Hugging Face's logo

MarkKisker
/
RoBERTa-base-RottenTomatoes_v2

Text Classification
Transformers
TensorBoard
Safetensors
roberta
Inference Endpoints
Model card Files Metrics Community
MarkKisker commited on
Commit
9ec2f44
·
verified ·
1 Parent(s): cc735f6

Training in progress, epoch 1

Browse files
Files changed (18) hide show
  1. config.json +32 -0
  2. logs/events.out.tfevents.1709907916.DESKTOP-U0CPD9M.13936.1 +3 -0
  3. model.safetensors +3 -0
  4. tmp-checkpoint-1067/config.json +32 -0
  5. tmp-checkpoint-1067/model.safetensors +3 -0
  6. tmp-checkpoint-1067/optimizer.pt +3 -0
  7. tmp-checkpoint-1067/rng_state.pth +3 -0
  8. tmp-checkpoint-1067/scheduler.pt +3 -0
  9. tmp-checkpoint-1067/trainer_state.json +775 -0
  10. tmp-checkpoint-1067/training_args.bin +3 -0
  11. tmp-checkpoint-5335/config.json +32 -0
  12. tmp-checkpoint-5335/model.safetensors +3 -0
  13. tmp-checkpoint-5335/optimizer.pt +3 -0
  14. tmp-checkpoint-5335/rng_state.pth +3 -0
  15. tmp-checkpoint-5335/scheduler.pt +3 -0
  16. tmp-checkpoint-5335/trainer_state.json +3812 -0
  17. tmp-checkpoint-5335/training_args.bin +3 -0
  18. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base",
3
+ "architectures": [
4
+ "RobertaForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "neg",
15
+ "1": "pos"
16
+ },
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 3072,
19
+ "layer_norm_eps": 1e-05,
20
+ "max_position_embeddings": 514,
21
+ "model_type": "roberta",
22
+ "num_attention_heads": 12,
23
+ "num_hidden_layers": 12,
24
+ "pad_token_id": 1,
25
+ "position_embedding_type": "absolute",
26
+ "problem_type": "single_label_classification",
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.38.2",
29
+ "type_vocab_size": 1,
30
+ "use_cache": true,
31
+ "vocab_size": 50265
32
+ }
logs/events.out.tfevents.1709907916.DESKTOP-U0CPD9M.13936.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf3cb64657e95eb088e8a85eb54416c9430cf6771bd1864b2bd371bc559b5119
3
+ size 146918
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d8773cb8b07b3f97005edaf1661a5c9113d4377b12c6c726392e0e81468c80e
3
+ size 498612824
tmp-checkpoint-1067/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base",
3
+ "architectures": [
4
+ "RobertaForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "neg",
15
+ "1": "pos"
16
+ },
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 3072,
19
+ "layer_norm_eps": 1e-05,
20
+ "max_position_embeddings": 514,
21
+ "model_type": "roberta",
22
+ "num_attention_heads": 12,
23
+ "num_hidden_layers": 12,
24
+ "pad_token_id": 1,
25
+ "position_embedding_type": "absolute",
26
+ "problem_type": "single_label_classification",
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.38.2",
29
+ "type_vocab_size": 1,
30
+ "use_cache": true,
31
+ "vocab_size": 50265
32
+ }
tmp-checkpoint-1067/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d8773cb8b07b3f97005edaf1661a5c9113d4377b12c6c726392e0e81468c80e
3
+ size 498612824
tmp-checkpoint-1067/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fbac00785a0ee4936130c3d53f6930d52380220f3003afa61236cbb22e39fa55
3
+ size 997345530
tmp-checkpoint-1067/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d335a0c30b8fbab50a79e273ebee2f52a5ebd0b7b19052d2ece2b472cc7f677
3
+ size 14244
tmp-checkpoint-1067/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76578fd31f01f2bf7d4b6e42a46719f15dc06a3254eb11aa10709f32807be7f2
3
+ size 1064
tmp-checkpoint-1067/trainer_state.json ADDED
@@ -0,0 +1,775 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7780232429504395,
3
+ "best_model_checkpoint": "MarkKisker/RoBERTa-base-RottenTomatoes_v2\\checkpoint-1067",
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 1067,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "grad_norm": 0.03423001989722252,
14
+ "learning_rate": 1.0000000000000002e-06,
15
+ "loss": 0.0018,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.02,
20
+ "grad_norm": 0.04875793680548668,
21
+ "learning_rate": 2.0000000000000003e-06,
22
+ "loss": 0.0014,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.03,
27
+ "grad_norm": 20.525543212890625,
28
+ "learning_rate": 3e-06,
29
+ "loss": 0.1788,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.04,
34
+ "grad_norm": 0.02198684774339199,
35
+ "learning_rate": 4.000000000000001e-06,
36
+ "loss": 0.0014,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.05,
41
+ "grad_norm": 0.09893256425857544,
42
+ "learning_rate": 5e-06,
43
+ "loss": 0.1576,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.06,
48
+ "grad_norm": 0.04567793011665344,
49
+ "learning_rate": 6e-06,
50
+ "loss": 0.1448,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.07,
55
+ "grad_norm": 0.7623605728149414,
56
+ "learning_rate": 7.000000000000001e-06,
57
+ "loss": 0.0016,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.07,
62
+ "grad_norm": 0.023505745455622673,
63
+ "learning_rate": 8.000000000000001e-06,
64
+ "loss": 0.0017,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.08,
69
+ "grad_norm": 0.022804420441389084,
70
+ "learning_rate": 9e-06,
71
+ "loss": 0.1304,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.09,
76
+ "grad_norm": 0.03228422999382019,
77
+ "learning_rate": 1e-05,
78
+ "loss": 0.1446,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.1,
83
+ "grad_norm": 0.23935572803020477,
84
+ "learning_rate": 1.1000000000000001e-05,
85
+ "loss": 0.0425,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.11,
90
+ "grad_norm": 7.51793098449707,
91
+ "learning_rate": 1.2e-05,
92
+ "loss": 0.0024,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.12,
97
+ "grad_norm": 0.023124126717448235,
98
+ "learning_rate": 1.3000000000000001e-05,
99
+ "loss": 0.0228,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.13,
104
+ "grad_norm": 17.817167282104492,
105
+ "learning_rate": 1.4000000000000001e-05,
106
+ "loss": 0.0753,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.14,
111
+ "grad_norm": 0.015741823241114616,
112
+ "learning_rate": 1.5e-05,
113
+ "loss": 0.1112,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.15,
118
+ "grad_norm": 0.26574474573135376,
119
+ "learning_rate": 1.6000000000000003e-05,
120
+ "loss": 0.0092,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.16,
125
+ "grad_norm": 4.9772748947143555,
126
+ "learning_rate": 1.7000000000000003e-05,
127
+ "loss": 0.2384,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.17,
132
+ "grad_norm": 0.061539579182863235,
133
+ "learning_rate": 1.8e-05,
134
+ "loss": 0.0012,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.18,
139
+ "grad_norm": 0.08193587511777878,
140
+ "learning_rate": 1.9e-05,
141
+ "loss": 0.0812,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.19,
146
+ "grad_norm": 0.023329803720116615,
147
+ "learning_rate": 2e-05,
148
+ "loss": 0.1662,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 0.2,
153
+ "grad_norm": 0.031778186559677124,
154
+ "learning_rate": 2.1e-05,
155
+ "loss": 0.0013,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 0.21,
160
+ "grad_norm": 0.023606792092323303,
161
+ "learning_rate": 2.2000000000000003e-05,
162
+ "loss": 0.0547,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 0.22,
167
+ "grad_norm": 0.019992610439658165,
168
+ "learning_rate": 2.3000000000000003e-05,
169
+ "loss": 0.0011,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 0.22,
174
+ "grad_norm": 0.09242820739746094,
175
+ "learning_rate": 2.4e-05,
176
+ "loss": 0.0012,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 0.23,
181
+ "grad_norm": 0.01790749281644821,
182
+ "learning_rate": 2.5e-05,
183
+ "loss": 0.0803,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 0.24,
188
+ "grad_norm": 0.013409961014986038,
189
+ "learning_rate": 2.6000000000000002e-05,
190
+ "loss": 0.0633,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.25,
195
+ "grad_norm": 0.045994311571121216,
196
+ "learning_rate": 2.7000000000000002e-05,
197
+ "loss": 0.1399,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 0.26,
202
+ "grad_norm": 0.030622974038124084,
203
+ "learning_rate": 2.8000000000000003e-05,
204
+ "loss": 0.0015,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 0.27,
209
+ "grad_norm": 0.052308339625597,
210
+ "learning_rate": 2.9e-05,
211
+ "loss": 0.0852,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.28,
216
+ "grad_norm": 0.09171419590711594,
217
+ "learning_rate": 3e-05,
218
+ "loss": 0.0022,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 0.29,
223
+ "grad_norm": 0.024435508996248245,
224
+ "learning_rate": 3.1e-05,
225
+ "loss": 0.0859,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 0.3,
230
+ "grad_norm": 330.25262451171875,
231
+ "learning_rate": 3.2000000000000005e-05,
232
+ "loss": 0.0497,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 0.31,
237
+ "grad_norm": 0.021127384155988693,
238
+ "learning_rate": 3.3e-05,
239
+ "loss": 0.0818,
240
+ "step": 330
241
+ },
242
+ {
243
+ "epoch": 0.32,
244
+ "grad_norm": 0.030799318104982376,
245
+ "learning_rate": 3.4000000000000007e-05,
246
+ "loss": 0.0009,
247
+ "step": 340
248
+ },
249
+ {
250
+ "epoch": 0.33,
251
+ "grad_norm": 0.0609976127743721,
252
+ "learning_rate": 3.5e-05,
253
+ "loss": 0.0016,
254
+ "step": 350
255
+ },
256
+ {
257
+ "epoch": 0.34,
258
+ "grad_norm": 0.013851546682417393,
259
+ "learning_rate": 3.6e-05,
260
+ "loss": 0.2128,
261
+ "step": 360
262
+ },
263
+ {
264
+ "epoch": 0.35,
265
+ "grad_norm": 0.012113348580896854,
266
+ "learning_rate": 3.7e-05,
267
+ "loss": 0.119,
268
+ "step": 370
269
+ },
270
+ {
271
+ "epoch": 0.36,
272
+ "grad_norm": 0.0172914806753397,
273
+ "learning_rate": 3.8e-05,
274
+ "loss": 0.2769,
275
+ "step": 380
276
+ },
277
+ {
278
+ "epoch": 0.37,
279
+ "grad_norm": 0.047122351825237274,
280
+ "learning_rate": 3.9000000000000006e-05,
281
+ "loss": 0.0015,
282
+ "step": 390
283
+ },
284
+ {
285
+ "epoch": 0.37,
286
+ "grad_norm": 0.029489964246749878,
287
+ "learning_rate": 4e-05,
288
+ "loss": 0.0019,
289
+ "step": 400
290
+ },
291
+ {
292
+ "epoch": 0.38,
293
+ "grad_norm": 0.02585042454302311,
294
+ "learning_rate": 4.1e-05,
295
+ "loss": 0.0759,
296
+ "step": 410
297
+ },
298
+ {
299
+ "epoch": 0.39,
300
+ "grad_norm": 0.044562604278326035,
301
+ "learning_rate": 4.2e-05,
302
+ "loss": 0.0025,
303
+ "step": 420
304
+ },
305
+ {
306
+ "epoch": 0.4,
307
+ "grad_norm": 0.14535053074359894,
308
+ "learning_rate": 4.3e-05,
309
+ "loss": 0.0856,
310
+ "step": 430
311
+ },
312
+ {
313
+ "epoch": 0.41,
314
+ "grad_norm": 14.468594551086426,
315
+ "learning_rate": 4.4000000000000006e-05,
316
+ "loss": 0.1576,
317
+ "step": 440
318
+ },
319
+ {
320
+ "epoch": 0.42,
321
+ "grad_norm": 3.1556191444396973,
322
+ "learning_rate": 4.5e-05,
323
+ "loss": 0.1671,
324
+ "step": 450
325
+ },
326
+ {
327
+ "epoch": 0.43,
328
+ "grad_norm": 3.5952117443084717,
329
+ "learning_rate": 4.600000000000001e-05,
330
+ "loss": 0.1422,
331
+ "step": 460
332
+ },
333
+ {
334
+ "epoch": 0.44,
335
+ "grad_norm": 0.10417389869689941,
336
+ "learning_rate": 4.7e-05,
337
+ "loss": 0.0705,
338
+ "step": 470
339
+ },
340
+ {
341
+ "epoch": 0.45,
342
+ "grad_norm": 24.245695114135742,
343
+ "learning_rate": 4.8e-05,
344
+ "loss": 0.0586,
345
+ "step": 480
346
+ },
347
+ {
348
+ "epoch": 0.46,
349
+ "grad_norm": 0.6337321400642395,
350
+ "learning_rate": 4.9e-05,
351
+ "loss": 0.1252,
352
+ "step": 490
353
+ },
354
+ {
355
+ "epoch": 0.47,
356
+ "grad_norm": 62.256656646728516,
357
+ "learning_rate": 5e-05,
358
+ "loss": 0.1297,
359
+ "step": 500
360
+ },
361
+ {
362
+ "epoch": 0.48,
363
+ "grad_norm": 0.2806699573993683,
364
+ "learning_rate": 4.9896587383660806e-05,
365
+ "loss": 0.1789,
366
+ "step": 510
367
+ },
368
+ {
369
+ "epoch": 0.49,
370
+ "grad_norm": 0.05291756987571716,
371
+ "learning_rate": 4.9793174767321616e-05,
372
+ "loss": 0.1434,
373
+ "step": 520
374
+ },
375
+ {
376
+ "epoch": 0.5,
377
+ "grad_norm": 0.12910176813602448,
378
+ "learning_rate": 4.968976215098242e-05,
379
+ "loss": 0.2909,
380
+ "step": 530
381
+ },
382
+ {
383
+ "epoch": 0.51,
384
+ "grad_norm": 9.672201156616211,
385
+ "learning_rate": 4.958634953464323e-05,
386
+ "loss": 0.178,
387
+ "step": 540
388
+ },
389
+ {
390
+ "epoch": 0.52,
391
+ "grad_norm": 0.1485089659690857,
392
+ "learning_rate": 4.948293691830403e-05,
393
+ "loss": 0.0731,
394
+ "step": 550
395
+ },
396
+ {
397
+ "epoch": 0.52,
398
+ "grad_norm": 0.11018037796020508,
399
+ "learning_rate": 4.937952430196484e-05,
400
+ "loss": 0.1337,
401
+ "step": 560
402
+ },
403
+ {
404
+ "epoch": 0.53,
405
+ "grad_norm": 0.047798193991184235,
406
+ "learning_rate": 4.9276111685625646e-05,
407
+ "loss": 0.0028,
408
+ "step": 570
409
+ },
410
+ {
411
+ "epoch": 0.54,
412
+ "grad_norm": 3.2003190517425537,
413
+ "learning_rate": 4.9172699069286456e-05,
414
+ "loss": 0.2401,
415
+ "step": 580
416
+ },
417
+ {
418
+ "epoch": 0.55,
419
+ "grad_norm": 0.033508703112602234,
420
+ "learning_rate": 4.906928645294726e-05,
421
+ "loss": 0.057,
422
+ "step": 590
423
+ },
424
+ {
425
+ "epoch": 0.56,
426
+ "grad_norm": 0.07519116997718811,
427
+ "learning_rate": 4.896587383660807e-05,
428
+ "loss": 0.1506,
429
+ "step": 600
430
+ },
431
+ {
432
+ "epoch": 0.57,
433
+ "grad_norm": 0.10883668810129166,
434
+ "learning_rate": 4.886246122026887e-05,
435
+ "loss": 0.0799,
436
+ "step": 610
437
+ },
438
+ {
439
+ "epoch": 0.58,
440
+ "grad_norm": 3.15360689163208,
441
+ "learning_rate": 4.8759048603929683e-05,
442
+ "loss": 0.0904,
443
+ "step": 620
444
+ },
445
+ {
446
+ "epoch": 0.59,
447
+ "grad_norm": 2.191103219985962,
448
+ "learning_rate": 4.865563598759049e-05,
449
+ "loss": 0.0639,
450
+ "step": 630
451
+ },
452
+ {
453
+ "epoch": 0.6,
454
+ "grad_norm": 0.05860808119177818,
455
+ "learning_rate": 4.855222337125129e-05,
456
+ "loss": 0.2293,
457
+ "step": 640
458
+ },
459
+ {
460
+ "epoch": 0.61,
461
+ "grad_norm": 180.98397827148438,
462
+ "learning_rate": 4.84488107549121e-05,
463
+ "loss": 0.1808,
464
+ "step": 650
465
+ },
466
+ {
467
+ "epoch": 0.62,
468
+ "grad_norm": 127.35352325439453,
469
+ "learning_rate": 4.8345398138572904e-05,
470
+ "loss": 0.1102,
471
+ "step": 660
472
+ },
473
+ {
474
+ "epoch": 0.63,
475
+ "grad_norm": 0.03393542766571045,
476
+ "learning_rate": 4.8241985522233714e-05,
477
+ "loss": 0.2701,
478
+ "step": 670
479
+ },
480
+ {
481
+ "epoch": 0.64,
482
+ "grad_norm": 0.3284960985183716,
483
+ "learning_rate": 4.813857290589452e-05,
484
+ "loss": 0.0668,
485
+ "step": 680
486
+ },
487
+ {
488
+ "epoch": 0.65,
489
+ "grad_norm": 0.05796672776341438,
490
+ "learning_rate": 4.803516028955533e-05,
491
+ "loss": 0.1625,
492
+ "step": 690
493
+ },
494
+ {
495
+ "epoch": 0.66,
496
+ "grad_norm": 0.607434093952179,
497
+ "learning_rate": 4.793174767321613e-05,
498
+ "loss": 0.3021,
499
+ "step": 700
500
+ },
501
+ {
502
+ "epoch": 0.67,
503
+ "grad_norm": 0.1597072184085846,
504
+ "learning_rate": 4.782833505687694e-05,
505
+ "loss": 0.1627,
506
+ "step": 710
507
+ },
508
+ {
509
+ "epoch": 0.67,
510
+ "grad_norm": 0.0897730141878128,
511
+ "learning_rate": 4.772492244053775e-05,
512
+ "loss": 0.1379,
513
+ "step": 720
514
+ },
515
+ {
516
+ "epoch": 0.68,
517
+ "grad_norm": 70.59358978271484,
518
+ "learning_rate": 4.7621509824198554e-05,
519
+ "loss": 0.571,
520
+ "step": 730
521
+ },
522
+ {
523
+ "epoch": 0.69,
524
+ "grad_norm": 7.284711837768555,
525
+ "learning_rate": 4.7518097207859365e-05,
526
+ "loss": 0.4225,
527
+ "step": 740
528
+ },
529
+ {
530
+ "epoch": 0.7,
531
+ "grad_norm": 1.0496598482131958,
532
+ "learning_rate": 4.741468459152017e-05,
533
+ "loss": 0.0981,
534
+ "step": 750
535
+ },
536
+ {
537
+ "epoch": 0.71,
538
+ "grad_norm": 0.06985878944396973,
539
+ "learning_rate": 4.731127197518098e-05,
540
+ "loss": 0.1593,
541
+ "step": 760
542
+ },
543
+ {
544
+ "epoch": 0.72,
545
+ "grad_norm": 3.0135066509246826,
546
+ "learning_rate": 4.720785935884178e-05,
547
+ "loss": 0.3379,
548
+ "step": 770
549
+ },
550
+ {
551
+ "epoch": 0.73,
552
+ "grad_norm": 0.7335708141326904,
553
+ "learning_rate": 4.710444674250259e-05,
554
+ "loss": 0.4362,
555
+ "step": 780
556
+ },
557
+ {
558
+ "epoch": 0.74,
559
+ "grad_norm": 47.25736618041992,
560
+ "learning_rate": 4.7001034126163395e-05,
561
+ "loss": 0.3113,
562
+ "step": 790
563
+ },
564
+ {
565
+ "epoch": 0.75,
566
+ "grad_norm": 0.23570404946804047,
567
+ "learning_rate": 4.6897621509824205e-05,
568
+ "loss": 0.2591,
569
+ "step": 800
570
+ },
571
+ {
572
+ "epoch": 0.76,
573
+ "grad_norm": 1.6712744235992432,
574
+ "learning_rate": 4.679420889348501e-05,
575
+ "loss": 0.3156,
576
+ "step": 810
577
+ },
578
+ {
579
+ "epoch": 0.77,
580
+ "grad_norm": 3.803595781326294,
581
+ "learning_rate": 4.669079627714581e-05,
582
+ "loss": 0.2374,
583
+ "step": 820
584
+ },
585
+ {
586
+ "epoch": 0.78,
587
+ "grad_norm": 0.053344208747148514,
588
+ "learning_rate": 4.658738366080662e-05,
589
+ "loss": 0.0993,
590
+ "step": 830
591
+ },
592
+ {
593
+ "epoch": 0.79,
594
+ "grad_norm": 0.0863012745976448,
595
+ "learning_rate": 4.6483971044467425e-05,
596
+ "loss": 0.2164,
597
+ "step": 840
598
+ },
599
+ {
600
+ "epoch": 0.8,
601
+ "grad_norm": 50.73778533935547,
602
+ "learning_rate": 4.6380558428128236e-05,
603
+ "loss": 0.7427,
604
+ "step": 850
605
+ },
606
+ {
607
+ "epoch": 0.81,
608
+ "grad_norm": 1.3867453336715698,
609
+ "learning_rate": 4.627714581178904e-05,
610
+ "loss": 0.4358,
611
+ "step": 860
612
+ },
613
+ {
614
+ "epoch": 0.82,
615
+ "grad_norm": 2.2237284183502197,
616
+ "learning_rate": 4.617373319544985e-05,
617
+ "loss": 0.1522,
618
+ "step": 870
619
+ },
620
+ {
621
+ "epoch": 0.82,
622
+ "grad_norm": 189.51943969726562,
623
+ "learning_rate": 4.607032057911065e-05,
624
+ "loss": 0.3018,
625
+ "step": 880
626
+ },
627
+ {
628
+ "epoch": 0.83,
629
+ "grad_norm": 76.69792938232422,
630
+ "learning_rate": 4.596690796277146e-05,
631
+ "loss": 0.4078,
632
+ "step": 890
633
+ },
634
+ {
635
+ "epoch": 0.84,
636
+ "grad_norm": 11.872729301452637,
637
+ "learning_rate": 4.5863495346432266e-05,
638
+ "loss": 0.3047,
639
+ "step": 900
640
+ },
641
+ {
642
+ "epoch": 0.85,
643
+ "grad_norm": 2.8548457622528076,
644
+ "learning_rate": 4.5760082730093076e-05,
645
+ "loss": 0.5353,
646
+ "step": 910
647
+ },
648
+ {
649
+ "epoch": 0.86,
650
+ "grad_norm": 1.8873672485351562,
651
+ "learning_rate": 4.565667011375388e-05,
652
+ "loss": 0.1907,
653
+ "step": 920
654
+ },
655
+ {
656
+ "epoch": 0.87,
657
+ "grad_norm": 0.31615641713142395,
658
+ "learning_rate": 4.555325749741469e-05,
659
+ "loss": 0.0634,
660
+ "step": 930
661
+ },
662
+ {
663
+ "epoch": 0.88,
664
+ "grad_norm": 44.98077392578125,
665
+ "learning_rate": 4.544984488107549e-05,
666
+ "loss": 0.2581,
667
+ "step": 940
668
+ },
669
+ {
670
+ "epoch": 0.89,
671
+ "grad_norm": 0.426419734954834,
672
+ "learning_rate": 4.5346432264736296e-05,
673
+ "loss": 0.4992,
674
+ "step": 950
675
+ },
676
+ {
677
+ "epoch": 0.9,
678
+ "grad_norm": 0.7562969923019409,
679
+ "learning_rate": 4.5243019648397106e-05,
680
+ "loss": 0.3383,
681
+ "step": 960
682
+ },
683
+ {
684
+ "epoch": 0.91,
685
+ "grad_norm": 0.20380929112434387,
686
+ "learning_rate": 4.513960703205791e-05,
687
+ "loss": 0.1892,
688
+ "step": 970
689
+ },
690
+ {
691
+ "epoch": 0.92,
692
+ "grad_norm": 2.884596586227417,
693
+ "learning_rate": 4.503619441571872e-05,
694
+ "loss": 0.3536,
695
+ "step": 980
696
+ },
697
+ {
698
+ "epoch": 0.93,
699
+ "grad_norm": 14.720122337341309,
700
+ "learning_rate": 4.493278179937952e-05,
701
+ "loss": 0.4856,
702
+ "step": 990
703
+ },
704
+ {
705
+ "epoch": 0.94,
706
+ "grad_norm": 0.2174040526151657,
707
+ "learning_rate": 4.4829369183040333e-05,
708
+ "loss": 0.1159,
709
+ "step": 1000
710
+ },
711
+ {
712
+ "epoch": 0.95,
713
+ "grad_norm": 0.5567955374717712,
714
+ "learning_rate": 4.472595656670114e-05,
715
+ "loss": 0.5414,
716
+ "step": 1010
717
+ },
718
+ {
719
+ "epoch": 0.96,
720
+ "grad_norm": 0.8856528997421265,
721
+ "learning_rate": 4.462254395036195e-05,
722
+ "loss": 0.2257,
723
+ "step": 1020
724
+ },
725
+ {
726
+ "epoch": 0.97,
727
+ "grad_norm": 0.36583012342453003,
728
+ "learning_rate": 4.451913133402275e-05,
729
+ "loss": 0.0835,
730
+ "step": 1030
731
+ },
732
+ {
733
+ "epoch": 0.97,
734
+ "grad_norm": 2.9916436672210693,
735
+ "learning_rate": 4.441571871768356e-05,
736
+ "loss": 0.2258,
737
+ "step": 1040
738
+ },
739
+ {
740
+ "epoch": 0.98,
741
+ "grad_norm": 0.16345125436782837,
742
+ "learning_rate": 4.4312306101344364e-05,
743
+ "loss": 0.2466,
744
+ "step": 1050
745
+ },
746
+ {
747
+ "epoch": 0.99,
748
+ "grad_norm": 0.1895245909690857,
749
+ "learning_rate": 4.420889348500517e-05,
750
+ "loss": 0.4954,
751
+ "step": 1060
752
+ },
753
+ {
754
+ "epoch": 1.0,
755
+ "eval_accuracy": 0.8030018761726079,
756
+ "eval_f1": 0.8080438756855576,
757
+ "eval_loss": 0.7780232429504395,
758
+ "eval_precision": 0.7864768683274022,
759
+ "eval_recall": 0.8308270676691729,
760
+ "eval_runtime": 1.4434,
761
+ "eval_samples_per_second": 369.265,
762
+ "eval_steps_per_second": 46.418,
763
+ "step": 1067
764
+ }
765
+ ],
766
+ "logging_steps": 10,
767
+ "max_steps": 5335,
768
+ "num_input_tokens_seen": 0,
769
+ "num_train_epochs": 5,
770
+ "save_steps": 500,
771
+ "total_flos": 350677703472000.0,
772
+ "train_batch_size": 8,
773
+ "trial_name": null,
774
+ "trial_params": null
775
+ }
tmp-checkpoint-1067/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:845e2d48921e73f25ae02136aa40e96ecac5011c603a274ba4507a46d1806916
3
+ size 4984
tmp-checkpoint-5335/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base",
3
+ "architectures": [
4
+ "RobertaForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "neg",
15
+ "1": "pos"
16
+ },
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 3072,
19
+ "layer_norm_eps": 1e-05,
20
+ "max_position_embeddings": 514,
21
+ "model_type": "roberta",
22
+ "num_attention_heads": 12,
23
+ "num_hidden_layers": 12,
24
+ "pad_token_id": 1,
25
+ "position_embedding_type": "absolute",
26
+ "problem_type": "single_label_classification",
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.38.2",
29
+ "type_vocab_size": 1,
30
+ "use_cache": true,
31
+ "vocab_size": 50265
32
+ }
tmp-checkpoint-5335/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e67f896e8cbfa1f0d805804546478e28e3efa999578716e7c1023bcec72e750a
3
+ size 498612824
tmp-checkpoint-5335/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f20a4eee54d9a2b59194a4254989aa19711f9228074bec54d8f543e2820d3d74
3
+ size 997345530
tmp-checkpoint-5335/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ddb0ff9a10358fb591bb85ce5679f7ec7cf2ed23c4b0ca93a659c9f4ad7476a
3
+ size 14244
tmp-checkpoint-5335/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d853e56321f1267d767a3b750ac3b8bb40cf66f3f684730778d2e9dbe473afd8
3
+ size 1064
tmp-checkpoint-5335/trainer_state.json ADDED
@@ -0,0 +1,3812 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.3931359052658081,
3
+ "best_model_checkpoint": "MarkKisker/RoBERTa-base-RottenTomatoes_v2\\checkpoint-2134",
4
+ "epoch": 5.0,
5
+ "eval_steps": 500,
6
+ "global_step": 5335,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "grad_norm": 3.744861602783203,
14
+ "learning_rate": 1.0000000000000002e-06,
15
+ "loss": 0.7101,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.02,
20
+ "grad_norm": 4.64638090133667,
21
+ "learning_rate": 2.0000000000000003e-06,
22
+ "loss": 0.7158,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.03,
27
+ "grad_norm": 1.5597444772720337,
28
+ "learning_rate": 3e-06,
29
+ "loss": 0.6905,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.04,
34
+ "grad_norm": 2.208892345428467,
35
+ "learning_rate": 4.000000000000001e-06,
36
+ "loss": 0.6901,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.05,
41
+ "grad_norm": 4.314438819885254,
42
+ "learning_rate": 5e-06,
43
+ "loss": 0.6825,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.06,
48
+ "grad_norm": 3.1201136112213135,
49
+ "learning_rate": 6e-06,
50
+ "loss": 0.7093,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.07,
55
+ "grad_norm": 1.3445158004760742,
56
+ "learning_rate": 7.000000000000001e-06,
57
+ "loss": 0.691,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.07,
62
+ "grad_norm": 2.153444766998291,
63
+ "learning_rate": 8.000000000000001e-06,
64
+ "loss": 0.7013,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.08,
69
+ "grad_norm": 1.514472246170044,
70
+ "learning_rate": 9e-06,
71
+ "loss": 0.6843,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.09,
76
+ "grad_norm": 7.173806190490723,
77
+ "learning_rate": 1e-05,
78
+ "loss": 0.6641,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.1,
83
+ "grad_norm": 11.494062423706055,
84
+ "learning_rate": 1.1000000000000001e-05,
85
+ "loss": 0.5903,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.11,
90
+ "grad_norm": 28.968042373657227,
91
+ "learning_rate": 1.2e-05,
92
+ "loss": 0.4994,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.12,
97
+ "grad_norm": 44.565860748291016,
98
+ "learning_rate": 1.3000000000000001e-05,
99
+ "loss": 0.4509,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.13,
104
+ "grad_norm": 28.26981544494629,
105
+ "learning_rate": 1.4000000000000001e-05,
106
+ "loss": 0.4284,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.14,
111
+ "grad_norm": 8.198760032653809,
112
+ "learning_rate": 1.5e-05,
113
+ "loss": 0.4876,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.15,
118
+ "grad_norm": 50.528987884521484,
119
+ "learning_rate": 1.6000000000000003e-05,
120
+ "loss": 0.4672,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.16,
125
+ "grad_norm": 16.226484298706055,
126
+ "learning_rate": 1.7000000000000003e-05,
127
+ "loss": 0.6434,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.17,
132
+ "grad_norm": 14.364855766296387,
133
+ "learning_rate": 1.8e-05,
134
+ "loss": 0.5164,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.18,
139
+ "grad_norm": 7.223922252655029,
140
+ "learning_rate": 1.9e-05,
141
+ "loss": 0.4885,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.19,
146
+ "grad_norm": 8.672370910644531,
147
+ "learning_rate": 2e-05,
148
+ "loss": 0.4994,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 0.2,
153
+ "grad_norm": 52.54035949707031,
154
+ "learning_rate": 2.1e-05,
155
+ "loss": 0.462,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 0.21,
160
+ "grad_norm": 0.855399489402771,
161
+ "learning_rate": 2.2000000000000003e-05,
162
+ "loss": 0.217,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 0.22,
167
+ "grad_norm": 66.37751007080078,
168
+ "learning_rate": 2.3000000000000003e-05,
169
+ "loss": 0.6573,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 0.22,
174
+ "grad_norm": 6.453437805175781,
175
+ "learning_rate": 2.4e-05,
176
+ "loss": 0.5052,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 0.23,
181
+ "grad_norm": 7.456045627593994,
182
+ "learning_rate": 2.5e-05,
183
+ "loss": 0.3942,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 0.24,
188
+ "grad_norm": 16.87609100341797,
189
+ "learning_rate": 2.6000000000000002e-05,
190
+ "loss": 0.4712,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.25,
195
+ "grad_norm": 22.84710693359375,
196
+ "learning_rate": 2.7000000000000002e-05,
197
+ "loss": 0.4618,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 0.26,
202
+ "grad_norm": 1.8799303770065308,
203
+ "learning_rate": 2.8000000000000003e-05,
204
+ "loss": 0.59,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 0.27,
209
+ "grad_norm": 80.71158599853516,
210
+ "learning_rate": 2.9e-05,
211
+ "loss": 0.392,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.28,
216
+ "grad_norm": 2.51127290725708,
217
+ "learning_rate": 3e-05,
218
+ "loss": 0.4768,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 0.29,
223
+ "grad_norm": 4.434652805328369,
224
+ "learning_rate": 3.1e-05,
225
+ "loss": 0.3732,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 0.3,
230
+ "grad_norm": 12.18094253540039,
231
+ "learning_rate": 3.2000000000000005e-05,
232
+ "loss": 0.4689,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 0.31,
237
+ "grad_norm": 22.681018829345703,
238
+ "learning_rate": 3.3e-05,
239
+ "loss": 0.4172,
240
+ "step": 330
241
+ },
242
+ {
243
+ "epoch": 0.32,
244
+ "grad_norm": 12.466619491577148,
245
+ "learning_rate": 3.4000000000000007e-05,
246
+ "loss": 0.4661,
247
+ "step": 340
248
+ },
249
+ {
250
+ "epoch": 0.33,
251
+ "grad_norm": 7.371677398681641,
252
+ "learning_rate": 3.5e-05,
253
+ "loss": 0.3976,
254
+ "step": 350
255
+ },
256
+ {
257
+ "epoch": 0.34,
258
+ "grad_norm": 4.3304595947265625,
259
+ "learning_rate": 3.6e-05,
260
+ "loss": 0.6109,
261
+ "step": 360
262
+ },
263
+ {
264
+ "epoch": 0.35,
265
+ "grad_norm": 165.4158172607422,
266
+ "learning_rate": 3.7e-05,
267
+ "loss": 0.3678,
268
+ "step": 370
269
+ },
270
+ {
271
+ "epoch": 0.36,
272
+ "grad_norm": 4.8646016120910645,
273
+ "learning_rate": 3.8e-05,
274
+ "loss": 0.6374,
275
+ "step": 380
276
+ },
277
+ {
278
+ "epoch": 0.37,
279
+ "grad_norm": 1.7399684190750122,
280
+ "learning_rate": 3.9000000000000006e-05,
281
+ "loss": 0.4887,
282
+ "step": 390
283
+ },
284
+ {
285
+ "epoch": 0.37,
286
+ "grad_norm": 16.256572723388672,
287
+ "learning_rate": 4e-05,
288
+ "loss": 0.5702,
289
+ "step": 400
290
+ },
291
+ {
292
+ "epoch": 0.38,
293
+ "grad_norm": 45.110679626464844,
294
+ "learning_rate": 4.1e-05,
295
+ "loss": 0.5779,
296
+ "step": 410
297
+ },
298
+ {
299
+ "epoch": 0.39,
300
+ "grad_norm": 10.500941276550293,
301
+ "learning_rate": 4.2e-05,
302
+ "loss": 0.6049,
303
+ "step": 420
304
+ },
305
+ {
306
+ "epoch": 0.4,
307
+ "grad_norm": 46.1656494140625,
308
+ "learning_rate": 4.3e-05,
309
+ "loss": 0.5598,
310
+ "step": 430
311
+ },
312
+ {
313
+ "epoch": 0.41,
314
+ "grad_norm": 11.01528549194336,
315
+ "learning_rate": 4.4000000000000006e-05,
316
+ "loss": 0.4441,
317
+ "step": 440
318
+ },
319
+ {
320
+ "epoch": 0.42,
321
+ "grad_norm": 11.193294525146484,
322
+ "learning_rate": 4.5e-05,
323
+ "loss": 0.5585,
324
+ "step": 450
325
+ },
326
+ {
327
+ "epoch": 0.43,
328
+ "grad_norm": 18.2698917388916,
329
+ "learning_rate": 4.600000000000001e-05,
330
+ "loss": 0.4241,
331
+ "step": 460
332
+ },
333
+ {
334
+ "epoch": 0.44,
335
+ "grad_norm": 166.50296020507812,
336
+ "learning_rate": 4.7e-05,
337
+ "loss": 0.3354,
338
+ "step": 470
339
+ },
340
+ {
341
+ "epoch": 0.45,
342
+ "grad_norm": 128.95260620117188,
343
+ "learning_rate": 4.8e-05,
344
+ "loss": 0.5357,
345
+ "step": 480
346
+ },
347
+ {
348
+ "epoch": 0.46,
349
+ "grad_norm": 5.128959655761719,
350
+ "learning_rate": 4.9e-05,
351
+ "loss": 0.4072,
352
+ "step": 490
353
+ },
354
+ {
355
+ "epoch": 0.47,
356
+ "grad_norm": 3.268726348876953,
357
+ "learning_rate": 5e-05,
358
+ "loss": 0.474,
359
+ "step": 500
360
+ },
361
+ {
362
+ "epoch": 0.48,
363
+ "grad_norm": 37.61716842651367,
364
+ "learning_rate": 4.9896587383660806e-05,
365
+ "loss": 0.6837,
366
+ "step": 510
367
+ },
368
+ {
369
+ "epoch": 0.49,
370
+ "grad_norm": 5.7842302322387695,
371
+ "learning_rate": 4.9793174767321616e-05,
372
+ "loss": 0.4518,
373
+ "step": 520
374
+ },
375
+ {
376
+ "epoch": 0.5,
377
+ "grad_norm": 17.824024200439453,
378
+ "learning_rate": 4.968976215098242e-05,
379
+ "loss": 0.5325,
380
+ "step": 530
381
+ },
382
+ {
383
+ "epoch": 0.51,
384
+ "grad_norm": 13.727099418640137,
385
+ "learning_rate": 4.958634953464323e-05,
386
+ "loss": 0.4236,
387
+ "step": 540
388
+ },
389
+ {
390
+ "epoch": 0.52,
391
+ "grad_norm": 11.013121604919434,
392
+ "learning_rate": 4.948293691830403e-05,
393
+ "loss": 0.3925,
394
+ "step": 550
395
+ },
396
+ {
397
+ "epoch": 0.52,
398
+ "grad_norm": 10.758121490478516,
399
+ "learning_rate": 4.937952430196484e-05,
400
+ "loss": 0.4354,
401
+ "step": 560
402
+ },
403
+ {
404
+ "epoch": 0.53,
405
+ "grad_norm": 63.49281692504883,
406
+ "learning_rate": 4.9276111685625646e-05,
407
+ "loss": 0.456,
408
+ "step": 570
409
+ },
410
+ {
411
+ "epoch": 0.54,
412
+ "grad_norm": 14.087475776672363,
413
+ "learning_rate": 4.9172699069286456e-05,
414
+ "loss": 0.514,
415
+ "step": 580
416
+ },
417
+ {
418
+ "epoch": 0.55,
419
+ "grad_norm": 2.470353841781616,
420
+ "learning_rate": 4.906928645294726e-05,
421
+ "loss": 0.5102,
422
+ "step": 590
423
+ },
424
+ {
425
+ "epoch": 0.56,
426
+ "grad_norm": 43.60289001464844,
427
+ "learning_rate": 4.896587383660807e-05,
428
+ "loss": 0.4577,
429
+ "step": 600
430
+ },
431
+ {
432
+ "epoch": 0.57,
433
+ "grad_norm": 13.848435401916504,
434
+ "learning_rate": 4.886246122026887e-05,
435
+ "loss": 0.3769,
436
+ "step": 610
437
+ },
438
+ {
439
+ "epoch": 0.58,
440
+ "grad_norm": 85.09272003173828,
441
+ "learning_rate": 4.8759048603929683e-05,
442
+ "loss": 0.4745,
443
+ "step": 620
444
+ },
445
+ {
446
+ "epoch": 0.59,
447
+ "grad_norm": 2.0452070236206055,
448
+ "learning_rate": 4.865563598759049e-05,
449
+ "loss": 0.3618,
450
+ "step": 630
451
+ },
452
+ {
453
+ "epoch": 0.6,
454
+ "grad_norm": 5.641032695770264,
455
+ "learning_rate": 4.855222337125129e-05,
456
+ "loss": 0.5703,
457
+ "step": 640
458
+ },
459
+ {
460
+ "epoch": 0.61,
461
+ "grad_norm": 19.266847610473633,
462
+ "learning_rate": 4.84488107549121e-05,
463
+ "loss": 0.5826,
464
+ "step": 650
465
+ },
466
+ {
467
+ "epoch": 0.62,
468
+ "grad_norm": 73.6524658203125,
469
+ "learning_rate": 4.8345398138572904e-05,
470
+ "loss": 0.5455,
471
+ "step": 660
472
+ },
473
+ {
474
+ "epoch": 0.63,
475
+ "grad_norm": 1.602294683456421,
476
+ "learning_rate": 4.8241985522233714e-05,
477
+ "loss": 0.5497,
478
+ "step": 670
479
+ },
480
+ {
481
+ "epoch": 0.64,
482
+ "grad_norm": 22.117345809936523,
483
+ "learning_rate": 4.813857290589452e-05,
484
+ "loss": 0.4378,
485
+ "step": 680
486
+ },
487
+ {
488
+ "epoch": 0.65,
489
+ "grad_norm": 110.3412094116211,
490
+ "learning_rate": 4.803516028955533e-05,
491
+ "loss": 0.4393,
492
+ "step": 690
493
+ },
494
+ {
495
+ "epoch": 0.66,
496
+ "grad_norm": 3.794416904449463,
497
+ "learning_rate": 4.793174767321613e-05,
498
+ "loss": 0.5334,
499
+ "step": 700
500
+ },
501
+ {
502
+ "epoch": 0.67,
503
+ "grad_norm": 21.734609603881836,
504
+ "learning_rate": 4.782833505687694e-05,
505
+ "loss": 0.5833,
506
+ "step": 710
507
+ },
508
+ {
509
+ "epoch": 0.67,
510
+ "grad_norm": 15.466120719909668,
511
+ "learning_rate": 4.772492244053775e-05,
512
+ "loss": 0.6138,
513
+ "step": 720
514
+ },
515
+ {
516
+ "epoch": 0.68,
517
+ "grad_norm": 15.851093292236328,
518
+ "learning_rate": 4.7621509824198554e-05,
519
+ "loss": 0.474,
520
+ "step": 730
521
+ },
522
+ {
523
+ "epoch": 0.69,
524
+ "grad_norm": 11.810195922851562,
525
+ "learning_rate": 4.7518097207859365e-05,
526
+ "loss": 0.5048,
527
+ "step": 740
528
+ },
529
+ {
530
+ "epoch": 0.7,
531
+ "grad_norm": 7.658618450164795,
532
+ "learning_rate": 4.741468459152017e-05,
533
+ "loss": 0.3721,
534
+ "step": 750
535
+ },
536
+ {
537
+ "epoch": 0.71,
538
+ "grad_norm": 5.516314506530762,
539
+ "learning_rate": 4.731127197518098e-05,
540
+ "loss": 0.4057,
541
+ "step": 760
542
+ },
543
+ {
544
+ "epoch": 0.72,
545
+ "grad_norm": 29.502197265625,
546
+ "learning_rate": 4.720785935884178e-05,
547
+ "loss": 0.4228,
548
+ "step": 770
549
+ },
550
+ {
551
+ "epoch": 0.73,
552
+ "grad_norm": 5.364101409912109,
553
+ "learning_rate": 4.710444674250259e-05,
554
+ "loss": 0.6065,
555
+ "step": 780
556
+ },
557
+ {
558
+ "epoch": 0.74,
559
+ "grad_norm": 3.6527419090270996,
560
+ "learning_rate": 4.7001034126163395e-05,
561
+ "loss": 0.4702,
562
+ "step": 790
563
+ },
564
+ {
565
+ "epoch": 0.75,
566
+ "grad_norm": 234.80238342285156,
567
+ "learning_rate": 4.6897621509824205e-05,
568
+ "loss": 0.4136,
569
+ "step": 800
570
+ },
571
+ {
572
+ "epoch": 0.76,
573
+ "grad_norm": 3.4541990756988525,
574
+ "learning_rate": 4.679420889348501e-05,
575
+ "loss": 0.4633,
576
+ "step": 810
577
+ },
578
+ {
579
+ "epoch": 0.77,
580
+ "grad_norm": 10.891709327697754,
581
+ "learning_rate": 4.669079627714581e-05,
582
+ "loss": 0.4773,
583
+ "step": 820
584
+ },
585
+ {
586
+ "epoch": 0.78,
587
+ "grad_norm": 2.062467575073242,
588
+ "learning_rate": 4.658738366080662e-05,
589
+ "loss": 0.4229,
590
+ "step": 830
591
+ },
592
+ {
593
+ "epoch": 0.79,
594
+ "grad_norm": 1.5813032388687134,
595
+ "learning_rate": 4.6483971044467425e-05,
596
+ "loss": 0.4163,
597
+ "step": 840
598
+ },
599
+ {
600
+ "epoch": 0.8,
601
+ "grad_norm": 2.950200080871582,
602
+ "learning_rate": 4.6380558428128236e-05,
603
+ "loss": 0.4312,
604
+ "step": 850
605
+ },
606
+ {
607
+ "epoch": 0.81,
608
+ "grad_norm": 15.027694702148438,
609
+ "learning_rate": 4.627714581178904e-05,
610
+ "loss": 0.4475,
611
+ "step": 860
612
+ },
613
+ {
614
+ "epoch": 0.82,
615
+ "grad_norm": 6.54260778427124,
616
+ "learning_rate": 4.617373319544985e-05,
617
+ "loss": 0.4871,
618
+ "step": 870
619
+ },
620
+ {
621
+ "epoch": 0.82,
622
+ "grad_norm": 40.894622802734375,
623
+ "learning_rate": 4.607032057911065e-05,
624
+ "loss": 0.7644,
625
+ "step": 880
626
+ },
627
+ {
628
+ "epoch": 0.83,
629
+ "grad_norm": 40.53754425048828,
630
+ "learning_rate": 4.596690796277146e-05,
631
+ "loss": 0.6225,
632
+ "step": 890
633
+ },
634
+ {
635
+ "epoch": 0.84,
636
+ "grad_norm": 2.118955373764038,
637
+ "learning_rate": 4.5863495346432266e-05,
638
+ "loss": 0.38,
639
+ "step": 900
640
+ },
641
+ {
642
+ "epoch": 0.85,
643
+ "grad_norm": 4.862216472625732,
644
+ "learning_rate": 4.5760082730093076e-05,
645
+ "loss": 0.4339,
646
+ "step": 910
647
+ },
648
+ {
649
+ "epoch": 0.86,
650
+ "grad_norm": 32.316802978515625,
651
+ "learning_rate": 4.565667011375388e-05,
652
+ "loss": 0.6216,
653
+ "step": 920
654
+ },
655
+ {
656
+ "epoch": 0.87,
657
+ "grad_norm": 8.157034873962402,
658
+ "learning_rate": 4.555325749741469e-05,
659
+ "loss": 0.3952,
660
+ "step": 930
661
+ },
662
+ {
663
+ "epoch": 0.88,
664
+ "grad_norm": 34.675384521484375,
665
+ "learning_rate": 4.544984488107549e-05,
666
+ "loss": 0.4752,
667
+ "step": 940
668
+ },
669
+ {
670
+ "epoch": 0.89,
671
+ "grad_norm": 1.4986870288848877,
672
+ "learning_rate": 4.5346432264736296e-05,
673
+ "loss": 0.5754,
674
+ "step": 950
675
+ },
676
+ {
677
+ "epoch": 0.9,
678
+ "grad_norm": 5.362887859344482,
679
+ "learning_rate": 4.5243019648397106e-05,
680
+ "loss": 0.4518,
681
+ "step": 960
682
+ },
683
+ {
684
+ "epoch": 0.91,
685
+ "grad_norm": 8.203104972839355,
686
+ "learning_rate": 4.513960703205791e-05,
687
+ "loss": 0.4244,
688
+ "step": 970
689
+ },
690
+ {
691
+ "epoch": 0.92,
692
+ "grad_norm": 146.74620056152344,
693
+ "learning_rate": 4.503619441571872e-05,
694
+ "loss": 0.4484,
695
+ "step": 980
696
+ },
697
+ {
698
+ "epoch": 0.93,
699
+ "grad_norm": 38.64397048950195,
700
+ "learning_rate": 4.493278179937952e-05,
701
+ "loss": 0.5375,
702
+ "step": 990
703
+ },
704
+ {
705
+ "epoch": 0.94,
706
+ "grad_norm": 9.168529510498047,
707
+ "learning_rate": 4.4829369183040333e-05,
708
+ "loss": 0.4305,
709
+ "step": 1000
710
+ },
711
+ {
712
+ "epoch": 0.95,
713
+ "grad_norm": 11.166984558105469,
714
+ "learning_rate": 4.472595656670114e-05,
715
+ "loss": 0.5018,
716
+ "step": 1010
717
+ },
718
+ {
719
+ "epoch": 0.96,
720
+ "grad_norm": 2.078636646270752,
721
+ "learning_rate": 4.462254395036195e-05,
722
+ "loss": 0.5365,
723
+ "step": 1020
724
+ },
725
+ {
726
+ "epoch": 0.97,
727
+ "grad_norm": 2.5594077110290527,
728
+ "learning_rate": 4.451913133402275e-05,
729
+ "loss": 0.2867,
730
+ "step": 1030
731
+ },
732
+ {
733
+ "epoch": 0.97,
734
+ "grad_norm": 40.97541046142578,
735
+ "learning_rate": 4.441571871768356e-05,
736
+ "loss": 0.5615,
737
+ "step": 1040
738
+ },
739
+ {
740
+ "epoch": 0.98,
741
+ "grad_norm": 5.333548545837402,
742
+ "learning_rate": 4.4312306101344364e-05,
743
+ "loss": 0.371,
744
+ "step": 1050
745
+ },
746
+ {
747
+ "epoch": 0.99,
748
+ "grad_norm": 29.254602432250977,
749
+ "learning_rate": 4.420889348500517e-05,
750
+ "loss": 0.7758,
751
+ "step": 1060
752
+ },
753
+ {
754
+ "epoch": 1.0,
755
+ "eval_accuracy": 0.8180112570356473,
756
+ "eval_f1": 0.831304347826087,
757
+ "eval_loss": 0.4565158784389496,
758
+ "eval_precision": 0.7734627831715211,
759
+ "eval_recall": 0.8984962406015038,
760
+ "eval_runtime": 1.5762,
761
+ "eval_samples_per_second": 338.164,
762
+ "eval_steps_per_second": 42.508,
763
+ "step": 1067
764
+ },
765
+ {
766
+ "epoch": 1.0,
767
+ "grad_norm": 4.774388313293457,
768
+ "learning_rate": 4.410548086866598e-05,
769
+ "loss": 0.4325,
770
+ "step": 1070
771
+ },
772
+ {
773
+ "epoch": 1.01,
774
+ "grad_norm": 25.7803897857666,
775
+ "learning_rate": 4.400206825232678e-05,
776
+ "loss": 0.3303,
777
+ "step": 1080
778
+ },
779
+ {
780
+ "epoch": 1.02,
781
+ "grad_norm": 4.480411529541016,
782
+ "learning_rate": 4.389865563598759e-05,
783
+ "loss": 0.425,
784
+ "step": 1090
785
+ },
786
+ {
787
+ "epoch": 1.03,
788
+ "grad_norm": 2.4002604484558105,
789
+ "learning_rate": 4.3795243019648394e-05,
790
+ "loss": 0.386,
791
+ "step": 1100
792
+ },
793
+ {
794
+ "epoch": 1.04,
795
+ "grad_norm": 61.24464416503906,
796
+ "learning_rate": 4.369183040330921e-05,
797
+ "loss": 0.277,
798
+ "step": 1110
799
+ },
800
+ {
801
+ "epoch": 1.05,
802
+ "grad_norm": 25.068614959716797,
803
+ "learning_rate": 4.3588417786970015e-05,
804
+ "loss": 0.354,
805
+ "step": 1120
806
+ },
807
+ {
808
+ "epoch": 1.06,
809
+ "grad_norm": 1.0058590173721313,
810
+ "learning_rate": 4.3485005170630825e-05,
811
+ "loss": 0.3038,
812
+ "step": 1130
813
+ },
814
+ {
815
+ "epoch": 1.07,
816
+ "grad_norm": 408.1586608886719,
817
+ "learning_rate": 4.338159255429163e-05,
818
+ "loss": 0.5157,
819
+ "step": 1140
820
+ },
821
+ {
822
+ "epoch": 1.08,
823
+ "grad_norm": 6.401839256286621,
824
+ "learning_rate": 4.327817993795243e-05,
825
+ "loss": 0.4993,
826
+ "step": 1150
827
+ },
828
+ {
829
+ "epoch": 1.09,
830
+ "grad_norm": 6.1133317947387695,
831
+ "learning_rate": 4.317476732161324e-05,
832
+ "loss": 0.4571,
833
+ "step": 1160
834
+ },
835
+ {
836
+ "epoch": 1.1,
837
+ "grad_norm": 2.3999035358428955,
838
+ "learning_rate": 4.3071354705274045e-05,
839
+ "loss": 0.4561,
840
+ "step": 1170
841
+ },
842
+ {
843
+ "epoch": 1.11,
844
+ "grad_norm": 12.716032028198242,
845
+ "learning_rate": 4.2967942088934855e-05,
846
+ "loss": 0.2751,
847
+ "step": 1180
848
+ },
849
+ {
850
+ "epoch": 1.12,
851
+ "grad_norm": 36.668941497802734,
852
+ "learning_rate": 4.286452947259566e-05,
853
+ "loss": 0.328,
854
+ "step": 1190
855
+ },
856
+ {
857
+ "epoch": 1.12,
858
+ "grad_norm": 8.840494155883789,
859
+ "learning_rate": 4.276111685625647e-05,
860
+ "loss": 0.4708,
861
+ "step": 1200
862
+ },
863
+ {
864
+ "epoch": 1.13,
865
+ "grad_norm": 2.9924654960632324,
866
+ "learning_rate": 4.265770423991727e-05,
867
+ "loss": 0.2318,
868
+ "step": 1210
869
+ },
870
+ {
871
+ "epoch": 1.14,
872
+ "grad_norm": 0.4895036816596985,
873
+ "learning_rate": 4.255429162357808e-05,
874
+ "loss": 0.3037,
875
+ "step": 1220
876
+ },
877
+ {
878
+ "epoch": 1.15,
879
+ "grad_norm": 44.88699722290039,
880
+ "learning_rate": 4.2450879007238886e-05,
881
+ "loss": 0.3585,
882
+ "step": 1230
883
+ },
884
+ {
885
+ "epoch": 1.16,
886
+ "grad_norm": 5.701717376708984,
887
+ "learning_rate": 4.2347466390899696e-05,
888
+ "loss": 0.5978,
889
+ "step": 1240
890
+ },
891
+ {
892
+ "epoch": 1.17,
893
+ "grad_norm": 1.0451136827468872,
894
+ "learning_rate": 4.22440537745605e-05,
895
+ "loss": 0.8103,
896
+ "step": 1250
897
+ },
898
+ {
899
+ "epoch": 1.18,
900
+ "grad_norm": 2.085301399230957,
901
+ "learning_rate": 4.21406411582213e-05,
902
+ "loss": 0.3038,
903
+ "step": 1260
904
+ },
905
+ {
906
+ "epoch": 1.19,
907
+ "grad_norm": 66.31008911132812,
908
+ "learning_rate": 4.203722854188211e-05,
909
+ "loss": 0.5508,
910
+ "step": 1270
911
+ },
912
+ {
913
+ "epoch": 1.2,
914
+ "grad_norm": 19.349102020263672,
915
+ "learning_rate": 4.1933815925542916e-05,
916
+ "loss": 0.2065,
917
+ "step": 1280
918
+ },
919
+ {
920
+ "epoch": 1.21,
921
+ "grad_norm": 2.037997007369995,
922
+ "learning_rate": 4.1830403309203726e-05,
923
+ "loss": 0.4177,
924
+ "step": 1290
925
+ },
926
+ {
927
+ "epoch": 1.22,
928
+ "grad_norm": 0.6965357065200806,
929
+ "learning_rate": 4.172699069286453e-05,
930
+ "loss": 0.3867,
931
+ "step": 1300
932
+ },
933
+ {
934
+ "epoch": 1.23,
935
+ "grad_norm": 0.722122848033905,
936
+ "learning_rate": 4.162357807652534e-05,
937
+ "loss": 0.4882,
938
+ "step": 1310
939
+ },
940
+ {
941
+ "epoch": 1.24,
942
+ "grad_norm": 9.377044677734375,
943
+ "learning_rate": 4.152016546018614e-05,
944
+ "loss": 0.4661,
945
+ "step": 1320
946
+ },
947
+ {
948
+ "epoch": 1.25,
949
+ "grad_norm": 0.40828177332878113,
950
+ "learning_rate": 4.141675284384695e-05,
951
+ "loss": 0.4306,
952
+ "step": 1330
953
+ },
954
+ {
955
+ "epoch": 1.26,
956
+ "grad_norm": 4.25218391418457,
957
+ "learning_rate": 4.1313340227507756e-05,
958
+ "loss": 0.4858,
959
+ "step": 1340
960
+ },
961
+ {
962
+ "epoch": 1.27,
963
+ "grad_norm": 3.3388795852661133,
964
+ "learning_rate": 4.1209927611168567e-05,
965
+ "loss": 0.5039,
966
+ "step": 1350
967
+ },
968
+ {
969
+ "epoch": 1.27,
970
+ "grad_norm": 6.391818523406982,
971
+ "learning_rate": 4.110651499482937e-05,
972
+ "loss": 0.4893,
973
+ "step": 1360
974
+ },
975
+ {
976
+ "epoch": 1.28,
977
+ "grad_norm": 27.01968765258789,
978
+ "learning_rate": 4.100310237849018e-05,
979
+ "loss": 0.4502,
980
+ "step": 1370
981
+ },
982
+ {
983
+ "epoch": 1.29,
984
+ "grad_norm": 29.210132598876953,
985
+ "learning_rate": 4.0899689762150983e-05,
986
+ "loss": 0.3657,
987
+ "step": 1380
988
+ },
989
+ {
990
+ "epoch": 1.3,
991
+ "grad_norm": 54.26066970825195,
992
+ "learning_rate": 4.079627714581179e-05,
993
+ "loss": 0.2958,
994
+ "step": 1390
995
+ },
996
+ {
997
+ "epoch": 1.31,
998
+ "grad_norm": 36.26032257080078,
999
+ "learning_rate": 4.06928645294726e-05,
1000
+ "loss": 0.2768,
1001
+ "step": 1400
1002
+ },
1003
+ {
1004
+ "epoch": 1.32,
1005
+ "grad_norm": 1.0523189306259155,
1006
+ "learning_rate": 4.05894519131334e-05,
1007
+ "loss": 0.3634,
1008
+ "step": 1410
1009
+ },
1010
+ {
1011
+ "epoch": 1.33,
1012
+ "grad_norm": 191.5191192626953,
1013
+ "learning_rate": 4.048603929679421e-05,
1014
+ "loss": 0.3442,
1015
+ "step": 1420
1016
+ },
1017
+ {
1018
+ "epoch": 1.34,
1019
+ "grad_norm": 0.6843079924583435,
1020
+ "learning_rate": 4.0382626680455014e-05,
1021
+ "loss": 0.4025,
1022
+ "step": 1430
1023
+ },
1024
+ {
1025
+ "epoch": 1.35,
1026
+ "grad_norm": 0.2217019647359848,
1027
+ "learning_rate": 4.0279214064115824e-05,
1028
+ "loss": 0.2445,
1029
+ "step": 1440
1030
+ },
1031
+ {
1032
+ "epoch": 1.36,
1033
+ "grad_norm": 7.919774055480957,
1034
+ "learning_rate": 4.017580144777663e-05,
1035
+ "loss": 0.548,
1036
+ "step": 1450
1037
+ },
1038
+ {
1039
+ "epoch": 1.37,
1040
+ "grad_norm": 1.322029709815979,
1041
+ "learning_rate": 4.007238883143744e-05,
1042
+ "loss": 0.3518,
1043
+ "step": 1460
1044
+ },
1045
+ {
1046
+ "epoch": 1.38,
1047
+ "grad_norm": 0.4105088412761688,
1048
+ "learning_rate": 3.996897621509824e-05,
1049
+ "loss": 0.3859,
1050
+ "step": 1470
1051
+ },
1052
+ {
1053
+ "epoch": 1.39,
1054
+ "grad_norm": 0.31217288970947266,
1055
+ "learning_rate": 3.986556359875905e-05,
1056
+ "loss": 0.3287,
1057
+ "step": 1480
1058
+ },
1059
+ {
1060
+ "epoch": 1.4,
1061
+ "grad_norm": 39.67789077758789,
1062
+ "learning_rate": 3.9762150982419854e-05,
1063
+ "loss": 0.4649,
1064
+ "step": 1490
1065
+ },
1066
+ {
1067
+ "epoch": 1.41,
1068
+ "grad_norm": 17.59617042541504,
1069
+ "learning_rate": 3.965873836608066e-05,
1070
+ "loss": 0.4667,
1071
+ "step": 1500
1072
+ },
1073
+ {
1074
+ "epoch": 1.42,
1075
+ "grad_norm": 14.470635414123535,
1076
+ "learning_rate": 3.9555325749741475e-05,
1077
+ "loss": 0.4061,
1078
+ "step": 1510
1079
+ },
1080
+ {
1081
+ "epoch": 1.42,
1082
+ "grad_norm": 62.437618255615234,
1083
+ "learning_rate": 3.945191313340228e-05,
1084
+ "loss": 0.466,
1085
+ "step": 1520
1086
+ },
1087
+ {
1088
+ "epoch": 1.43,
1089
+ "grad_norm": 23.588979721069336,
1090
+ "learning_rate": 3.934850051706309e-05,
1091
+ "loss": 0.3333,
1092
+ "step": 1530
1093
+ },
1094
+ {
1095
+ "epoch": 1.44,
1096
+ "grad_norm": 3.491382598876953,
1097
+ "learning_rate": 3.924508790072389e-05,
1098
+ "loss": 0.2536,
1099
+ "step": 1540
1100
+ },
1101
+ {
1102
+ "epoch": 1.45,
1103
+ "grad_norm": 0.3138933479785919,
1104
+ "learning_rate": 3.91416752843847e-05,
1105
+ "loss": 0.3297,
1106
+ "step": 1550
1107
+ },
1108
+ {
1109
+ "epoch": 1.46,
1110
+ "grad_norm": 4.9050211906433105,
1111
+ "learning_rate": 3.9038262668045505e-05,
1112
+ "loss": 0.3643,
1113
+ "step": 1560
1114
+ },
1115
+ {
1116
+ "epoch": 1.47,
1117
+ "grad_norm": 13.311896324157715,
1118
+ "learning_rate": 3.8934850051706315e-05,
1119
+ "loss": 0.2224,
1120
+ "step": 1570
1121
+ },
1122
+ {
1123
+ "epoch": 1.48,
1124
+ "grad_norm": 136.0790252685547,
1125
+ "learning_rate": 3.883143743536712e-05,
1126
+ "loss": 0.5125,
1127
+ "step": 1580
1128
+ },
1129
+ {
1130
+ "epoch": 1.49,
1131
+ "grad_norm": 1.9536207914352417,
1132
+ "learning_rate": 3.872802481902792e-05,
1133
+ "loss": 0.3984,
1134
+ "step": 1590
1135
+ },
1136
+ {
1137
+ "epoch": 1.5,
1138
+ "grad_norm": 15.781469345092773,
1139
+ "learning_rate": 3.862461220268873e-05,
1140
+ "loss": 0.4071,
1141
+ "step": 1600
1142
+ },
1143
+ {
1144
+ "epoch": 1.51,
1145
+ "grad_norm": 4.010928153991699,
1146
+ "learning_rate": 3.8521199586349535e-05,
1147
+ "loss": 0.3088,
1148
+ "step": 1610
1149
+ },
1150
+ {
1151
+ "epoch": 1.52,
1152
+ "grad_norm": 15.730233192443848,
1153
+ "learning_rate": 3.8417786970010346e-05,
1154
+ "loss": 0.4905,
1155
+ "step": 1620
1156
+ },
1157
+ {
1158
+ "epoch": 1.53,
1159
+ "grad_norm": 18.30491065979004,
1160
+ "learning_rate": 3.831437435367115e-05,
1161
+ "loss": 0.366,
1162
+ "step": 1630
1163
+ },
1164
+ {
1165
+ "epoch": 1.54,
1166
+ "grad_norm": 0.39613863825798035,
1167
+ "learning_rate": 3.821096173733196e-05,
1168
+ "loss": 0.3205,
1169
+ "step": 1640
1170
+ },
1171
+ {
1172
+ "epoch": 1.55,
1173
+ "grad_norm": 51.20091247558594,
1174
+ "learning_rate": 3.810754912099276e-05,
1175
+ "loss": 0.5628,
1176
+ "step": 1650
1177
+ },
1178
+ {
1179
+ "epoch": 1.56,
1180
+ "grad_norm": 17.219799041748047,
1181
+ "learning_rate": 3.800413650465357e-05,
1182
+ "loss": 0.6222,
1183
+ "step": 1660
1184
+ },
1185
+ {
1186
+ "epoch": 1.57,
1187
+ "grad_norm": 41.373634338378906,
1188
+ "learning_rate": 3.7900723888314376e-05,
1189
+ "loss": 0.293,
1190
+ "step": 1670
1191
+ },
1192
+ {
1193
+ "epoch": 1.57,
1194
+ "grad_norm": 77.48905181884766,
1195
+ "learning_rate": 3.7797311271975186e-05,
1196
+ "loss": 0.3165,
1197
+ "step": 1680
1198
+ },
1199
+ {
1200
+ "epoch": 1.58,
1201
+ "grad_norm": 94.227783203125,
1202
+ "learning_rate": 3.769389865563599e-05,
1203
+ "loss": 0.455,
1204
+ "step": 1690
1205
+ },
1206
+ {
1207
+ "epoch": 1.59,
1208
+ "grad_norm": 21.350830078125,
1209
+ "learning_rate": 3.759048603929679e-05,
1210
+ "loss": 0.4369,
1211
+ "step": 1700
1212
+ },
1213
+ {
1214
+ "epoch": 1.6,
1215
+ "grad_norm": 17.99802589416504,
1216
+ "learning_rate": 3.74870734229576e-05,
1217
+ "loss": 0.4513,
1218
+ "step": 1710
1219
+ },
1220
+ {
1221
+ "epoch": 1.61,
1222
+ "grad_norm": 2.999746322631836,
1223
+ "learning_rate": 3.7383660806618406e-05,
1224
+ "loss": 0.3328,
1225
+ "step": 1720
1226
+ },
1227
+ {
1228
+ "epoch": 1.62,
1229
+ "grad_norm": 1.245676040649414,
1230
+ "learning_rate": 3.7280248190279217e-05,
1231
+ "loss": 0.5112,
1232
+ "step": 1730
1233
+ },
1234
+ {
1235
+ "epoch": 1.63,
1236
+ "grad_norm": 0.8955879211425781,
1237
+ "learning_rate": 3.717683557394002e-05,
1238
+ "loss": 0.2696,
1239
+ "step": 1740
1240
+ },
1241
+ {
1242
+ "epoch": 1.64,
1243
+ "grad_norm": 14.041428565979004,
1244
+ "learning_rate": 3.707342295760083e-05,
1245
+ "loss": 0.3887,
1246
+ "step": 1750
1247
+ },
1248
+ {
1249
+ "epoch": 1.65,
1250
+ "grad_norm": 10.70749568939209,
1251
+ "learning_rate": 3.6970010341261633e-05,
1252
+ "loss": 0.5087,
1253
+ "step": 1760
1254
+ },
1255
+ {
1256
+ "epoch": 1.66,
1257
+ "grad_norm": 26.841726303100586,
1258
+ "learning_rate": 3.6866597724922444e-05,
1259
+ "loss": 0.5083,
1260
+ "step": 1770
1261
+ },
1262
+ {
1263
+ "epoch": 1.67,
1264
+ "grad_norm": 9.196104049682617,
1265
+ "learning_rate": 3.676318510858325e-05,
1266
+ "loss": 0.3208,
1267
+ "step": 1780
1268
+ },
1269
+ {
1270
+ "epoch": 1.68,
1271
+ "grad_norm": 15.969416618347168,
1272
+ "learning_rate": 3.665977249224406e-05,
1273
+ "loss": 0.3257,
1274
+ "step": 1790
1275
+ },
1276
+ {
1277
+ "epoch": 1.69,
1278
+ "grad_norm": 27.856632232666016,
1279
+ "learning_rate": 3.655635987590486e-05,
1280
+ "loss": 0.3663,
1281
+ "step": 1800
1282
+ },
1283
+ {
1284
+ "epoch": 1.7,
1285
+ "grad_norm": 1.3687829971313477,
1286
+ "learning_rate": 3.645294725956567e-05,
1287
+ "loss": 0.3494,
1288
+ "step": 1810
1289
+ },
1290
+ {
1291
+ "epoch": 1.71,
1292
+ "grad_norm": 24.28046989440918,
1293
+ "learning_rate": 3.6349534643226474e-05,
1294
+ "loss": 0.4308,
1295
+ "step": 1820
1296
+ },
1297
+ {
1298
+ "epoch": 1.72,
1299
+ "grad_norm": 0.7270779013633728,
1300
+ "learning_rate": 3.624612202688728e-05,
1301
+ "loss": 0.3686,
1302
+ "step": 1830
1303
+ },
1304
+ {
1305
+ "epoch": 1.72,
1306
+ "grad_norm": 0.5241263508796692,
1307
+ "learning_rate": 3.614270941054809e-05,
1308
+ "loss": 0.3533,
1309
+ "step": 1840
1310
+ },
1311
+ {
1312
+ "epoch": 1.73,
1313
+ "grad_norm": 120.1414794921875,
1314
+ "learning_rate": 3.603929679420889e-05,
1315
+ "loss": 0.1999,
1316
+ "step": 1850
1317
+ },
1318
+ {
1319
+ "epoch": 1.74,
1320
+ "grad_norm": 26.968687057495117,
1321
+ "learning_rate": 3.59358841778697e-05,
1322
+ "loss": 0.3365,
1323
+ "step": 1860
1324
+ },
1325
+ {
1326
+ "epoch": 1.75,
1327
+ "grad_norm": 31.189796447753906,
1328
+ "learning_rate": 3.5832471561530504e-05,
1329
+ "loss": 0.5752,
1330
+ "step": 1870
1331
+ },
1332
+ {
1333
+ "epoch": 1.76,
1334
+ "grad_norm": 3.3193278312683105,
1335
+ "learning_rate": 3.5729058945191315e-05,
1336
+ "loss": 0.4337,
1337
+ "step": 1880
1338
+ },
1339
+ {
1340
+ "epoch": 1.77,
1341
+ "grad_norm": 17.27393913269043,
1342
+ "learning_rate": 3.562564632885212e-05,
1343
+ "loss": 0.4251,
1344
+ "step": 1890
1345
+ },
1346
+ {
1347
+ "epoch": 1.78,
1348
+ "grad_norm": 44.38993835449219,
1349
+ "learning_rate": 3.552223371251293e-05,
1350
+ "loss": 0.3813,
1351
+ "step": 1900
1352
+ },
1353
+ {
1354
+ "epoch": 1.79,
1355
+ "grad_norm": 14.854008674621582,
1356
+ "learning_rate": 3.541882109617373e-05,
1357
+ "loss": 0.349,
1358
+ "step": 1910
1359
+ },
1360
+ {
1361
+ "epoch": 1.8,
1362
+ "grad_norm": 6.35029935836792,
1363
+ "learning_rate": 3.531540847983454e-05,
1364
+ "loss": 0.3621,
1365
+ "step": 1920
1366
+ },
1367
+ {
1368
+ "epoch": 1.81,
1369
+ "grad_norm": 38.28048324584961,
1370
+ "learning_rate": 3.521199586349535e-05,
1371
+ "loss": 0.3143,
1372
+ "step": 1930
1373
+ },
1374
+ {
1375
+ "epoch": 1.82,
1376
+ "grad_norm": 3.274022340774536,
1377
+ "learning_rate": 3.5108583247156155e-05,
1378
+ "loss": 0.3147,
1379
+ "step": 1940
1380
+ },
1381
+ {
1382
+ "epoch": 1.83,
1383
+ "grad_norm": 24.270750045776367,
1384
+ "learning_rate": 3.5005170630816965e-05,
1385
+ "loss": 0.486,
1386
+ "step": 1950
1387
+ },
1388
+ {
1389
+ "epoch": 1.84,
1390
+ "grad_norm": 26.76897621154785,
1391
+ "learning_rate": 3.490175801447777e-05,
1392
+ "loss": 0.3353,
1393
+ "step": 1960
1394
+ },
1395
+ {
1396
+ "epoch": 1.85,
1397
+ "grad_norm": 28.255657196044922,
1398
+ "learning_rate": 3.479834539813858e-05,
1399
+ "loss": 0.3074,
1400
+ "step": 1970
1401
+ },
1402
+ {
1403
+ "epoch": 1.86,
1404
+ "grad_norm": 14.969733238220215,
1405
+ "learning_rate": 3.469493278179938e-05,
1406
+ "loss": 0.3648,
1407
+ "step": 1980
1408
+ },
1409
+ {
1410
+ "epoch": 1.87,
1411
+ "grad_norm": 13.978160858154297,
1412
+ "learning_rate": 3.459152016546019e-05,
1413
+ "loss": 0.3108,
1414
+ "step": 1990
1415
+ },
1416
+ {
1417
+ "epoch": 1.87,
1418
+ "grad_norm": 4.878652095794678,
1419
+ "learning_rate": 3.4488107549120996e-05,
1420
+ "loss": 0.3236,
1421
+ "step": 2000
1422
+ },
1423
+ {
1424
+ "epoch": 1.88,
1425
+ "grad_norm": 9.4730806350708,
1426
+ "learning_rate": 3.4384694932781806e-05,
1427
+ "loss": 0.3036,
1428
+ "step": 2010
1429
+ },
1430
+ {
1431
+ "epoch": 1.89,
1432
+ "grad_norm": 71.49247741699219,
1433
+ "learning_rate": 3.428128231644261e-05,
1434
+ "loss": 0.3228,
1435
+ "step": 2020
1436
+ },
1437
+ {
1438
+ "epoch": 1.9,
1439
+ "grad_norm": 11.899968147277832,
1440
+ "learning_rate": 3.417786970010341e-05,
1441
+ "loss": 0.422,
1442
+ "step": 2030
1443
+ },
1444
+ {
1445
+ "epoch": 1.91,
1446
+ "grad_norm": 10.609174728393555,
1447
+ "learning_rate": 3.407445708376422e-05,
1448
+ "loss": 0.2226,
1449
+ "step": 2040
1450
+ },
1451
+ {
1452
+ "epoch": 1.92,
1453
+ "grad_norm": 0.9701952338218689,
1454
+ "learning_rate": 3.3971044467425026e-05,
1455
+ "loss": 0.3561,
1456
+ "step": 2050
1457
+ },
1458
+ {
1459
+ "epoch": 1.93,
1460
+ "grad_norm": 16.32933235168457,
1461
+ "learning_rate": 3.3867631851085836e-05,
1462
+ "loss": 0.3124,
1463
+ "step": 2060
1464
+ },
1465
+ {
1466
+ "epoch": 1.94,
1467
+ "grad_norm": 1.6818139553070068,
1468
+ "learning_rate": 3.376421923474664e-05,
1469
+ "loss": 0.274,
1470
+ "step": 2070
1471
+ },
1472
+ {
1473
+ "epoch": 1.95,
1474
+ "grad_norm": 2.9835000038146973,
1475
+ "learning_rate": 3.366080661840745e-05,
1476
+ "loss": 0.3897,
1477
+ "step": 2080
1478
+ },
1479
+ {
1480
+ "epoch": 1.96,
1481
+ "grad_norm": 0.44764024019241333,
1482
+ "learning_rate": 3.355739400206825e-05,
1483
+ "loss": 0.3541,
1484
+ "step": 2090
1485
+ },
1486
+ {
1487
+ "epoch": 1.97,
1488
+ "grad_norm": 34.775390625,
1489
+ "learning_rate": 3.345398138572906e-05,
1490
+ "loss": 0.3058,
1491
+ "step": 2100
1492
+ },
1493
+ {
1494
+ "epoch": 1.98,
1495
+ "grad_norm": 10.107741355895996,
1496
+ "learning_rate": 3.3350568769389867e-05,
1497
+ "loss": 0.2805,
1498
+ "step": 2110
1499
+ },
1500
+ {
1501
+ "epoch": 1.99,
1502
+ "grad_norm": 9.582823753356934,
1503
+ "learning_rate": 3.324715615305068e-05,
1504
+ "loss": 0.3837,
1505
+ "step": 2120
1506
+ },
1507
+ {
1508
+ "epoch": 2.0,
1509
+ "grad_norm": 86.71916198730469,
1510
+ "learning_rate": 3.314374353671148e-05,
1511
+ "loss": 0.4249,
1512
+ "step": 2130
1513
+ },
1514
+ {
1515
+ "epoch": 2.0,
1516
+ "eval_accuracy": 0.8555347091932458,
1517
+ "eval_f1": 0.8560747663551402,
1518
+ "eval_loss": 0.3931359052658081,
1519
+ "eval_precision": 0.8513011152416357,
1520
+ "eval_recall": 0.8609022556390977,
1521
+ "eval_runtime": 1.5356,
1522
+ "eval_samples_per_second": 347.106,
1523
+ "eval_steps_per_second": 43.632,
1524
+ "step": 2134
1525
+ },
1526
+ {
1527
+ "epoch": 2.01,
1528
+ "grad_norm": 2.3712680339813232,
1529
+ "learning_rate": 3.3040330920372283e-05,
1530
+ "loss": 0.1953,
1531
+ "step": 2140
1532
+ },
1533
+ {
1534
+ "epoch": 2.01,
1535
+ "grad_norm": 28.71424102783203,
1536
+ "learning_rate": 3.2936918304033094e-05,
1537
+ "loss": 0.1712,
1538
+ "step": 2150
1539
+ },
1540
+ {
1541
+ "epoch": 2.02,
1542
+ "grad_norm": 0.18505160510540009,
1543
+ "learning_rate": 3.28335056876939e-05,
1544
+ "loss": 0.3299,
1545
+ "step": 2160
1546
+ },
1547
+ {
1548
+ "epoch": 2.03,
1549
+ "grad_norm": 74.10963439941406,
1550
+ "learning_rate": 3.273009307135471e-05,
1551
+ "loss": 0.2948,
1552
+ "step": 2170
1553
+ },
1554
+ {
1555
+ "epoch": 2.04,
1556
+ "grad_norm": 1.3842068910598755,
1557
+ "learning_rate": 3.262668045501551e-05,
1558
+ "loss": 0.3453,
1559
+ "step": 2180
1560
+ },
1561
+ {
1562
+ "epoch": 2.05,
1563
+ "grad_norm": 10.536425590515137,
1564
+ "learning_rate": 3.252326783867632e-05,
1565
+ "loss": 0.23,
1566
+ "step": 2190
1567
+ },
1568
+ {
1569
+ "epoch": 2.06,
1570
+ "grad_norm": 28.119539260864258,
1571
+ "learning_rate": 3.2419855222337124e-05,
1572
+ "loss": 0.2037,
1573
+ "step": 2200
1574
+ },
1575
+ {
1576
+ "epoch": 2.07,
1577
+ "grad_norm": 0.34018927812576294,
1578
+ "learning_rate": 3.2316442605997934e-05,
1579
+ "loss": 0.1796,
1580
+ "step": 2210
1581
+ },
1582
+ {
1583
+ "epoch": 2.08,
1584
+ "grad_norm": 2.8183701038360596,
1585
+ "learning_rate": 3.221302998965874e-05,
1586
+ "loss": 0.5003,
1587
+ "step": 2220
1588
+ },
1589
+ {
1590
+ "epoch": 2.09,
1591
+ "grad_norm": 0.478957861661911,
1592
+ "learning_rate": 3.210961737331955e-05,
1593
+ "loss": 0.14,
1594
+ "step": 2230
1595
+ },
1596
+ {
1597
+ "epoch": 2.1,
1598
+ "grad_norm": 2.9625442028045654,
1599
+ "learning_rate": 3.200620475698035e-05,
1600
+ "loss": 0.3229,
1601
+ "step": 2240
1602
+ },
1603
+ {
1604
+ "epoch": 2.11,
1605
+ "grad_norm": 0.5814473032951355,
1606
+ "learning_rate": 3.190279214064116e-05,
1607
+ "loss": 0.1369,
1608
+ "step": 2250
1609
+ },
1610
+ {
1611
+ "epoch": 2.12,
1612
+ "grad_norm": 1.5362516641616821,
1613
+ "learning_rate": 3.1799379524301965e-05,
1614
+ "loss": 0.1266,
1615
+ "step": 2260
1616
+ },
1617
+ {
1618
+ "epoch": 2.13,
1619
+ "grad_norm": 30.91729736328125,
1620
+ "learning_rate": 3.169596690796277e-05,
1621
+ "loss": 0.4261,
1622
+ "step": 2270
1623
+ },
1624
+ {
1625
+ "epoch": 2.14,
1626
+ "grad_norm": 6.7020673751831055,
1627
+ "learning_rate": 3.159255429162358e-05,
1628
+ "loss": 0.3651,
1629
+ "step": 2280
1630
+ },
1631
+ {
1632
+ "epoch": 2.15,
1633
+ "grad_norm": 3.024595022201538,
1634
+ "learning_rate": 3.148914167528438e-05,
1635
+ "loss": 0.2789,
1636
+ "step": 2290
1637
+ },
1638
+ {
1639
+ "epoch": 2.16,
1640
+ "grad_norm": 387.9389953613281,
1641
+ "learning_rate": 3.138572905894519e-05,
1642
+ "loss": 0.4301,
1643
+ "step": 2300
1644
+ },
1645
+ {
1646
+ "epoch": 2.16,
1647
+ "grad_norm": 0.2992786467075348,
1648
+ "learning_rate": 3.1282316442605995e-05,
1649
+ "loss": 0.276,
1650
+ "step": 2310
1651
+ },
1652
+ {
1653
+ "epoch": 2.17,
1654
+ "grad_norm": 4.288240432739258,
1655
+ "learning_rate": 3.117890382626681e-05,
1656
+ "loss": 0.0722,
1657
+ "step": 2320
1658
+ },
1659
+ {
1660
+ "epoch": 2.18,
1661
+ "grad_norm": 0.20439238846302032,
1662
+ "learning_rate": 3.1075491209927615e-05,
1663
+ "loss": 0.1687,
1664
+ "step": 2330
1665
+ },
1666
+ {
1667
+ "epoch": 2.19,
1668
+ "grad_norm": 0.17426259815692902,
1669
+ "learning_rate": 3.097207859358842e-05,
1670
+ "loss": 0.3444,
1671
+ "step": 2340
1672
+ },
1673
+ {
1674
+ "epoch": 2.2,
1675
+ "grad_norm": 2.802089214324951,
1676
+ "learning_rate": 3.086866597724923e-05,
1677
+ "loss": 0.4715,
1678
+ "step": 2350
1679
+ },
1680
+ {
1681
+ "epoch": 2.21,
1682
+ "grad_norm": 0.27714866399765015,
1683
+ "learning_rate": 3.076525336091003e-05,
1684
+ "loss": 0.2581,
1685
+ "step": 2360
1686
+ },
1687
+ {
1688
+ "epoch": 2.22,
1689
+ "grad_norm": 9.906682014465332,
1690
+ "learning_rate": 3.066184074457084e-05,
1691
+ "loss": 0.1077,
1692
+ "step": 2370
1693
+ },
1694
+ {
1695
+ "epoch": 2.23,
1696
+ "grad_norm": 17.114538192749023,
1697
+ "learning_rate": 3.0558428128231646e-05,
1698
+ "loss": 0.3108,
1699
+ "step": 2380
1700
+ },
1701
+ {
1702
+ "epoch": 2.24,
1703
+ "grad_norm": 5.7600178718566895,
1704
+ "learning_rate": 3.0455015511892452e-05,
1705
+ "loss": 0.3055,
1706
+ "step": 2390
1707
+ },
1708
+ {
1709
+ "epoch": 2.25,
1710
+ "grad_norm": 0.660796582698822,
1711
+ "learning_rate": 3.035160289555326e-05,
1712
+ "loss": 0.1998,
1713
+ "step": 2400
1714
+ },
1715
+ {
1716
+ "epoch": 2.26,
1717
+ "grad_norm": 0.19979248940944672,
1718
+ "learning_rate": 3.0248190279214066e-05,
1719
+ "loss": 0.3523,
1720
+ "step": 2410
1721
+ },
1722
+ {
1723
+ "epoch": 2.27,
1724
+ "grad_norm": 0.6000668406486511,
1725
+ "learning_rate": 3.0144777662874873e-05,
1726
+ "loss": 0.1009,
1727
+ "step": 2420
1728
+ },
1729
+ {
1730
+ "epoch": 2.28,
1731
+ "grad_norm": 0.15889067947864532,
1732
+ "learning_rate": 3.004136504653568e-05,
1733
+ "loss": 0.0861,
1734
+ "step": 2430
1735
+ },
1736
+ {
1737
+ "epoch": 2.29,
1738
+ "grad_norm": 0.32534798979759216,
1739
+ "learning_rate": 2.9937952430196486e-05,
1740
+ "loss": 0.1119,
1741
+ "step": 2440
1742
+ },
1743
+ {
1744
+ "epoch": 2.3,
1745
+ "grad_norm": 6.747263431549072,
1746
+ "learning_rate": 2.9834539813857293e-05,
1747
+ "loss": 0.2019,
1748
+ "step": 2450
1749
+ },
1750
+ {
1751
+ "epoch": 2.31,
1752
+ "grad_norm": 0.0872272402048111,
1753
+ "learning_rate": 2.97311271975181e-05,
1754
+ "loss": 0.3013,
1755
+ "step": 2460
1756
+ },
1757
+ {
1758
+ "epoch": 2.31,
1759
+ "grad_norm": 0.418996125459671,
1760
+ "learning_rate": 2.9627714581178906e-05,
1761
+ "loss": 0.3679,
1762
+ "step": 2470
1763
+ },
1764
+ {
1765
+ "epoch": 2.32,
1766
+ "grad_norm": 34.88379669189453,
1767
+ "learning_rate": 2.9524301964839713e-05,
1768
+ "loss": 0.3748,
1769
+ "step": 2480
1770
+ },
1771
+ {
1772
+ "epoch": 2.33,
1773
+ "grad_norm": 0.8229822516441345,
1774
+ "learning_rate": 2.942088934850052e-05,
1775
+ "loss": 0.3046,
1776
+ "step": 2490
1777
+ },
1778
+ {
1779
+ "epoch": 2.34,
1780
+ "grad_norm": 1.254721999168396,
1781
+ "learning_rate": 2.9317476732161327e-05,
1782
+ "loss": 0.4162,
1783
+ "step": 2500
1784
+ },
1785
+ {
1786
+ "epoch": 2.35,
1787
+ "grad_norm": 0.805443286895752,
1788
+ "learning_rate": 2.921406411582213e-05,
1789
+ "loss": 0.2995,
1790
+ "step": 2510
1791
+ },
1792
+ {
1793
+ "epoch": 2.36,
1794
+ "grad_norm": 0.21225698292255402,
1795
+ "learning_rate": 2.9110651499482937e-05,
1796
+ "loss": 0.2081,
1797
+ "step": 2520
1798
+ },
1799
+ {
1800
+ "epoch": 2.37,
1801
+ "grad_norm": 1.794771671295166,
1802
+ "learning_rate": 2.9007238883143744e-05,
1803
+ "loss": 0.2845,
1804
+ "step": 2530
1805
+ },
1806
+ {
1807
+ "epoch": 2.38,
1808
+ "grad_norm": 0.45042118430137634,
1809
+ "learning_rate": 2.890382626680455e-05,
1810
+ "loss": 0.1312,
1811
+ "step": 2540
1812
+ },
1813
+ {
1814
+ "epoch": 2.39,
1815
+ "grad_norm": 2.9366912841796875,
1816
+ "learning_rate": 2.8800413650465357e-05,
1817
+ "loss": 0.2553,
1818
+ "step": 2550
1819
+ },
1820
+ {
1821
+ "epoch": 2.4,
1822
+ "grad_norm": 0.3621287941932678,
1823
+ "learning_rate": 2.8697001034126164e-05,
1824
+ "loss": 0.419,
1825
+ "step": 2560
1826
+ },
1827
+ {
1828
+ "epoch": 2.41,
1829
+ "grad_norm": 0.33347466588020325,
1830
+ "learning_rate": 2.859358841778697e-05,
1831
+ "loss": 0.1815,
1832
+ "step": 2570
1833
+ },
1834
+ {
1835
+ "epoch": 2.42,
1836
+ "grad_norm": 2.7859878540039062,
1837
+ "learning_rate": 2.8490175801447777e-05,
1838
+ "loss": 0.2801,
1839
+ "step": 2580
1840
+ },
1841
+ {
1842
+ "epoch": 2.43,
1843
+ "grad_norm": 0.30300045013427734,
1844
+ "learning_rate": 2.8386763185108584e-05,
1845
+ "loss": 0.1433,
1846
+ "step": 2590
1847
+ },
1848
+ {
1849
+ "epoch": 2.44,
1850
+ "grad_norm": 98.14810943603516,
1851
+ "learning_rate": 2.828335056876939e-05,
1852
+ "loss": 0.2292,
1853
+ "step": 2600
1854
+ },
1855
+ {
1856
+ "epoch": 2.45,
1857
+ "grad_norm": 4.406097412109375,
1858
+ "learning_rate": 2.8179937952430198e-05,
1859
+ "loss": 0.1548,
1860
+ "step": 2610
1861
+ },
1862
+ {
1863
+ "epoch": 2.46,
1864
+ "grad_norm": 5.7536234855651855,
1865
+ "learning_rate": 2.8076525336091004e-05,
1866
+ "loss": 0.1126,
1867
+ "step": 2620
1868
+ },
1869
+ {
1870
+ "epoch": 2.46,
1871
+ "grad_norm": 36.3065299987793,
1872
+ "learning_rate": 2.7973112719751808e-05,
1873
+ "loss": 0.3183,
1874
+ "step": 2630
1875
+ },
1876
+ {
1877
+ "epoch": 2.47,
1878
+ "grad_norm": 34.32221603393555,
1879
+ "learning_rate": 2.7869700103412615e-05,
1880
+ "loss": 0.342,
1881
+ "step": 2640
1882
+ },
1883
+ {
1884
+ "epoch": 2.48,
1885
+ "grad_norm": 0.14024418592453003,
1886
+ "learning_rate": 2.776628748707342e-05,
1887
+ "loss": 0.3198,
1888
+ "step": 2650
1889
+ },
1890
+ {
1891
+ "epoch": 2.49,
1892
+ "grad_norm": 0.32867252826690674,
1893
+ "learning_rate": 2.7662874870734228e-05,
1894
+ "loss": 0.34,
1895
+ "step": 2660
1896
+ },
1897
+ {
1898
+ "epoch": 2.5,
1899
+ "grad_norm": 4.173491954803467,
1900
+ "learning_rate": 2.7559462254395035e-05,
1901
+ "loss": 0.2842,
1902
+ "step": 2670
1903
+ },
1904
+ {
1905
+ "epoch": 2.51,
1906
+ "grad_norm": 167.7888641357422,
1907
+ "learning_rate": 2.745604963805584e-05,
1908
+ "loss": 0.2955,
1909
+ "step": 2680
1910
+ },
1911
+ {
1912
+ "epoch": 2.52,
1913
+ "grad_norm": 0.7819741368293762,
1914
+ "learning_rate": 2.735263702171665e-05,
1915
+ "loss": 0.0981,
1916
+ "step": 2690
1917
+ },
1918
+ {
1919
+ "epoch": 2.53,
1920
+ "grad_norm": 3.086498498916626,
1921
+ "learning_rate": 2.7249224405377455e-05,
1922
+ "loss": 0.2673,
1923
+ "step": 2700
1924
+ },
1925
+ {
1926
+ "epoch": 2.54,
1927
+ "grad_norm": 0.086996890604496,
1928
+ "learning_rate": 2.7145811789038262e-05,
1929
+ "loss": 0.2067,
1930
+ "step": 2710
1931
+ },
1932
+ {
1933
+ "epoch": 2.55,
1934
+ "grad_norm": 0.09881442785263062,
1935
+ "learning_rate": 2.7042399172699072e-05,
1936
+ "loss": 0.365,
1937
+ "step": 2720
1938
+ },
1939
+ {
1940
+ "epoch": 2.56,
1941
+ "grad_norm": 43.90235137939453,
1942
+ "learning_rate": 2.693898655635988e-05,
1943
+ "loss": 0.1421,
1944
+ "step": 2730
1945
+ },
1946
+ {
1947
+ "epoch": 2.57,
1948
+ "grad_norm": 0.06362267583608627,
1949
+ "learning_rate": 2.6835573940020685e-05,
1950
+ "loss": 0.3048,
1951
+ "step": 2740
1952
+ },
1953
+ {
1954
+ "epoch": 2.58,
1955
+ "grad_norm": 0.08752346783876419,
1956
+ "learning_rate": 2.6732161323681492e-05,
1957
+ "loss": 0.0187,
1958
+ "step": 2750
1959
+ },
1960
+ {
1961
+ "epoch": 2.59,
1962
+ "grad_norm": 0.5300313234329224,
1963
+ "learning_rate": 2.66287487073423e-05,
1964
+ "loss": 0.1348,
1965
+ "step": 2760
1966
+ },
1967
+ {
1968
+ "epoch": 2.6,
1969
+ "grad_norm": 0.04888344556093216,
1970
+ "learning_rate": 2.6525336091003106e-05,
1971
+ "loss": 0.1468,
1972
+ "step": 2770
1973
+ },
1974
+ {
1975
+ "epoch": 2.61,
1976
+ "grad_norm": 13.05820369720459,
1977
+ "learning_rate": 2.6421923474663913e-05,
1978
+ "loss": 0.4363,
1979
+ "step": 2780
1980
+ },
1981
+ {
1982
+ "epoch": 2.61,
1983
+ "grad_norm": 3.226611852645874,
1984
+ "learning_rate": 2.631851085832472e-05,
1985
+ "loss": 0.2585,
1986
+ "step": 2790
1987
+ },
1988
+ {
1989
+ "epoch": 2.62,
1990
+ "grad_norm": 127.92499542236328,
1991
+ "learning_rate": 2.6215098241985526e-05,
1992
+ "loss": 0.2791,
1993
+ "step": 2800
1994
+ },
1995
+ {
1996
+ "epoch": 2.63,
1997
+ "grad_norm": 4.131041526794434,
1998
+ "learning_rate": 2.6111685625646333e-05,
1999
+ "loss": 0.3342,
2000
+ "step": 2810
2001
+ },
2002
+ {
2003
+ "epoch": 2.64,
2004
+ "grad_norm": 9.45926570892334,
2005
+ "learning_rate": 2.600827300930714e-05,
2006
+ "loss": 0.4095,
2007
+ "step": 2820
2008
+ },
2009
+ {
2010
+ "epoch": 2.65,
2011
+ "grad_norm": 0.3252735435962677,
2012
+ "learning_rate": 2.5904860392967943e-05,
2013
+ "loss": 0.1566,
2014
+ "step": 2830
2015
+ },
2016
+ {
2017
+ "epoch": 2.66,
2018
+ "grad_norm": 17.8929386138916,
2019
+ "learning_rate": 2.580144777662875e-05,
2020
+ "loss": 0.2914,
2021
+ "step": 2840
2022
+ },
2023
+ {
2024
+ "epoch": 2.67,
2025
+ "grad_norm": 0.4428044557571411,
2026
+ "learning_rate": 2.5698035160289556e-05,
2027
+ "loss": 0.1937,
2028
+ "step": 2850
2029
+ },
2030
+ {
2031
+ "epoch": 2.68,
2032
+ "grad_norm": 0.196681946516037,
2033
+ "learning_rate": 2.5594622543950363e-05,
2034
+ "loss": 0.1334,
2035
+ "step": 2860
2036
+ },
2037
+ {
2038
+ "epoch": 2.69,
2039
+ "grad_norm": 14.465700149536133,
2040
+ "learning_rate": 2.549120992761117e-05,
2041
+ "loss": 0.4305,
2042
+ "step": 2870
2043
+ },
2044
+ {
2045
+ "epoch": 2.7,
2046
+ "grad_norm": 0.25671684741973877,
2047
+ "learning_rate": 2.5387797311271977e-05,
2048
+ "loss": 0.3097,
2049
+ "step": 2880
2050
+ },
2051
+ {
2052
+ "epoch": 2.71,
2053
+ "grad_norm": 8.047564506530762,
2054
+ "learning_rate": 2.5284384694932783e-05,
2055
+ "loss": 0.483,
2056
+ "step": 2890
2057
+ },
2058
+ {
2059
+ "epoch": 2.72,
2060
+ "grad_norm": 12.367530822753906,
2061
+ "learning_rate": 2.518097207859359e-05,
2062
+ "loss": 0.1829,
2063
+ "step": 2900
2064
+ },
2065
+ {
2066
+ "epoch": 2.73,
2067
+ "grad_norm": 0.40190449357032776,
2068
+ "learning_rate": 2.5077559462254397e-05,
2069
+ "loss": 0.2601,
2070
+ "step": 2910
2071
+ },
2072
+ {
2073
+ "epoch": 2.74,
2074
+ "grad_norm": 0.18794836103916168,
2075
+ "learning_rate": 2.4974146845915204e-05,
2076
+ "loss": 0.1182,
2077
+ "step": 2920
2078
+ },
2079
+ {
2080
+ "epoch": 2.75,
2081
+ "grad_norm": 4.731940269470215,
2082
+ "learning_rate": 2.487073422957601e-05,
2083
+ "loss": 0.4801,
2084
+ "step": 2930
2085
+ },
2086
+ {
2087
+ "epoch": 2.76,
2088
+ "grad_norm": 72.07913208007812,
2089
+ "learning_rate": 2.4767321613236817e-05,
2090
+ "loss": 0.1826,
2091
+ "step": 2940
2092
+ },
2093
+ {
2094
+ "epoch": 2.76,
2095
+ "grad_norm": 0.34144583344459534,
2096
+ "learning_rate": 2.466390899689762e-05,
2097
+ "loss": 0.2458,
2098
+ "step": 2950
2099
+ },
2100
+ {
2101
+ "epoch": 2.77,
2102
+ "grad_norm": 2.847762107849121,
2103
+ "learning_rate": 2.4560496380558427e-05,
2104
+ "loss": 0.2508,
2105
+ "step": 2960
2106
+ },
2107
+ {
2108
+ "epoch": 2.78,
2109
+ "grad_norm": 1.0531044006347656,
2110
+ "learning_rate": 2.4457083764219234e-05,
2111
+ "loss": 0.3881,
2112
+ "step": 2970
2113
+ },
2114
+ {
2115
+ "epoch": 2.79,
2116
+ "grad_norm": 0.8281335830688477,
2117
+ "learning_rate": 2.435367114788004e-05,
2118
+ "loss": 0.221,
2119
+ "step": 2980
2120
+ },
2121
+ {
2122
+ "epoch": 2.8,
2123
+ "grad_norm": 12.635905265808105,
2124
+ "learning_rate": 2.4250258531540848e-05,
2125
+ "loss": 0.2008,
2126
+ "step": 2990
2127
+ },
2128
+ {
2129
+ "epoch": 2.81,
2130
+ "grad_norm": 0.37540680170059204,
2131
+ "learning_rate": 2.4146845915201654e-05,
2132
+ "loss": 0.3256,
2133
+ "step": 3000
2134
+ },
2135
+ {
2136
+ "epoch": 2.82,
2137
+ "grad_norm": 0.306692510843277,
2138
+ "learning_rate": 2.404343329886246e-05,
2139
+ "loss": 0.2346,
2140
+ "step": 3010
2141
+ },
2142
+ {
2143
+ "epoch": 2.83,
2144
+ "grad_norm": 0.23924803733825684,
2145
+ "learning_rate": 2.394002068252327e-05,
2146
+ "loss": 0.2269,
2147
+ "step": 3020
2148
+ },
2149
+ {
2150
+ "epoch": 2.84,
2151
+ "grad_norm": 43.663082122802734,
2152
+ "learning_rate": 2.3836608066184078e-05,
2153
+ "loss": 0.2109,
2154
+ "step": 3030
2155
+ },
2156
+ {
2157
+ "epoch": 2.85,
2158
+ "grad_norm": 3.9084434509277344,
2159
+ "learning_rate": 2.3733195449844885e-05,
2160
+ "loss": 0.4289,
2161
+ "step": 3040
2162
+ },
2163
+ {
2164
+ "epoch": 2.86,
2165
+ "grad_norm": 1.4723049402236938,
2166
+ "learning_rate": 2.3629782833505688e-05,
2167
+ "loss": 0.2405,
2168
+ "step": 3050
2169
+ },
2170
+ {
2171
+ "epoch": 2.87,
2172
+ "grad_norm": 3.2489137649536133,
2173
+ "learning_rate": 2.3526370217166495e-05,
2174
+ "loss": 0.2141,
2175
+ "step": 3060
2176
+ },
2177
+ {
2178
+ "epoch": 2.88,
2179
+ "grad_norm": 10.305293083190918,
2180
+ "learning_rate": 2.3422957600827302e-05,
2181
+ "loss": 0.4549,
2182
+ "step": 3070
2183
+ },
2184
+ {
2185
+ "epoch": 2.89,
2186
+ "grad_norm": 18.369117736816406,
2187
+ "learning_rate": 2.331954498448811e-05,
2188
+ "loss": 0.3059,
2189
+ "step": 3080
2190
+ },
2191
+ {
2192
+ "epoch": 2.9,
2193
+ "grad_norm": 24.559249877929688,
2194
+ "learning_rate": 2.3216132368148915e-05,
2195
+ "loss": 0.1978,
2196
+ "step": 3090
2197
+ },
2198
+ {
2199
+ "epoch": 2.91,
2200
+ "grad_norm": 0.5222854018211365,
2201
+ "learning_rate": 2.3112719751809722e-05,
2202
+ "loss": 0.1677,
2203
+ "step": 3100
2204
+ },
2205
+ {
2206
+ "epoch": 2.91,
2207
+ "grad_norm": 7.490653038024902,
2208
+ "learning_rate": 2.300930713547053e-05,
2209
+ "loss": 0.3634,
2210
+ "step": 3110
2211
+ },
2212
+ {
2213
+ "epoch": 2.92,
2214
+ "grad_norm": 0.39714524149894714,
2215
+ "learning_rate": 2.2905894519131335e-05,
2216
+ "loss": 0.2277,
2217
+ "step": 3120
2218
+ },
2219
+ {
2220
+ "epoch": 2.93,
2221
+ "grad_norm": 0.16488385200500488,
2222
+ "learning_rate": 2.2802481902792142e-05,
2223
+ "loss": 0.1301,
2224
+ "step": 3130
2225
+ },
2226
+ {
2227
+ "epoch": 2.94,
2228
+ "grad_norm": 3.3054399490356445,
2229
+ "learning_rate": 2.269906928645295e-05,
2230
+ "loss": 0.3157,
2231
+ "step": 3140
2232
+ },
2233
+ {
2234
+ "epoch": 2.95,
2235
+ "grad_norm": 0.1860102266073227,
2236
+ "learning_rate": 2.2595656670113756e-05,
2237
+ "loss": 0.1737,
2238
+ "step": 3150
2239
+ },
2240
+ {
2241
+ "epoch": 2.96,
2242
+ "grad_norm": 6.283505916595459,
2243
+ "learning_rate": 2.2492244053774563e-05,
2244
+ "loss": 0.1935,
2245
+ "step": 3160
2246
+ },
2247
+ {
2248
+ "epoch": 2.97,
2249
+ "grad_norm": 0.3177931010723114,
2250
+ "learning_rate": 2.2388831437435366e-05,
2251
+ "loss": 0.2413,
2252
+ "step": 3170
2253
+ },
2254
+ {
2255
+ "epoch": 2.98,
2256
+ "grad_norm": 9.307920455932617,
2257
+ "learning_rate": 2.2285418821096173e-05,
2258
+ "loss": 0.2362,
2259
+ "step": 3180
2260
+ },
2261
+ {
2262
+ "epoch": 2.99,
2263
+ "grad_norm": 20.813262939453125,
2264
+ "learning_rate": 2.218200620475698e-05,
2265
+ "loss": 0.2896,
2266
+ "step": 3190
2267
+ },
2268
+ {
2269
+ "epoch": 3.0,
2270
+ "grad_norm": 0.17595571279525757,
2271
+ "learning_rate": 2.2078593588417786e-05,
2272
+ "loss": 0.1038,
2273
+ "step": 3200
2274
+ },
2275
+ {
2276
+ "epoch": 3.0,
2277
+ "eval_accuracy": 0.8667917448405253,
2278
+ "eval_f1": 0.864244741873805,
2279
+ "eval_loss": 0.5857056975364685,
2280
+ "eval_precision": 0.8793774319066148,
2281
+ "eval_recall": 0.849624060150376,
2282
+ "eval_runtime": 1.5831,
2283
+ "eval_samples_per_second": 336.679,
2284
+ "eval_steps_per_second": 42.322,
2285
+ "step": 3201
2286
+ },
2287
+ {
2288
+ "epoch": 3.01,
2289
+ "grad_norm": 0.1377851516008377,
2290
+ "learning_rate": 2.1975180972078593e-05,
2291
+ "loss": 0.0617,
2292
+ "step": 3210
2293
+ },
2294
+ {
2295
+ "epoch": 3.02,
2296
+ "grad_norm": 8.743398666381836,
2297
+ "learning_rate": 2.1871768355739403e-05,
2298
+ "loss": 0.2428,
2299
+ "step": 3220
2300
+ },
2301
+ {
2302
+ "epoch": 3.03,
2303
+ "grad_norm": 1.2507226467132568,
2304
+ "learning_rate": 2.176835573940021e-05,
2305
+ "loss": 0.187,
2306
+ "step": 3230
2307
+ },
2308
+ {
2309
+ "epoch": 3.04,
2310
+ "grad_norm": 0.17167934775352478,
2311
+ "learning_rate": 2.1664943123061017e-05,
2312
+ "loss": 0.1137,
2313
+ "step": 3240
2314
+ },
2315
+ {
2316
+ "epoch": 3.05,
2317
+ "grad_norm": 5.110489845275879,
2318
+ "learning_rate": 2.1561530506721823e-05,
2319
+ "loss": 0.0637,
2320
+ "step": 3250
2321
+ },
2322
+ {
2323
+ "epoch": 3.06,
2324
+ "grad_norm": 0.11725780367851257,
2325
+ "learning_rate": 2.145811789038263e-05,
2326
+ "loss": 0.1933,
2327
+ "step": 3260
2328
+ },
2329
+ {
2330
+ "epoch": 3.06,
2331
+ "grad_norm": 0.11883195489645004,
2332
+ "learning_rate": 2.1354705274043433e-05,
2333
+ "loss": 0.1123,
2334
+ "step": 3270
2335
+ },
2336
+ {
2337
+ "epoch": 3.07,
2338
+ "grad_norm": 7.5306315422058105,
2339
+ "learning_rate": 2.125129265770424e-05,
2340
+ "loss": 0.1364,
2341
+ "step": 3280
2342
+ },
2343
+ {
2344
+ "epoch": 3.08,
2345
+ "grad_norm": 0.7140617370605469,
2346
+ "learning_rate": 2.1147880041365047e-05,
2347
+ "loss": 0.148,
2348
+ "step": 3290
2349
+ },
2350
+ {
2351
+ "epoch": 3.09,
2352
+ "grad_norm": 0.9037841558456421,
2353
+ "learning_rate": 2.1044467425025854e-05,
2354
+ "loss": 0.0663,
2355
+ "step": 3300
2356
+ },
2357
+ {
2358
+ "epoch": 3.1,
2359
+ "grad_norm": 0.05070972070097923,
2360
+ "learning_rate": 2.094105480868666e-05,
2361
+ "loss": 0.0595,
2362
+ "step": 3310
2363
+ },
2364
+ {
2365
+ "epoch": 3.11,
2366
+ "grad_norm": 0.06952597945928574,
2367
+ "learning_rate": 2.0837642192347467e-05,
2368
+ "loss": 0.0802,
2369
+ "step": 3320
2370
+ },
2371
+ {
2372
+ "epoch": 3.12,
2373
+ "grad_norm": 0.09682538360357285,
2374
+ "learning_rate": 2.0734229576008274e-05,
2375
+ "loss": 0.0756,
2376
+ "step": 3330
2377
+ },
2378
+ {
2379
+ "epoch": 3.13,
2380
+ "grad_norm": 7.944037914276123,
2381
+ "learning_rate": 2.063081695966908e-05,
2382
+ "loss": 0.1648,
2383
+ "step": 3340
2384
+ },
2385
+ {
2386
+ "epoch": 3.14,
2387
+ "grad_norm": 0.049520719796419144,
2388
+ "learning_rate": 2.0527404343329888e-05,
2389
+ "loss": 0.2814,
2390
+ "step": 3350
2391
+ },
2392
+ {
2393
+ "epoch": 3.15,
2394
+ "grad_norm": 0.12401607632637024,
2395
+ "learning_rate": 2.0423991726990694e-05,
2396
+ "loss": 0.1665,
2397
+ "step": 3360
2398
+ },
2399
+ {
2400
+ "epoch": 3.16,
2401
+ "grad_norm": 3.3132777214050293,
2402
+ "learning_rate": 2.03205791106515e-05,
2403
+ "loss": 0.2163,
2404
+ "step": 3370
2405
+ },
2406
+ {
2407
+ "epoch": 3.17,
2408
+ "grad_norm": 0.06092935800552368,
2409
+ "learning_rate": 2.0217166494312308e-05,
2410
+ "loss": 0.055,
2411
+ "step": 3380
2412
+ },
2413
+ {
2414
+ "epoch": 3.18,
2415
+ "grad_norm": 0.09410246461629868,
2416
+ "learning_rate": 2.011375387797311e-05,
2417
+ "loss": 0.0564,
2418
+ "step": 3390
2419
+ },
2420
+ {
2421
+ "epoch": 3.19,
2422
+ "grad_norm": 0.09851285070180893,
2423
+ "learning_rate": 2.0010341261633918e-05,
2424
+ "loss": 0.0996,
2425
+ "step": 3400
2426
+ },
2427
+ {
2428
+ "epoch": 3.2,
2429
+ "grad_norm": 0.15248961746692657,
2430
+ "learning_rate": 1.9906928645294725e-05,
2431
+ "loss": 0.202,
2432
+ "step": 3410
2433
+ },
2434
+ {
2435
+ "epoch": 3.21,
2436
+ "grad_norm": 0.06347668915987015,
2437
+ "learning_rate": 1.980351602895553e-05,
2438
+ "loss": 0.0841,
2439
+ "step": 3420
2440
+ },
2441
+ {
2442
+ "epoch": 3.21,
2443
+ "grad_norm": 193.50746154785156,
2444
+ "learning_rate": 1.970010341261634e-05,
2445
+ "loss": 0.4587,
2446
+ "step": 3430
2447
+ },
2448
+ {
2449
+ "epoch": 3.22,
2450
+ "grad_norm": 0.1516660451889038,
2451
+ "learning_rate": 1.959669079627715e-05,
2452
+ "loss": 0.0655,
2453
+ "step": 3440
2454
+ },
2455
+ {
2456
+ "epoch": 3.23,
2457
+ "grad_norm": 0.42711296677589417,
2458
+ "learning_rate": 1.9493278179937955e-05,
2459
+ "loss": 0.1857,
2460
+ "step": 3450
2461
+ },
2462
+ {
2463
+ "epoch": 3.24,
2464
+ "grad_norm": 0.11023885011672974,
2465
+ "learning_rate": 1.9389865563598762e-05,
2466
+ "loss": 0.0828,
2467
+ "step": 3460
2468
+ },
2469
+ {
2470
+ "epoch": 3.25,
2471
+ "grad_norm": 0.061248745769262314,
2472
+ "learning_rate": 1.928645294725957e-05,
2473
+ "loss": 0.1167,
2474
+ "step": 3470
2475
+ },
2476
+ {
2477
+ "epoch": 3.26,
2478
+ "grad_norm": 0.06203962489962578,
2479
+ "learning_rate": 1.9183040330920375e-05,
2480
+ "loss": 0.1089,
2481
+ "step": 3480
2482
+ },
2483
+ {
2484
+ "epoch": 3.27,
2485
+ "grad_norm": 19.638856887817383,
2486
+ "learning_rate": 1.907962771458118e-05,
2487
+ "loss": 0.1296,
2488
+ "step": 3490
2489
+ },
2490
+ {
2491
+ "epoch": 3.28,
2492
+ "grad_norm": 0.05891411006450653,
2493
+ "learning_rate": 1.8976215098241985e-05,
2494
+ "loss": 0.0137,
2495
+ "step": 3500
2496
+ },
2497
+ {
2498
+ "epoch": 3.29,
2499
+ "grad_norm": 0.10509899258613586,
2500
+ "learning_rate": 1.8872802481902792e-05,
2501
+ "loss": 0.0542,
2502
+ "step": 3510
2503
+ },
2504
+ {
2505
+ "epoch": 3.3,
2506
+ "grad_norm": 3.27602481842041,
2507
+ "learning_rate": 1.87693898655636e-05,
2508
+ "loss": 0.1655,
2509
+ "step": 3520
2510
+ },
2511
+ {
2512
+ "epoch": 3.31,
2513
+ "grad_norm": 0.08927629142999649,
2514
+ "learning_rate": 1.8665977249224406e-05,
2515
+ "loss": 0.0737,
2516
+ "step": 3530
2517
+ },
2518
+ {
2519
+ "epoch": 3.32,
2520
+ "grad_norm": 0.2776730954647064,
2521
+ "learning_rate": 1.8562564632885213e-05,
2522
+ "loss": 0.1569,
2523
+ "step": 3540
2524
+ },
2525
+ {
2526
+ "epoch": 3.33,
2527
+ "grad_norm": 3.598266124725342,
2528
+ "learning_rate": 1.845915201654602e-05,
2529
+ "loss": 0.3042,
2530
+ "step": 3550
2531
+ },
2532
+ {
2533
+ "epoch": 3.34,
2534
+ "grad_norm": 3.0984160900115967,
2535
+ "learning_rate": 1.8355739400206826e-05,
2536
+ "loss": 0.1836,
2537
+ "step": 3560
2538
+ },
2539
+ {
2540
+ "epoch": 3.35,
2541
+ "grad_norm": 0.09482892602682114,
2542
+ "learning_rate": 1.8252326783867633e-05,
2543
+ "loss": 0.0684,
2544
+ "step": 3570
2545
+ },
2546
+ {
2547
+ "epoch": 3.36,
2548
+ "grad_norm": 4.153615474700928,
2549
+ "learning_rate": 1.814891416752844e-05,
2550
+ "loss": 0.188,
2551
+ "step": 3580
2552
+ },
2553
+ {
2554
+ "epoch": 3.36,
2555
+ "grad_norm": 241.65493774414062,
2556
+ "learning_rate": 1.8045501551189246e-05,
2557
+ "loss": 0.1741,
2558
+ "step": 3590
2559
+ },
2560
+ {
2561
+ "epoch": 3.37,
2562
+ "grad_norm": 0.13454625010490417,
2563
+ "learning_rate": 1.7942088934850053e-05,
2564
+ "loss": 0.0606,
2565
+ "step": 3600
2566
+ },
2567
+ {
2568
+ "epoch": 3.38,
2569
+ "grad_norm": 0.12659063935279846,
2570
+ "learning_rate": 1.7838676318510856e-05,
2571
+ "loss": 0.0743,
2572
+ "step": 3610
2573
+ },
2574
+ {
2575
+ "epoch": 3.39,
2576
+ "grad_norm": 0.12111490219831467,
2577
+ "learning_rate": 1.7735263702171663e-05,
2578
+ "loss": 0.0042,
2579
+ "step": 3620
2580
+ },
2581
+ {
2582
+ "epoch": 3.4,
2583
+ "grad_norm": 3.1387064456939697,
2584
+ "learning_rate": 1.7631851085832473e-05,
2585
+ "loss": 0.3181,
2586
+ "step": 3630
2587
+ },
2588
+ {
2589
+ "epoch": 3.41,
2590
+ "grad_norm": 66.55416870117188,
2591
+ "learning_rate": 1.752843846949328e-05,
2592
+ "loss": 0.1883,
2593
+ "step": 3640
2594
+ },
2595
+ {
2596
+ "epoch": 3.42,
2597
+ "grad_norm": 9.684922218322754,
2598
+ "learning_rate": 1.7425025853154087e-05,
2599
+ "loss": 0.129,
2600
+ "step": 3650
2601
+ },
2602
+ {
2603
+ "epoch": 3.43,
2604
+ "grad_norm": 0.1191384568810463,
2605
+ "learning_rate": 1.7321613236814894e-05,
2606
+ "loss": 0.1684,
2607
+ "step": 3660
2608
+ },
2609
+ {
2610
+ "epoch": 3.44,
2611
+ "grad_norm": 0.09780813008546829,
2612
+ "learning_rate": 1.72182006204757e-05,
2613
+ "loss": 0.0052,
2614
+ "step": 3670
2615
+ },
2616
+ {
2617
+ "epoch": 3.45,
2618
+ "grad_norm": 3.4690449237823486,
2619
+ "learning_rate": 1.7114788004136507e-05,
2620
+ "loss": 0.2599,
2621
+ "step": 3680
2622
+ },
2623
+ {
2624
+ "epoch": 3.46,
2625
+ "grad_norm": 0.1335831582546234,
2626
+ "learning_rate": 1.7011375387797314e-05,
2627
+ "loss": 0.2327,
2628
+ "step": 3690
2629
+ },
2630
+ {
2631
+ "epoch": 3.47,
2632
+ "grad_norm": 0.1467386782169342,
2633
+ "learning_rate": 1.690796277145812e-05,
2634
+ "loss": 0.1308,
2635
+ "step": 3700
2636
+ },
2637
+ {
2638
+ "epoch": 3.48,
2639
+ "grad_norm": 373.8140869140625,
2640
+ "learning_rate": 1.6804550155118924e-05,
2641
+ "loss": 0.2381,
2642
+ "step": 3710
2643
+ },
2644
+ {
2645
+ "epoch": 3.49,
2646
+ "grad_norm": 0.08402853459119797,
2647
+ "learning_rate": 1.670113753877973e-05,
2648
+ "loss": 0.0955,
2649
+ "step": 3720
2650
+ },
2651
+ {
2652
+ "epoch": 3.5,
2653
+ "grad_norm": 171.42713928222656,
2654
+ "learning_rate": 1.6597724922440538e-05,
2655
+ "loss": 0.127,
2656
+ "step": 3730
2657
+ },
2658
+ {
2659
+ "epoch": 3.51,
2660
+ "grad_norm": 0.11677587032318115,
2661
+ "learning_rate": 1.6494312306101344e-05,
2662
+ "loss": 0.0415,
2663
+ "step": 3740
2664
+ },
2665
+ {
2666
+ "epoch": 3.51,
2667
+ "grad_norm": 0.0933157354593277,
2668
+ "learning_rate": 1.639089968976215e-05,
2669
+ "loss": 0.1434,
2670
+ "step": 3750
2671
+ },
2672
+ {
2673
+ "epoch": 3.52,
2674
+ "grad_norm": 0.06422335654497147,
2675
+ "learning_rate": 1.6287487073422958e-05,
2676
+ "loss": 0.2715,
2677
+ "step": 3760
2678
+ },
2679
+ {
2680
+ "epoch": 3.53,
2681
+ "grad_norm": 0.08810001611709595,
2682
+ "learning_rate": 1.6184074457083765e-05,
2683
+ "loss": 0.077,
2684
+ "step": 3770
2685
+ },
2686
+ {
2687
+ "epoch": 3.54,
2688
+ "grad_norm": 0.08342567831277847,
2689
+ "learning_rate": 1.608066184074457e-05,
2690
+ "loss": 0.2652,
2691
+ "step": 3780
2692
+ },
2693
+ {
2694
+ "epoch": 3.55,
2695
+ "grad_norm": 0.10808378458023071,
2696
+ "learning_rate": 1.5977249224405378e-05,
2697
+ "loss": 0.0818,
2698
+ "step": 3790
2699
+ },
2700
+ {
2701
+ "epoch": 3.56,
2702
+ "grad_norm": 0.06099862605333328,
2703
+ "learning_rate": 1.5873836608066185e-05,
2704
+ "loss": 0.195,
2705
+ "step": 3800
2706
+ },
2707
+ {
2708
+ "epoch": 3.57,
2709
+ "grad_norm": 17.37858772277832,
2710
+ "learning_rate": 1.577042399172699e-05,
2711
+ "loss": 0.2626,
2712
+ "step": 3810
2713
+ },
2714
+ {
2715
+ "epoch": 3.58,
2716
+ "grad_norm": 0.15915565192699432,
2717
+ "learning_rate": 1.56670113753878e-05,
2718
+ "loss": 0.0859,
2719
+ "step": 3820
2720
+ },
2721
+ {
2722
+ "epoch": 3.59,
2723
+ "grad_norm": 13.893326759338379,
2724
+ "learning_rate": 1.5563598759048605e-05,
2725
+ "loss": 0.1901,
2726
+ "step": 3830
2727
+ },
2728
+ {
2729
+ "epoch": 3.6,
2730
+ "grad_norm": 0.19214822351932526,
2731
+ "learning_rate": 1.5460186142709412e-05,
2732
+ "loss": 0.1584,
2733
+ "step": 3840
2734
+ },
2735
+ {
2736
+ "epoch": 3.61,
2737
+ "grad_norm": 0.07823023200035095,
2738
+ "learning_rate": 1.535677352637022e-05,
2739
+ "loss": 0.1354,
2740
+ "step": 3850
2741
+ },
2742
+ {
2743
+ "epoch": 3.62,
2744
+ "grad_norm": 0.2098669558763504,
2745
+ "learning_rate": 1.5253360910031025e-05,
2746
+ "loss": 0.2055,
2747
+ "step": 3860
2748
+ },
2749
+ {
2750
+ "epoch": 3.63,
2751
+ "grad_norm": 3.7674732208251953,
2752
+ "learning_rate": 1.5149948293691832e-05,
2753
+ "loss": 0.1174,
2754
+ "step": 3870
2755
+ },
2756
+ {
2757
+ "epoch": 3.64,
2758
+ "grad_norm": 0.2199217975139618,
2759
+ "learning_rate": 1.5046535677352639e-05,
2760
+ "loss": 0.2572,
2761
+ "step": 3880
2762
+ },
2763
+ {
2764
+ "epoch": 3.65,
2765
+ "grad_norm": 13.79918098449707,
2766
+ "learning_rate": 1.4943123061013446e-05,
2767
+ "loss": 0.1485,
2768
+ "step": 3890
2769
+ },
2770
+ {
2771
+ "epoch": 3.66,
2772
+ "grad_norm": 57.75919723510742,
2773
+ "learning_rate": 1.483971044467425e-05,
2774
+ "loss": 0.2868,
2775
+ "step": 3900
2776
+ },
2777
+ {
2778
+ "epoch": 3.66,
2779
+ "grad_norm": 0.26549792289733887,
2780
+ "learning_rate": 1.4736297828335057e-05,
2781
+ "loss": 0.1935,
2782
+ "step": 3910
2783
+ },
2784
+ {
2785
+ "epoch": 3.67,
2786
+ "grad_norm": 0.11525550484657288,
2787
+ "learning_rate": 1.4632885211995864e-05,
2788
+ "loss": 0.1079,
2789
+ "step": 3920
2790
+ },
2791
+ {
2792
+ "epoch": 3.68,
2793
+ "grad_norm": 0.11162678152322769,
2794
+ "learning_rate": 1.4529472595656671e-05,
2795
+ "loss": 0.2091,
2796
+ "step": 3930
2797
+ },
2798
+ {
2799
+ "epoch": 3.69,
2800
+ "grad_norm": 0.18426716327667236,
2801
+ "learning_rate": 1.4426059979317478e-05,
2802
+ "loss": 0.188,
2803
+ "step": 3940
2804
+ },
2805
+ {
2806
+ "epoch": 3.7,
2807
+ "grad_norm": 0.4274304509162903,
2808
+ "learning_rate": 1.4322647362978284e-05,
2809
+ "loss": 0.272,
2810
+ "step": 3950
2811
+ },
2812
+ {
2813
+ "epoch": 3.71,
2814
+ "grad_norm": 0.33479902148246765,
2815
+ "learning_rate": 1.421923474663909e-05,
2816
+ "loss": 0.0622,
2817
+ "step": 3960
2818
+ },
2819
+ {
2820
+ "epoch": 3.72,
2821
+ "grad_norm": 0.14429700374603271,
2822
+ "learning_rate": 1.4115822130299896e-05,
2823
+ "loss": 0.3126,
2824
+ "step": 3970
2825
+ },
2826
+ {
2827
+ "epoch": 3.73,
2828
+ "grad_norm": 17.044048309326172,
2829
+ "learning_rate": 1.4012409513960703e-05,
2830
+ "loss": 0.1594,
2831
+ "step": 3980
2832
+ },
2833
+ {
2834
+ "epoch": 3.74,
2835
+ "grad_norm": 0.23532900214195251,
2836
+ "learning_rate": 1.390899689762151e-05,
2837
+ "loss": 0.1205,
2838
+ "step": 3990
2839
+ },
2840
+ {
2841
+ "epoch": 3.75,
2842
+ "grad_norm": 0.15815971791744232,
2843
+ "learning_rate": 1.3805584281282317e-05,
2844
+ "loss": 0.1607,
2845
+ "step": 4000
2846
+ },
2847
+ {
2848
+ "epoch": 3.76,
2849
+ "grad_norm": 0.1729772984981537,
2850
+ "learning_rate": 1.3702171664943123e-05,
2851
+ "loss": 0.0638,
2852
+ "step": 4010
2853
+ },
2854
+ {
2855
+ "epoch": 3.77,
2856
+ "grad_norm": 10.028935432434082,
2857
+ "learning_rate": 1.3598759048603928e-05,
2858
+ "loss": 0.1221,
2859
+ "step": 4020
2860
+ },
2861
+ {
2862
+ "epoch": 3.78,
2863
+ "grad_norm": 0.1198321133852005,
2864
+ "learning_rate": 1.3495346432264739e-05,
2865
+ "loss": 0.2659,
2866
+ "step": 4030
2867
+ },
2868
+ {
2869
+ "epoch": 3.79,
2870
+ "grad_norm": 0.061706289649009705,
2871
+ "learning_rate": 1.3391933815925545e-05,
2872
+ "loss": 0.0927,
2873
+ "step": 4040
2874
+ },
2875
+ {
2876
+ "epoch": 3.8,
2877
+ "grad_norm": 18.801136016845703,
2878
+ "learning_rate": 1.3288521199586352e-05,
2879
+ "loss": 0.3013,
2880
+ "step": 4050
2881
+ },
2882
+ {
2883
+ "epoch": 3.81,
2884
+ "grad_norm": 3.1690914630889893,
2885
+ "learning_rate": 1.3185108583247157e-05,
2886
+ "loss": 0.1682,
2887
+ "step": 4060
2888
+ },
2889
+ {
2890
+ "epoch": 3.81,
2891
+ "grad_norm": 1.2492997646331787,
2892
+ "learning_rate": 1.3081695966907964e-05,
2893
+ "loss": 0.1623,
2894
+ "step": 4070
2895
+ },
2896
+ {
2897
+ "epoch": 3.82,
2898
+ "grad_norm": 0.12981970608234406,
2899
+ "learning_rate": 1.297828335056877e-05,
2900
+ "loss": 0.091,
2901
+ "step": 4080
2902
+ },
2903
+ {
2904
+ "epoch": 3.83,
2905
+ "grad_norm": 1.0257341861724854,
2906
+ "learning_rate": 1.2874870734229577e-05,
2907
+ "loss": 0.3142,
2908
+ "step": 4090
2909
+ },
2910
+ {
2911
+ "epoch": 3.84,
2912
+ "grad_norm": 0.25799766182899475,
2913
+ "learning_rate": 1.2771458117890384e-05,
2914
+ "loss": 0.1454,
2915
+ "step": 4100
2916
+ },
2917
+ {
2918
+ "epoch": 3.85,
2919
+ "grad_norm": 40.982479095458984,
2920
+ "learning_rate": 1.2668045501551191e-05,
2921
+ "loss": 0.1361,
2922
+ "step": 4110
2923
+ },
2924
+ {
2925
+ "epoch": 3.86,
2926
+ "grad_norm": 0.11332831531763077,
2927
+ "learning_rate": 1.2564632885211996e-05,
2928
+ "loss": 0.0518,
2929
+ "step": 4120
2930
+ },
2931
+ {
2932
+ "epoch": 3.87,
2933
+ "grad_norm": 0.04680272936820984,
2934
+ "learning_rate": 1.2461220268872803e-05,
2935
+ "loss": 0.1327,
2936
+ "step": 4130
2937
+ },
2938
+ {
2939
+ "epoch": 3.88,
2940
+ "grad_norm": 0.40334421396255493,
2941
+ "learning_rate": 1.235780765253361e-05,
2942
+ "loss": 0.2707,
2943
+ "step": 4140
2944
+ },
2945
+ {
2946
+ "epoch": 3.89,
2947
+ "grad_norm": 0.09748423099517822,
2948
+ "learning_rate": 1.2254395036194416e-05,
2949
+ "loss": 0.081,
2950
+ "step": 4150
2951
+ },
2952
+ {
2953
+ "epoch": 3.9,
2954
+ "grad_norm": 0.09192584455013275,
2955
+ "learning_rate": 1.2150982419855223e-05,
2956
+ "loss": 0.2279,
2957
+ "step": 4160
2958
+ },
2959
+ {
2960
+ "epoch": 3.91,
2961
+ "grad_norm": 0.05917090177536011,
2962
+ "learning_rate": 1.204756980351603e-05,
2963
+ "loss": 0.0764,
2964
+ "step": 4170
2965
+ },
2966
+ {
2967
+ "epoch": 3.92,
2968
+ "grad_norm": 10.152327537536621,
2969
+ "learning_rate": 1.1944157187176836e-05,
2970
+ "loss": 0.1237,
2971
+ "step": 4180
2972
+ },
2973
+ {
2974
+ "epoch": 3.93,
2975
+ "grad_norm": 0.07578109949827194,
2976
+ "learning_rate": 1.1840744570837643e-05,
2977
+ "loss": 0.136,
2978
+ "step": 4190
2979
+ },
2980
+ {
2981
+ "epoch": 3.94,
2982
+ "grad_norm": 0.13006837666034698,
2983
+ "learning_rate": 1.173733195449845e-05,
2984
+ "loss": 0.0878,
2985
+ "step": 4200
2986
+ },
2987
+ {
2988
+ "epoch": 3.95,
2989
+ "grad_norm": 0.2893955111503601,
2990
+ "learning_rate": 1.1633919338159257e-05,
2991
+ "loss": 0.1322,
2992
+ "step": 4210
2993
+ },
2994
+ {
2995
+ "epoch": 3.96,
2996
+ "grad_norm": 0.04363380745053291,
2997
+ "learning_rate": 1.1530506721820064e-05,
2998
+ "loss": 0.0566,
2999
+ "step": 4220
3000
+ },
3001
+ {
3002
+ "epoch": 3.96,
3003
+ "grad_norm": 17.78626823425293,
3004
+ "learning_rate": 1.1427094105480869e-05,
3005
+ "loss": 0.1348,
3006
+ "step": 4230
3007
+ },
3008
+ {
3009
+ "epoch": 3.97,
3010
+ "grad_norm": 0.08856278657913208,
3011
+ "learning_rate": 1.1323681489141675e-05,
3012
+ "loss": 0.0527,
3013
+ "step": 4240
3014
+ },
3015
+ {
3016
+ "epoch": 3.98,
3017
+ "grad_norm": 3.122804880142212,
3018
+ "learning_rate": 1.1220268872802482e-05,
3019
+ "loss": 0.1003,
3020
+ "step": 4250
3021
+ },
3022
+ {
3023
+ "epoch": 3.99,
3024
+ "grad_norm": 0.032136429101228714,
3025
+ "learning_rate": 1.1116856256463289e-05,
3026
+ "loss": 0.1428,
3027
+ "step": 4260
3028
+ },
3029
+ {
3030
+ "epoch": 4.0,
3031
+ "eval_accuracy": 0.8799249530956847,
3032
+ "eval_f1": 0.8801498127340824,
3033
+ "eval_loss": 0.6459251046180725,
3034
+ "eval_precision": 0.8768656716417911,
3035
+ "eval_recall": 0.8834586466165414,
3036
+ "eval_runtime": 1.6082,
3037
+ "eval_samples_per_second": 331.423,
3038
+ "eval_steps_per_second": 41.661,
3039
+ "step": 4268
3040
+ },
3041
+ {
3042
+ "epoch": 4.0,
3043
+ "grad_norm": 12.53495979309082,
3044
+ "learning_rate": 1.1013443640124096e-05,
3045
+ "loss": 0.2006,
3046
+ "step": 4270
3047
+ },
3048
+ {
3049
+ "epoch": 4.01,
3050
+ "grad_norm": 0.36368077993392944,
3051
+ "learning_rate": 1.0910031023784902e-05,
3052
+ "loss": 0.0682,
3053
+ "step": 4280
3054
+ },
3055
+ {
3056
+ "epoch": 4.02,
3057
+ "grad_norm": 0.036783114075660706,
3058
+ "learning_rate": 1.0806618407445709e-05,
3059
+ "loss": 0.0578,
3060
+ "step": 4290
3061
+ },
3062
+ {
3063
+ "epoch": 4.03,
3064
+ "grad_norm": 0.032453957945108414,
3065
+ "learning_rate": 1.0703205791106516e-05,
3066
+ "loss": 0.0511,
3067
+ "step": 4300
3068
+ },
3069
+ {
3070
+ "epoch": 4.04,
3071
+ "grad_norm": 0.0314195491373539,
3072
+ "learning_rate": 1.0599793174767323e-05,
3073
+ "loss": 0.0016,
3074
+ "step": 4310
3075
+ },
3076
+ {
3077
+ "epoch": 4.05,
3078
+ "grad_norm": 0.34390100836753845,
3079
+ "learning_rate": 1.049638055842813e-05,
3080
+ "loss": 0.063,
3081
+ "step": 4320
3082
+ },
3083
+ {
3084
+ "epoch": 4.06,
3085
+ "grad_norm": 0.08080201596021652,
3086
+ "learning_rate": 1.0392967942088936e-05,
3087
+ "loss": 0.0625,
3088
+ "step": 4330
3089
+ },
3090
+ {
3091
+ "epoch": 4.07,
3092
+ "grad_norm": 0.07401977479457855,
3093
+ "learning_rate": 1.0289555325749741e-05,
3094
+ "loss": 0.1792,
3095
+ "step": 4340
3096
+ },
3097
+ {
3098
+ "epoch": 4.08,
3099
+ "grad_norm": 0.02694222889840603,
3100
+ "learning_rate": 1.0186142709410548e-05,
3101
+ "loss": 0.064,
3102
+ "step": 4350
3103
+ },
3104
+ {
3105
+ "epoch": 4.09,
3106
+ "grad_norm": 5.257876396179199,
3107
+ "learning_rate": 1.0082730093071355e-05,
3108
+ "loss": 0.094,
3109
+ "step": 4360
3110
+ },
3111
+ {
3112
+ "epoch": 4.1,
3113
+ "grad_norm": 0.13338564336299896,
3114
+ "learning_rate": 9.979317476732161e-06,
3115
+ "loss": 0.0827,
3116
+ "step": 4370
3117
+ },
3118
+ {
3119
+ "epoch": 4.1,
3120
+ "grad_norm": 0.04957311972975731,
3121
+ "learning_rate": 9.875904860392968e-06,
3122
+ "loss": 0.224,
3123
+ "step": 4380
3124
+ },
3125
+ {
3126
+ "epoch": 4.11,
3127
+ "grad_norm": 0.3861539363861084,
3128
+ "learning_rate": 9.772492244053775e-06,
3129
+ "loss": 0.0612,
3130
+ "step": 4390
3131
+ },
3132
+ {
3133
+ "epoch": 4.12,
3134
+ "grad_norm": 0.1312892585992813,
3135
+ "learning_rate": 9.669079627714582e-06,
3136
+ "loss": 0.1828,
3137
+ "step": 4400
3138
+ },
3139
+ {
3140
+ "epoch": 4.13,
3141
+ "grad_norm": 0.11486046761274338,
3142
+ "learning_rate": 9.565667011375389e-06,
3143
+ "loss": 0.0034,
3144
+ "step": 4410
3145
+ },
3146
+ {
3147
+ "epoch": 4.14,
3148
+ "grad_norm": 0.09139719605445862,
3149
+ "learning_rate": 9.462254395036195e-06,
3150
+ "loss": 0.079,
3151
+ "step": 4420
3152
+ },
3153
+ {
3154
+ "epoch": 4.15,
3155
+ "grad_norm": 0.033867135643959045,
3156
+ "learning_rate": 9.358841778697002e-06,
3157
+ "loss": 0.0016,
3158
+ "step": 4430
3159
+ },
3160
+ {
3161
+ "epoch": 4.16,
3162
+ "grad_norm": 0.30136638879776,
3163
+ "learning_rate": 9.255429162357807e-06,
3164
+ "loss": 0.0799,
3165
+ "step": 4440
3166
+ },
3167
+ {
3168
+ "epoch": 4.17,
3169
+ "grad_norm": 0.03786735609173775,
3170
+ "learning_rate": 9.152016546018614e-06,
3171
+ "loss": 0.2652,
3172
+ "step": 4450
3173
+ },
3174
+ {
3175
+ "epoch": 4.18,
3176
+ "grad_norm": 0.047059349715709686,
3177
+ "learning_rate": 9.04860392967942e-06,
3178
+ "loss": 0.0014,
3179
+ "step": 4460
3180
+ },
3181
+ {
3182
+ "epoch": 4.19,
3183
+ "grad_norm": 0.054901909083127975,
3184
+ "learning_rate": 8.945191313340227e-06,
3185
+ "loss": 0.0509,
3186
+ "step": 4470
3187
+ },
3188
+ {
3189
+ "epoch": 4.2,
3190
+ "grad_norm": 227.820556640625,
3191
+ "learning_rate": 8.841778697001036e-06,
3192
+ "loss": 0.0222,
3193
+ "step": 4480
3194
+ },
3195
+ {
3196
+ "epoch": 4.21,
3197
+ "grad_norm": 0.043635886162519455,
3198
+ "learning_rate": 8.738366080661841e-06,
3199
+ "loss": 0.0011,
3200
+ "step": 4490
3201
+ },
3202
+ {
3203
+ "epoch": 4.22,
3204
+ "grad_norm": 0.0816226452589035,
3205
+ "learning_rate": 8.634953464322648e-06,
3206
+ "loss": 0.0873,
3207
+ "step": 4500
3208
+ },
3209
+ {
3210
+ "epoch": 4.23,
3211
+ "grad_norm": 0.07973457872867584,
3212
+ "learning_rate": 8.531540847983454e-06,
3213
+ "loss": 0.0016,
3214
+ "step": 4510
3215
+ },
3216
+ {
3217
+ "epoch": 4.24,
3218
+ "grad_norm": 0.03563067317008972,
3219
+ "learning_rate": 8.428128231644261e-06,
3220
+ "loss": 0.0758,
3221
+ "step": 4520
3222
+ },
3223
+ {
3224
+ "epoch": 4.25,
3225
+ "grad_norm": 0.09928788989782333,
3226
+ "learning_rate": 8.324715615305068e-06,
3227
+ "loss": 0.0595,
3228
+ "step": 4530
3229
+ },
3230
+ {
3231
+ "epoch": 4.25,
3232
+ "grad_norm": 0.027137843891978264,
3233
+ "learning_rate": 8.221302998965875e-06,
3234
+ "loss": 0.0263,
3235
+ "step": 4540
3236
+ },
3237
+ {
3238
+ "epoch": 4.26,
3239
+ "grad_norm": 0.021461045369505882,
3240
+ "learning_rate": 8.11789038262668e-06,
3241
+ "loss": 0.0635,
3242
+ "step": 4550
3243
+ },
3244
+ {
3245
+ "epoch": 4.27,
3246
+ "grad_norm": 0.026392286643385887,
3247
+ "learning_rate": 8.014477766287486e-06,
3248
+ "loss": 0.2159,
3249
+ "step": 4560
3250
+ },
3251
+ {
3252
+ "epoch": 4.28,
3253
+ "grad_norm": 15.914426803588867,
3254
+ "learning_rate": 7.911065149948293e-06,
3255
+ "loss": 0.0597,
3256
+ "step": 4570
3257
+ },
3258
+ {
3259
+ "epoch": 4.29,
3260
+ "grad_norm": 0.028651505708694458,
3261
+ "learning_rate": 7.807652533609102e-06,
3262
+ "loss": 0.0014,
3263
+ "step": 4580
3264
+ },
3265
+ {
3266
+ "epoch": 4.3,
3267
+ "grad_norm": 0.07976741343736649,
3268
+ "learning_rate": 7.704239917269908e-06,
3269
+ "loss": 0.0019,
3270
+ "step": 4590
3271
+ },
3272
+ {
3273
+ "epoch": 4.31,
3274
+ "grad_norm": 0.7245219945907593,
3275
+ "learning_rate": 7.600827300930714e-06,
3276
+ "loss": 0.1196,
3277
+ "step": 4600
3278
+ },
3279
+ {
3280
+ "epoch": 4.32,
3281
+ "grad_norm": 3.421379327774048,
3282
+ "learning_rate": 7.497414684591521e-06,
3283
+ "loss": 0.1707,
3284
+ "step": 4610
3285
+ },
3286
+ {
3287
+ "epoch": 4.33,
3288
+ "grad_norm": 54.59723663330078,
3289
+ "learning_rate": 7.394002068252327e-06,
3290
+ "loss": 0.1257,
3291
+ "step": 4620
3292
+ },
3293
+ {
3294
+ "epoch": 4.34,
3295
+ "grad_norm": 0.02159339189529419,
3296
+ "learning_rate": 7.290589451913134e-06,
3297
+ "loss": 0.0011,
3298
+ "step": 4630
3299
+ },
3300
+ {
3301
+ "epoch": 4.35,
3302
+ "grad_norm": 0.02801990881562233,
3303
+ "learning_rate": 7.1871768355739405e-06,
3304
+ "loss": 0.0013,
3305
+ "step": 4640
3306
+ },
3307
+ {
3308
+ "epoch": 4.36,
3309
+ "grad_norm": 0.12835495173931122,
3310
+ "learning_rate": 7.0837642192347465e-06,
3311
+ "loss": 0.1343,
3312
+ "step": 4650
3313
+ },
3314
+ {
3315
+ "epoch": 4.37,
3316
+ "grad_norm": 0.06059630960226059,
3317
+ "learning_rate": 6.980351602895553e-06,
3318
+ "loss": 0.0029,
3319
+ "step": 4660
3320
+ },
3321
+ {
3322
+ "epoch": 4.38,
3323
+ "grad_norm": 0.021486638113856316,
3324
+ "learning_rate": 6.87693898655636e-06,
3325
+ "loss": 0.0872,
3326
+ "step": 4670
3327
+ },
3328
+ {
3329
+ "epoch": 4.39,
3330
+ "grad_norm": 0.024658987298607826,
3331
+ "learning_rate": 6.773526370217166e-06,
3332
+ "loss": 0.0013,
3333
+ "step": 4680
3334
+ },
3335
+ {
3336
+ "epoch": 4.4,
3337
+ "grad_norm": 0.020257245749235153,
3338
+ "learning_rate": 6.670113753877974e-06,
3339
+ "loss": 0.0012,
3340
+ "step": 4690
3341
+ },
3342
+ {
3343
+ "epoch": 4.4,
3344
+ "grad_norm": 0.022628970444202423,
3345
+ "learning_rate": 6.56670113753878e-06,
3346
+ "loss": 0.0879,
3347
+ "step": 4700
3348
+ },
3349
+ {
3350
+ "epoch": 4.41,
3351
+ "grad_norm": 0.17840439081192017,
3352
+ "learning_rate": 6.463288521199587e-06,
3353
+ "loss": 0.1423,
3354
+ "step": 4710
3355
+ },
3356
+ {
3357
+ "epoch": 4.42,
3358
+ "grad_norm": 248.1060333251953,
3359
+ "learning_rate": 6.359875904860394e-06,
3360
+ "loss": 0.1144,
3361
+ "step": 4720
3362
+ },
3363
+ {
3364
+ "epoch": 4.43,
3365
+ "grad_norm": 3.23333740234375,
3366
+ "learning_rate": 6.2564632885212e-06,
3367
+ "loss": 0.0923,
3368
+ "step": 4730
3369
+ },
3370
+ {
3371
+ "epoch": 4.44,
3372
+ "grad_norm": 0.022127114236354828,
3373
+ "learning_rate": 6.153050672182006e-06,
3374
+ "loss": 0.1237,
3375
+ "step": 4740
3376
+ },
3377
+ {
3378
+ "epoch": 4.45,
3379
+ "grad_norm": 0.03299199789762497,
3380
+ "learning_rate": 6.049638055842813e-06,
3381
+ "loss": 0.0742,
3382
+ "step": 4750
3383
+ },
3384
+ {
3385
+ "epoch": 4.46,
3386
+ "grad_norm": 0.041477978229522705,
3387
+ "learning_rate": 5.94622543950362e-06,
3388
+ "loss": 0.0012,
3389
+ "step": 4760
3390
+ },
3391
+ {
3392
+ "epoch": 4.47,
3393
+ "grad_norm": 0.28576919436454773,
3394
+ "learning_rate": 5.842812823164427e-06,
3395
+ "loss": 0.0737,
3396
+ "step": 4770
3397
+ },
3398
+ {
3399
+ "epoch": 4.48,
3400
+ "grad_norm": 0.03505215421319008,
3401
+ "learning_rate": 5.739400206825233e-06,
3402
+ "loss": 0.168,
3403
+ "step": 4780
3404
+ },
3405
+ {
3406
+ "epoch": 4.49,
3407
+ "grad_norm": 0.023467887192964554,
3408
+ "learning_rate": 5.635987590486039e-06,
3409
+ "loss": 0.0011,
3410
+ "step": 4790
3411
+ },
3412
+ {
3413
+ "epoch": 4.5,
3414
+ "grad_norm": 0.0256861113011837,
3415
+ "learning_rate": 5.532574974146846e-06,
3416
+ "loss": 0.1276,
3417
+ "step": 4800
3418
+ },
3419
+ {
3420
+ "epoch": 4.51,
3421
+ "grad_norm": 89.85589599609375,
3422
+ "learning_rate": 5.429162357807653e-06,
3423
+ "loss": 0.0627,
3424
+ "step": 4810
3425
+ },
3426
+ {
3427
+ "epoch": 4.52,
3428
+ "grad_norm": 0.019946351647377014,
3429
+ "learning_rate": 5.32574974146846e-06,
3430
+ "loss": 0.0012,
3431
+ "step": 4820
3432
+ },
3433
+ {
3434
+ "epoch": 4.53,
3435
+ "grad_norm": 0.1361633539199829,
3436
+ "learning_rate": 5.222337125129266e-06,
3437
+ "loss": 0.0013,
3438
+ "step": 4830
3439
+ },
3440
+ {
3441
+ "epoch": 4.54,
3442
+ "grad_norm": 0.06002867594361305,
3443
+ "learning_rate": 5.118924508790072e-06,
3444
+ "loss": 0.0014,
3445
+ "step": 4840
3446
+ },
3447
+ {
3448
+ "epoch": 4.55,
3449
+ "grad_norm": 0.02299017459154129,
3450
+ "learning_rate": 5.015511892450879e-06,
3451
+ "loss": 0.0895,
3452
+ "step": 4850
3453
+ },
3454
+ {
3455
+ "epoch": 4.55,
3456
+ "grad_norm": 0.01912473328411579,
3457
+ "learning_rate": 4.912099276111686e-06,
3458
+ "loss": 0.0163,
3459
+ "step": 4860
3460
+ },
3461
+ {
3462
+ "epoch": 4.56,
3463
+ "grad_norm": 0.020831523463129997,
3464
+ "learning_rate": 4.8086866597724926e-06,
3465
+ "loss": 0.001,
3466
+ "step": 4870
3467
+ },
3468
+ {
3469
+ "epoch": 4.57,
3470
+ "grad_norm": 0.017449194565415382,
3471
+ "learning_rate": 4.705274043433299e-06,
3472
+ "loss": 0.0637,
3473
+ "step": 4880
3474
+ },
3475
+ {
3476
+ "epoch": 4.58,
3477
+ "grad_norm": 0.14713767170906067,
3478
+ "learning_rate": 4.601861427094105e-06,
3479
+ "loss": 0.0035,
3480
+ "step": 4890
3481
+ },
3482
+ {
3483
+ "epoch": 4.59,
3484
+ "grad_norm": 3.178408145904541,
3485
+ "learning_rate": 4.498448810754912e-06,
3486
+ "loss": 0.091,
3487
+ "step": 4900
3488
+ },
3489
+ {
3490
+ "epoch": 4.6,
3491
+ "grad_norm": 0.0492083765566349,
3492
+ "learning_rate": 4.39503619441572e-06,
3493
+ "loss": 0.24,
3494
+ "step": 4910
3495
+ },
3496
+ {
3497
+ "epoch": 4.61,
3498
+ "grad_norm": 0.01996445097029209,
3499
+ "learning_rate": 4.2916235780765255e-06,
3500
+ "loss": 0.1126,
3501
+ "step": 4920
3502
+ },
3503
+ {
3504
+ "epoch": 4.62,
3505
+ "grad_norm": 0.021161267533898354,
3506
+ "learning_rate": 4.188210961737332e-06,
3507
+ "loss": 0.0011,
3508
+ "step": 4930
3509
+ },
3510
+ {
3511
+ "epoch": 4.63,
3512
+ "grad_norm": 0.019828295335173607,
3513
+ "learning_rate": 4.084798345398139e-06,
3514
+ "loss": 0.0025,
3515
+ "step": 4940
3516
+ },
3517
+ {
3518
+ "epoch": 4.64,
3519
+ "grad_norm": 0.021803775802254677,
3520
+ "learning_rate": 3.981385729058945e-06,
3521
+ "loss": 0.1434,
3522
+ "step": 4950
3523
+ },
3524
+ {
3525
+ "epoch": 4.65,
3526
+ "grad_norm": 0.13735263049602509,
3527
+ "learning_rate": 3.8779731127197525e-06,
3528
+ "loss": 0.1412,
3529
+ "step": 4960
3530
+ },
3531
+ {
3532
+ "epoch": 4.66,
3533
+ "grad_norm": 0.24444641172885895,
3534
+ "learning_rate": 3.774560496380559e-06,
3535
+ "loss": 0.0046,
3536
+ "step": 4970
3537
+ },
3538
+ {
3539
+ "epoch": 4.67,
3540
+ "grad_norm": 0.023776857182383537,
3541
+ "learning_rate": 3.671147880041365e-06,
3542
+ "loss": 0.1191,
3543
+ "step": 4980
3544
+ },
3545
+ {
3546
+ "epoch": 4.68,
3547
+ "grad_norm": 0.024539895355701447,
3548
+ "learning_rate": 3.5677352637021715e-06,
3549
+ "loss": 0.1884,
3550
+ "step": 4990
3551
+ },
3552
+ {
3553
+ "epoch": 4.69,
3554
+ "grad_norm": 0.01984153315424919,
3555
+ "learning_rate": 3.4643226473629783e-06,
3556
+ "loss": 0.1128,
3557
+ "step": 5000
3558
+ },
3559
+ {
3560
+ "epoch": 4.7,
3561
+ "grad_norm": 2.897346258163452,
3562
+ "learning_rate": 3.3609100310237855e-06,
3563
+ "loss": 0.0782,
3564
+ "step": 5010
3565
+ },
3566
+ {
3567
+ "epoch": 4.7,
3568
+ "grad_norm": 0.04367586970329285,
3569
+ "learning_rate": 3.257497414684592e-06,
3570
+ "loss": 0.1729,
3571
+ "step": 5020
3572
+ },
3573
+ {
3574
+ "epoch": 4.71,
3575
+ "grad_norm": 0.022409643977880478,
3576
+ "learning_rate": 3.154084798345398e-06,
3577
+ "loss": 0.0013,
3578
+ "step": 5030
3579
+ },
3580
+ {
3581
+ "epoch": 4.72,
3582
+ "grad_norm": 0.02230711840093136,
3583
+ "learning_rate": 3.050672182006205e-06,
3584
+ "loss": 0.0605,
3585
+ "step": 5040
3586
+ },
3587
+ {
3588
+ "epoch": 4.73,
3589
+ "grad_norm": 0.08053728193044662,
3590
+ "learning_rate": 2.9472595656670117e-06,
3591
+ "loss": 0.0581,
3592
+ "step": 5050
3593
+ },
3594
+ {
3595
+ "epoch": 4.74,
3596
+ "grad_norm": 0.382304847240448,
3597
+ "learning_rate": 2.843846949327818e-06,
3598
+ "loss": 0.0396,
3599
+ "step": 5060
3600
+ },
3601
+ {
3602
+ "epoch": 4.75,
3603
+ "grad_norm": 0.0988391637802124,
3604
+ "learning_rate": 2.7404343329886247e-06,
3605
+ "loss": 0.0291,
3606
+ "step": 5070
3607
+ },
3608
+ {
3609
+ "epoch": 4.76,
3610
+ "grad_norm": 0.025470012798905373,
3611
+ "learning_rate": 2.6370217166494315e-06,
3612
+ "loss": 0.0012,
3613
+ "step": 5080
3614
+ },
3615
+ {
3616
+ "epoch": 4.77,
3617
+ "grad_norm": 0.01994398981332779,
3618
+ "learning_rate": 2.533609100310238e-06,
3619
+ "loss": 0.0021,
3620
+ "step": 5090
3621
+ },
3622
+ {
3623
+ "epoch": 4.78,
3624
+ "grad_norm": 0.062303327023983,
3625
+ "learning_rate": 2.4301964839710446e-06,
3626
+ "loss": 0.1794,
3627
+ "step": 5100
3628
+ },
3629
+ {
3630
+ "epoch": 4.79,
3631
+ "grad_norm": 0.02839079312980175,
3632
+ "learning_rate": 2.3267838676318514e-06,
3633
+ "loss": 0.0827,
3634
+ "step": 5110
3635
+ },
3636
+ {
3637
+ "epoch": 4.8,
3638
+ "grad_norm": 0.02093564346432686,
3639
+ "learning_rate": 2.2233712512926577e-06,
3640
+ "loss": 0.1439,
3641
+ "step": 5120
3642
+ },
3643
+ {
3644
+ "epoch": 4.81,
3645
+ "grad_norm": 0.024724125862121582,
3646
+ "learning_rate": 2.1199586349534644e-06,
3647
+ "loss": 0.2626,
3648
+ "step": 5130
3649
+ },
3650
+ {
3651
+ "epoch": 4.82,
3652
+ "grad_norm": 0.025263186544179916,
3653
+ "learning_rate": 2.0165460186142708e-06,
3654
+ "loss": 0.0011,
3655
+ "step": 5140
3656
+ },
3657
+ {
3658
+ "epoch": 4.83,
3659
+ "grad_norm": 0.026100702583789825,
3660
+ "learning_rate": 1.913133402275078e-06,
3661
+ "loss": 0.0012,
3662
+ "step": 5150
3663
+ },
3664
+ {
3665
+ "epoch": 4.84,
3666
+ "grad_norm": 0.42885711789131165,
3667
+ "learning_rate": 1.8097207859358843e-06,
3668
+ "loss": 0.0014,
3669
+ "step": 5160
3670
+ },
3671
+ {
3672
+ "epoch": 4.85,
3673
+ "grad_norm": 0.09349822252988815,
3674
+ "learning_rate": 1.7063081695966906e-06,
3675
+ "loss": 0.0014,
3676
+ "step": 5170
3677
+ },
3678
+ {
3679
+ "epoch": 4.85,
3680
+ "grad_norm": 0.020912567153573036,
3681
+ "learning_rate": 1.6028955532574976e-06,
3682
+ "loss": 0.2671,
3683
+ "step": 5180
3684
+ },
3685
+ {
3686
+ "epoch": 4.86,
3687
+ "grad_norm": 0.0201337318867445,
3688
+ "learning_rate": 1.499482936918304e-06,
3689
+ "loss": 0.0653,
3690
+ "step": 5190
3691
+ },
3692
+ {
3693
+ "epoch": 4.87,
3694
+ "grad_norm": 0.12457222491502762,
3695
+ "learning_rate": 1.3960703205791107e-06,
3696
+ "loss": 0.0855,
3697
+ "step": 5200
3698
+ },
3699
+ {
3700
+ "epoch": 4.88,
3701
+ "grad_norm": 0.06253860145807266,
3702
+ "learning_rate": 1.2926577042399172e-06,
3703
+ "loss": 0.0576,
3704
+ "step": 5210
3705
+ },
3706
+ {
3707
+ "epoch": 4.89,
3708
+ "grad_norm": 0.033292561769485474,
3709
+ "learning_rate": 1.189245087900724e-06,
3710
+ "loss": 0.1763,
3711
+ "step": 5220
3712
+ },
3713
+ {
3714
+ "epoch": 4.9,
3715
+ "grad_norm": 5.004137992858887,
3716
+ "learning_rate": 1.0858324715615305e-06,
3717
+ "loss": 0.0863,
3718
+ "step": 5230
3719
+ },
3720
+ {
3721
+ "epoch": 4.91,
3722
+ "grad_norm": 10.605862617492676,
3723
+ "learning_rate": 9.82419855222337e-07,
3724
+ "loss": 0.1183,
3725
+ "step": 5240
3726
+ },
3727
+ {
3728
+ "epoch": 4.92,
3729
+ "grad_norm": 0.02466098591685295,
3730
+ "learning_rate": 8.790072388831437e-07,
3731
+ "loss": 0.0013,
3732
+ "step": 5250
3733
+ },
3734
+ {
3735
+ "epoch": 4.93,
3736
+ "grad_norm": 47.83183288574219,
3737
+ "learning_rate": 7.755946225439504e-07,
3738
+ "loss": 0.1172,
3739
+ "step": 5260
3740
+ },
3741
+ {
3742
+ "epoch": 4.94,
3743
+ "grad_norm": 0.05490906536579132,
3744
+ "learning_rate": 6.72182006204757e-07,
3745
+ "loss": 0.0862,
3746
+ "step": 5270
3747
+ },
3748
+ {
3749
+ "epoch": 4.95,
3750
+ "grad_norm": 0.09379950165748596,
3751
+ "learning_rate": 5.687693898655636e-07,
3752
+ "loss": 0.0013,
3753
+ "step": 5280
3754
+ },
3755
+ {
3756
+ "epoch": 4.96,
3757
+ "grad_norm": 0.023985665291547775,
3758
+ "learning_rate": 4.6535677352637023e-07,
3759
+ "loss": 0.0016,
3760
+ "step": 5290
3761
+ },
3762
+ {
3763
+ "epoch": 4.97,
3764
+ "grad_norm": 0.18879929184913635,
3765
+ "learning_rate": 3.619441571871769e-07,
3766
+ "loss": 0.0607,
3767
+ "step": 5300
3768
+ },
3769
+ {
3770
+ "epoch": 4.98,
3771
+ "grad_norm": 0.043404027819633484,
3772
+ "learning_rate": 2.585315408479835e-07,
3773
+ "loss": 0.0933,
3774
+ "step": 5310
3775
+ },
3776
+ {
3777
+ "epoch": 4.99,
3778
+ "grad_norm": 10.754716873168945,
3779
+ "learning_rate": 1.5511892450879008e-07,
3780
+ "loss": 0.0017,
3781
+ "step": 5320
3782
+ },
3783
+ {
3784
+ "epoch": 5.0,
3785
+ "grad_norm": 0.027110379189252853,
3786
+ "learning_rate": 5.1706308169596694e-08,
3787
+ "loss": 0.0889,
3788
+ "step": 5330
3789
+ },
3790
+ {
3791
+ "epoch": 5.0,
3792
+ "eval_accuracy": 0.8761726078799249,
3793
+ "eval_f1": 0.8735632183908046,
3794
+ "eval_loss": 0.7370651364326477,
3795
+ "eval_precision": 0.890625,
3796
+ "eval_recall": 0.8571428571428571,
3797
+ "eval_runtime": 1.5139,
3798
+ "eval_samples_per_second": 352.061,
3799
+ "eval_steps_per_second": 44.255,
3800
+ "step": 5335
3801
+ }
3802
+ ],
3803
+ "logging_steps": 10,
3804
+ "max_steps": 5335,
3805
+ "num_input_tokens_seen": 0,
3806
+ "num_train_epochs": 5,
3807
+ "save_steps": 500,
3808
+ "total_flos": 1753388517360000.0,
3809
+ "train_batch_size": 8,
3810
+ "trial_name": null,
3811
+ "trial_params": null
3812
+ }
tmp-checkpoint-5335/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:845e2d48921e73f25ae02136aa40e96ecac5011c603a274ba4507a46d1806916
3
+ size 4984
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:845e2d48921e73f25ae02136aa40e96ecac5011c603a274ba4507a46d1806916
3
+ size 4984