+ "learning_rate": 1.2738609637368049e-05,
2199
+ "loss": 0.2958,
2200
+ "step": 55000
2201
+ },
2202
+ {
2203
+ "epoch": 2.24,
2204
+ "eval_accuracy": 0.8064257028112449,
2205
+ "eval_loss": 0.6207754611968994,
2206
+ "eval_runtime": 9.0338,
2207
+ "eval_samples_per_second": 275.632,
2208
+ "step": 55000
2209
+ },
2210
+ {
2211
+ "epoch": 2.25,
2212
+ "learning_rate": 1.2567685828365146e-05,
2213
+ "loss": 0.2888,
2214
+ "step": 55250
2215
+ },
2216
+ {
2217
+ "epoch": 2.26,
2218
+ "learning_rate": 1.239676201936225e-05,
2219
+ "loss": 0.2996,
2220
+ "step": 55500
2221
+ },
2222
+ {
2223
+ "epoch": 2.26,
2224
+ "eval_accuracy": 0.8128514056224899,
2225
+ "eval_loss": 0.5894312262535095,
2226
+ "eval_runtime": 8.9904,
2227
+ "eval_samples_per_second": 276.961,
2228
+ "step": 55500
2229
+ },
2230
+ {
2231
+ "epoch": 2.27,
2232
+ "learning_rate": 1.2225838210359351e-05,
2233
+ "loss": 0.2994,
2234
+ "step": 55750
2235
+ },
2236
+ {
2237
+ "epoch": 2.28,
2238
+ "learning_rate": 1.205491440135645e-05,
2239
+ "loss": 0.288,
2240
+ "step": 56000
2241
+ },
2242
+ {
2243
+ "epoch": 2.28,
2244
+ "eval_accuracy": 0.8052208835341366,
2245
+ "eval_loss": 0.6171417832374573,
2246
+ "eval_runtime": 9.2622,
2247
+ "eval_samples_per_second": 268.834,
2248
+ "step": 56000
2249
+ },
2250
+ {
2251
+ "epoch": 2.29,
2252
+ "learning_rate": 1.1883990592353553e-05,
2253
+ "loss": 0.2902,
2254
+ "step": 56250
2255
+ },
2256
+ {
2257
+ "epoch": 2.3,
2258
+ "learning_rate": 1.1713066783350654e-05,
2259
+ "loss": 0.3005,
2260
+ "step": 56500
2261
+ },
2262
+ {
2263
+ "epoch": 2.3,
2264
+ "eval_accuracy": 0.8112449799196787,
2265
+ "eval_loss": 0.5888203978538513,
2266
+ "eval_runtime": 8.9926,
2267
+ "eval_samples_per_second": 276.894,
2268
+ "step": 56500
2269
+ },
2270
+ {
2271
+ "epoch": 2.31,
2272
+ "learning_rate": 1.1542142974347755e-05,
2273
+ "loss": 0.3011,
2274
+ "step": 56750
2275
+ },
2276
+ {
2277
+ "epoch": 2.32,
2278
+ "learning_rate": 1.1371219165344857e-05,
2279
+ "loss": 0.3082,
2280
+ "step": 57000
2281
+ },
2282
+ {
2283
+ "epoch": 2.32,
2284
+ "eval_accuracy": 0.8076305220883534,
2285
+ "eval_loss": 0.6049151420593262,
2286
+ "eval_runtime": 9.2353,
2287
+ "eval_samples_per_second": 269.619,
2288
+ "step": 57000
2289
+ },
2290
+ {
2291
+ "epoch": 2.33,
2292
+ "learning_rate": 1.1200295356341956e-05,
2293
+ "loss": 0.273,
2294
+ "step": 57250
2295
+ },
2296
+ {
2297
+ "epoch": 2.34,
2298
+ "learning_rate": 1.1029371547339059e-05,
2299
+ "loss": 0.2773,
2300
+ "step": 57500
2301
+ },
2302
+ {
2303
+ "epoch": 2.34,
2304
+ "eval_accuracy": 0.810843373493976,
2305
+ "eval_loss": 0.6248819231987,
2306
+ "eval_runtime": 9.0359,
2307
+ "eval_samples_per_second": 275.569,
2308
+ "step": 57500
2309
+ },
2310
+ {
2311
+ "epoch": 2.35,
2312
+ "learning_rate": 1.085844773833616e-05,
2313
+ "loss": 0.3164,
2314
+ "step": 57750
2315
+ },
2316
+ {
2317
+ "epoch": 2.36,
2318
+ "learning_rate": 1.068752392933326e-05,
2319
+ "loss": 0.2824,
2320
+ "step": 58000
2321
+ },
2322
+ {
2323
+ "epoch": 2.36,
2324
+ "eval_accuracy": 0.8136546184738955,
2325
+ "eval_loss": 0.5784918069839478,
2326
+ "eval_runtime": 9.2382,
2327
+ "eval_samples_per_second": 269.533,
2328
+ "step": 58000
2329
+ },
2330
+ {
2331
+ "epoch": 2.37,
2332
+ "learning_rate": 1.0516600120330363e-05,
2333
+ "loss": 0.2992,
2334
+ "step": 58250
2335
+ },
2336
+ {
2337
+ "epoch": 2.38,
2338
+ "learning_rate": 1.0345676311327464e-05,
2339
+ "loss": 0.293,
2340
+ "step": 58500
2341
+ },
2342
+ {
2343
+ "epoch": 2.38,
2344
+ "eval_accuracy": 0.8072289156626506,
2345
+ "eval_loss": 0.611821711063385,
2346
+ "eval_runtime": 9.0825,
2347
+ "eval_samples_per_second": 274.153,
2348
+ "step": 58500
2349
+ },
2350
+ {
2351
+ "epoch": 2.39,
2352
+ "learning_rate": 1.0174752502324563e-05,
2353
+ "loss": 0.2742,
2354
+ "step": 58750
2355
+ },
2356
+ {
2357
+ "epoch": 2.4,
2358
+ "learning_rate": 1.0003828693321665e-05,
2359
+ "loss": 0.2927,
2360
+ "step": 59000
2361
+ },
2362
+ {
2363
+ "epoch": 2.4,
2364
+ "eval_accuracy": 0.804417670682731,
2365
+ "eval_loss": 0.6135737895965576,
2366
+ "eval_runtime": 9.1843,
2367
+ "eval_samples_per_second": 271.114,
2368
+ "step": 59000
2369
+ },
2370
+ {
2371
+ "epoch": 2.41,
2372
+ "learning_rate": 9.832904884318766e-06,
2373
+ "loss": 0.282,
2374
+ "step": 59250
2375
+ },
2376
+ {
2377
+ "epoch": 2.42,
2378
+ "learning_rate": 9.661981075315867e-06,
2379
+ "loss": 0.3021,
2380
+ "step": 59500
2381
+ },
2382
+ {
2383
+ "epoch": 2.42,
2384
+ "eval_accuracy": 0.8088353413654619,
2385
+ "eval_loss": 0.5996263027191162,
2386
+ "eval_runtime": 9.1543,
2387
+ "eval_samples_per_second": 272.003,
2388
+ "step": 59500
2389
+ },
2390
+ {
2391
+ "epoch": 2.43,
2392
+ "learning_rate": 9.49105726631297e-06,
2393
+ "loss": 0.2931,
2394
+ "step": 59750
2395
+ },
2396
+ {
2397
+ "epoch": 2.44,
2398
+ "learning_rate": 9.320133457310069e-06,
2399
+ "loss": 0.2745,
2400
+ "step": 60000
2401
+ },
2402
+ {
2403
+ "epoch": 2.44,
2404
+ "eval_accuracy": 0.810843373493976,
2405
+ "eval_loss": 0.5868379473686218,
2406
+ "eval_runtime": 9.1158,
2407
+ "eval_samples_per_second": 273.152,
2408
+ "step": 60000
2409
+ },
2410
+ {
2411
+ "epoch": 2.45,
2412
+ "learning_rate": 9.149209648307171e-06,
2413
+ "loss": 0.2897,
2414
+ "step": 60250
2415
+ },
2416
+ {
2417
+ "epoch": 2.46,
2418
+ "learning_rate": 8.978285839304272e-06,
2419
+ "loss": 0.2919,
2420
+ "step": 60500
2421
+ },
2422
+ {
2423
+ "epoch": 2.46,
2424
+ "eval_accuracy": 0.8096385542168675,
2425
+ "eval_loss": 0.5863232016563416,
2426
+ "eval_runtime": 9.7927,
2427
+ "eval_samples_per_second": 254.27,
2428
+ "step": 60500
2429
+ },
2430
+ {
2431
+ "epoch": 2.48,
2432
+ "learning_rate": 8.807362030301373e-06,
2433
+ "loss": 0.2903,
2434
+ "step": 60750
2435
+ },
2436
+ {
2437
+ "epoch": 2.49,
2438
+ "learning_rate": 8.636438221298475e-06,
2439
+ "loss": 0.2662,
2440
+ "step": 61000
2441
+ },
2442
+ {
2443
+ "epoch": 2.49,
2444
+ "eval_accuracy": 0.804417670682731,
2445
+ "eval_loss": 0.636022686958313,
2446
+ "eval_runtime": 10.4967,
2447
+ "eval_samples_per_second": 237.217,
2448
+ "step": 61000
2449
+ },
2450
+ {
2451
+ "epoch": 2.5,
2452
+ "learning_rate": 8.465514412295575e-06,
2453
+ "loss": 0.2908,
2454
+ "step": 61250
2455
+ },
2456
+ {
2457
+ "epoch": 2.51,
2458
+ "learning_rate": 8.294590603292677e-06,
2459
+ "loss": 0.2977,
2460
+ "step": 61500
2461
+ },
2462
+ {
2463
+ "epoch": 2.51,
2464
+ "eval_accuracy": 0.8104417670682731,
2465
+ "eval_loss": 0.596953272819519,
2466
+ "eval_runtime": 9.4896,
2467
+ "eval_samples_per_second": 262.393,
2468
+ "step": 61500
2469
+ },
2470
+ {
2471
+ "epoch": 2.52,
2472
+ "learning_rate": 8.123666794289778e-06,
2473
+ "loss": 0.2723,
2474
+ "step": 61750
2475
+ },
2476
+ {
2477
+ "epoch": 2.53,
2478
+ "learning_rate": 7.952742985286879e-06,
2479
+ "loss": 0.2785,
2480
+ "step": 62000
2481
+ },
2482
+ {
2483
+ "epoch": 2.53,
2484
+ "eval_accuracy": 0.8164658634538152,
2485
+ "eval_loss": 0.5756428241729736,
2486
+ "eval_runtime": 9.105,
2487
+ "eval_samples_per_second": 273.476,
2488
+ "step": 62000
2489
+ }
2490
+ ],
2491
+ "max_steps": 73632,
2492
+ "num_train_epochs": 3,
2493
+ "total_flos": 55970352570961800,
2494
+ "trial_name": null,
2495
+ "trial_params": null
2496
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:622587664ec8a08158962cb8292ad2c26e3a3c89c6279169fcd8caee360cb2a6
3
+ size 2095
vocab.txt ADDED
The diff for this file is too large to render. See raw diff