EC2 Default User commited on
Commit
6554ccb
1 Parent(s): 299aa66
config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/tmp/mnli_distil_output/checkpoint-6000",
3
+ "activation": "gelu",
4
+ "architectures": [
5
+ "DistilBertForSequenceClassification"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "dim": 768,
9
+ "dropout": 0.1,
10
+ "finetuning_task": "mnli",
11
+ "hidden_dim": 3072,
12
+ "id2label": {
13
+ "0": "LABEL_0",
14
+ "1": "LABEL_1",
15
+ "2": "LABEL_2"
16
+ },
17
+ "initializer_range": 0.02,
18
+ "label2id": {
19
+ "LABEL_0": 0,
20
+ "LABEL_1": 1,
21
+ "LABEL_2": 2
22
+ },
23
+ "max_position_embeddings": 512,
24
+ "model_type": "distilbert",
25
+ "n_heads": 12,
26
+ "n_layers": 6,
27
+ "pad_token_id": 0,
28
+ "qa_dropout": 0.1,
29
+ "seq_classif_dropout": 0.2,
30
+ "sinusoidal_pos_embds": false,
31
+ "tie_weights_": true,
32
+ "transformers_version": "4.3.2",
33
+ "vocab_size": 30522,
34
+ "xla_device": true
35
+ }
eval_results_mnli-mm.txt ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ epoch = 5.0
2
+ eval_accuracy = 0.8206875508543532
3
+ eval_loss = 0.8706700205802917
4
+ eval_runtime = 17.8278
5
+ eval_samples_per_second = 551.498
eval_results_mnli.txt ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ epoch = 5.0
2
+ eval_accuracy = 0.8221090168110036
3
+ eval_loss = 0.8824708461761475
4
+ eval_runtime = 18.1167
5
+ eval_samples_per_second = 541.765
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e9c34cf80a78f251571afee70b7139ce1c5137a4c35a5a984ae519337e2092d0
3
+ size 267866263
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "model_max_length": 512, "name_or_path": "distilbert-base-uncased"}
train_results.txt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ epoch = 5.0
2
+ train_runtime = 14248.9446
3
+ train_samples_per_second = 8.613
trainer_state.json ADDED
@@ -0,0 +1,1493 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 5.0,
5
+ "global_step": 122720,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.08,
12
+ "learning_rate": 1.9456757931334205e-05,
13
+ "loss": 0.8613,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.16,
18
+ "learning_rate": 1.8913515862668405e-05,
19
+ "loss": 0.7137,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.24,
24
+ "learning_rate": 1.837027379400261e-05,
25
+ "loss": 0.664,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.33,
30
+ "learning_rate": 1.7827031725336812e-05,
31
+ "loss": 0.6249,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.41,
36
+ "learning_rate": 1.7283789656671015e-05,
37
+ "loss": 0.6086,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.49,
42
+ "learning_rate": 1.6740547588005215e-05,
43
+ "loss": 0.5861,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.57,
48
+ "learning_rate": 1.619730551933942e-05,
49
+ "loss": 0.5828,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.65,
54
+ "learning_rate": 1.5654063450673622e-05,
55
+ "loss": 0.5786,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.73,
60
+ "learning_rate": 1.5110821382007822e-05,
61
+ "loss": 0.5533,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.81,
66
+ "learning_rate": 1.4567579313342026e-05,
67
+ "loss": 0.582,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.9,
72
+ "learning_rate": 1.402433724467623e-05,
73
+ "loss": 0.5415,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.98,
78
+ "learning_rate": 1.3481095176010431e-05,
79
+ "loss": 0.5371,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.26,
84
+ "learning_rate": 1.8940677966101697e-05,
85
+ "loss": 0.5397,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.29,
90
+ "learning_rate": 1.8859191655801828e-05,
91
+ "loss": 0.5201,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.31,
96
+ "learning_rate": 1.8777705345501956e-05,
97
+ "loss": 0.5129,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.33,
102
+ "learning_rate": 1.8696219035202087e-05,
103
+ "loss": 0.5152,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.35,
108
+ "learning_rate": 1.861473272490222e-05,
109
+ "loss": 0.5041,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.37,
114
+ "learning_rate": 1.853324641460235e-05,
115
+ "loss": 0.5183,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.39,
120
+ "learning_rate": 1.8451760104302477e-05,
121
+ "loss": 0.5183,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.41,
126
+ "learning_rate": 1.837027379400261e-05,
127
+ "loss": 0.5126,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.43,
132
+ "learning_rate": 1.828878748370274e-05,
133
+ "loss": 0.5179,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.45,
138
+ "learning_rate": 1.820730117340287e-05,
139
+ "loss": 0.4975,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.47,
144
+ "learning_rate": 1.8125814863103e-05,
145
+ "loss": 0.5099,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.49,
150
+ "learning_rate": 1.804432855280313e-05,
151
+ "loss": 0.4978,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.51,
156
+ "learning_rate": 1.796284224250326e-05,
157
+ "loss": 0.4882,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.53,
162
+ "learning_rate": 1.788135593220339e-05,
163
+ "loss": 0.4891,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.55,
168
+ "learning_rate": 1.7799869621903524e-05,
169
+ "loss": 0.4992,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.57,
174
+ "learning_rate": 1.771838331160365e-05,
175
+ "loss": 0.4922,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.59,
180
+ "learning_rate": 1.7636897001303783e-05,
181
+ "loss": 0.4901,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.61,
186
+ "learning_rate": 1.7555410691003914e-05,
187
+ "loss": 0.4972,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.63,
192
+ "learning_rate": 1.7473924380704045e-05,
193
+ "loss": 0.476,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.65,
198
+ "learning_rate": 1.7392438070404173e-05,
199
+ "loss": 0.4918,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.67,
204
+ "learning_rate": 1.7310951760104304e-05,
205
+ "loss": 0.4856,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.69,
210
+ "learning_rate": 1.7229465449804435e-05,
211
+ "loss": 0.4855,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.71,
216
+ "learning_rate": 1.7147979139504566e-05,
217
+ "loss": 0.485,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.73,
222
+ "learning_rate": 1.7066492829204694e-05,
223
+ "loss": 0.4855,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.75,
228
+ "learning_rate": 1.6985006518904825e-05,
229
+ "loss": 0.4817,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.77,
234
+ "learning_rate": 1.6903520208604957e-05,
235
+ "loss": 0.5006,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.79,
240
+ "learning_rate": 1.6822033898305084e-05,
241
+ "loss": 0.4814,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.81,
246
+ "learning_rate": 1.6740547588005215e-05,
247
+ "loss": 0.4944,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.84,
252
+ "learning_rate": 1.6659061277705347e-05,
253
+ "loss": 0.4805,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.86,
258
+ "learning_rate": 1.6577574967405478e-05,
259
+ "loss": 0.4927,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.88,
264
+ "learning_rate": 1.6496088657105606e-05,
265
+ "loss": 0.4699,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.9,
270
+ "learning_rate": 1.6414602346805737e-05,
271
+ "loss": 0.4839,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.92,
276
+ "learning_rate": 1.6333116036505868e-05,
277
+ "loss": 0.4795,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.94,
282
+ "learning_rate": 1.6251629726206e-05,
283
+ "loss": 0.4888,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.96,
288
+ "learning_rate": 1.617014341590613e-05,
289
+ "loss": 0.4744,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.98,
294
+ "learning_rate": 1.608865710560626e-05,
295
+ "loss": 0.473,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 1.0,
300
+ "learning_rate": 1.600717079530639e-05,
301
+ "loss": 0.4739,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 1.02,
306
+ "learning_rate": 1.592568448500652e-05,
307
+ "loss": 0.3848,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 1.04,
312
+ "learning_rate": 1.5844198174706652e-05,
313
+ "loss": 0.3872,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 1.06,
318
+ "learning_rate": 1.576271186440678e-05,
319
+ "loss": 0.3878,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 1.08,
324
+ "learning_rate": 1.568122555410691e-05,
325
+ "loss": 0.3813,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 1.1,
330
+ "learning_rate": 1.5599739243807042e-05,
331
+ "loss": 0.3873,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 1.12,
336
+ "learning_rate": 1.5518252933507173e-05,
337
+ "loss": 0.3923,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 1.14,
342
+ "learning_rate": 1.54367666232073e-05,
343
+ "loss": 0.3901,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 1.16,
348
+ "learning_rate": 1.5355280312907432e-05,
349
+ "loss": 0.3889,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 1.18,
354
+ "learning_rate": 1.5273794002607563e-05,
355
+ "loss": 0.38,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 1.2,
360
+ "learning_rate": 1.5192307692307693e-05,
361
+ "loss": 0.388,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 1.22,
366
+ "learning_rate": 1.5110821382007822e-05,
367
+ "loss": 0.3993,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 1.24,
372
+ "learning_rate": 1.5029335071707954e-05,
373
+ "loss": 0.3911,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 1.26,
378
+ "learning_rate": 1.4947848761408083e-05,
379
+ "loss": 0.393,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 1.28,
384
+ "learning_rate": 1.4866362451108216e-05,
385
+ "loss": 0.3804,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 1.3,
390
+ "learning_rate": 1.4784876140808346e-05,
391
+ "loss": 0.3983,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 1.32,
396
+ "learning_rate": 1.4703389830508477e-05,
397
+ "loss": 0.3978,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 1.34,
402
+ "learning_rate": 1.4621903520208606e-05,
403
+ "loss": 0.4031,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 1.36,
408
+ "learning_rate": 1.4540417209908737e-05,
409
+ "loss": 0.3871,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 1.39,
414
+ "learning_rate": 1.4458930899608867e-05,
415
+ "loss": 0.4003,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 1.41,
420
+ "learning_rate": 1.4377444589308998e-05,
421
+ "loss": 0.3789,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 1.43,
426
+ "learning_rate": 1.4295958279009128e-05,
427
+ "loss": 0.3932,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 1.45,
432
+ "learning_rate": 1.4214471968709259e-05,
433
+ "loss": 0.3867,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 1.47,
438
+ "learning_rate": 1.4132985658409388e-05,
439
+ "loss": 0.391,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 1.49,
444
+ "learning_rate": 1.4051499348109518e-05,
445
+ "loss": 0.3852,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 1.51,
450
+ "learning_rate": 1.3970013037809649e-05,
451
+ "loss": 0.3865,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 1.53,
456
+ "learning_rate": 1.3888526727509778e-05,
457
+ "loss": 0.3932,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 1.55,
462
+ "learning_rate": 1.380704041720991e-05,
463
+ "loss": 0.3957,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 1.57,
468
+ "learning_rate": 1.372555410691004e-05,
469
+ "loss": 0.3908,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 1.59,
474
+ "learning_rate": 1.364406779661017e-05,
475
+ "loss": 0.3984,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 1.61,
480
+ "learning_rate": 1.35625814863103e-05,
481
+ "loss": 0.394,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 1.63,
486
+ "learning_rate": 1.3481095176010431e-05,
487
+ "loss": 0.3937,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 1.65,
492
+ "learning_rate": 1.3399608865710562e-05,
493
+ "loss": 0.4022,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 1.67,
498
+ "learning_rate": 1.3318122555410693e-05,
499
+ "loss": 0.3872,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 1.69,
504
+ "learning_rate": 1.3236636245110823e-05,
505
+ "loss": 0.3815,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 1.71,
510
+ "learning_rate": 1.3155149934810954e-05,
511
+ "loss": 0.3853,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 1.73,
516
+ "learning_rate": 1.3073663624511084e-05,
517
+ "loss": 0.3913,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 1.75,
522
+ "learning_rate": 1.2992177314211213e-05,
523
+ "loss": 0.3913,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 1.77,
528
+ "learning_rate": 1.2910691003911344e-05,
529
+ "loss": 0.3931,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 1.79,
534
+ "learning_rate": 1.2829204693611474e-05,
535
+ "loss": 0.3929,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 1.81,
540
+ "learning_rate": 1.2747718383311605e-05,
541
+ "loss": 0.3772,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 1.83,
546
+ "learning_rate": 1.2666232073011735e-05,
547
+ "loss": 0.3892,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 1.85,
552
+ "learning_rate": 1.2584745762711866e-05,
553
+ "loss": 0.3939,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 1.87,
558
+ "learning_rate": 1.2503259452411995e-05,
559
+ "loss": 0.3979,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 1.89,
564
+ "learning_rate": 1.2421773142112126e-05,
565
+ "loss": 0.3779,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 1.91,
570
+ "learning_rate": 1.2340286831812256e-05,
571
+ "loss": 0.3921,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 1.94,
576
+ "learning_rate": 1.2258800521512385e-05,
577
+ "loss": 0.3848,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 1.96,
582
+ "learning_rate": 1.2177314211212517e-05,
583
+ "loss": 0.3868,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 1.98,
588
+ "learning_rate": 1.2095827900912646e-05,
589
+ "loss": 0.3853,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 2.0,
594
+ "learning_rate": 1.2014341590612777e-05,
595
+ "loss": 0.3865,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 2.02,
600
+ "learning_rate": 1.193285528031291e-05,
601
+ "loss": 0.2994,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 2.04,
606
+ "learning_rate": 1.185136897001304e-05,
607
+ "loss": 0.2864,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 2.06,
612
+ "learning_rate": 1.176988265971317e-05,
613
+ "loss": 0.2874,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 2.08,
618
+ "learning_rate": 1.16883963494133e-05,
619
+ "loss": 0.2832,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 2.1,
624
+ "learning_rate": 1.160691003911343e-05,
625
+ "loss": 0.27,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 2.12,
630
+ "learning_rate": 1.1525423728813561e-05,
631
+ "loss": 0.2907,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 2.14,
636
+ "learning_rate": 1.144393741851369e-05,
637
+ "loss": 0.2772,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 2.16,
642
+ "learning_rate": 1.1362451108213822e-05,
643
+ "loss": 0.2823,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 2.18,
648
+ "learning_rate": 1.1280964797913951e-05,
649
+ "loss": 0.2755,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 2.2,
654
+ "learning_rate": 1.1199478487614082e-05,
655
+ "loss": 0.2722,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 2.22,
660
+ "learning_rate": 1.1117992177314212e-05,
661
+ "loss": 0.2835,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 2.24,
666
+ "learning_rate": 1.1036505867014341e-05,
667
+ "loss": 0.2825,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 2.26,
672
+ "learning_rate": 1.0955019556714473e-05,
673
+ "loss": 0.2781,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 2.28,
678
+ "learning_rate": 1.0873533246414602e-05,
679
+ "loss": 0.2866,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 2.3,
684
+ "learning_rate": 1.0792046936114733e-05,
685
+ "loss": 0.2839,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 2.32,
690
+ "learning_rate": 1.0710560625814863e-05,
691
+ "loss": 0.2915,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 2.34,
696
+ "learning_rate": 1.0629074315514994e-05,
697
+ "loss": 0.2876,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 2.36,
702
+ "learning_rate": 1.0547588005215125e-05,
703
+ "loss": 0.2768,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 2.38,
708
+ "learning_rate": 1.0466101694915256e-05,
709
+ "loss": 0.2825,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 2.4,
714
+ "learning_rate": 1.0384615384615386e-05,
715
+ "loss": 0.264,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 2.42,
720
+ "learning_rate": 1.0303129074315517e-05,
721
+ "loss": 0.2975,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 2.44,
726
+ "learning_rate": 1.0221642764015647e-05,
727
+ "loss": 0.2829,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 2.46,
732
+ "learning_rate": 1.0140156453715778e-05,
733
+ "loss": 0.2979,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 2.49,
738
+ "learning_rate": 1.0058670143415907e-05,
739
+ "loss": 0.2833,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 2.51,
744
+ "learning_rate": 9.977183833116037e-06,
745
+ "loss": 0.2856,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 2.53,
750
+ "learning_rate": 9.895697522816168e-06,
751
+ "loss": 0.291,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 2.55,
756
+ "learning_rate": 9.814211212516298e-06,
757
+ "loss": 0.2841,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 2.57,
762
+ "learning_rate": 9.732724902216429e-06,
763
+ "loss": 0.2806,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 2.59,
768
+ "learning_rate": 9.651238591916558e-06,
769
+ "loss": 0.2952,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 2.61,
774
+ "learning_rate": 9.56975228161669e-06,
775
+ "loss": 0.2876,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 2.63,
780
+ "learning_rate": 9.488265971316819e-06,
781
+ "loss": 0.285,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 2.65,
786
+ "learning_rate": 9.40677966101695e-06,
787
+ "loss": 0.28,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 2.67,
792
+ "learning_rate": 9.325293350717081e-06,
793
+ "loss": 0.2823,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 2.69,
798
+ "learning_rate": 9.24380704041721e-06,
799
+ "loss": 0.2918,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 2.71,
804
+ "learning_rate": 9.162320730117342e-06,
805
+ "loss": 0.2919,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 2.73,
810
+ "learning_rate": 9.080834419817471e-06,
811
+ "loss": 0.2888,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 2.75,
816
+ "learning_rate": 8.999348109517601e-06,
817
+ "loss": 0.2677,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 2.77,
822
+ "learning_rate": 8.917861799217732e-06,
823
+ "loss": 0.2879,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 2.79,
828
+ "learning_rate": 8.836375488917862e-06,
829
+ "loss": 0.2856,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 2.81,
834
+ "learning_rate": 8.754889178617993e-06,
835
+ "loss": 0.2923,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 2.83,
840
+ "learning_rate": 8.673402868318124e-06,
841
+ "loss": 0.2949,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 2.85,
846
+ "learning_rate": 8.591916558018254e-06,
847
+ "loss": 0.2826,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 2.87,
852
+ "learning_rate": 8.510430247718385e-06,
853
+ "loss": 0.2716,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 2.89,
858
+ "learning_rate": 8.428943937418514e-06,
859
+ "loss": 0.2858,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 2.91,
864
+ "learning_rate": 8.347457627118645e-06,
865
+ "loss": 0.284,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 2.93,
870
+ "learning_rate": 8.265971316818775e-06,
871
+ "loss": 0.2869,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 2.95,
876
+ "learning_rate": 8.184485006518904e-06,
877
+ "loss": 0.2937,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 2.97,
882
+ "learning_rate": 8.102998696219036e-06,
883
+ "loss": 0.2837,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 2.99,
888
+ "learning_rate": 8.021512385919165e-06,
889
+ "loss": 0.2783,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 3.01,
894
+ "learning_rate": 7.940026075619296e-06,
895
+ "loss": 0.2272,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 3.04,
900
+ "learning_rate": 7.858539765319428e-06,
901
+ "loss": 0.1973,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 3.06,
906
+ "learning_rate": 7.777053455019557e-06,
907
+ "loss": 0.2154,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 3.08,
912
+ "learning_rate": 7.695567144719688e-06,
913
+ "loss": 0.1972,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 3.1,
918
+ "learning_rate": 7.614080834419818e-06,
919
+ "loss": 0.213,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 3.12,
924
+ "learning_rate": 7.532594524119948e-06,
925
+ "loss": 0.1933,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 3.14,
930
+ "learning_rate": 7.451108213820078e-06,
931
+ "loss": 0.2167,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 3.16,
936
+ "learning_rate": 7.369621903520209e-06,
937
+ "loss": 0.1957,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 3.18,
942
+ "learning_rate": 7.288135593220339e-06,
943
+ "loss": 0.2187,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 3.2,
948
+ "learning_rate": 7.20664928292047e-06,
949
+ "loss": 0.2148,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 3.22,
954
+ "learning_rate": 7.125162972620601e-06,
955
+ "loss": 0.2187,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 3.24,
960
+ "learning_rate": 7.043676662320731e-06,
961
+ "loss": 0.2128,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 3.26,
966
+ "learning_rate": 6.962190352020861e-06,
967
+ "loss": 0.207,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 3.28,
972
+ "learning_rate": 6.880704041720992e-06,
973
+ "loss": 0.2116,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 3.3,
978
+ "learning_rate": 6.799217731421122e-06,
979
+ "loss": 0.2049,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 3.32,
984
+ "learning_rate": 6.7177314211212515e-06,
985
+ "loss": 0.2031,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 3.34,
990
+ "learning_rate": 6.636245110821382e-06,
991
+ "loss": 0.216,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 3.36,
996
+ "learning_rate": 6.554758800521513e-06,
997
+ "loss": 0.2082,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 3.38,
1002
+ "learning_rate": 6.473272490221643e-06,
1003
+ "loss": 0.2197,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 3.4,
1008
+ "learning_rate": 6.391786179921774e-06,
1009
+ "loss": 0.2109,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 3.42,
1014
+ "learning_rate": 6.310299869621904e-06,
1015
+ "loss": 0.2187,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 3.44,
1020
+ "learning_rate": 6.2288135593220344e-06,
1021
+ "loss": 0.2206,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 3.46,
1026
+ "learning_rate": 6.147327249022165e-06,
1027
+ "loss": 0.2038,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 3.48,
1032
+ "learning_rate": 6.065840938722295e-06,
1033
+ "loss": 0.2197,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 3.5,
1038
+ "learning_rate": 5.9843546284224255e-06,
1039
+ "loss": 0.2222,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 3.52,
1044
+ "learning_rate": 5.902868318122556e-06,
1045
+ "loss": 0.2109,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 3.54,
1050
+ "learning_rate": 5.821382007822687e-06,
1051
+ "loss": 0.2128,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 3.57,
1056
+ "learning_rate": 5.739895697522817e-06,
1057
+ "loss": 0.236,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 3.59,
1062
+ "learning_rate": 5.658409387222948e-06,
1063
+ "loss": 0.1958,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 3.61,
1068
+ "learning_rate": 5.576923076923077e-06,
1069
+ "loss": 0.2258,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 3.63,
1074
+ "learning_rate": 5.4954367666232076e-06,
1075
+ "loss": 0.2214,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 3.65,
1080
+ "learning_rate": 5.413950456323338e-06,
1081
+ "loss": 0.2241,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 3.67,
1086
+ "learning_rate": 5.332464146023468e-06,
1087
+ "loss": 0.2093,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 3.69,
1092
+ "learning_rate": 5.250977835723599e-06,
1093
+ "loss": 0.2215,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 3.71,
1098
+ "learning_rate": 5.169491525423729e-06,
1099
+ "loss": 0.2118,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 3.73,
1104
+ "learning_rate": 5.08800521512386e-06,
1105
+ "loss": 0.2216,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 3.75,
1110
+ "learning_rate": 5.0065189048239905e-06,
1111
+ "loss": 0.2036,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 3.77,
1116
+ "learning_rate": 4.92503259452412e-06,
1117
+ "loss": 0.2186,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 3.79,
1122
+ "learning_rate": 4.843546284224251e-06,
1123
+ "loss": 0.2201,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 3.81,
1128
+ "learning_rate": 4.7620599739243815e-06,
1129
+ "loss": 0.2036,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 3.83,
1134
+ "learning_rate": 4.680573663624511e-06,
1135
+ "loss": 0.2319,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 3.85,
1140
+ "learning_rate": 4.599087353324641e-06,
1141
+ "loss": 0.2126,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 3.87,
1146
+ "learning_rate": 4.5176010430247726e-06,
1147
+ "loss": 0.2246,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 3.89,
1152
+ "learning_rate": 4.436114732724903e-06,
1153
+ "loss": 0.2269,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 3.91,
1158
+ "learning_rate": 4.354628422425033e-06,
1159
+ "loss": 0.2074,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 3.93,
1164
+ "learning_rate": 4.273142112125163e-06,
1165
+ "loss": 0.2219,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 3.95,
1170
+ "learning_rate": 4.191655801825294e-06,
1171
+ "loss": 0.1998,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 3.97,
1176
+ "learning_rate": 4.110169491525424e-06,
1177
+ "loss": 0.2124,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 3.99,
1182
+ "learning_rate": 4.028683181225555e-06,
1183
+ "loss": 0.2225,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 4.01,
1188
+ "learning_rate": 3.947196870925685e-06,
1189
+ "loss": 0.1845,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 4.03,
1194
+ "learning_rate": 3.865710560625815e-06,
1195
+ "loss": 0.156,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 4.05,
1200
+ "learning_rate": 3.7842242503259457e-06,
1201
+ "loss": 0.1695,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 4.07,
1206
+ "learning_rate": 3.702737940026076e-06,
1207
+ "loss": 0.1729,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 4.09,
1212
+ "learning_rate": 3.6212516297262064e-06,
1213
+ "loss": 0.159,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 4.12,
1218
+ "learning_rate": 3.5397653194263363e-06,
1219
+ "loss": 0.1735,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 4.14,
1224
+ "learning_rate": 3.4582790091264675e-06,
1225
+ "loss": 0.1683,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 4.16,
1230
+ "learning_rate": 3.3767926988265974e-06,
1231
+ "loss": 0.1734,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 4.18,
1236
+ "learning_rate": 3.2953063885267278e-06,
1237
+ "loss": 0.1575,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 4.2,
1242
+ "learning_rate": 3.213820078226858e-06,
1243
+ "loss": 0.1643,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 4.22,
1248
+ "learning_rate": 3.1323337679269885e-06,
1249
+ "loss": 0.1626,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 4.24,
1254
+ "learning_rate": 3.0508474576271192e-06,
1255
+ "loss": 0.1631,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 4.26,
1260
+ "learning_rate": 2.969361147327249e-06,
1261
+ "loss": 0.1731,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 4.28,
1266
+ "learning_rate": 2.8878748370273795e-06,
1267
+ "loss": 0.1729,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 4.3,
1272
+ "learning_rate": 2.80638852672751e-06,
1273
+ "loss": 0.1658,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 4.32,
1278
+ "learning_rate": 2.7249022164276406e-06,
1279
+ "loss": 0.1641,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 4.34,
1284
+ "learning_rate": 2.643415906127771e-06,
1285
+ "loss": 0.1716,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 4.36,
1290
+ "learning_rate": 2.5619295958279013e-06,
1291
+ "loss": 0.1765,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 4.38,
1296
+ "learning_rate": 2.4804432855280312e-06,
1297
+ "loss": 0.1713,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 4.4,
1302
+ "learning_rate": 2.398956975228162e-06,
1303
+ "loss": 0.1716,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 4.42,
1308
+ "learning_rate": 2.3174706649282924e-06,
1309
+ "loss": 0.1754,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 4.44,
1314
+ "learning_rate": 2.2359843546284227e-06,
1315
+ "loss": 0.1551,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 4.46,
1320
+ "learning_rate": 2.154498044328553e-06,
1321
+ "loss": 0.1704,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 4.48,
1326
+ "learning_rate": 2.0730117340286834e-06,
1327
+ "loss": 0.1744,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 4.5,
1332
+ "learning_rate": 1.9915254237288137e-06,
1333
+ "loss": 0.1608,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 4.52,
1338
+ "learning_rate": 1.910039113428944e-06,
1339
+ "loss": 0.1731,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 4.54,
1344
+ "learning_rate": 1.8285528031290744e-06,
1345
+ "loss": 0.1635,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 4.56,
1350
+ "learning_rate": 1.7470664928292048e-06,
1351
+ "loss": 0.1539,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 4.58,
1356
+ "learning_rate": 1.6655801825293353e-06,
1357
+ "loss": 0.1608,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 4.6,
1362
+ "learning_rate": 1.5840938722294655e-06,
1363
+ "loss": 0.1732,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 4.62,
1368
+ "learning_rate": 1.502607561929596e-06,
1369
+ "loss": 0.1554,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 4.64,
1374
+ "learning_rate": 1.4211212516297262e-06,
1375
+ "loss": 0.1719,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 4.67,
1380
+ "learning_rate": 1.3396349413298567e-06,
1381
+ "loss": 0.1605,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 4.69,
1386
+ "learning_rate": 1.258148631029987e-06,
1387
+ "loss": 0.1698,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 4.71,
1392
+ "learning_rate": 1.1766623207301174e-06,
1393
+ "loss": 0.1686,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 4.73,
1398
+ "learning_rate": 1.0951760104302478e-06,
1399
+ "loss": 0.1651,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 4.75,
1404
+ "learning_rate": 1.0136897001303781e-06,
1405
+ "loss": 0.1647,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 4.77,
1410
+ "learning_rate": 9.322033898305086e-07,
1411
+ "loss": 0.1735,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 4.79,
1416
+ "learning_rate": 8.507170795306389e-07,
1417
+ "loss": 0.1578,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 4.81,
1422
+ "learning_rate": 7.692307692307694e-07,
1423
+ "loss": 0.1729,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 4.83,
1428
+ "learning_rate": 6.877444589308997e-07,
1429
+ "loss": 0.1682,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 4.85,
1434
+ "learning_rate": 6.0625814863103e-07,
1435
+ "loss": 0.1585,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 4.87,
1440
+ "learning_rate": 5.247718383311604e-07,
1441
+ "loss": 0.1605,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 4.89,
1446
+ "learning_rate": 4.432855280312908e-07,
1447
+ "loss": 0.1629,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 4.91,
1452
+ "learning_rate": 3.6179921773142114e-07,
1453
+ "loss": 0.1504,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 4.93,
1458
+ "learning_rate": 2.803129074315515e-07,
1459
+ "loss": 0.1706,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 4.95,
1464
+ "learning_rate": 1.988265971316819e-07,
1465
+ "loss": 0.16,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 4.97,
1470
+ "learning_rate": 1.1734028683181226e-07,
1471
+ "loss": 0.1641,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 4.99,
1476
+ "learning_rate": 3.585397653194264e-08,
1477
+ "loss": 0.1709,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 5.0,
1482
+ "step": 122720,
1483
+ "total_flos": 98499428530398720,
1484
+ "train_runtime": 14248.9446,
1485
+ "train_samples_per_second": 8.613
1486
+ }
1487
+ ],
1488
+ "max_steps": 122720,
1489
+ "num_train_epochs": 5,
1490
+ "total_flos": 98499428530398720,
1491
+ "trial_name": null,
1492
+ "trial_params": null
1493
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4260388ec322db77beec97fb72f7af3c54e948a8396b8ce3270a6cc6aa8a7604
3
+ size 2159
vocab.txt ADDED
The diff for this file is too large to render. See raw diff