Hugging Face's logo

tarabfish
/
amzn_yelp

Safetensors
Model card Files
xet
 Community
tarabfish commited on
Commit
97c34c5
·
verified ·
1 Parent(s): 0a1bffd

Upload folder using huggingface_hub

Browse files
Files changed (25) hide show
  1. checkpoints/checkpoint-5300/config.json +26 -0
  2. checkpoints/checkpoint-5300/model.safetensors +3 -0
  3. checkpoints/checkpoint-5300/optimizer.pt +3 -0
  4. checkpoints/checkpoint-5300/rng_state.pth +3 -0
  5. checkpoints/checkpoint-5300/scheduler.pt +3 -0
  6. checkpoints/checkpoint-5300/special_tokens_map.json +7 -0
  7. checkpoints/checkpoint-5300/tokenizer.json +0 -0
  8. checkpoints/checkpoint-5300/tokenizer_config.json +56 -0
  9. checkpoints/checkpoint-5300/trainer_state.json +581 -0
  10. checkpoints/checkpoint-5300/training_args.bin +3 -0
  11. checkpoints/checkpoint-5300/vocab.txt +0 -0
  12. checkpoints/checkpoint-9600/config.json +26 -0
  13. checkpoints/checkpoint-9600/model.safetensors +3 -0
  14. checkpoints/checkpoint-9600/optimizer.pt +3 -0
  15. checkpoints/checkpoint-9600/rng_state.pth +3 -0
  16. checkpoints/checkpoint-9600/scheduler.pt +3 -0
  17. checkpoints/checkpoint-9600/special_tokens_map.json +7 -0
  18. checkpoints/checkpoint-9600/tokenizer.json +0 -0
  19. checkpoints/checkpoint-9600/tokenizer_config.json +56 -0
  20. checkpoints/checkpoint-9600/trainer_state.json +1031 -0
  21. checkpoints/checkpoint-9600/training_args.bin +3 -0
  22. checkpoints/checkpoint-9600/vocab.txt +0 -0
  23. final_model.pt +3 -0
  24. final_model_meta.json +1 -0
  25. test_outputs.jsonl +96 -0
checkpoints/checkpoint-5300/config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 3072,
13
+ "layer_norm_eps": 1e-12,
14
+ "max_position_embeddings": 512,
15
+ "model_type": "bert",
16
+ "num_attention_heads": 12,
17
+ "num_hidden_layers": 12,
18
+ "pad_token_id": 0,
19
+ "position_embedding_type": "absolute",
20
+ "problem_type": "single_label_classification",
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.51.3",
23
+ "type_vocab_size": 2,
24
+ "use_cache": true,
25
+ "vocab_size": 30522
26
+ }
checkpoints/checkpoint-5300/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a94e9fca5fe42fd4faa0639854ccc7a4c6b4694b39a9ee4546535237970236a4
3
+ size 437958648
checkpoints/checkpoint-5300/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8075779ee5ca6c70c114ee662b19e2677400a5ec66cab67db1275d73c87a32ee
3
+ size 876038394
checkpoints/checkpoint-5300/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ccf68acb16b42df8c6cbf2989e34213e24c5f5b43afb3e67c67c0314224eb79c
3
+ size 14244
checkpoints/checkpoint-5300/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9136b9fcc9e87fdf00a3ae59416e1668e25e5ef391855c73cb3229ecaaf47fdb
3
+ size 1064
checkpoints/checkpoint-5300/special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
checkpoints/checkpoint-5300/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoints/checkpoint-5300/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 512,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
checkpoints/checkpoint-5300/trainer_state.json ADDED
@@ -0,0 +1,581 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 5300,
3
+ "best_metric": 0.94375,
4
+ "best_model_checkpoint": "../merge_models/amzn_yelp/task2/checkpoints/checkpoint-5300",
5
+ "epoch": 2.2083333333333335,
6
+ "eval_steps": 100,
7
+ "global_step": 5300,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.041666666666666664,
14
+ "eval_accuracy": 0.871875,
15
+ "eval_loss": 0.7828950881958008,
16
+ "eval_runtime": 2.8057,
17
+ "eval_samples_per_second": 114.053,
18
+ "eval_steps_per_second": 57.027,
19
+ "step": 100
20
+ },
21
+ {
22
+ "epoch": 0.08333333333333333,
23
+ "eval_accuracy": 0.88125,
24
+ "eval_loss": 0.6410558819770813,
25
+ "eval_runtime": 2.9454,
26
+ "eval_samples_per_second": 108.645,
27
+ "eval_steps_per_second": 54.323,
28
+ "step": 200
29
+ },
30
+ {
31
+ "epoch": 0.125,
32
+ "eval_accuracy": 0.90625,
33
+ "eval_loss": 0.4724481701850891,
34
+ "eval_runtime": 2.6489,
35
+ "eval_samples_per_second": 120.804,
36
+ "eval_steps_per_second": 60.402,
37
+ "step": 300
38
+ },
39
+ {
40
+ "epoch": 0.16666666666666666,
41
+ "eval_accuracy": 0.88125,
42
+ "eval_loss": 0.46632274985313416,
43
+ "eval_runtime": 2.743,
44
+ "eval_samples_per_second": 116.66,
45
+ "eval_steps_per_second": 58.33,
46
+ "step": 400
47
+ },
48
+ {
49
+ "epoch": 0.20833333333333334,
50
+ "grad_norm": 0.26992425322532654,
51
+ "learning_rate": 3.792083333333334e-05,
52
+ "loss": 0.5589,
53
+ "step": 500
54
+ },
55
+ {
56
+ "epoch": 0.20833333333333334,
57
+ "eval_accuracy": 0.928125,
58
+ "eval_loss": 0.336311399936676,
59
+ "eval_runtime": 2.7645,
60
+ "eval_samples_per_second": 115.753,
61
+ "eval_steps_per_second": 57.876,
62
+ "step": 500
63
+ },
64
+ {
65
+ "epoch": 0.25,
66
+ "eval_accuracy": 0.875,
67
+ "eval_loss": 0.5237163305282593,
68
+ "eval_runtime": 2.9888,
69
+ "eval_samples_per_second": 107.065,
70
+ "eval_steps_per_second": 53.532,
71
+ "step": 600
72
+ },
73
+ {
74
+ "epoch": 0.2916666666666667,
75
+ "eval_accuracy": 0.865625,
76
+ "eval_loss": 0.612436056137085,
77
+ "eval_runtime": 3.0505,
78
+ "eval_samples_per_second": 104.901,
79
+ "eval_steps_per_second": 52.45,
80
+ "step": 700
81
+ },
82
+ {
83
+ "epoch": 0.3333333333333333,
84
+ "eval_accuracy": 0.9,
85
+ "eval_loss": 0.47742635011672974,
86
+ "eval_runtime": 2.8541,
87
+ "eval_samples_per_second": 112.12,
88
+ "eval_steps_per_second": 56.06,
89
+ "step": 800
90
+ },
91
+ {
92
+ "epoch": 0.375,
93
+ "eval_accuracy": 0.925,
94
+ "eval_loss": 0.4391939640045166,
95
+ "eval_runtime": 2.7428,
96
+ "eval_samples_per_second": 116.667,
97
+ "eval_steps_per_second": 58.334,
98
+ "step": 900
99
+ },
100
+ {
101
+ "epoch": 0.4166666666666667,
102
+ "grad_norm": 0.19945169985294342,
103
+ "learning_rate": 3.5837500000000004e-05,
104
+ "loss": 0.5193,
105
+ "step": 1000
106
+ },
107
+ {
108
+ "epoch": 0.4166666666666667,
109
+ "eval_accuracy": 0.925,
110
+ "eval_loss": 0.3529580235481262,
111
+ "eval_runtime": 2.7396,
112
+ "eval_samples_per_second": 116.807,
113
+ "eval_steps_per_second": 58.404,
114
+ "step": 1000
115
+ },
116
+ {
117
+ "epoch": 0.4583333333333333,
118
+ "eval_accuracy": 0.925,
119
+ "eval_loss": 0.32948821783065796,
120
+ "eval_runtime": 3.0406,
121
+ "eval_samples_per_second": 105.242,
122
+ "eval_steps_per_second": 52.621,
123
+ "step": 1100
124
+ },
125
+ {
126
+ "epoch": 0.5,
127
+ "eval_accuracy": 0.9,
128
+ "eval_loss": 0.5077411532402039,
129
+ "eval_runtime": 3.0926,
130
+ "eval_samples_per_second": 103.474,
131
+ "eval_steps_per_second": 51.737,
132
+ "step": 1200
133
+ },
134
+ {
135
+ "epoch": 0.5416666666666666,
136
+ "eval_accuracy": 0.9,
137
+ "eval_loss": 0.4512642025947571,
138
+ "eval_runtime": 2.7715,
139
+ "eval_samples_per_second": 115.459,
140
+ "eval_steps_per_second": 57.73,
141
+ "step": 1300
142
+ },
143
+ {
144
+ "epoch": 0.5833333333333334,
145
+ "eval_accuracy": 0.89375,
146
+ "eval_loss": 0.43066683411598206,
147
+ "eval_runtime": 2.8379,
148
+ "eval_samples_per_second": 112.76,
149
+ "eval_steps_per_second": 56.38,
150
+ "step": 1400
151
+ },
152
+ {
153
+ "epoch": 0.625,
154
+ "grad_norm": 0.08616471290588379,
155
+ "learning_rate": 3.375416666666667e-05,
156
+ "loss": 0.5088,
157
+ "step": 1500
158
+ },
159
+ {
160
+ "epoch": 0.625,
161
+ "eval_accuracy": 0.9125,
162
+ "eval_loss": 0.37902718782424927,
163
+ "eval_runtime": 2.8742,
164
+ "eval_samples_per_second": 111.336,
165
+ "eval_steps_per_second": 55.668,
166
+ "step": 1500
167
+ },
168
+ {
169
+ "epoch": 0.6666666666666666,
170
+ "eval_accuracy": 0.8875,
171
+ "eval_loss": 0.435788631439209,
172
+ "eval_runtime": 3.0691,
173
+ "eval_samples_per_second": 104.266,
174
+ "eval_steps_per_second": 52.133,
175
+ "step": 1600
176
+ },
177
+ {
178
+ "epoch": 0.7083333333333334,
179
+ "eval_accuracy": 0.878125,
180
+ "eval_loss": 0.6496487855911255,
181
+ "eval_runtime": 3.019,
182
+ "eval_samples_per_second": 105.995,
183
+ "eval_steps_per_second": 52.998,
184
+ "step": 1700
185
+ },
186
+ {
187
+ "epoch": 0.75,
188
+ "eval_accuracy": 0.896875,
189
+ "eval_loss": 0.5356670618057251,
190
+ "eval_runtime": 2.7479,
191
+ "eval_samples_per_second": 116.452,
192
+ "eval_steps_per_second": 58.226,
193
+ "step": 1800
194
+ },
195
+ {
196
+ "epoch": 0.7916666666666666,
197
+ "eval_accuracy": 0.903125,
198
+ "eval_loss": 0.5577458143234253,
199
+ "eval_runtime": 2.7431,
200
+ "eval_samples_per_second": 116.657,
201
+ "eval_steps_per_second": 58.329,
202
+ "step": 1900
203
+ },
204
+ {
205
+ "epoch": 0.8333333333333334,
206
+ "grad_norm": 0.2015516757965088,
207
+ "learning_rate": 3.1670833333333335e-05,
208
+ "loss": 0.5272,
209
+ "step": 2000
210
+ },
211
+ {
212
+ "epoch": 0.8333333333333334,
213
+ "eval_accuracy": 0.9,
214
+ "eval_loss": 0.5010732412338257,
215
+ "eval_runtime": 2.8418,
216
+ "eval_samples_per_second": 112.606,
217
+ "eval_steps_per_second": 56.303,
218
+ "step": 2000
219
+ },
220
+ {
221
+ "epoch": 0.875,
222
+ "eval_accuracy": 0.871875,
223
+ "eval_loss": 0.5578997135162354,
224
+ "eval_runtime": 2.9909,
225
+ "eval_samples_per_second": 106.99,
226
+ "eval_steps_per_second": 53.495,
227
+ "step": 2100
228
+ },
229
+ {
230
+ "epoch": 0.9166666666666666,
231
+ "eval_accuracy": 0.9125,
232
+ "eval_loss": 0.49204739928245544,
233
+ "eval_runtime": 2.9421,
234
+ "eval_samples_per_second": 108.768,
235
+ "eval_steps_per_second": 54.384,
236
+ "step": 2200
237
+ },
238
+ {
239
+ "epoch": 0.9583333333333334,
240
+ "eval_accuracy": 0.921875,
241
+ "eval_loss": 0.42611804604530334,
242
+ "eval_runtime": 2.8842,
243
+ "eval_samples_per_second": 110.949,
244
+ "eval_steps_per_second": 55.475,
245
+ "step": 2300
246
+ },
247
+ {
248
+ "epoch": 1.0,
249
+ "eval_accuracy": 0.903125,
250
+ "eval_loss": 0.5209008455276489,
251
+ "eval_runtime": 2.8447,
252
+ "eval_samples_per_second": 112.49,
253
+ "eval_steps_per_second": 56.245,
254
+ "step": 2400
255
+ },
256
+ {
257
+ "epoch": 1.0416666666666667,
258
+ "grad_norm": 0.026173487305641174,
259
+ "learning_rate": 2.9587500000000004e-05,
260
+ "loss": 0.3682,
261
+ "step": 2500
262
+ },
263
+ {
264
+ "epoch": 1.0416666666666667,
265
+ "eval_accuracy": 0.909375,
266
+ "eval_loss": 0.558965802192688,
267
+ "eval_runtime": 3.0287,
268
+ "eval_samples_per_second": 105.655,
269
+ "eval_steps_per_second": 52.828,
270
+ "step": 2500
271
+ },
272
+ {
273
+ "epoch": 1.0833333333333333,
274
+ "eval_accuracy": 0.909375,
275
+ "eval_loss": 0.5337815284729004,
276
+ "eval_runtime": 2.8886,
277
+ "eval_samples_per_second": 110.781,
278
+ "eval_steps_per_second": 55.391,
279
+ "step": 2600
280
+ },
281
+ {
282
+ "epoch": 1.125,
283
+ "eval_accuracy": 0.91875,
284
+ "eval_loss": 0.43373221158981323,
285
+ "eval_runtime": 2.927,
286
+ "eval_samples_per_second": 109.328,
287
+ "eval_steps_per_second": 54.664,
288
+ "step": 2700
289
+ },
290
+ {
291
+ "epoch": 1.1666666666666667,
292
+ "eval_accuracy": 0.90625,
293
+ "eval_loss": 0.562017560005188,
294
+ "eval_runtime": 3.1459,
295
+ "eval_samples_per_second": 101.72,
296
+ "eval_steps_per_second": 50.86,
297
+ "step": 2800
298
+ },
299
+ {
300
+ "epoch": 1.2083333333333333,
301
+ "eval_accuracy": 0.915625,
302
+ "eval_loss": 0.48360759019851685,
303
+ "eval_runtime": 2.7861,
304
+ "eval_samples_per_second": 114.855,
305
+ "eval_steps_per_second": 57.427,
306
+ "step": 2900
307
+ },
308
+ {
309
+ "epoch": 1.25,
310
+ "grad_norm": 0.03431430086493492,
311
+ "learning_rate": 2.750416666666667e-05,
312
+ "loss": 0.2665,
313
+ "step": 3000
314
+ },
315
+ {
316
+ "epoch": 1.25,
317
+ "eval_accuracy": 0.903125,
318
+ "eval_loss": 0.5819778442382812,
319
+ "eval_runtime": 2.7837,
320
+ "eval_samples_per_second": 114.956,
321
+ "eval_steps_per_second": 57.478,
322
+ "step": 3000
323
+ },
324
+ {
325
+ "epoch": 1.2916666666666667,
326
+ "eval_accuracy": 0.90625,
327
+ "eval_loss": 0.549329400062561,
328
+ "eval_runtime": 3.1117,
329
+ "eval_samples_per_second": 102.837,
330
+ "eval_steps_per_second": 51.418,
331
+ "step": 3100
332
+ },
333
+ {
334
+ "epoch": 1.3333333333333333,
335
+ "eval_accuracy": 0.909375,
336
+ "eval_loss": 0.5409368276596069,
337
+ "eval_runtime": 3.0275,
338
+ "eval_samples_per_second": 105.698,
339
+ "eval_steps_per_second": 52.849,
340
+ "step": 3200
341
+ },
342
+ {
343
+ "epoch": 1.375,
344
+ "eval_accuracy": 0.91875,
345
+ "eval_loss": 0.5061095952987671,
346
+ "eval_runtime": 2.7533,
347
+ "eval_samples_per_second": 116.222,
348
+ "eval_steps_per_second": 58.111,
349
+ "step": 3300
350
+ },
351
+ {
352
+ "epoch": 1.4166666666666667,
353
+ "eval_accuracy": 0.915625,
354
+ "eval_loss": 0.5100991129875183,
355
+ "eval_runtime": 2.8582,
356
+ "eval_samples_per_second": 111.958,
357
+ "eval_steps_per_second": 55.979,
358
+ "step": 3400
359
+ },
360
+ {
361
+ "epoch": 1.4583333333333333,
362
+ "grad_norm": 0.04029667377471924,
363
+ "learning_rate": 2.5420833333333336e-05,
364
+ "loss": 0.2556,
365
+ "step": 3500
366
+ },
367
+ {
368
+ "epoch": 1.4583333333333333,
369
+ "eval_accuracy": 0.91875,
370
+ "eval_loss": 0.48856163024902344,
371
+ "eval_runtime": 2.8545,
372
+ "eval_samples_per_second": 112.104,
373
+ "eval_steps_per_second": 56.052,
374
+ "step": 3500
375
+ },
376
+ {
377
+ "epoch": 1.5,
378
+ "eval_accuracy": 0.91875,
379
+ "eval_loss": 0.5146762728691101,
380
+ "eval_runtime": 3.0905,
381
+ "eval_samples_per_second": 103.544,
382
+ "eval_steps_per_second": 51.772,
383
+ "step": 3600
384
+ },
385
+ {
386
+ "epoch": 1.5416666666666665,
387
+ "eval_accuracy": 0.915625,
388
+ "eval_loss": 0.5367563962936401,
389
+ "eval_runtime": 2.9095,
390
+ "eval_samples_per_second": 109.983,
391
+ "eval_steps_per_second": 54.992,
392
+ "step": 3700
393
+ },
394
+ {
395
+ "epoch": 1.5833333333333335,
396
+ "eval_accuracy": 0.909375,
397
+ "eval_loss": 0.4761125445365906,
398
+ "eval_runtime": 2.8026,
399
+ "eval_samples_per_second": 114.178,
400
+ "eval_steps_per_second": 57.089,
401
+ "step": 3800
402
+ },
403
+ {
404
+ "epoch": 1.625,
405
+ "eval_accuracy": 0.90625,
406
+ "eval_loss": 0.4483864903450012,
407
+ "eval_runtime": 3.0492,
408
+ "eval_samples_per_second": 104.944,
409
+ "eval_steps_per_second": 52.472,
410
+ "step": 3900
411
+ },
412
+ {
413
+ "epoch": 1.6666666666666665,
414
+ "grad_norm": 107.47529602050781,
415
+ "learning_rate": 2.3337500000000005e-05,
416
+ "loss": 0.2449,
417
+ "step": 4000
418
+ },
419
+ {
420
+ "epoch": 1.6666666666666665,
421
+ "eval_accuracy": 0.928125,
422
+ "eval_loss": 0.35619640350341797,
423
+ "eval_runtime": 2.9753,
424
+ "eval_samples_per_second": 107.553,
425
+ "eval_steps_per_second": 53.777,
426
+ "step": 4000
427
+ },
428
+ {
429
+ "epoch": 1.7083333333333335,
430
+ "eval_accuracy": 0.921875,
431
+ "eval_loss": 0.38642579317092896,
432
+ "eval_runtime": 3.0163,
433
+ "eval_samples_per_second": 106.091,
434
+ "eval_steps_per_second": 53.046,
435
+ "step": 4100
436
+ },
437
+ {
438
+ "epoch": 1.75,
439
+ "eval_accuracy": 0.90625,
440
+ "eval_loss": 0.5521766543388367,
441
+ "eval_runtime": 2.8851,
442
+ "eval_samples_per_second": 110.915,
443
+ "eval_steps_per_second": 55.457,
444
+ "step": 4200
445
+ },
446
+ {
447
+ "epoch": 1.7916666666666665,
448
+ "eval_accuracy": 0.890625,
449
+ "eval_loss": 0.4897506833076477,
450
+ "eval_runtime": 2.805,
451
+ "eval_samples_per_second": 114.082,
452
+ "eval_steps_per_second": 57.041,
453
+ "step": 4300
454
+ },
455
+ {
456
+ "epoch": 1.8333333333333335,
457
+ "eval_accuracy": 0.921875,
458
+ "eval_loss": 0.4268820881843567,
459
+ "eval_runtime": 3.0937,
460
+ "eval_samples_per_second": 103.437,
461
+ "eval_steps_per_second": 51.718,
462
+ "step": 4400
463
+ },
464
+ {
465
+ "epoch": 1.875,
466
+ "grad_norm": 0.026214903220534325,
467
+ "learning_rate": 2.125416666666667e-05,
468
+ "loss": 0.2134,
469
+ "step": 4500
470
+ },
471
+ {
472
+ "epoch": 1.875,
473
+ "eval_accuracy": 0.928125,
474
+ "eval_loss": 0.43541350960731506,
475
+ "eval_runtime": 2.9558,
476
+ "eval_samples_per_second": 108.262,
477
+ "eval_steps_per_second": 54.131,
478
+ "step": 4500
479
+ },
480
+ {
481
+ "epoch": 1.9166666666666665,
482
+ "eval_accuracy": 0.91875,
483
+ "eval_loss": 0.4590047299861908,
484
+ "eval_runtime": 3.1134,
485
+ "eval_samples_per_second": 102.782,
486
+ "eval_steps_per_second": 51.391,
487
+ "step": 4600
488
+ },
489
+ {
490
+ "epoch": 1.9583333333333335,
491
+ "eval_accuracy": 0.921875,
492
+ "eval_loss": 0.42555299401283264,
493
+ "eval_runtime": 2.7577,
494
+ "eval_samples_per_second": 116.038,
495
+ "eval_steps_per_second": 58.019,
496
+ "step": 4700
497
+ },
498
+ {
499
+ "epoch": 2.0,
500
+ "eval_accuracy": 0.921875,
501
+ "eval_loss": 0.46792250871658325,
502
+ "eval_runtime": 3.199,
503
+ "eval_samples_per_second": 100.03,
504
+ "eval_steps_per_second": 50.015,
505
+ "step": 4800
506
+ },
507
+ {
508
+ "epoch": 2.0416666666666665,
509
+ "eval_accuracy": 0.934375,
510
+ "eval_loss": 0.40328049659729004,
511
+ "eval_runtime": 2.8034,
512
+ "eval_samples_per_second": 114.146,
513
+ "eval_steps_per_second": 57.073,
514
+ "step": 4900
515
+ },
516
+ {
517
+ "epoch": 2.0833333333333335,
518
+ "grad_norm": 0.021490707993507385,
519
+ "learning_rate": 1.9170833333333333e-05,
520
+ "loss": 0.2505,
521
+ "step": 5000
522
+ },
523
+ {
524
+ "epoch": 2.0833333333333335,
525
+ "eval_accuracy": 0.91875,
526
+ "eval_loss": 0.4490671753883362,
527
+ "eval_runtime": 2.8849,
528
+ "eval_samples_per_second": 110.924,
529
+ "eval_steps_per_second": 55.462,
530
+ "step": 5000
531
+ },
532
+ {
533
+ "epoch": 2.125,
534
+ "eval_accuracy": 0.925,
535
+ "eval_loss": 0.41796278953552246,
536
+ "eval_runtime": 3.1312,
537
+ "eval_samples_per_second": 102.196,
538
+ "eval_steps_per_second": 51.098,
539
+ "step": 5100
540
+ },
541
+ {
542
+ "epoch": 2.1666666666666665,
543
+ "eval_accuracy": 0.925,
544
+ "eval_loss": 0.3970945477485657,
545
+ "eval_runtime": 2.9082,
546
+ "eval_samples_per_second": 110.032,
547
+ "eval_steps_per_second": 55.016,
548
+ "step": 5200
549
+ },
550
+ {
551
+ "epoch": 2.2083333333333335,
552
+ "eval_accuracy": 0.94375,
553
+ "eval_loss": 0.3905588984489441,
554
+ "eval_runtime": 2.972,
555
+ "eval_samples_per_second": 107.671,
556
+ "eval_steps_per_second": 53.835,
557
+ "step": 5300
558
+ }
559
+ ],
560
+ "logging_steps": 500,
561
+ "max_steps": 9600,
562
+ "num_input_tokens_seen": 0,
563
+ "num_train_epochs": 4,
564
+ "save_steps": 100,
565
+ "stateful_callbacks": {
566
+ "TrainerControl": {
567
+ "args": {
568
+ "should_epoch_stop": false,
569
+ "should_evaluate": false,
570
+ "should_log": false,
571
+ "should_save": true,
572
+ "should_training_stop": false
573
+ },
574
+ "attributes": {}
575
+ }
576
+ },
577
+ "total_flos": 718532653304280.0,
578
+ "train_batch_size": 2,
579
+ "trial_name": null,
580
+ "trial_params": null
581
+ }
checkpoints/checkpoint-5300/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65494a2338792a7eb98d28ceffbee46e49ecedf50dfe5b401590a45c130902ae
3
+ size 5304
checkpoints/checkpoint-5300/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoints/checkpoint-9600/config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 3072,
13
+ "layer_norm_eps": 1e-12,
14
+ "max_position_embeddings": 512,
15
+ "model_type": "bert",
16
+ "num_attention_heads": 12,
17
+ "num_hidden_layers": 12,
18
+ "pad_token_id": 0,
19
+ "position_embedding_type": "absolute",
20
+ "problem_type": "single_label_classification",
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.51.3",
23
+ "type_vocab_size": 2,
24
+ "use_cache": true,
25
+ "vocab_size": 30522
26
+ }
checkpoints/checkpoint-9600/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6309c2d1e3641872fbf02c22b043f8ba23a03ffd3d222edff88ad69dbbb9c6de
3
+ size 437958648
checkpoints/checkpoint-9600/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2eff799c2f9c3df2f2d9f2994fa0caa2f35d5e41ca4324748a8fdb0791b30452
3
+ size 876038394
checkpoints/checkpoint-9600/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5cf02a5b5d7c80c61d5768256c8d689bb502772d19a15c35eb350226afeac75
3
+ size 14244
checkpoints/checkpoint-9600/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70b4b49382b81ae82c82f376964b1763ea6f54b6fa49526da8f199bd9ab93c7a
3
+ size 1064
checkpoints/checkpoint-9600/special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
checkpoints/checkpoint-9600/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoints/checkpoint-9600/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 512,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
checkpoints/checkpoint-9600/trainer_state.json ADDED
@@ -0,0 +1,1031 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 5300,
3
+ "best_metric": 0.94375,
4
+ "best_model_checkpoint": "../merge_models/amzn_yelp/task2/checkpoints/checkpoint-5300",
5
+ "epoch": 4.0,
6
+ "eval_steps": 100,
7
+ "global_step": 9600,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.041666666666666664,
14
+ "eval_accuracy": 0.871875,
15
+ "eval_loss": 0.7828950881958008,
16
+ "eval_runtime": 2.8057,
17
+ "eval_samples_per_second": 114.053,
18
+ "eval_steps_per_second": 57.027,
19
+ "step": 100
20
+ },
21
+ {
22
+ "epoch": 0.08333333333333333,
23
+ "eval_accuracy": 0.88125,
24
+ "eval_loss": 0.6410558819770813,
25
+ "eval_runtime": 2.9454,
26
+ "eval_samples_per_second": 108.645,
27
+ "eval_steps_per_second": 54.323,
28
+ "step": 200
29
+ },
30
+ {
31
+ "epoch": 0.125,
32
+ "eval_accuracy": 0.90625,
33
+ "eval_loss": 0.4724481701850891,
34
+ "eval_runtime": 2.6489,
35
+ "eval_samples_per_second": 120.804,
36
+ "eval_steps_per_second": 60.402,
37
+ "step": 300
38
+ },
39
+ {
40
+ "epoch": 0.16666666666666666,
41
+ "eval_accuracy": 0.88125,
42
+ "eval_loss": 0.46632274985313416,
43
+ "eval_runtime": 2.743,
44
+ "eval_samples_per_second": 116.66,
45
+ "eval_steps_per_second": 58.33,
46
+ "step": 400
47
+ },
48
+ {
49
+ "epoch": 0.20833333333333334,
50
+ "grad_norm": 0.26992425322532654,
51
+ "learning_rate": 3.792083333333334e-05,
52
+ "loss": 0.5589,
53
+ "step": 500
54
+ },
55
+ {
56
+ "epoch": 0.20833333333333334,
57
+ "eval_accuracy": 0.928125,
58
+ "eval_loss": 0.336311399936676,
59
+ "eval_runtime": 2.7645,
60
+ "eval_samples_per_second": 115.753,
61
+ "eval_steps_per_second": 57.876,
62
+ "step": 500
63
+ },
64
+ {
65
+ "epoch": 0.25,
66
+ "eval_accuracy": 0.875,
67
+ "eval_loss": 0.5237163305282593,
68
+ "eval_runtime": 2.9888,
69
+ "eval_samples_per_second": 107.065,
70
+ "eval_steps_per_second": 53.532,
71
+ "step": 600
72
+ },
73
+ {
74
+ "epoch": 0.2916666666666667,
75
+ "eval_accuracy": 0.865625,
76
+ "eval_loss": 0.612436056137085,
77
+ "eval_runtime": 3.0505,
78
+ "eval_samples_per_second": 104.901,
79
+ "eval_steps_per_second": 52.45,
80
+ "step": 700
81
+ },
82
+ {
83
+ "epoch": 0.3333333333333333,
84
+ "eval_accuracy": 0.9,
85
+ "eval_loss": 0.47742635011672974,
86
+ "eval_runtime": 2.8541,
87
+ "eval_samples_per_second": 112.12,
88
+ "eval_steps_per_second": 56.06,
89
+ "step": 800
90
+ },
91
+ {
92
+ "epoch": 0.375,
93
+ "eval_accuracy": 0.925,
94
+ "eval_loss": 0.4391939640045166,
95
+ "eval_runtime": 2.7428,
96
+ "eval_samples_per_second": 116.667,
97
+ "eval_steps_per_second": 58.334,
98
+ "step": 900
99
+ },
100
+ {
101
+ "epoch": 0.4166666666666667,
102
+ "grad_norm": 0.19945169985294342,
103
+ "learning_rate": 3.5837500000000004e-05,
104
+ "loss": 0.5193,
105
+ "step": 1000
106
+ },
107
+ {
108
+ "epoch": 0.4166666666666667,
109
+ "eval_accuracy": 0.925,
110
+ "eval_loss": 0.3529580235481262,
111
+ "eval_runtime": 2.7396,
112
+ "eval_samples_per_second": 116.807,
113
+ "eval_steps_per_second": 58.404,
114
+ "step": 1000
115
+ },
116
+ {
117
+ "epoch": 0.4583333333333333,
118
+ "eval_accuracy": 0.925,
119
+ "eval_loss": 0.32948821783065796,
120
+ "eval_runtime": 3.0406,
121
+ "eval_samples_per_second": 105.242,
122
+ "eval_steps_per_second": 52.621,
123
+ "step": 1100
124
+ },
125
+ {
126
+ "epoch": 0.5,
127
+ "eval_accuracy": 0.9,
128
+ "eval_loss": 0.5077411532402039,
129
+ "eval_runtime": 3.0926,
130
+ "eval_samples_per_second": 103.474,
131
+ "eval_steps_per_second": 51.737,
132
+ "step": 1200
133
+ },
134
+ {
135
+ "epoch": 0.5416666666666666,
136
+ "eval_accuracy": 0.9,
137
+ "eval_loss": 0.4512642025947571,
138
+ "eval_runtime": 2.7715,
139
+ "eval_samples_per_second": 115.459,
140
+ "eval_steps_per_second": 57.73,
141
+ "step": 1300
142
+ },
143
+ {
144
+ "epoch": 0.5833333333333334,
145
+ "eval_accuracy": 0.89375,
146
+ "eval_loss": 0.43066683411598206,
147
+ "eval_runtime": 2.8379,
148
+ "eval_samples_per_second": 112.76,
149
+ "eval_steps_per_second": 56.38,
150
+ "step": 1400
151
+ },
152
+ {
153
+ "epoch": 0.625,
154
+ "grad_norm": 0.08616471290588379,
155
+ "learning_rate": 3.375416666666667e-05,
156
+ "loss": 0.5088,
157
+ "step": 1500
158
+ },
159
+ {
160
+ "epoch": 0.625,
161
+ "eval_accuracy": 0.9125,
162
+ "eval_loss": 0.37902718782424927,
163
+ "eval_runtime": 2.8742,
164
+ "eval_samples_per_second": 111.336,
165
+ "eval_steps_per_second": 55.668,
166
+ "step": 1500
167
+ },
168
+ {
169
+ "epoch": 0.6666666666666666,
170
+ "eval_accuracy": 0.8875,
171
+ "eval_loss": 0.435788631439209,
172
+ "eval_runtime": 3.0691,
173
+ "eval_samples_per_second": 104.266,
174
+ "eval_steps_per_second": 52.133,
175
+ "step": 1600
176
+ },
177
+ {
178
+ "epoch": 0.7083333333333334,
179
+ "eval_accuracy": 0.878125,
180
+ "eval_loss": 0.6496487855911255,
181
+ "eval_runtime": 3.019,
182
+ "eval_samples_per_second": 105.995,
183
+ "eval_steps_per_second": 52.998,
184
+ "step": 1700
185
+ },
186
+ {
187
+ "epoch": 0.75,
188
+ "eval_accuracy": 0.896875,
189
+ "eval_loss": 0.5356670618057251,
190
+ "eval_runtime": 2.7479,
191
+ "eval_samples_per_second": 116.452,
192
+ "eval_steps_per_second": 58.226,
193
+ "step": 1800
194
+ },
195
+ {
196
+ "epoch": 0.7916666666666666,
197
+ "eval_accuracy": 0.903125,
198
+ "eval_loss": 0.5577458143234253,
199
+ "eval_runtime": 2.7431,
200
+ "eval_samples_per_second": 116.657,
201
+ "eval_steps_per_second": 58.329,
202
+ "step": 1900
203
+ },
204
+ {
205
+ "epoch": 0.8333333333333334,
206
+ "grad_norm": 0.2015516757965088,
207
+ "learning_rate": 3.1670833333333335e-05,
208
+ "loss": 0.5272,
209
+ "step": 2000
210
+ },
211
+ {
212
+ "epoch": 0.8333333333333334,
213
+ "eval_accuracy": 0.9,
214
+ "eval_loss": 0.5010732412338257,
215
+ "eval_runtime": 2.8418,
216
+ "eval_samples_per_second": 112.606,
217
+ "eval_steps_per_second": 56.303,
218
+ "step": 2000
219
+ },
220
+ {
221
+ "epoch": 0.875,
222
+ "eval_accuracy": 0.871875,
223
+ "eval_loss": 0.5578997135162354,
224
+ "eval_runtime": 2.9909,
225
+ "eval_samples_per_second": 106.99,
226
+ "eval_steps_per_second": 53.495,
227
+ "step": 2100
228
+ },
229
+ {
230
+ "epoch": 0.9166666666666666,
231
+ "eval_accuracy": 0.9125,
232
+ "eval_loss": 0.49204739928245544,
233
+ "eval_runtime": 2.9421,
234
+ "eval_samples_per_second": 108.768,
235
+ "eval_steps_per_second": 54.384,
236
+ "step": 2200
237
+ },
238
+ {
239
+ "epoch": 0.9583333333333334,
240
+ "eval_accuracy": 0.921875,
241
+ "eval_loss": 0.42611804604530334,
242
+ "eval_runtime": 2.8842,
243
+ "eval_samples_per_second": 110.949,
244
+ "eval_steps_per_second": 55.475,
245
+ "step": 2300
246
+ },
247
+ {
248
+ "epoch": 1.0,
249
+ "eval_accuracy": 0.903125,
250
+ "eval_loss": 0.5209008455276489,
251
+ "eval_runtime": 2.8447,
252
+ "eval_samples_per_second": 112.49,
253
+ "eval_steps_per_second": 56.245,
254
+ "step": 2400
255
+ },
256
+ {
257
+ "epoch": 1.0416666666666667,
258
+ "grad_norm": 0.026173487305641174,
259
+ "learning_rate": 2.9587500000000004e-05,
260
+ "loss": 0.3682,
261
+ "step": 2500
262
+ },
263
+ {
264
+ "epoch": 1.0416666666666667,
265
+ "eval_accuracy": 0.909375,
266
+ "eval_loss": 0.558965802192688,
267
+ "eval_runtime": 3.0287,
268
+ "eval_samples_per_second": 105.655,
269
+ "eval_steps_per_second": 52.828,
270
+ "step": 2500
271
+ },
272
+ {
273
+ "epoch": 1.0833333333333333,
274
+ "eval_accuracy": 0.909375,
275
+ "eval_loss": 0.5337815284729004,
276
+ "eval_runtime": 2.8886,
277
+ "eval_samples_per_second": 110.781,
278
+ "eval_steps_per_second": 55.391,
279
+ "step": 2600
280
+ },
281
+ {
282
+ "epoch": 1.125,
283
+ "eval_accuracy": 0.91875,
284
+ "eval_loss": 0.43373221158981323,
285
+ "eval_runtime": 2.927,
286
+ "eval_samples_per_second": 109.328,
287
+ "eval_steps_per_second": 54.664,
288
+ "step": 2700
289
+ },
290
+ {
291
+ "epoch": 1.1666666666666667,
292
+ "eval_accuracy": 0.90625,
293
+ "eval_loss": 0.562017560005188,
294
+ "eval_runtime": 3.1459,
295
+ "eval_samples_per_second": 101.72,
296
+ "eval_steps_per_second": 50.86,
297
+ "step": 2800
298
+ },
299
+ {
300
+ "epoch": 1.2083333333333333,
301
+ "eval_accuracy": 0.915625,
302
+ "eval_loss": 0.48360759019851685,
303
+ "eval_runtime": 2.7861,
304
+ "eval_samples_per_second": 114.855,
305
+ "eval_steps_per_second": 57.427,
306
+ "step": 2900
307
+ },
308
+ {
309
+ "epoch": 1.25,
310
+ "grad_norm": 0.03431430086493492,
311
+ "learning_rate": 2.750416666666667e-05,
312
+ "loss": 0.2665,
313
+ "step": 3000
314
+ },
315
+ {
316
+ "epoch": 1.25,
317
+ "eval_accuracy": 0.903125,
318
+ "eval_loss": 0.5819778442382812,
319
+ "eval_runtime": 2.7837,
320
+ "eval_samples_per_second": 114.956,
321
+ "eval_steps_per_second": 57.478,
322
+ "step": 3000
323
+ },
324
+ {
325
+ "epoch": 1.2916666666666667,
326
+ "eval_accuracy": 0.90625,
327
+ "eval_loss": 0.549329400062561,
328
+ "eval_runtime": 3.1117,
329
+ "eval_samples_per_second": 102.837,
330
+ "eval_steps_per_second": 51.418,
331
+ "step": 3100
332
+ },
333
+ {
334
+ "epoch": 1.3333333333333333,
335
+ "eval_accuracy": 0.909375,
336
+ "eval_loss": 0.5409368276596069,
337
+ "eval_runtime": 3.0275,
338
+ "eval_samples_per_second": 105.698,
339
+ "eval_steps_per_second": 52.849,
340
+ "step": 3200
341
+ },
342
+ {
343
+ "epoch": 1.375,
344
+ "eval_accuracy": 0.91875,
345
+ "eval_loss": 0.5061095952987671,
346
+ "eval_runtime": 2.7533,
347
+ "eval_samples_per_second": 116.222,
348
+ "eval_steps_per_second": 58.111,
349
+ "step": 3300
350
+ },
351
+ {
352
+ "epoch": 1.4166666666666667,
353
+ "eval_accuracy": 0.915625,
354
+ "eval_loss": 0.5100991129875183,
355
+ "eval_runtime": 2.8582,
356
+ "eval_samples_per_second": 111.958,
357
+ "eval_steps_per_second": 55.979,
358
+ "step": 3400
359
+ },
360
+ {
361
+ "epoch": 1.4583333333333333,
362
+ "grad_norm": 0.04029667377471924,
363
+ "learning_rate": 2.5420833333333336e-05,
364
+ "loss": 0.2556,
365
+ "step": 3500
366
+ },
367
+ {
368
+ "epoch": 1.4583333333333333,
369
+ "eval_accuracy": 0.91875,
370
+ "eval_loss": 0.48856163024902344,
371
+ "eval_runtime": 2.8545,
372
+ "eval_samples_per_second": 112.104,
373
+ "eval_steps_per_second": 56.052,
374
+ "step": 3500
375
+ },
376
+ {
377
+ "epoch": 1.5,
378
+ "eval_accuracy": 0.91875,
379
+ "eval_loss": 0.5146762728691101,
380
+ "eval_runtime": 3.0905,
381
+ "eval_samples_per_second": 103.544,
382
+ "eval_steps_per_second": 51.772,
383
+ "step": 3600
384
+ },
385
+ {
386
+ "epoch": 1.5416666666666665,
387
+ "eval_accuracy": 0.915625,
388
+ "eval_loss": 0.5367563962936401,
389
+ "eval_runtime": 2.9095,
390
+ "eval_samples_per_second": 109.983,
391
+ "eval_steps_per_second": 54.992,
392
+ "step": 3700
393
+ },
394
+ {
395
+ "epoch": 1.5833333333333335,
396
+ "eval_accuracy": 0.909375,
397
+ "eval_loss": 0.4761125445365906,
398
+ "eval_runtime": 2.8026,
399
+ "eval_samples_per_second": 114.178,
400
+ "eval_steps_per_second": 57.089,
401
+ "step": 3800
402
+ },
403
+ {
404
+ "epoch": 1.625,
405
+ "eval_accuracy": 0.90625,
406
+ "eval_loss": 0.4483864903450012,
407
+ "eval_runtime": 3.0492,
408
+ "eval_samples_per_second": 104.944,
409
+ "eval_steps_per_second": 52.472,
410
+ "step": 3900
411
+ },
412
+ {
413
+ "epoch": 1.6666666666666665,
414
+ "grad_norm": 107.47529602050781,
415
+ "learning_rate": 2.3337500000000005e-05,
416
+ "loss": 0.2449,
417
+ "step": 4000
418
+ },
419
+ {
420
+ "epoch": 1.6666666666666665,
421
+ "eval_accuracy": 0.928125,
422
+ "eval_loss": 0.35619640350341797,
423
+ "eval_runtime": 2.9753,
424
+ "eval_samples_per_second": 107.553,
425
+ "eval_steps_per_second": 53.777,
426
+ "step": 4000
427
+ },
428
+ {
429
+ "epoch": 1.7083333333333335,
430
+ "eval_accuracy": 0.921875,
431
+ "eval_loss": 0.38642579317092896,
432
+ "eval_runtime": 3.0163,
433
+ "eval_samples_per_second": 106.091,
434
+ "eval_steps_per_second": 53.046,
435
+ "step": 4100
436
+ },
437
+ {
438
+ "epoch": 1.75,
439
+ "eval_accuracy": 0.90625,
440
+ "eval_loss": 0.5521766543388367,
441
+ "eval_runtime": 2.8851,
442
+ "eval_samples_per_second": 110.915,
443
+ "eval_steps_per_second": 55.457,
444
+ "step": 4200
445
+ },
446
+ {
447
+ "epoch": 1.7916666666666665,
448
+ "eval_accuracy": 0.890625,
449
+ "eval_loss": 0.4897506833076477,
450
+ "eval_runtime": 2.805,
451
+ "eval_samples_per_second": 114.082,
452
+ "eval_steps_per_second": 57.041,
453
+ "step": 4300
454
+ },
455
+ {
456
+ "epoch": 1.8333333333333335,
457
+ "eval_accuracy": 0.921875,
458
+ "eval_loss": 0.4268820881843567,
459
+ "eval_runtime": 3.0937,
460
+ "eval_samples_per_second": 103.437,
461
+ "eval_steps_per_second": 51.718,
462
+ "step": 4400
463
+ },
464
+ {
465
+ "epoch": 1.875,
466
+ "grad_norm": 0.026214903220534325,
467
+ "learning_rate": 2.125416666666667e-05,
468
+ "loss": 0.2134,
469
+ "step": 4500
470
+ },
471
+ {
472
+ "epoch": 1.875,
473
+ "eval_accuracy": 0.928125,
474
+ "eval_loss": 0.43541350960731506,
475
+ "eval_runtime": 2.9558,
476
+ "eval_samples_per_second": 108.262,
477
+ "eval_steps_per_second": 54.131,
478
+ "step": 4500
479
+ },
480
+ {
481
+ "epoch": 1.9166666666666665,
482
+ "eval_accuracy": 0.91875,
483
+ "eval_loss": 0.4590047299861908,
484
+ "eval_runtime": 3.1134,
485
+ "eval_samples_per_second": 102.782,
486
+ "eval_steps_per_second": 51.391,
487
+ "step": 4600
488
+ },
489
+ {
490
+ "epoch": 1.9583333333333335,
491
+ "eval_accuracy": 0.921875,
492
+ "eval_loss": 0.42555299401283264,
493
+ "eval_runtime": 2.7577,
494
+ "eval_samples_per_second": 116.038,
495
+ "eval_steps_per_second": 58.019,
496
+ "step": 4700
497
+ },
498
+ {
499
+ "epoch": 2.0,
500
+ "eval_accuracy": 0.921875,
501
+ "eval_loss": 0.46792250871658325,
502
+ "eval_runtime": 3.199,
503
+ "eval_samples_per_second": 100.03,
504
+ "eval_steps_per_second": 50.015,
505
+ "step": 4800
506
+ },
507
+ {
508
+ "epoch": 2.0416666666666665,
509
+ "eval_accuracy": 0.934375,
510
+ "eval_loss": 0.40328049659729004,
511
+ "eval_runtime": 2.8034,
512
+ "eval_samples_per_second": 114.146,
513
+ "eval_steps_per_second": 57.073,
514
+ "step": 4900
515
+ },
516
+ {
517
+ "epoch": 2.0833333333333335,
518
+ "grad_norm": 0.021490707993507385,
519
+ "learning_rate": 1.9170833333333333e-05,
520
+ "loss": 0.2505,
521
+ "step": 5000
522
+ },
523
+ {
524
+ "epoch": 2.0833333333333335,
525
+ "eval_accuracy": 0.91875,
526
+ "eval_loss": 0.4490671753883362,
527
+ "eval_runtime": 2.8849,
528
+ "eval_samples_per_second": 110.924,
529
+ "eval_steps_per_second": 55.462,
530
+ "step": 5000
531
+ },
532
+ {
533
+ "epoch": 2.125,
534
+ "eval_accuracy": 0.925,
535
+ "eval_loss": 0.41796278953552246,
536
+ "eval_runtime": 3.1312,
537
+ "eval_samples_per_second": 102.196,
538
+ "eval_steps_per_second": 51.098,
539
+ "step": 5100
540
+ },
541
+ {
542
+ "epoch": 2.1666666666666665,
543
+ "eval_accuracy": 0.925,
544
+ "eval_loss": 0.3970945477485657,
545
+ "eval_runtime": 2.9082,
546
+ "eval_samples_per_second": 110.032,
547
+ "eval_steps_per_second": 55.016,
548
+ "step": 5200
549
+ },
550
+ {
551
+ "epoch": 2.2083333333333335,
552
+ "eval_accuracy": 0.94375,
553
+ "eval_loss": 0.3905588984489441,
554
+ "eval_runtime": 2.972,
555
+ "eval_samples_per_second": 107.671,
556
+ "eval_steps_per_second": 53.835,
557
+ "step": 5300
558
+ },
559
+ {
560
+ "epoch": 2.25,
561
+ "eval_accuracy": 0.9375,
562
+ "eval_loss": 0.3661550283432007,
563
+ "eval_runtime": 2.7487,
564
+ "eval_samples_per_second": 116.419,
565
+ "eval_steps_per_second": 58.21,
566
+ "step": 5400
567
+ },
568
+ {
569
+ "epoch": 2.2916666666666665,
570
+ "grad_norm": 0.016279272735118866,
571
+ "learning_rate": 1.7087500000000002e-05,
572
+ "loss": 0.1376,
573
+ "step": 5500
574
+ },
575
+ {
576
+ "epoch": 2.2916666666666665,
577
+ "eval_accuracy": 0.928125,
578
+ "eval_loss": 0.42674118280410767,
579
+ "eval_runtime": 2.824,
580
+ "eval_samples_per_second": 113.315,
581
+ "eval_steps_per_second": 56.658,
582
+ "step": 5500
583
+ },
584
+ {
585
+ "epoch": 2.3333333333333335,
586
+ "eval_accuracy": 0.93125,
587
+ "eval_loss": 0.47028741240501404,
588
+ "eval_runtime": 2.9991,
589
+ "eval_samples_per_second": 106.699,
590
+ "eval_steps_per_second": 53.35,
591
+ "step": 5600
592
+ },
593
+ {
594
+ "epoch": 2.375,
595
+ "eval_accuracy": 0.925,
596
+ "eval_loss": 0.48885369300842285,
597
+ "eval_runtime": 3.0056,
598
+ "eval_samples_per_second": 106.467,
599
+ "eval_steps_per_second": 53.233,
600
+ "step": 5700
601
+ },
602
+ {
603
+ "epoch": 2.4166666666666665,
604
+ "eval_accuracy": 0.925,
605
+ "eval_loss": 0.4400312006473541,
606
+ "eval_runtime": 2.7602,
607
+ "eval_samples_per_second": 115.934,
608
+ "eval_steps_per_second": 57.967,
609
+ "step": 5800
610
+ },
611
+ {
612
+ "epoch": 2.4583333333333335,
613
+ "eval_accuracy": 0.925,
614
+ "eval_loss": 0.4716396927833557,
615
+ "eval_runtime": 2.8432,
616
+ "eval_samples_per_second": 112.548,
617
+ "eval_steps_per_second": 56.274,
618
+ "step": 5900
619
+ },
620
+ {
621
+ "epoch": 2.5,
622
+ "grad_norm": 0.02649284340441227,
623
+ "learning_rate": 1.5004166666666668e-05,
624
+ "loss": 0.1487,
625
+ "step": 6000
626
+ },
627
+ {
628
+ "epoch": 2.5,
629
+ "eval_accuracy": 0.9125,
630
+ "eval_loss": 0.45153647661209106,
631
+ "eval_runtime": 2.8676,
632
+ "eval_samples_per_second": 111.59,
633
+ "eval_steps_per_second": 55.795,
634
+ "step": 6000
635
+ },
636
+ {
637
+ "epoch": 2.5416666666666665,
638
+ "eval_accuracy": 0.925,
639
+ "eval_loss": 0.45402488112449646,
640
+ "eval_runtime": 3.1616,
641
+ "eval_samples_per_second": 101.214,
642
+ "eval_steps_per_second": 50.607,
643
+ "step": 6100
644
+ },
645
+ {
646
+ "epoch": 2.5833333333333335,
647
+ "eval_accuracy": 0.909375,
648
+ "eval_loss": 0.5484554767608643,
649
+ "eval_runtime": 3.133,
650
+ "eval_samples_per_second": 102.137,
651
+ "eval_steps_per_second": 51.069,
652
+ "step": 6200
653
+ },
654
+ {
655
+ "epoch": 2.625,
656
+ "eval_accuracy": 0.915625,
657
+ "eval_loss": 0.541451632976532,
658
+ "eval_runtime": 2.8683,
659
+ "eval_samples_per_second": 111.564,
660
+ "eval_steps_per_second": 55.782,
661
+ "step": 6300
662
+ },
663
+ {
664
+ "epoch": 2.6666666666666665,
665
+ "eval_accuracy": 0.9,
666
+ "eval_loss": 0.619576096534729,
667
+ "eval_runtime": 2.8756,
668
+ "eval_samples_per_second": 111.281,
669
+ "eval_steps_per_second": 55.641,
670
+ "step": 6400
671
+ },
672
+ {
673
+ "epoch": 2.7083333333333335,
674
+ "grad_norm": 0.024532603099942207,
675
+ "learning_rate": 1.2920833333333333e-05,
676
+ "loss": 0.1248,
677
+ "step": 6500
678
+ },
679
+ {
680
+ "epoch": 2.7083333333333335,
681
+ "eval_accuracy": 0.93125,
682
+ "eval_loss": 0.4348299503326416,
683
+ "eval_runtime": 2.7745,
684
+ "eval_samples_per_second": 115.336,
685
+ "eval_steps_per_second": 57.668,
686
+ "step": 6500
687
+ },
688
+ {
689
+ "epoch": 2.75,
690
+ "eval_accuracy": 0.921875,
691
+ "eval_loss": 0.4980931878089905,
692
+ "eval_runtime": 2.9938,
693
+ "eval_samples_per_second": 106.888,
694
+ "eval_steps_per_second": 53.444,
695
+ "step": 6600
696
+ },
697
+ {
698
+ "epoch": 2.7916666666666665,
699
+ "eval_accuracy": 0.921875,
700
+ "eval_loss": 0.4891650080680847,
701
+ "eval_runtime": 2.9254,
702
+ "eval_samples_per_second": 109.387,
703
+ "eval_steps_per_second": 54.694,
704
+ "step": 6700
705
+ },
706
+ {
707
+ "epoch": 2.8333333333333335,
708
+ "eval_accuracy": 0.925,
709
+ "eval_loss": 0.5331800580024719,
710
+ "eval_runtime": 2.8311,
711
+ "eval_samples_per_second": 113.03,
712
+ "eval_steps_per_second": 56.515,
713
+ "step": 6800
714
+ },
715
+ {
716
+ "epoch": 2.875,
717
+ "eval_accuracy": 0.91875,
718
+ "eval_loss": 0.5215120315551758,
719
+ "eval_runtime": 2.7313,
720
+ "eval_samples_per_second": 117.159,
721
+ "eval_steps_per_second": 58.58,
722
+ "step": 6900
723
+ },
724
+ {
725
+ "epoch": 2.9166666666666665,
726
+ "grad_norm": 0.022552967071533203,
727
+ "learning_rate": 1.08375e-05,
728
+ "loss": 0.1273,
729
+ "step": 7000
730
+ },
731
+ {
732
+ "epoch": 2.9166666666666665,
733
+ "eval_accuracy": 0.921875,
734
+ "eval_loss": 0.5357473492622375,
735
+ "eval_runtime": 2.872,
736
+ "eval_samples_per_second": 111.419,
737
+ "eval_steps_per_second": 55.71,
738
+ "step": 7000
739
+ },
740
+ {
741
+ "epoch": 2.9583333333333335,
742
+ "eval_accuracy": 0.91875,
743
+ "eval_loss": 0.5796685218811035,
744
+ "eval_runtime": 3.1783,
745
+ "eval_samples_per_second": 100.682,
746
+ "eval_steps_per_second": 50.341,
747
+ "step": 7100
748
+ },
749
+ {
750
+ "epoch": 3.0,
751
+ "eval_accuracy": 0.921875,
752
+ "eval_loss": 0.4970870614051819,
753
+ "eval_runtime": 3.0019,
754
+ "eval_samples_per_second": 106.598,
755
+ "eval_steps_per_second": 53.299,
756
+ "step": 7200
757
+ },
758
+ {
759
+ "epoch": 3.0416666666666665,
760
+ "eval_accuracy": 0.925,
761
+ "eval_loss": 0.46576887369155884,
762
+ "eval_runtime": 2.83,
763
+ "eval_samples_per_second": 113.073,
764
+ "eval_steps_per_second": 56.536,
765
+ "step": 7300
766
+ },
767
+ {
768
+ "epoch": 3.0833333333333335,
769
+ "eval_accuracy": 0.91875,
770
+ "eval_loss": 0.46547237038612366,
771
+ "eval_runtime": 2.873,
772
+ "eval_samples_per_second": 111.381,
773
+ "eval_steps_per_second": 55.691,
774
+ "step": 7400
775
+ },
776
+ {
777
+ "epoch": 3.125,
778
+ "grad_norm": 0.019446508958935738,
779
+ "learning_rate": 8.754166666666667e-06,
780
+ "loss": 0.1225,
781
+ "step": 7500
782
+ },
783
+ {
784
+ "epoch": 3.125,
785
+ "eval_accuracy": 0.93125,
786
+ "eval_loss": 0.4488906264305115,
787
+ "eval_runtime": 2.7426,
788
+ "eval_samples_per_second": 116.676,
789
+ "eval_steps_per_second": 58.338,
790
+ "step": 7500
791
+ },
792
+ {
793
+ "epoch": 3.1666666666666665,
794
+ "eval_accuracy": 0.93125,
795
+ "eval_loss": 0.4806918203830719,
796
+ "eval_runtime": 2.884,
797
+ "eval_samples_per_second": 110.955,
798
+ "eval_steps_per_second": 55.478,
799
+ "step": 7600
800
+ },
801
+ {
802
+ "epoch": 3.2083333333333335,
803
+ "eval_accuracy": 0.90625,
804
+ "eval_loss": 0.5847142934799194,
805
+ "eval_runtime": 2.8473,
806
+ "eval_samples_per_second": 112.387,
807
+ "eval_steps_per_second": 56.193,
808
+ "step": 7700
809
+ },
810
+ {
811
+ "epoch": 3.25,
812
+ "eval_accuracy": 0.909375,
813
+ "eval_loss": 0.5970549583435059,
814
+ "eval_runtime": 3.0567,
815
+ "eval_samples_per_second": 104.688,
816
+ "eval_steps_per_second": 52.344,
817
+ "step": 7800
818
+ },
819
+ {
820
+ "epoch": 3.2916666666666665,
821
+ "eval_accuracy": 0.925,
822
+ "eval_loss": 0.4819498658180237,
823
+ "eval_runtime": 2.9925,
824
+ "eval_samples_per_second": 106.933,
825
+ "eval_steps_per_second": 53.467,
826
+ "step": 7900
827
+ },
828
+ {
829
+ "epoch": 3.3333333333333335,
830
+ "grad_norm": 0.016704484820365906,
831
+ "learning_rate": 6.670833333333334e-06,
832
+ "loss": 0.074,
833
+ "step": 8000
834
+ },
835
+ {
836
+ "epoch": 3.3333333333333335,
837
+ "eval_accuracy": 0.928125,
838
+ "eval_loss": 0.4904124140739441,
839
+ "eval_runtime": 3.0803,
840
+ "eval_samples_per_second": 103.886,
841
+ "eval_steps_per_second": 51.943,
842
+ "step": 8000
843
+ },
844
+ {
845
+ "epoch": 3.375,
846
+ "eval_accuracy": 0.928125,
847
+ "eval_loss": 0.4956900179386139,
848
+ "eval_runtime": 2.7659,
849
+ "eval_samples_per_second": 115.694,
850
+ "eval_steps_per_second": 57.847,
851
+ "step": 8100
852
+ },
853
+ {
854
+ "epoch": 3.4166666666666665,
855
+ "eval_accuracy": 0.90625,
856
+ "eval_loss": 0.5745713114738464,
857
+ "eval_runtime": 2.9216,
858
+ "eval_samples_per_second": 109.529,
859
+ "eval_steps_per_second": 54.764,
860
+ "step": 8200
861
+ },
862
+ {
863
+ "epoch": 3.4583333333333335,
864
+ "eval_accuracy": 0.915625,
865
+ "eval_loss": 0.5323331952095032,
866
+ "eval_runtime": 3.1843,
867
+ "eval_samples_per_second": 100.494,
868
+ "eval_steps_per_second": 50.247,
869
+ "step": 8300
870
+ },
871
+ {
872
+ "epoch": 3.5,
873
+ "eval_accuracy": 0.915625,
874
+ "eval_loss": 0.5545214414596558,
875
+ "eval_runtime": 3.0076,
876
+ "eval_samples_per_second": 106.398,
877
+ "eval_steps_per_second": 53.199,
878
+ "step": 8400
879
+ },
880
+ {
881
+ "epoch": 3.5416666666666665,
882
+ "grad_norm": 0.013942297548055649,
883
+ "learning_rate": 4.5875000000000005e-06,
884
+ "loss": 0.0776,
885
+ "step": 8500
886
+ },
887
+ {
888
+ "epoch": 3.5416666666666665,
889
+ "eval_accuracy": 0.9375,
890
+ "eval_loss": 0.43759164214134216,
891
+ "eval_runtime": 3.0869,
892
+ "eval_samples_per_second": 103.665,
893
+ "eval_steps_per_second": 51.832,
894
+ "step": 8500
895
+ },
896
+ {
897
+ "epoch": 3.5833333333333335,
898
+ "eval_accuracy": 0.9375,
899
+ "eval_loss": 0.43747997283935547,
900
+ "eval_runtime": 2.8407,
901
+ "eval_samples_per_second": 112.65,
902
+ "eval_steps_per_second": 56.325,
903
+ "step": 8600
904
+ },
905
+ {
906
+ "epoch": 3.625,
907
+ "eval_accuracy": 0.9375,
908
+ "eval_loss": 0.43234896659851074,
909
+ "eval_runtime": 2.8979,
910
+ "eval_samples_per_second": 110.425,
911
+ "eval_steps_per_second": 55.213,
912
+ "step": 8700
913
+ },
914
+ {
915
+ "epoch": 3.6666666666666665,
916
+ "eval_accuracy": 0.934375,
917
+ "eval_loss": 0.4293266832828522,
918
+ "eval_runtime": 3.1221,
919
+ "eval_samples_per_second": 102.496,
920
+ "eval_steps_per_second": 51.248,
921
+ "step": 8800
922
+ },
923
+ {
924
+ "epoch": 3.7083333333333335,
925
+ "eval_accuracy": 0.9375,
926
+ "eval_loss": 0.4287172853946686,
927
+ "eval_runtime": 2.9021,
928
+ "eval_samples_per_second": 110.265,
929
+ "eval_steps_per_second": 55.132,
930
+ "step": 8900
931
+ },
932
+ {
933
+ "epoch": 3.75,
934
+ "grad_norm": 0.009651360101997852,
935
+ "learning_rate": 2.504166666666667e-06,
936
+ "loss": 0.0718,
937
+ "step": 9000
938
+ },
939
+ {
940
+ "epoch": 3.75,
941
+ "eval_accuracy": 0.9375,
942
+ "eval_loss": 0.43642768263816833,
943
+ "eval_runtime": 2.9997,
944
+ "eval_samples_per_second": 106.679,
945
+ "eval_steps_per_second": 53.339,
946
+ "step": 9000
947
+ },
948
+ {
949
+ "epoch": 3.7916666666666665,
950
+ "eval_accuracy": 0.934375,
951
+ "eval_loss": 0.4443955421447754,
952
+ "eval_runtime": 2.7499,
953
+ "eval_samples_per_second": 116.37,
954
+ "eval_steps_per_second": 58.185,
955
+ "step": 9100
956
+ },
957
+ {
958
+ "epoch": 3.8333333333333335,
959
+ "eval_accuracy": 0.9375,
960
+ "eval_loss": 0.43916934728622437,
961
+ "eval_runtime": 2.8399,
962
+ "eval_samples_per_second": 112.68,
963
+ "eval_steps_per_second": 56.34,
964
+ "step": 9200
965
+ },
966
+ {
967
+ "epoch": 3.875,
968
+ "eval_accuracy": 0.934375,
969
+ "eval_loss": 0.44628801941871643,
970
+ "eval_runtime": 2.7737,
971
+ "eval_samples_per_second": 115.367,
972
+ "eval_steps_per_second": 57.684,
973
+ "step": 9300
974
+ },
975
+ {
976
+ "epoch": 3.9166666666666665,
977
+ "eval_accuracy": 0.9375,
978
+ "eval_loss": 0.4474868178367615,
979
+ "eval_runtime": 2.8338,
980
+ "eval_samples_per_second": 112.924,
981
+ "eval_steps_per_second": 56.462,
982
+ "step": 9400
983
+ },
984
+ {
985
+ "epoch": 3.9583333333333335,
986
+ "grad_norm": 0.013326543383300304,
987
+ "learning_rate": 4.2083333333333336e-07,
988
+ "loss": 0.0505,
989
+ "step": 9500
990
+ },
991
+ {
992
+ "epoch": 3.9583333333333335,
993
+ "eval_accuracy": 0.9375,
994
+ "eval_loss": 0.448885440826416,
995
+ "eval_runtime": 3.0868,
996
+ "eval_samples_per_second": 103.668,
997
+ "eval_steps_per_second": 51.834,
998
+ "step": 9500
999
+ },
1000
+ {
1001
+ "epoch": 4.0,
1002
+ "eval_accuracy": 0.9375,
1003
+ "eval_loss": 0.44648313522338867,
1004
+ "eval_runtime": 3.1112,
1005
+ "eval_samples_per_second": 102.855,
1006
+ "eval_steps_per_second": 51.427,
1007
+ "step": 9600
1008
+ }
1009
+ ],
1010
+ "logging_steps": 500,
1011
+ "max_steps": 9600,
1012
+ "num_input_tokens_seen": 0,
1013
+ "num_train_epochs": 4,
1014
+ "save_steps": 100,
1015
+ "stateful_callbacks": {
1016
+ "TrainerControl": {
1017
+ "args": {
1018
+ "should_epoch_stop": false,
1019
+ "should_evaluate": false,
1020
+ "should_log": false,
1021
+ "should_save": true,
1022
+ "should_training_stop": true
1023
+ },
1024
+ "attributes": {}
1025
+ }
1026
+ },
1027
+ "total_flos": 1303231196078040.0,
1028
+ "train_batch_size": 2,
1029
+ "trial_name": null,
1030
+ "trial_params": null
1031
+ }
checkpoints/checkpoint-9600/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65494a2338792a7eb98d28ceffbee46e49ecedf50dfe5b401590a45c130902ae
3
+ size 5304
checkpoints/checkpoint-9600/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
final_model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4841b3f5997fb9b7affa793d77470a9e90ed9513e37df98628f64b87fdb9449
3
+ size 438017556
final_model_meta.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"model_name": "google-bert/bert-base-uncased"}
test_outputs.jsonl ADDED
@@ -0,0 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"step": 100, "loss": 0.7828950881958008, "accuracy": 0.871875, "runtime": 2.8057, "samples_per_second": 114.053, "steps_per_second": 57.027}
2
+ {"step": 200, "loss": 0.6410558819770813, "accuracy": 0.88125, "runtime": 2.9454, "samples_per_second": 108.645, "steps_per_second": 54.323}
3
+ {"step": 300, "loss": 0.4724481701850891, "accuracy": 0.90625, "runtime": 2.6489, "samples_per_second": 120.804, "steps_per_second": 60.402}
4
+ {"step": 400, "loss": 0.46632274985313416, "accuracy": 0.88125, "runtime": 2.743, "samples_per_second": 116.66, "steps_per_second": 58.33}
5
+ {"step": 500, "loss": 0.336311399936676, "accuracy": 0.928125, "runtime": 2.7645, "samples_per_second": 115.753, "steps_per_second": 57.876}
6
+ {"step": 600, "loss": 0.5237163305282593, "accuracy": 0.875, "runtime": 2.9888, "samples_per_second": 107.065, "steps_per_second": 53.532}
7
+ {"step": 700, "loss": 0.612436056137085, "accuracy": 0.865625, "runtime": 3.0505, "samples_per_second": 104.901, "steps_per_second": 52.45}
8
+ {"step": 800, "loss": 0.47742635011672974, "accuracy": 0.9, "runtime": 2.8541, "samples_per_second": 112.12, "steps_per_second": 56.06}
9
+ {"step": 900, "loss": 0.4391939640045166, "accuracy": 0.925, "runtime": 2.7428, "samples_per_second": 116.667, "steps_per_second": 58.334}
10
+ {"step": 1000, "loss": 0.3529580235481262, "accuracy": 0.925, "runtime": 2.7396, "samples_per_second": 116.807, "steps_per_second": 58.404}
11
+ {"step": 1100, "loss": 0.32948821783065796, "accuracy": 0.925, "runtime": 3.0406, "samples_per_second": 105.242, "steps_per_second": 52.621}
12
+ {"step": 1200, "loss": 0.5077411532402039, "accuracy": 0.9, "runtime": 3.0926, "samples_per_second": 103.474, "steps_per_second": 51.737}
13
+ {"step": 1300, "loss": 0.4512642025947571, "accuracy": 0.9, "runtime": 2.7715, "samples_per_second": 115.459, "steps_per_second": 57.73}
14
+ {"step": 1400, "loss": 0.43066683411598206, "accuracy": 0.89375, "runtime": 2.8379, "samples_per_second": 112.76, "steps_per_second": 56.38}
15
+ {"step": 1500, "loss": 0.37902718782424927, "accuracy": 0.9125, "runtime": 2.8742, "samples_per_second": 111.336, "steps_per_second": 55.668}
16
+ {"step": 1600, "loss": 0.435788631439209, "accuracy": 0.8875, "runtime": 3.0691, "samples_per_second": 104.266, "steps_per_second": 52.133}
17
+ {"step": 1700, "loss": 0.6496487855911255, "accuracy": 0.878125, "runtime": 3.019, "samples_per_second": 105.995, "steps_per_second": 52.998}
18
+ {"step": 1800, "loss": 0.5356670618057251, "accuracy": 0.896875, "runtime": 2.7479, "samples_per_second": 116.452, "steps_per_second": 58.226}
19
+ {"step": 1900, "loss": 0.5577458143234253, "accuracy": 0.903125, "runtime": 2.7431, "samples_per_second": 116.657, "steps_per_second": 58.329}
20
+ {"step": 2000, "loss": 0.5010732412338257, "accuracy": 0.9, "runtime": 2.8418, "samples_per_second": 112.606, "steps_per_second": 56.303}
21
+ {"step": 2100, "loss": 0.5578997135162354, "accuracy": 0.871875, "runtime": 2.9909, "samples_per_second": 106.99, "steps_per_second": 53.495}
22
+ {"step": 2200, "loss": 0.49204739928245544, "accuracy": 0.9125, "runtime": 2.9421, "samples_per_second": 108.768, "steps_per_second": 54.384}
23
+ {"step": 2300, "loss": 0.42611804604530334, "accuracy": 0.921875, "runtime": 2.8842, "samples_per_second": 110.949, "steps_per_second": 55.475}
24
+ {"step": 2400, "loss": 0.5209008455276489, "accuracy": 0.903125, "runtime": 2.8447, "samples_per_second": 112.49, "steps_per_second": 56.245}
25
+ {"step": 2500, "loss": 0.558965802192688, "accuracy": 0.909375, "runtime": 3.0287, "samples_per_second": 105.655, "steps_per_second": 52.828}
26
+ {"step": 2600, "loss": 0.5337815284729004, "accuracy": 0.909375, "runtime": 2.8886, "samples_per_second": 110.781, "steps_per_second": 55.391}
27
+ {"step": 2700, "loss": 0.43373221158981323, "accuracy": 0.91875, "runtime": 2.927, "samples_per_second": 109.328, "steps_per_second": 54.664}
28
+ {"step": 2800, "loss": 0.562017560005188, "accuracy": 0.90625, "runtime": 3.1459, "samples_per_second": 101.72, "steps_per_second": 50.86}
29
+ {"step": 2900, "loss": 0.48360759019851685, "accuracy": 0.915625, "runtime": 2.7861, "samples_per_second": 114.855, "steps_per_second": 57.427}
30
+ {"step": 3000, "loss": 0.5819778442382812, "accuracy": 0.903125, "runtime": 2.7837, "samples_per_second": 114.956, "steps_per_second": 57.478}
31
+ {"step": 3100, "loss": 0.549329400062561, "accuracy": 0.90625, "runtime": 3.1117, "samples_per_second": 102.837, "steps_per_second": 51.418}
32
+ {"step": 3200, "loss": 0.5409368276596069, "accuracy": 0.909375, "runtime": 3.0275, "samples_per_second": 105.698, "steps_per_second": 52.849}
33
+ {"step": 3300, "loss": 0.5061095952987671, "accuracy": 0.91875, "runtime": 2.7533, "samples_per_second": 116.222, "steps_per_second": 58.111}
34
+ {"step": 3400, "loss": 0.5100991129875183, "accuracy": 0.915625, "runtime": 2.8582, "samples_per_second": 111.958, "steps_per_second": 55.979}
35
+ {"step": 3500, "loss": 0.48856163024902344, "accuracy": 0.91875, "runtime": 2.8545, "samples_per_second": 112.104, "steps_per_second": 56.052}
36
+ {"step": 3600, "loss": 0.5146762728691101, "accuracy": 0.91875, "runtime": 3.0905, "samples_per_second": 103.544, "steps_per_second": 51.772}
37
+ {"step": 3700, "loss": 0.5367563962936401, "accuracy": 0.915625, "runtime": 2.9095, "samples_per_second": 109.983, "steps_per_second": 54.992}
38
+ {"step": 3800, "loss": 0.4761125445365906, "accuracy": 0.909375, "runtime": 2.8026, "samples_per_second": 114.178, "steps_per_second": 57.089}
39
+ {"step": 3900, "loss": 0.4483864903450012, "accuracy": 0.90625, "runtime": 3.0492, "samples_per_second": 104.944, "steps_per_second": 52.472}
40
+ {"step": 4000, "loss": 0.35619640350341797, "accuracy": 0.928125, "runtime": 2.9753, "samples_per_second": 107.553, "steps_per_second": 53.777}
41
+ {"step": 4100, "loss": 0.38642579317092896, "accuracy": 0.921875, "runtime": 3.0163, "samples_per_second": 106.091, "steps_per_second": 53.046}
42
+ {"step": 4200, "loss": 0.5521766543388367, "accuracy": 0.90625, "runtime": 2.8851, "samples_per_second": 110.915, "steps_per_second": 55.457}
43
+ {"step": 4300, "loss": 0.4897506833076477, "accuracy": 0.890625, "runtime": 2.805, "samples_per_second": 114.082, "steps_per_second": 57.041}
44
+ {"step": 4400, "loss": 0.4268820881843567, "accuracy": 0.921875, "runtime": 3.0937, "samples_per_second": 103.437, "steps_per_second": 51.718}
45
+ {"step": 4500, "loss": 0.43541350960731506, "accuracy": 0.928125, "runtime": 2.9558, "samples_per_second": 108.262, "steps_per_second": 54.131}
46
+ {"step": 4600, "loss": 0.4590047299861908, "accuracy": 0.91875, "runtime": 3.1134, "samples_per_second": 102.782, "steps_per_second": 51.391}
47
+ {"step": 4700, "loss": 0.42555299401283264, "accuracy": 0.921875, "runtime": 2.7577, "samples_per_second": 116.038, "steps_per_second": 58.019}
48
+ {"step": 4800, "loss": 0.46792250871658325, "accuracy": 0.921875, "runtime": 3.199, "samples_per_second": 100.03, "steps_per_second": 50.015}
49
+ {"step": 4900, "loss": 0.40328049659729004, "accuracy": 0.934375, "runtime": 2.8034, "samples_per_second": 114.146, "steps_per_second": 57.073}
50
+ {"step": 5000, "loss": 0.4490671753883362, "accuracy": 0.91875, "runtime": 2.8849, "samples_per_second": 110.924, "steps_per_second": 55.462}
51
+ {"step": 5100, "loss": 0.41796278953552246, "accuracy": 0.925, "runtime": 3.1312, "samples_per_second": 102.196, "steps_per_second": 51.098}
52
+ {"step": 5200, "loss": 0.3970945477485657, "accuracy": 0.925, "runtime": 2.9082, "samples_per_second": 110.032, "steps_per_second": 55.016}
53
+ {"step": 5300, "loss": 0.3905588984489441, "accuracy": 0.94375, "runtime": 2.972, "samples_per_second": 107.671, "steps_per_second": 53.835}
54
+ {"step": 5400, "loss": 0.3661550283432007, "accuracy": 0.9375, "runtime": 2.7487, "samples_per_second": 116.419, "steps_per_second": 58.21}
55
+ {"step": 5500, "loss": 0.42674118280410767, "accuracy": 0.928125, "runtime": 2.824, "samples_per_second": 113.315, "steps_per_second": 56.658}
56
+ {"step": 5600, "loss": 0.47028741240501404, "accuracy": 0.93125, "runtime": 2.9991, "samples_per_second": 106.699, "steps_per_second": 53.35}
57
+ {"step": 5700, "loss": 0.48885369300842285, "accuracy": 0.925, "runtime": 3.0056, "samples_per_second": 106.467, "steps_per_second": 53.233}
58
+ {"step": 5800, "loss": 0.4400312006473541, "accuracy": 0.925, "runtime": 2.7602, "samples_per_second": 115.934, "steps_per_second": 57.967}
59
+ {"step": 5900, "loss": 0.4716396927833557, "accuracy": 0.925, "runtime": 2.8432, "samples_per_second": 112.548, "steps_per_second": 56.274}
60
+ {"step": 6000, "loss": 0.45153647661209106, "accuracy": 0.9125, "runtime": 2.8676, "samples_per_second": 111.59, "steps_per_second": 55.795}
61
+ {"step": 6100, "loss": 0.45402488112449646, "accuracy": 0.925, "runtime": 3.1616, "samples_per_second": 101.214, "steps_per_second": 50.607}
62
+ {"step": 6200, "loss": 0.5484554767608643, "accuracy": 0.909375, "runtime": 3.133, "samples_per_second": 102.137, "steps_per_second": 51.069}
63
+ {"step": 6300, "loss": 0.541451632976532, "accuracy": 0.915625, "runtime": 2.8683, "samples_per_second": 111.564, "steps_per_second": 55.782}
64
+ {"step": 6400, "loss": 0.619576096534729, "accuracy": 0.9, "runtime": 2.8756, "samples_per_second": 111.281, "steps_per_second": 55.641}
65
+ {"step": 6500, "loss": 0.4348299503326416, "accuracy": 0.93125, "runtime": 2.7745, "samples_per_second": 115.336, "steps_per_second": 57.668}
66
+ {"step": 6600, "loss": 0.4980931878089905, "accuracy": 0.921875, "runtime": 2.9938, "samples_per_second": 106.888, "steps_per_second": 53.444}
67
+ {"step": 6700, "loss": 0.4891650080680847, "accuracy": 0.921875, "runtime": 2.9254, "samples_per_second": 109.387, "steps_per_second": 54.694}
68
+ {"step": 6800, "loss": 0.5331800580024719, "accuracy": 0.925, "runtime": 2.8311, "samples_per_second": 113.03, "steps_per_second": 56.515}
69
+ {"step": 6900, "loss": 0.5215120315551758, "accuracy": 0.91875, "runtime": 2.7313, "samples_per_second": 117.159, "steps_per_second": 58.58}
70
+ {"step": 7000, "loss": 0.5357473492622375, "accuracy": 0.921875, "runtime": 2.872, "samples_per_second": 111.419, "steps_per_second": 55.71}
71
+ {"step": 7100, "loss": 0.5796685218811035, "accuracy": 0.91875, "runtime": 3.1783, "samples_per_second": 100.682, "steps_per_second": 50.341}
72
+ {"step": 7200, "loss": 0.4970870614051819, "accuracy": 0.921875, "runtime": 3.0019, "samples_per_second": 106.598, "steps_per_second": 53.299}
73
+ {"step": 7300, "loss": 0.46576887369155884, "accuracy": 0.925, "runtime": 2.83, "samples_per_second": 113.073, "steps_per_second": 56.536}
74
+ {"step": 7400, "loss": 0.46547237038612366, "accuracy": 0.91875, "runtime": 2.873, "samples_per_second": 111.381, "steps_per_second": 55.691}
75
+ {"step": 7500, "loss": 0.4488906264305115, "accuracy": 0.93125, "runtime": 2.7426, "samples_per_second": 116.676, "steps_per_second": 58.338}
76
+ {"step": 7600, "loss": 0.4806918203830719, "accuracy": 0.93125, "runtime": 2.884, "samples_per_second": 110.955, "steps_per_second": 55.478}
77
+ {"step": 7700, "loss": 0.5847142934799194, "accuracy": 0.90625, "runtime": 2.8473, "samples_per_second": 112.387, "steps_per_second": 56.193}
78
+ {"step": 7800, "loss": 0.5970549583435059, "accuracy": 0.909375, "runtime": 3.0567, "samples_per_second": 104.688, "steps_per_second": 52.344}
79
+ {"step": 7900, "loss": 0.4819498658180237, "accuracy": 0.925, "runtime": 2.9925, "samples_per_second": 106.933, "steps_per_second": 53.467}
80
+ {"step": 8000, "loss": 0.4904124140739441, "accuracy": 0.928125, "runtime": 3.0803, "samples_per_second": 103.886, "steps_per_second": 51.943}
81
+ {"step": 8100, "loss": 0.4956900179386139, "accuracy": 0.928125, "runtime": 2.7659, "samples_per_second": 115.694, "steps_per_second": 57.847}
82
+ {"step": 8200, "loss": 0.5745713114738464, "accuracy": 0.90625, "runtime": 2.9216, "samples_per_second": 109.529, "steps_per_second": 54.764}
83
+ {"step": 8300, "loss": 0.5323331952095032, "accuracy": 0.915625, "runtime": 3.1843, "samples_per_second": 100.494, "steps_per_second": 50.247}
84
+ {"step": 8400, "loss": 0.5545214414596558, "accuracy": 0.915625, "runtime": 3.0076, "samples_per_second": 106.398, "steps_per_second": 53.199}
85
+ {"step": 8500, "loss": 0.43759164214134216, "accuracy": 0.9375, "runtime": 3.0869, "samples_per_second": 103.665, "steps_per_second": 51.832}
86
+ {"step": 8600, "loss": 0.43747997283935547, "accuracy": 0.9375, "runtime": 2.8407, "samples_per_second": 112.65, "steps_per_second": 56.325}
87
+ {"step": 8700, "loss": 0.43234896659851074, "accuracy": 0.9375, "runtime": 2.8979, "samples_per_second": 110.425, "steps_per_second": 55.213}
88
+ {"step": 8800, "loss": 0.4293266832828522, "accuracy": 0.934375, "runtime": 3.1221, "samples_per_second": 102.496, "steps_per_second": 51.248}
89
+ {"step": 8900, "loss": 0.4287172853946686, "accuracy": 0.9375, "runtime": 2.9021, "samples_per_second": 110.265, "steps_per_second": 55.132}
90
+ {"step": 9000, "loss": 0.43642768263816833, "accuracy": 0.9375, "runtime": 2.9997, "samples_per_second": 106.679, "steps_per_second": 53.339}
91
+ {"step": 9100, "loss": 0.4443955421447754, "accuracy": 0.934375, "runtime": 2.7499, "samples_per_second": 116.37, "steps_per_second": 58.185}
92
+ {"step": 9200, "loss": 0.43916934728622437, "accuracy": 0.9375, "runtime": 2.8399, "samples_per_second": 112.68, "steps_per_second": 56.34}
93
+ {"step": 9300, "loss": 0.44628801941871643, "accuracy": 0.934375, "runtime": 2.7737, "samples_per_second": 115.367, "steps_per_second": 57.684}
94
+ {"step": 9400, "loss": 0.4474868178367615, "accuracy": 0.9375, "runtime": 2.8338, "samples_per_second": 112.924, "steps_per_second": 56.462}
95
+ {"step": 9500, "loss": 0.448885440826416, "accuracy": 0.9375, "runtime": 3.0868, "samples_per_second": 103.668, "steps_per_second": 51.834}
96
+ {"step": 9600, "loss": 0.44648313522338867, "accuracy": 0.9375, "runtime": 3.1112, "samples_per_second": 102.855, "steps_per_second": 51.427}