selmamalak commited on
Commit
06d3fe6
1 Parent(s): ec95a74

End of training

Browse files
Files changed (5) hide show
  1. README.md +5 -5
  2. all_results.json +16 -0
  3. eval_results.json +11 -0
  4. train_results.json +8 -0
  5. trainer_state.json +1571 -0
README.md CHANGED
@@ -23,11 +23,11 @@ should probably proofread and complete it, then remove this comment. -->
23
 
24
  This model is a fine-tuned version of [facebook/deit-base-patch16-224](https://huggingface.co/facebook/deit-base-patch16-224) on the medmnist-v2 dataset.
25
  It achieves the following results on the evaluation set:
26
- - Loss: 0.0745
27
- - Accuracy: 0.9870
28
- - Precision: 0.9877
29
- - Recall: 0.9863
30
- - F1: 0.9868
31
 
32
  ## Model description
33
 
 
23
 
24
  This model is a fine-tuned version of [facebook/deit-base-patch16-224](https://huggingface.co/facebook/deit-base-patch16-224) on the medmnist-v2 dataset.
25
  It achieves the following results on the evaluation set:
26
+ - Loss: 0.2795
27
+ - Accuracy: 0.9240
28
+ - Precision: 0.9199
29
+ - Recall: 0.9123
30
+ - F1: 0.9154
31
 
32
  ## Model description
33
 
all_results.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.99,
3
+ "eval_accuracy": 0.9240445089501693,
4
+ "eval_f1": 0.9153681639608009,
5
+ "eval_loss": 0.2795476019382477,
6
+ "eval_precision": 0.9198770454998186,
7
+ "eval_recall": 0.912344054362017,
8
+ "eval_runtime": 43.4915,
9
+ "eval_samples_per_second": 190.106,
10
+ "eval_steps_per_second": 11.887,
11
+ "total_flos": 1.0133154899356189e+19,
12
+ "train_loss": 0.5248493043072705,
13
+ "train_runtime": 1518.4893,
14
+ "train_samples_per_second": 85.611,
15
+ "train_steps_per_second": 1.337
16
+ }
eval_results.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.99,
3
+ "eval_accuracy": 0.9240445089501693,
4
+ "eval_f1": 0.9153681639608009,
5
+ "eval_loss": 0.2795476019382477,
6
+ "eval_precision": 0.9198770454998186,
7
+ "eval_recall": 0.912344054362017,
8
+ "eval_runtime": 43.4915,
9
+ "eval_samples_per_second": 190.106,
10
+ "eval_steps_per_second": 11.887
11
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.99,
3
+ "total_flos": 1.0133154899356189e+19,
4
+ "train_loss": 0.5248493043072705,
5
+ "train_runtime": 1518.4893,
6
+ "train_samples_per_second": 85.611,
7
+ "train_steps_per_second": 1.337
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1571 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9870401337792643,
3
+ "best_model_checkpoint": "deit-base-patch16-224-finetuned-lora-medmnistv2/checkpoint-2030",
4
+ "epoch": 9.98769987699877,
5
+ "eval_steps": 500,
6
+ "global_step": 2030,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.05,
13
+ "grad_norm": 3.303891658782959,
14
+ "learning_rate": 0.004975369458128079,
15
+ "loss": 1.8368,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.1,
20
+ "grad_norm": 2.481412649154663,
21
+ "learning_rate": 0.004950738916256157,
22
+ "loss": 1.2897,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.15,
27
+ "grad_norm": 2.0582990646362305,
28
+ "learning_rate": 0.0049261083743842365,
29
+ "loss": 1.0672,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.2,
34
+ "grad_norm": 2.254044532775879,
35
+ "learning_rate": 0.004901477832512316,
36
+ "loss": 0.9723,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.25,
41
+ "grad_norm": 2.3351266384124756,
42
+ "learning_rate": 0.004876847290640395,
43
+ "loss": 0.917,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.3,
48
+ "grad_norm": 1.9788981676101685,
49
+ "learning_rate": 0.004852216748768473,
50
+ "loss": 0.8483,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.34,
55
+ "grad_norm": 1.460303544998169,
56
+ "learning_rate": 0.004827586206896552,
57
+ "loss": 0.7848,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.39,
62
+ "grad_norm": 3.4298815727233887,
63
+ "learning_rate": 0.004802955665024631,
64
+ "loss": 0.8819,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.44,
69
+ "grad_norm": 1.531933069229126,
70
+ "learning_rate": 0.004778325123152709,
71
+ "loss": 0.7758,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.49,
76
+ "grad_norm": 1.608288049697876,
77
+ "learning_rate": 0.004753694581280788,
78
+ "loss": 0.7678,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.54,
83
+ "grad_norm": 1.9291895627975464,
84
+ "learning_rate": 0.004729064039408867,
85
+ "loss": 0.7848,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.59,
90
+ "grad_norm": 1.8717544078826904,
91
+ "learning_rate": 0.004704433497536946,
92
+ "loss": 0.7742,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.64,
97
+ "grad_norm": 4.476926803588867,
98
+ "learning_rate": 0.004679802955665025,
99
+ "loss": 0.8906,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.69,
104
+ "grad_norm": 2.5952930450439453,
105
+ "learning_rate": 0.004655172413793103,
106
+ "loss": 0.8464,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.74,
111
+ "grad_norm": 1.8514671325683594,
112
+ "learning_rate": 0.004630541871921182,
113
+ "loss": 0.9079,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.79,
118
+ "grad_norm": 1.962122917175293,
119
+ "learning_rate": 0.004605911330049261,
120
+ "loss": 0.81,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.84,
125
+ "grad_norm": 1.9573622941970825,
126
+ "learning_rate": 0.00458128078817734,
127
+ "loss": 0.8099,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.89,
132
+ "grad_norm": 1.8094934225082397,
133
+ "learning_rate": 0.004559113300492611,
134
+ "loss": 0.8105,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.93,
139
+ "grad_norm": 2.7324118614196777,
140
+ "learning_rate": 0.00453448275862069,
141
+ "loss": 0.8123,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.98,
146
+ "grad_norm": 2.355945348739624,
147
+ "learning_rate": 0.004509852216748769,
148
+ "loss": 0.7947,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 1.0,
153
+ "eval_accuracy": 0.8975752508361204,
154
+ "eval_f1": 0.863152897342088,
155
+ "eval_loss": 0.3122749328613281,
156
+ "eval_precision": 0.909035520710901,
157
+ "eval_recall": 0.8450098410817735,
158
+ "eval_runtime": 12.4051,
159
+ "eval_samples_per_second": 192.824,
160
+ "eval_steps_per_second": 12.092,
161
+ "step": 203
162
+ },
163
+ {
164
+ "epoch": 1.03,
165
+ "grad_norm": 1.4632035493850708,
166
+ "learning_rate": 0.004485221674876847,
167
+ "loss": 0.8797,
168
+ "step": 210
169
+ },
170
+ {
171
+ "epoch": 1.08,
172
+ "grad_norm": 2.434492349624634,
173
+ "learning_rate": 0.004460591133004926,
174
+ "loss": 0.6601,
175
+ "step": 220
176
+ },
177
+ {
178
+ "epoch": 1.13,
179
+ "grad_norm": 1.9131174087524414,
180
+ "learning_rate": 0.004435960591133005,
181
+ "loss": 0.7093,
182
+ "step": 230
183
+ },
184
+ {
185
+ "epoch": 1.18,
186
+ "grad_norm": 1.491714358329773,
187
+ "learning_rate": 0.004411330049261084,
188
+ "loss": 0.7247,
189
+ "step": 240
190
+ },
191
+ {
192
+ "epoch": 1.23,
193
+ "grad_norm": 1.9442235231399536,
194
+ "learning_rate": 0.004386699507389163,
195
+ "loss": 0.7182,
196
+ "step": 250
197
+ },
198
+ {
199
+ "epoch": 1.28,
200
+ "grad_norm": 1.9666441679000854,
201
+ "learning_rate": 0.004362068965517241,
202
+ "loss": 0.733,
203
+ "step": 260
204
+ },
205
+ {
206
+ "epoch": 1.33,
207
+ "grad_norm": 1.89641273021698,
208
+ "learning_rate": 0.00433743842364532,
209
+ "loss": 0.6678,
210
+ "step": 270
211
+ },
212
+ {
213
+ "epoch": 1.38,
214
+ "grad_norm": 1.3621200323104858,
215
+ "learning_rate": 0.004312807881773399,
216
+ "loss": 0.7066,
217
+ "step": 280
218
+ },
219
+ {
220
+ "epoch": 1.43,
221
+ "grad_norm": 1.8440511226654053,
222
+ "learning_rate": 0.004288177339901478,
223
+ "loss": 0.673,
224
+ "step": 290
225
+ },
226
+ {
227
+ "epoch": 1.48,
228
+ "grad_norm": 1.712856650352478,
229
+ "learning_rate": 0.0042635467980295565,
230
+ "loss": 0.7424,
231
+ "step": 300
232
+ },
233
+ {
234
+ "epoch": 1.53,
235
+ "grad_norm": 1.674052357673645,
236
+ "learning_rate": 0.004238916256157636,
237
+ "loss": 0.6104,
238
+ "step": 310
239
+ },
240
+ {
241
+ "epoch": 1.57,
242
+ "grad_norm": 1.5503976345062256,
243
+ "learning_rate": 0.004214285714285715,
244
+ "loss": 0.6868,
245
+ "step": 320
246
+ },
247
+ {
248
+ "epoch": 1.62,
249
+ "grad_norm": 1.0976303815841675,
250
+ "learning_rate": 0.004189655172413793,
251
+ "loss": 0.703,
252
+ "step": 330
253
+ },
254
+ {
255
+ "epoch": 1.67,
256
+ "grad_norm": 1.5332385301589966,
257
+ "learning_rate": 0.004165024630541872,
258
+ "loss": 0.6599,
259
+ "step": 340
260
+ },
261
+ {
262
+ "epoch": 1.72,
263
+ "grad_norm": 1.4168181419372559,
264
+ "learning_rate": 0.004140394088669951,
265
+ "loss": 0.7273,
266
+ "step": 350
267
+ },
268
+ {
269
+ "epoch": 1.77,
270
+ "grad_norm": 2.414102792739868,
271
+ "learning_rate": 0.00411576354679803,
272
+ "loss": 0.6551,
273
+ "step": 360
274
+ },
275
+ {
276
+ "epoch": 1.82,
277
+ "grad_norm": 1.9595593214035034,
278
+ "learning_rate": 0.004091133004926108,
279
+ "loss": 0.7608,
280
+ "step": 370
281
+ },
282
+ {
283
+ "epoch": 1.87,
284
+ "grad_norm": 1.0985585451126099,
285
+ "learning_rate": 0.0040665024630541875,
286
+ "loss": 0.6946,
287
+ "step": 380
288
+ },
289
+ {
290
+ "epoch": 1.92,
291
+ "grad_norm": 2.288224458694458,
292
+ "learning_rate": 0.004041871921182267,
293
+ "loss": 0.7381,
294
+ "step": 390
295
+ },
296
+ {
297
+ "epoch": 1.97,
298
+ "grad_norm": 1.385890245437622,
299
+ "learning_rate": 0.004017241379310345,
300
+ "loss": 0.6703,
301
+ "step": 400
302
+ },
303
+ {
304
+ "epoch": 2.0,
305
+ "eval_accuracy": 0.9607023411371237,
306
+ "eval_f1": 0.9535454483827537,
307
+ "eval_loss": 0.14003877341747284,
308
+ "eval_precision": 0.9589551276899428,
309
+ "eval_recall": 0.9543228146341872,
310
+ "eval_runtime": 12.5652,
311
+ "eval_samples_per_second": 190.367,
312
+ "eval_steps_per_second": 11.938,
313
+ "step": 406
314
+ },
315
+ {
316
+ "epoch": 2.02,
317
+ "grad_norm": 2.190495729446411,
318
+ "learning_rate": 0.003992610837438423,
319
+ "loss": 0.6558,
320
+ "step": 410
321
+ },
322
+ {
323
+ "epoch": 2.07,
324
+ "grad_norm": 2.319401979446411,
325
+ "learning_rate": 0.003967980295566502,
326
+ "loss": 0.618,
327
+ "step": 420
328
+ },
329
+ {
330
+ "epoch": 2.12,
331
+ "grad_norm": 1.5089792013168335,
332
+ "learning_rate": 0.003943349753694581,
333
+ "loss": 0.6622,
334
+ "step": 430
335
+ },
336
+ {
337
+ "epoch": 2.16,
338
+ "grad_norm": 1.2998738288879395,
339
+ "learning_rate": 0.00391871921182266,
340
+ "loss": 0.6039,
341
+ "step": 440
342
+ },
343
+ {
344
+ "epoch": 2.21,
345
+ "grad_norm": 1.5582971572875977,
346
+ "learning_rate": 0.003894088669950739,
347
+ "loss": 0.585,
348
+ "step": 450
349
+ },
350
+ {
351
+ "epoch": 2.26,
352
+ "grad_norm": 2.1443846225738525,
353
+ "learning_rate": 0.0038694581280788176,
354
+ "loss": 0.6739,
355
+ "step": 460
356
+ },
357
+ {
358
+ "epoch": 2.31,
359
+ "grad_norm": 1.1868767738342285,
360
+ "learning_rate": 0.0038448275862068967,
361
+ "loss": 0.6598,
362
+ "step": 470
363
+ },
364
+ {
365
+ "epoch": 2.36,
366
+ "grad_norm": 1.5321897268295288,
367
+ "learning_rate": 0.0038201970443349754,
368
+ "loss": 0.6058,
369
+ "step": 480
370
+ },
371
+ {
372
+ "epoch": 2.41,
373
+ "grad_norm": 1.2971707582473755,
374
+ "learning_rate": 0.0037955665024630545,
375
+ "loss": 0.6025,
376
+ "step": 490
377
+ },
378
+ {
379
+ "epoch": 2.46,
380
+ "grad_norm": 1.0405155420303345,
381
+ "learning_rate": 0.003770935960591133,
382
+ "loss": 0.6544,
383
+ "step": 500
384
+ },
385
+ {
386
+ "epoch": 2.51,
387
+ "grad_norm": 2.27400541305542,
388
+ "learning_rate": 0.0037463054187192118,
389
+ "loss": 0.639,
390
+ "step": 510
391
+ },
392
+ {
393
+ "epoch": 2.56,
394
+ "grad_norm": 1.8367363214492798,
395
+ "learning_rate": 0.003721674876847291,
396
+ "loss": 0.6922,
397
+ "step": 520
398
+ },
399
+ {
400
+ "epoch": 2.61,
401
+ "grad_norm": 1.6862225532531738,
402
+ "learning_rate": 0.0036970443349753695,
403
+ "loss": 0.6698,
404
+ "step": 530
405
+ },
406
+ {
407
+ "epoch": 2.66,
408
+ "grad_norm": 1.1783074140548706,
409
+ "learning_rate": 0.0036724137931034486,
410
+ "loss": 0.648,
411
+ "step": 540
412
+ },
413
+ {
414
+ "epoch": 2.71,
415
+ "grad_norm": 1.327495813369751,
416
+ "learning_rate": 0.0036477832512315273,
417
+ "loss": 0.5485,
418
+ "step": 550
419
+ },
420
+ {
421
+ "epoch": 2.76,
422
+ "grad_norm": 1.2704271078109741,
423
+ "learning_rate": 0.0036231527093596064,
424
+ "loss": 0.6125,
425
+ "step": 560
426
+ },
427
+ {
428
+ "epoch": 2.8,
429
+ "grad_norm": 1.412690281867981,
430
+ "learning_rate": 0.003598522167487685,
431
+ "loss": 0.5872,
432
+ "step": 570
433
+ },
434
+ {
435
+ "epoch": 2.85,
436
+ "grad_norm": 1.5771632194519043,
437
+ "learning_rate": 0.0035738916256157637,
438
+ "loss": 0.541,
439
+ "step": 580
440
+ },
441
+ {
442
+ "epoch": 2.9,
443
+ "grad_norm": 1.2916010618209839,
444
+ "learning_rate": 0.0035492610837438428,
445
+ "loss": 0.6637,
446
+ "step": 590
447
+ },
448
+ {
449
+ "epoch": 2.95,
450
+ "grad_norm": 1.0224180221557617,
451
+ "learning_rate": 0.003524630541871921,
452
+ "loss": 0.5941,
453
+ "step": 600
454
+ },
455
+ {
456
+ "epoch": 3.0,
457
+ "eval_accuracy": 0.9698996655518395,
458
+ "eval_f1": 0.9649463741495116,
459
+ "eval_loss": 0.11816500872373581,
460
+ "eval_precision": 0.9646998653209304,
461
+ "eval_recall": 0.9681395759866063,
462
+ "eval_runtime": 12.5344,
463
+ "eval_samples_per_second": 190.834,
464
+ "eval_steps_per_second": 11.967,
465
+ "step": 609
466
+ },
467
+ {
468
+ "epoch": 3.0,
469
+ "grad_norm": 1.8427205085754395,
470
+ "learning_rate": 0.0034999999999999996,
471
+ "loss": 0.6179,
472
+ "step": 610
473
+ },
474
+ {
475
+ "epoch": 3.05,
476
+ "grad_norm": 1.1675821542739868,
477
+ "learning_rate": 0.0034753694581280787,
478
+ "loss": 0.6263,
479
+ "step": 620
480
+ },
481
+ {
482
+ "epoch": 3.1,
483
+ "grad_norm": 1.6908611059188843,
484
+ "learning_rate": 0.0034507389162561574,
485
+ "loss": 0.7175,
486
+ "step": 630
487
+ },
488
+ {
489
+ "epoch": 3.15,
490
+ "grad_norm": 1.5712032318115234,
491
+ "learning_rate": 0.0034261083743842365,
492
+ "loss": 0.6474,
493
+ "step": 640
494
+ },
495
+ {
496
+ "epoch": 3.2,
497
+ "grad_norm": 1.8690963983535767,
498
+ "learning_rate": 0.003401477832512315,
499
+ "loss": 0.5849,
500
+ "step": 650
501
+ },
502
+ {
503
+ "epoch": 3.25,
504
+ "grad_norm": 1.6917773485183716,
505
+ "learning_rate": 0.0033768472906403942,
506
+ "loss": 0.5954,
507
+ "step": 660
508
+ },
509
+ {
510
+ "epoch": 3.3,
511
+ "grad_norm": 1.4844752550125122,
512
+ "learning_rate": 0.003352216748768473,
513
+ "loss": 0.6284,
514
+ "step": 670
515
+ },
516
+ {
517
+ "epoch": 3.35,
518
+ "grad_norm": 1.444581389427185,
519
+ "learning_rate": 0.003327586206896552,
520
+ "loss": 0.5529,
521
+ "step": 680
522
+ },
523
+ {
524
+ "epoch": 3.39,
525
+ "grad_norm": 1.3921010494232178,
526
+ "learning_rate": 0.0033029556650246306,
527
+ "loss": 0.6022,
528
+ "step": 690
529
+ },
530
+ {
531
+ "epoch": 3.44,
532
+ "grad_norm": 1.3489701747894287,
533
+ "learning_rate": 0.0032783251231527093,
534
+ "loss": 0.6314,
535
+ "step": 700
536
+ },
537
+ {
538
+ "epoch": 3.49,
539
+ "grad_norm": 1.7876464128494263,
540
+ "learning_rate": 0.0032536945812807884,
541
+ "loss": 0.5322,
542
+ "step": 710
543
+ },
544
+ {
545
+ "epoch": 3.54,
546
+ "grad_norm": 1.2738828659057617,
547
+ "learning_rate": 0.003229064039408867,
548
+ "loss": 0.5869,
549
+ "step": 720
550
+ },
551
+ {
552
+ "epoch": 3.59,
553
+ "grad_norm": 1.5368149280548096,
554
+ "learning_rate": 0.003204433497536946,
555
+ "loss": 0.5659,
556
+ "step": 730
557
+ },
558
+ {
559
+ "epoch": 3.64,
560
+ "grad_norm": 1.890324354171753,
561
+ "learning_rate": 0.0031798029556650248,
562
+ "loss": 0.5506,
563
+ "step": 740
564
+ },
565
+ {
566
+ "epoch": 3.69,
567
+ "grad_norm": 1.553797721862793,
568
+ "learning_rate": 0.003155172413793104,
569
+ "loss": 0.6645,
570
+ "step": 750
571
+ },
572
+ {
573
+ "epoch": 3.74,
574
+ "grad_norm": 1.3873250484466553,
575
+ "learning_rate": 0.0031305418719211825,
576
+ "loss": 0.5013,
577
+ "step": 760
578
+ },
579
+ {
580
+ "epoch": 3.79,
581
+ "grad_norm": 1.6613869667053223,
582
+ "learning_rate": 0.0031059113300492616,
583
+ "loss": 0.4614,
584
+ "step": 770
585
+ },
586
+ {
587
+ "epoch": 3.84,
588
+ "grad_norm": 1.3628942966461182,
589
+ "learning_rate": 0.00308128078817734,
590
+ "loss": 0.5874,
591
+ "step": 780
592
+ },
593
+ {
594
+ "epoch": 3.89,
595
+ "grad_norm": 1.5102113485336304,
596
+ "learning_rate": 0.0030566502463054185,
597
+ "loss": 0.5564,
598
+ "step": 790
599
+ },
600
+ {
601
+ "epoch": 3.94,
602
+ "grad_norm": 1.4211273193359375,
603
+ "learning_rate": 0.0030320197044334976,
604
+ "loss": 0.6018,
605
+ "step": 800
606
+ },
607
+ {
608
+ "epoch": 3.99,
609
+ "grad_norm": 1.8017326593399048,
610
+ "learning_rate": 0.0030073891625615762,
611
+ "loss": 0.5837,
612
+ "step": 810
613
+ },
614
+ {
615
+ "epoch": 4.0,
616
+ "eval_accuracy": 0.967809364548495,
617
+ "eval_f1": 0.9551154966770515,
618
+ "eval_loss": 0.10157252848148346,
619
+ "eval_precision": 0.9557533496633682,
620
+ "eval_recall": 0.9586131276038764,
621
+ "eval_runtime": 12.6536,
622
+ "eval_samples_per_second": 189.038,
623
+ "eval_steps_per_second": 11.854,
624
+ "step": 813
625
+ },
626
+ {
627
+ "epoch": 4.03,
628
+ "grad_norm": 1.3088445663452148,
629
+ "learning_rate": 0.002982758620689655,
630
+ "loss": 0.5057,
631
+ "step": 820
632
+ },
633
+ {
634
+ "epoch": 4.08,
635
+ "grad_norm": 1.200412631034851,
636
+ "learning_rate": 0.002958128078817734,
637
+ "loss": 0.5485,
638
+ "step": 830
639
+ },
640
+ {
641
+ "epoch": 4.13,
642
+ "grad_norm": 1.6468169689178467,
643
+ "learning_rate": 0.0029334975369458127,
644
+ "loss": 0.5171,
645
+ "step": 840
646
+ },
647
+ {
648
+ "epoch": 4.18,
649
+ "grad_norm": 1.0748703479766846,
650
+ "learning_rate": 0.0029088669950738917,
651
+ "loss": 0.5664,
652
+ "step": 850
653
+ },
654
+ {
655
+ "epoch": 4.23,
656
+ "grad_norm": 1.396888256072998,
657
+ "learning_rate": 0.0028842364532019704,
658
+ "loss": 0.4641,
659
+ "step": 860
660
+ },
661
+ {
662
+ "epoch": 4.28,
663
+ "grad_norm": 1.0845372676849365,
664
+ "learning_rate": 0.0028596059113300495,
665
+ "loss": 0.5789,
666
+ "step": 870
667
+ },
668
+ {
669
+ "epoch": 4.33,
670
+ "grad_norm": 1.4134384393692017,
671
+ "learning_rate": 0.002834975369458128,
672
+ "loss": 0.4361,
673
+ "step": 880
674
+ },
675
+ {
676
+ "epoch": 4.38,
677
+ "grad_norm": 0.7656651735305786,
678
+ "learning_rate": 0.002810344827586207,
679
+ "loss": 0.5938,
680
+ "step": 890
681
+ },
682
+ {
683
+ "epoch": 4.43,
684
+ "grad_norm": 1.459712028503418,
685
+ "learning_rate": 0.002785714285714286,
686
+ "loss": 0.5146,
687
+ "step": 900
688
+ },
689
+ {
690
+ "epoch": 4.48,
691
+ "grad_norm": 1.2046053409576416,
692
+ "learning_rate": 0.0027610837438423646,
693
+ "loss": 0.4882,
694
+ "step": 910
695
+ },
696
+ {
697
+ "epoch": 4.53,
698
+ "grad_norm": 1.1301757097244263,
699
+ "learning_rate": 0.0027364532019704436,
700
+ "loss": 0.4728,
701
+ "step": 920
702
+ },
703
+ {
704
+ "epoch": 4.58,
705
+ "grad_norm": 1.255055546760559,
706
+ "learning_rate": 0.0027118226600985223,
707
+ "loss": 0.4384,
708
+ "step": 930
709
+ },
710
+ {
711
+ "epoch": 4.62,
712
+ "grad_norm": 1.3792164325714111,
713
+ "learning_rate": 0.0026871921182266014,
714
+ "loss": 0.5357,
715
+ "step": 940
716
+ },
717
+ {
718
+ "epoch": 4.67,
719
+ "grad_norm": 1.3066402673721313,
720
+ "learning_rate": 0.00266256157635468,
721
+ "loss": 0.5361,
722
+ "step": 950
723
+ },
724
+ {
725
+ "epoch": 4.72,
726
+ "grad_norm": 1.2377945184707642,
727
+ "learning_rate": 0.002637931034482759,
728
+ "loss": 0.5334,
729
+ "step": 960
730
+ },
731
+ {
732
+ "epoch": 4.77,
733
+ "grad_norm": 1.3673447370529175,
734
+ "learning_rate": 0.0026133004926108374,
735
+ "loss": 0.4896,
736
+ "step": 970
737
+ },
738
+ {
739
+ "epoch": 4.82,
740
+ "grad_norm": 1.8711413145065308,
741
+ "learning_rate": 0.002588669950738916,
742
+ "loss": 0.4729,
743
+ "step": 980
744
+ },
745
+ {
746
+ "epoch": 4.87,
747
+ "grad_norm": 1.1367807388305664,
748
+ "learning_rate": 0.002564039408866995,
749
+ "loss": 0.483,
750
+ "step": 990
751
+ },
752
+ {
753
+ "epoch": 4.92,
754
+ "grad_norm": 1.5432896614074707,
755
+ "learning_rate": 0.0025394088669950738,
756
+ "loss": 0.5477,
757
+ "step": 1000
758
+ },
759
+ {
760
+ "epoch": 4.97,
761
+ "grad_norm": 1.0067399740219116,
762
+ "learning_rate": 0.0025147783251231524,
763
+ "loss": 0.5193,
764
+ "step": 1010
765
+ },
766
+ {
767
+ "epoch": 5.0,
768
+ "eval_accuracy": 0.9790969899665551,
769
+ "eval_f1": 0.9675306891159455,
770
+ "eval_loss": 0.08001040667295456,
771
+ "eval_precision": 0.9700832487729493,
772
+ "eval_recall": 0.9684116828701858,
773
+ "eval_runtime": 12.5434,
774
+ "eval_samples_per_second": 190.698,
775
+ "eval_steps_per_second": 11.958,
776
+ "step": 1016
777
+ },
778
+ {
779
+ "epoch": 5.02,
780
+ "grad_norm": 1.8359023332595825,
781
+ "learning_rate": 0.0024901477832512315,
782
+ "loss": 0.5087,
783
+ "step": 1020
784
+ },
785
+ {
786
+ "epoch": 5.07,
787
+ "grad_norm": 1.3961881399154663,
788
+ "learning_rate": 0.00246551724137931,
789
+ "loss": 0.4565,
790
+ "step": 1030
791
+ },
792
+ {
793
+ "epoch": 5.12,
794
+ "grad_norm": 1.9095091819763184,
795
+ "learning_rate": 0.0024408866995073893,
796
+ "loss": 0.4432,
797
+ "step": 1040
798
+ },
799
+ {
800
+ "epoch": 5.17,
801
+ "grad_norm": 1.2952779531478882,
802
+ "learning_rate": 0.002416256157635468,
803
+ "loss": 0.4392,
804
+ "step": 1050
805
+ },
806
+ {
807
+ "epoch": 5.22,
808
+ "grad_norm": 1.309617042541504,
809
+ "learning_rate": 0.002391625615763547,
810
+ "loss": 0.5255,
811
+ "step": 1060
812
+ },
813
+ {
814
+ "epoch": 5.26,
815
+ "grad_norm": 1.497014045715332,
816
+ "learning_rate": 0.0023669950738916257,
817
+ "loss": 0.5122,
818
+ "step": 1070
819
+ },
820
+ {
821
+ "epoch": 5.31,
822
+ "grad_norm": 1.3211737871170044,
823
+ "learning_rate": 0.0023423645320197048,
824
+ "loss": 0.5529,
825
+ "step": 1080
826
+ },
827
+ {
828
+ "epoch": 5.36,
829
+ "grad_norm": 0.9946479797363281,
830
+ "learning_rate": 0.0023177339901477834,
831
+ "loss": 0.4708,
832
+ "step": 1090
833
+ },
834
+ {
835
+ "epoch": 5.41,
836
+ "grad_norm": 0.8456437587738037,
837
+ "learning_rate": 0.002293103448275862,
838
+ "loss": 0.4935,
839
+ "step": 1100
840
+ },
841
+ {
842
+ "epoch": 5.46,
843
+ "grad_norm": 1.495175838470459,
844
+ "learning_rate": 0.0022684729064039407,
845
+ "loss": 0.4975,
846
+ "step": 1110
847
+ },
848
+ {
849
+ "epoch": 5.51,
850
+ "grad_norm": 1.6447827816009521,
851
+ "learning_rate": 0.00224384236453202,
852
+ "loss": 0.4138,
853
+ "step": 1120
854
+ },
855
+ {
856
+ "epoch": 5.56,
857
+ "grad_norm": 0.8438058495521545,
858
+ "learning_rate": 0.0022192118226600985,
859
+ "loss": 0.4502,
860
+ "step": 1130
861
+ },
862
+ {
863
+ "epoch": 5.61,
864
+ "grad_norm": 1.3904708623886108,
865
+ "learning_rate": 0.0021945812807881776,
866
+ "loss": 0.4681,
867
+ "step": 1140
868
+ },
869
+ {
870
+ "epoch": 5.66,
871
+ "grad_norm": 1.498844861984253,
872
+ "learning_rate": 0.0021699507389162562,
873
+ "loss": 0.4637,
874
+ "step": 1150
875
+ },
876
+ {
877
+ "epoch": 5.71,
878
+ "grad_norm": 1.1716539859771729,
879
+ "learning_rate": 0.002145320197044335,
880
+ "loss": 0.5183,
881
+ "step": 1160
882
+ },
883
+ {
884
+ "epoch": 5.76,
885
+ "grad_norm": 0.7125697135925293,
886
+ "learning_rate": 0.002120689655172414,
887
+ "loss": 0.4307,
888
+ "step": 1170
889
+ },
890
+ {
891
+ "epoch": 5.81,
892
+ "grad_norm": 1.341647744178772,
893
+ "learning_rate": 0.0020960591133004926,
894
+ "loss": 0.4988,
895
+ "step": 1180
896
+ },
897
+ {
898
+ "epoch": 5.85,
899
+ "grad_norm": 1.4662394523620605,
900
+ "learning_rate": 0.0020714285714285717,
901
+ "loss": 0.4398,
902
+ "step": 1190
903
+ },
904
+ {
905
+ "epoch": 5.9,
906
+ "grad_norm": 1.7114837169647217,
907
+ "learning_rate": 0.0020467980295566504,
908
+ "loss": 0.4488,
909
+ "step": 1200
910
+ },
911
+ {
912
+ "epoch": 5.95,
913
+ "grad_norm": 1.0667368173599243,
914
+ "learning_rate": 0.002022167487684729,
915
+ "loss": 0.5513,
916
+ "step": 1210
917
+ },
918
+ {
919
+ "epoch": 6.0,
920
+ "eval_accuracy": 0.9862040133779264,
921
+ "eval_f1": 0.9840490701292556,
922
+ "eval_loss": 0.0578995905816555,
923
+ "eval_precision": 0.9830589209967975,
924
+ "eval_recall": 0.985517150491058,
925
+ "eval_runtime": 12.5187,
926
+ "eval_samples_per_second": 191.074,
927
+ "eval_steps_per_second": 11.982,
928
+ "step": 1219
929
+ },
930
+ {
931
+ "epoch": 6.0,
932
+ "grad_norm": 1.1651352643966675,
933
+ "learning_rate": 0.0019975369458128077,
934
+ "loss": 0.4321,
935
+ "step": 1220
936
+ },
937
+ {
938
+ "epoch": 6.05,
939
+ "grad_norm": 1.0694313049316406,
940
+ "learning_rate": 0.0019729064039408868,
941
+ "loss": 0.4343,
942
+ "step": 1230
943
+ },
944
+ {
945
+ "epoch": 6.1,
946
+ "grad_norm": 1.5686174631118774,
947
+ "learning_rate": 0.0019482758620689657,
948
+ "loss": 0.367,
949
+ "step": 1240
950
+ },
951
+ {
952
+ "epoch": 6.15,
953
+ "grad_norm": 0.7148666977882385,
954
+ "learning_rate": 0.0019236453201970443,
955
+ "loss": 0.4364,
956
+ "step": 1250
957
+ },
958
+ {
959
+ "epoch": 6.2,
960
+ "grad_norm": 1.4920200109481812,
961
+ "learning_rate": 0.0018990147783251232,
962
+ "loss": 0.4814,
963
+ "step": 1260
964
+ },
965
+ {
966
+ "epoch": 6.25,
967
+ "grad_norm": 1.0870678424835205,
968
+ "learning_rate": 0.001874384236453202,
969
+ "loss": 0.4145,
970
+ "step": 1270
971
+ },
972
+ {
973
+ "epoch": 6.3,
974
+ "grad_norm": 1.0466927289962769,
975
+ "learning_rate": 0.001849753694581281,
976
+ "loss": 0.4296,
977
+ "step": 1280
978
+ },
979
+ {
980
+ "epoch": 6.35,
981
+ "grad_norm": 0.9908223748207092,
982
+ "learning_rate": 0.0018251231527093596,
983
+ "loss": 0.4183,
984
+ "step": 1290
985
+ },
986
+ {
987
+ "epoch": 6.4,
988
+ "grad_norm": 0.6582946181297302,
989
+ "learning_rate": 0.0018004926108374385,
990
+ "loss": 0.4099,
991
+ "step": 1300
992
+ },
993
+ {
994
+ "epoch": 6.45,
995
+ "grad_norm": 1.3454304933547974,
996
+ "learning_rate": 0.0017758620689655171,
997
+ "loss": 0.367,
998
+ "step": 1310
999
+ },
1000
+ {
1001
+ "epoch": 6.49,
1002
+ "grad_norm": 1.3359636068344116,
1003
+ "learning_rate": 0.001751231527093596,
1004
+ "loss": 0.4025,
1005
+ "step": 1320
1006
+ },
1007
+ {
1008
+ "epoch": 6.54,
1009
+ "grad_norm": 1.2285734415054321,
1010
+ "learning_rate": 0.0017266009852216749,
1011
+ "loss": 0.4675,
1012
+ "step": 1330
1013
+ },
1014
+ {
1015
+ "epoch": 6.59,
1016
+ "grad_norm": 0.9923570156097412,
1017
+ "learning_rate": 0.0017019704433497537,
1018
+ "loss": 0.3958,
1019
+ "step": 1340
1020
+ },
1021
+ {
1022
+ "epoch": 6.64,
1023
+ "grad_norm": 0.8746837973594666,
1024
+ "learning_rate": 0.0016773399014778326,
1025
+ "loss": 0.4365,
1026
+ "step": 1350
1027
+ },
1028
+ {
1029
+ "epoch": 6.69,
1030
+ "grad_norm": 0.8892514705657959,
1031
+ "learning_rate": 0.0016527093596059115,
1032
+ "loss": 0.4296,
1033
+ "step": 1360
1034
+ },
1035
+ {
1036
+ "epoch": 6.74,
1037
+ "grad_norm": 1.2088005542755127,
1038
+ "learning_rate": 0.0016280788177339904,
1039
+ "loss": 0.3881,
1040
+ "step": 1370
1041
+ },
1042
+ {
1043
+ "epoch": 6.79,
1044
+ "grad_norm": 1.0085664987564087,
1045
+ "learning_rate": 0.0016034482758620688,
1046
+ "loss": 0.4745,
1047
+ "step": 1380
1048
+ },
1049
+ {
1050
+ "epoch": 6.84,
1051
+ "grad_norm": 1.0228571891784668,
1052
+ "learning_rate": 0.0015788177339901477,
1053
+ "loss": 0.3763,
1054
+ "step": 1390
1055
+ },
1056
+ {
1057
+ "epoch": 6.89,
1058
+ "grad_norm": 0.8100888133049011,
1059
+ "learning_rate": 0.0015541871921182266,
1060
+ "loss": 0.4198,
1061
+ "step": 1400
1062
+ },
1063
+ {
1064
+ "epoch": 6.94,
1065
+ "grad_norm": 1.2440354824066162,
1066
+ "learning_rate": 0.0015295566502463054,
1067
+ "loss": 0.4113,
1068
+ "step": 1410
1069
+ },
1070
+ {
1071
+ "epoch": 6.99,
1072
+ "grad_norm": 0.6661180257797241,
1073
+ "learning_rate": 0.0015049261083743843,
1074
+ "loss": 0.4343,
1075
+ "step": 1420
1076
+ },
1077
+ {
1078
+ "epoch": 7.0,
1079
+ "eval_accuracy": 0.9832775919732442,
1080
+ "eval_f1": 0.9834794316662396,
1081
+ "eval_loss": 0.07752905040979385,
1082
+ "eval_precision": 0.985796063365073,
1083
+ "eval_recall": 0.981752686688599,
1084
+ "eval_runtime": 12.5648,
1085
+ "eval_samples_per_second": 190.373,
1086
+ "eval_steps_per_second": 11.938,
1087
+ "step": 1422
1088
+ },
1089
+ {
1090
+ "epoch": 7.04,
1091
+ "grad_norm": 0.591243326663971,
1092
+ "learning_rate": 0.0014802955665024632,
1093
+ "loss": 0.3291,
1094
+ "step": 1430
1095
+ },
1096
+ {
1097
+ "epoch": 7.08,
1098
+ "grad_norm": 0.8764331936836243,
1099
+ "learning_rate": 0.001455665024630542,
1100
+ "loss": 0.3704,
1101
+ "step": 1440
1102
+ },
1103
+ {
1104
+ "epoch": 7.13,
1105
+ "grad_norm": 1.115868330001831,
1106
+ "learning_rate": 0.0014310344827586207,
1107
+ "loss": 0.3625,
1108
+ "step": 1450
1109
+ },
1110
+ {
1111
+ "epoch": 7.18,
1112
+ "grad_norm": 1.1736584901809692,
1113
+ "learning_rate": 0.0014064039408866996,
1114
+ "loss": 0.3571,
1115
+ "step": 1460
1116
+ },
1117
+ {
1118
+ "epoch": 7.23,
1119
+ "grad_norm": 0.9778345227241516,
1120
+ "learning_rate": 0.0013817733990147782,
1121
+ "loss": 0.3594,
1122
+ "step": 1470
1123
+ },
1124
+ {
1125
+ "epoch": 7.28,
1126
+ "grad_norm": 1.1396944522857666,
1127
+ "learning_rate": 0.0013571428571428571,
1128
+ "loss": 0.3615,
1129
+ "step": 1480
1130
+ },
1131
+ {
1132
+ "epoch": 7.33,
1133
+ "grad_norm": 1.2598211765289307,
1134
+ "learning_rate": 0.001332512315270936,
1135
+ "loss": 0.3802,
1136
+ "step": 1490
1137
+ },
1138
+ {
1139
+ "epoch": 7.38,
1140
+ "grad_norm": 1.1756126880645752,
1141
+ "learning_rate": 0.0013078817733990149,
1142
+ "loss": 0.4429,
1143
+ "step": 1500
1144
+ },
1145
+ {
1146
+ "epoch": 7.43,
1147
+ "grad_norm": 0.9109674096107483,
1148
+ "learning_rate": 0.0012832512315270935,
1149
+ "loss": 0.3578,
1150
+ "step": 1510
1151
+ },
1152
+ {
1153
+ "epoch": 7.48,
1154
+ "grad_norm": 0.7428516745567322,
1155
+ "learning_rate": 0.0012586206896551724,
1156
+ "loss": 0.3705,
1157
+ "step": 1520
1158
+ },
1159
+ {
1160
+ "epoch": 7.53,
1161
+ "grad_norm": 1.3957030773162842,
1162
+ "learning_rate": 0.0012339901477832513,
1163
+ "loss": 0.3769,
1164
+ "step": 1530
1165
+ },
1166
+ {
1167
+ "epoch": 7.58,
1168
+ "grad_norm": 1.0507686138153076,
1169
+ "learning_rate": 0.00120935960591133,
1170
+ "loss": 0.3525,
1171
+ "step": 1540
1172
+ },
1173
+ {
1174
+ "epoch": 7.63,
1175
+ "grad_norm": 0.8914185762405396,
1176
+ "learning_rate": 0.0011847290640394088,
1177
+ "loss": 0.4804,
1178
+ "step": 1550
1179
+ },
1180
+ {
1181
+ "epoch": 7.68,
1182
+ "grad_norm": 0.8193994760513306,
1183
+ "learning_rate": 0.0011600985221674877,
1184
+ "loss": 0.298,
1185
+ "step": 1560
1186
+ },
1187
+ {
1188
+ "epoch": 7.72,
1189
+ "grad_norm": 0.9263984560966492,
1190
+ "learning_rate": 0.0011354679802955665,
1191
+ "loss": 0.3142,
1192
+ "step": 1570
1193
+ },
1194
+ {
1195
+ "epoch": 7.77,
1196
+ "grad_norm": 1.8249924182891846,
1197
+ "learning_rate": 0.0011108374384236454,
1198
+ "loss": 0.4135,
1199
+ "step": 1580
1200
+ },
1201
+ {
1202
+ "epoch": 7.82,
1203
+ "grad_norm": 0.659723162651062,
1204
+ "learning_rate": 0.001086206896551724,
1205
+ "loss": 0.3844,
1206
+ "step": 1590
1207
+ },
1208
+ {
1209
+ "epoch": 7.87,
1210
+ "grad_norm": 0.7200958132743835,
1211
+ "learning_rate": 0.001061576354679803,
1212
+ "loss": 0.3627,
1213
+ "step": 1600
1214
+ },
1215
+ {
1216
+ "epoch": 7.92,
1217
+ "grad_norm": 1.055242657661438,
1218
+ "learning_rate": 0.0010369458128078818,
1219
+ "loss": 0.3522,
1220
+ "step": 1610
1221
+ },
1222
+ {
1223
+ "epoch": 7.97,
1224
+ "grad_norm": 1.0147466659545898,
1225
+ "learning_rate": 0.0010123152709359607,
1226
+ "loss": 0.3942,
1227
+ "step": 1620
1228
+ },
1229
+ {
1230
+ "epoch": 8.0,
1231
+ "eval_accuracy": 0.9832775919732442,
1232
+ "eval_f1": 0.9816549969154011,
1233
+ "eval_loss": 0.07823298126459122,
1234
+ "eval_precision": 0.9812899149698605,
1235
+ "eval_recall": 0.9827399465830431,
1236
+ "eval_runtime": 12.6409,
1237
+ "eval_samples_per_second": 189.227,
1238
+ "eval_steps_per_second": 11.866,
1239
+ "step": 1626
1240
+ },
1241
+ {
1242
+ "epoch": 8.02,
1243
+ "grad_norm": 1.2122970819473267,
1244
+ "learning_rate": 0.0009876847290640393,
1245
+ "loss": 0.2658,
1246
+ "step": 1630
1247
+ },
1248
+ {
1249
+ "epoch": 8.07,
1250
+ "grad_norm": 0.6683902740478516,
1251
+ "learning_rate": 0.0009630541871921182,
1252
+ "loss": 0.3499,
1253
+ "step": 1640
1254
+ },
1255
+ {
1256
+ "epoch": 8.12,
1257
+ "grad_norm": 1.0198993682861328,
1258
+ "learning_rate": 0.0009384236453201971,
1259
+ "loss": 0.4031,
1260
+ "step": 1650
1261
+ },
1262
+ {
1263
+ "epoch": 8.17,
1264
+ "grad_norm": 1.2388522624969482,
1265
+ "learning_rate": 0.0009137931034482759,
1266
+ "loss": 0.3164,
1267
+ "step": 1660
1268
+ },
1269
+ {
1270
+ "epoch": 8.22,
1271
+ "grad_norm": 1.2226431369781494,
1272
+ "learning_rate": 0.0008891625615763547,
1273
+ "loss": 0.299,
1274
+ "step": 1670
1275
+ },
1276
+ {
1277
+ "epoch": 8.27,
1278
+ "grad_norm": 1.0016721487045288,
1279
+ "learning_rate": 0.0008645320197044335,
1280
+ "loss": 0.3315,
1281
+ "step": 1680
1282
+ },
1283
+ {
1284
+ "epoch": 8.31,
1285
+ "grad_norm": 1.0766950845718384,
1286
+ "learning_rate": 0.0008399014778325123,
1287
+ "loss": 0.286,
1288
+ "step": 1690
1289
+ },
1290
+ {
1291
+ "epoch": 8.36,
1292
+ "grad_norm": 1.8925853967666626,
1293
+ "learning_rate": 0.0008152709359605911,
1294
+ "loss": 0.3745,
1295
+ "step": 1700
1296
+ },
1297
+ {
1298
+ "epoch": 8.41,
1299
+ "grad_norm": 0.8409897685050964,
1300
+ "learning_rate": 0.00079064039408867,
1301
+ "loss": 0.3276,
1302
+ "step": 1710
1303
+ },
1304
+ {
1305
+ "epoch": 8.46,
1306
+ "grad_norm": 1.1315199136734009,
1307
+ "learning_rate": 0.0007660098522167489,
1308
+ "loss": 0.3275,
1309
+ "step": 1720
1310
+ },
1311
+ {
1312
+ "epoch": 8.51,
1313
+ "grad_norm": 1.05160391330719,
1314
+ "learning_rate": 0.0007413793103448275,
1315
+ "loss": 0.3072,
1316
+ "step": 1730
1317
+ },
1318
+ {
1319
+ "epoch": 8.56,
1320
+ "grad_norm": 1.0058565139770508,
1321
+ "learning_rate": 0.0007167487684729064,
1322
+ "loss": 0.3413,
1323
+ "step": 1740
1324
+ },
1325
+ {
1326
+ "epoch": 8.61,
1327
+ "grad_norm": 0.9650315046310425,
1328
+ "learning_rate": 0.0006921182266009853,
1329
+ "loss": 0.365,
1330
+ "step": 1750
1331
+ },
1332
+ {
1333
+ "epoch": 8.66,
1334
+ "grad_norm": 0.6396649479866028,
1335
+ "learning_rate": 0.0006674876847290641,
1336
+ "loss": 0.3271,
1337
+ "step": 1760
1338
+ },
1339
+ {
1340
+ "epoch": 8.71,
1341
+ "grad_norm": 0.7196962833404541,
1342
+ "learning_rate": 0.0006428571428571428,
1343
+ "loss": 0.3683,
1344
+ "step": 1770
1345
+ },
1346
+ {
1347
+ "epoch": 8.76,
1348
+ "grad_norm": 0.8004360198974609,
1349
+ "learning_rate": 0.0006182266009852217,
1350
+ "loss": 0.3687,
1351
+ "step": 1780
1352
+ },
1353
+ {
1354
+ "epoch": 8.81,
1355
+ "grad_norm": 0.9620378613471985,
1356
+ "learning_rate": 0.0005935960591133005,
1357
+ "loss": 0.3298,
1358
+ "step": 1790
1359
+ },
1360
+ {
1361
+ "epoch": 8.86,
1362
+ "grad_norm": 0.7279284596443176,
1363
+ "learning_rate": 0.0005689655172413793,
1364
+ "loss": 0.3109,
1365
+ "step": 1800
1366
+ },
1367
+ {
1368
+ "epoch": 8.91,
1369
+ "grad_norm": 1.2889859676361084,
1370
+ "learning_rate": 0.0005443349753694581,
1371
+ "loss": 0.4205,
1372
+ "step": 1810
1373
+ },
1374
+ {
1375
+ "epoch": 8.95,
1376
+ "grad_norm": 0.8951707482337952,
1377
+ "learning_rate": 0.000519704433497537,
1378
+ "loss": 0.2971,
1379
+ "step": 1820
1380
+ },
1381
+ {
1382
+ "epoch": 9.0,
1383
+ "eval_accuracy": 0.9862040133779264,
1384
+ "eval_f1": 0.9872927641297526,
1385
+ "eval_loss": 0.08386523276567459,
1386
+ "eval_precision": 0.9884096313888006,
1387
+ "eval_recall": 0.9865651085702777,
1388
+ "eval_runtime": 12.5898,
1389
+ "eval_samples_per_second": 189.995,
1390
+ "eval_steps_per_second": 11.914,
1391
+ "step": 1829
1392
+ },
1393
+ {
1394
+ "epoch": 9.0,
1395
+ "grad_norm": 0.6368119120597839,
1396
+ "learning_rate": 0.0004950738916256157,
1397
+ "loss": 0.3692,
1398
+ "step": 1830
1399
+ },
1400
+ {
1401
+ "epoch": 9.05,
1402
+ "grad_norm": 0.5298008322715759,
1403
+ "learning_rate": 0.0004704433497536946,
1404
+ "loss": 0.3112,
1405
+ "step": 1840
1406
+ },
1407
+ {
1408
+ "epoch": 9.1,
1409
+ "grad_norm": 0.5324183702468872,
1410
+ "learning_rate": 0.0004458128078817734,
1411
+ "loss": 0.3389,
1412
+ "step": 1850
1413
+ },
1414
+ {
1415
+ "epoch": 9.15,
1416
+ "grad_norm": 0.6283653974533081,
1417
+ "learning_rate": 0.0004211822660098522,
1418
+ "loss": 0.3328,
1419
+ "step": 1860
1420
+ },
1421
+ {
1422
+ "epoch": 9.2,
1423
+ "grad_norm": 0.9130664467811584,
1424
+ "learning_rate": 0.0003965517241379311,
1425
+ "loss": 0.2774,
1426
+ "step": 1870
1427
+ },
1428
+ {
1429
+ "epoch": 9.25,
1430
+ "grad_norm": 0.569354772567749,
1431
+ "learning_rate": 0.00037192118226600984,
1432
+ "loss": 0.3055,
1433
+ "step": 1880
1434
+ },
1435
+ {
1436
+ "epoch": 9.3,
1437
+ "grad_norm": 0.91834557056427,
1438
+ "learning_rate": 0.0003472906403940887,
1439
+ "loss": 0.3108,
1440
+ "step": 1890
1441
+ },
1442
+ {
1443
+ "epoch": 9.35,
1444
+ "grad_norm": 1.2413830757141113,
1445
+ "learning_rate": 0.0003226600985221675,
1446
+ "loss": 0.2896,
1447
+ "step": 1900
1448
+ },
1449
+ {
1450
+ "epoch": 9.4,
1451
+ "grad_norm": 0.7163951396942139,
1452
+ "learning_rate": 0.0002980295566502463,
1453
+ "loss": 0.2883,
1454
+ "step": 1910
1455
+ },
1456
+ {
1457
+ "epoch": 9.45,
1458
+ "grad_norm": 1.121543288230896,
1459
+ "learning_rate": 0.0002733990147783251,
1460
+ "loss": 0.3061,
1461
+ "step": 1920
1462
+ },
1463
+ {
1464
+ "epoch": 9.5,
1465
+ "grad_norm": 0.8933872580528259,
1466
+ "learning_rate": 0.00024876847290640394,
1467
+ "loss": 0.3087,
1468
+ "step": 1930
1469
+ },
1470
+ {
1471
+ "epoch": 9.54,
1472
+ "grad_norm": 0.7040281295776367,
1473
+ "learning_rate": 0.00022413793103448276,
1474
+ "loss": 0.2723,
1475
+ "step": 1940
1476
+ },
1477
+ {
1478
+ "epoch": 9.59,
1479
+ "grad_norm": 0.3641883432865143,
1480
+ "learning_rate": 0.00019950738916256158,
1481
+ "loss": 0.2397,
1482
+ "step": 1950
1483
+ },
1484
+ {
1485
+ "epoch": 9.64,
1486
+ "grad_norm": 1.092432975769043,
1487
+ "learning_rate": 0.0001748768472906404,
1488
+ "loss": 0.2763,
1489
+ "step": 1960
1490
+ },
1491
+ {
1492
+ "epoch": 9.69,
1493
+ "grad_norm": 0.6717754602432251,
1494
+ "learning_rate": 0.00015024630541871922,
1495
+ "loss": 0.2921,
1496
+ "step": 1970
1497
+ },
1498
+ {
1499
+ "epoch": 9.74,
1500
+ "grad_norm": 0.8007289171218872,
1501
+ "learning_rate": 0.00012561576354679804,
1502
+ "loss": 0.3365,
1503
+ "step": 1980
1504
+ },
1505
+ {
1506
+ "epoch": 9.79,
1507
+ "grad_norm": 0.9652357697486877,
1508
+ "learning_rate": 0.00010098522167487686,
1509
+ "loss": 0.2998,
1510
+ "step": 1990
1511
+ },
1512
+ {
1513
+ "epoch": 9.84,
1514
+ "grad_norm": 0.6942909955978394,
1515
+ "learning_rate": 7.635467980295568e-05,
1516
+ "loss": 0.2917,
1517
+ "step": 2000
1518
+ },
1519
+ {
1520
+ "epoch": 9.89,
1521
+ "grad_norm": 0.5785544514656067,
1522
+ "learning_rate": 5.172413793103448e-05,
1523
+ "loss": 0.2662,
1524
+ "step": 2010
1525
+ },
1526
+ {
1527
+ "epoch": 9.94,
1528
+ "grad_norm": 0.7549653649330139,
1529
+ "learning_rate": 2.70935960591133e-05,
1530
+ "loss": 0.2989,
1531
+ "step": 2020
1532
+ },
1533
+ {
1534
+ "epoch": 9.99,
1535
+ "grad_norm": 1.3010107278823853,
1536
+ "learning_rate": 2.463054187192118e-06,
1537
+ "loss": 0.3242,
1538
+ "step": 2030
1539
+ },
1540
+ {
1541
+ "epoch": 9.99,
1542
+ "eval_accuracy": 0.9870401337792643,
1543
+ "eval_f1": 0.9868360981525407,
1544
+ "eval_loss": 0.07449871301651001,
1545
+ "eval_precision": 0.9876698762890729,
1546
+ "eval_recall": 0.9863397416476135,
1547
+ "eval_runtime": 12.5861,
1548
+ "eval_samples_per_second": 190.05,
1549
+ "eval_steps_per_second": 11.918,
1550
+ "step": 2030
1551
+ },
1552
+ {
1553
+ "epoch": 9.99,
1554
+ "step": 2030,
1555
+ "total_flos": 1.0133154899356189e+19,
1556
+ "train_loss": 0.5248493043072705,
1557
+ "train_runtime": 1518.4893,
1558
+ "train_samples_per_second": 85.611,
1559
+ "train_steps_per_second": 1.337
1560
+ }
1561
+ ],
1562
+ "logging_steps": 10,
1563
+ "max_steps": 2030,
1564
+ "num_input_tokens_seen": 0,
1565
+ "num_train_epochs": 10,
1566
+ "save_steps": 500,
1567
+ "total_flos": 1.0133154899356189e+19,
1568
+ "train_batch_size": 16,
1569
+ "trial_name": null,
1570
+ "trial_params": null
1571
+ }