jalaneunos commited on
Commit
47cc262
1 Parent(s): 9113858

End of training

Browse files
README.md CHANGED
@@ -22,7 +22,7 @@ model-index:
22
  metrics:
23
  - name: Accuracy
24
  type: accuracy
25
- value: 0.8397770811563915
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,8 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
32
 
33
  This model is a fine-tuned version of [WinKawaks/vit-tiny-patch16-224](https://huggingface.co/WinKawaks/vit-tiny-patch16-224) on the imagefolder dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 0.3540
36
- - Accuracy: 0.8398
37
 
38
  ## Model description
39
 
 
22
  metrics:
23
  - name: Accuracy
24
  type: accuracy
25
+ value: 0.842911877394636
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
32
 
33
  This model is a fine-tuned version of [WinKawaks/vit-tiny-patch16-224](https://huggingface.co/WinKawaks/vit-tiny-patch16-224) on the imagefolder dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: 0.3542
36
+ - Accuracy: 0.8429
37
 
38
  ## Model description
39
 
all_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_accuracy": 0.842911877394636,
4
+ "eval_loss": 0.3542243540287018,
5
+ "eval_runtime": 5.8665,
6
+ "eval_samples_per_second": 489.386,
7
+ "eval_steps_per_second": 15.341,
8
+ "total_flos": 1.2892209231838003e+18,
9
+ "train_loss": 0.45288851414576614,
10
+ "train_runtime": 810.9556,
11
+ "train_samples_per_second": 318.599,
12
+ "train_steps_per_second": 2.491
13
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_accuracy": 0.842911877394636,
4
+ "eval_loss": 0.3542243540287018,
5
+ "eval_runtime": 5.8665,
6
+ "eval_samples_per_second": 489.386,
7
+ "eval_steps_per_second": 15.341
8
+ }
runs/Mar18_16-33-47_e3106b2fe063/events.out.tfevents.1710780454.e3106b2fe063.997.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a277391455f7a0a3384c8dc36f81b4f7ce45d62948466c7922e98b7e6dd1be3
3
+ size 411
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "total_flos": 1.2892209231838003e+18,
4
+ "train_loss": 0.45288851414576614,
5
+ "train_runtime": 810.9556,
6
+ "train_samples_per_second": 318.599,
7
+ "train_steps_per_second": 2.491
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1534 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.842911877394636,
3
+ "best_model_checkpoint": "vit-tiny-patch16-224-winkawaks/checkpoint-1818",
4
+ "epoch": 10.0,
5
+ "eval_steps": 500,
6
+ "global_step": 2020,
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": 11.655478477478027,
14
+ "learning_rate": 2.4752475247524753e-06,
15
+ "loss": 0.8812,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.1,
20
+ "grad_norm": 6.654660224914551,
21
+ "learning_rate": 4.950495049504951e-06,
22
+ "loss": 0.7531,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.15,
27
+ "grad_norm": 4.010664939880371,
28
+ "learning_rate": 7.4257425742574256e-06,
29
+ "loss": 0.7312,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.2,
34
+ "grad_norm": 3.320401191711426,
35
+ "learning_rate": 9.900990099009901e-06,
36
+ "loss": 0.6753,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.25,
41
+ "grad_norm": 4.002676010131836,
42
+ "learning_rate": 1.2376237623762377e-05,
43
+ "loss": 0.6656,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.3,
48
+ "grad_norm": 5.1362690925598145,
49
+ "learning_rate": 1.4851485148514851e-05,
50
+ "loss": 0.636,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.35,
55
+ "grad_norm": 4.089080333709717,
56
+ "learning_rate": 1.7326732673267325e-05,
57
+ "loss": 0.6258,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.4,
62
+ "grad_norm": 2.782809257507324,
63
+ "learning_rate": 1.9801980198019803e-05,
64
+ "loss": 0.6074,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.45,
69
+ "grad_norm": 4.684001922607422,
70
+ "learning_rate": 2.227722772277228e-05,
71
+ "loss": 0.6114,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.5,
76
+ "grad_norm": 8.801894187927246,
77
+ "learning_rate": 2.4752475247524754e-05,
78
+ "loss": 0.5888,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.54,
83
+ "grad_norm": 10.810859680175781,
84
+ "learning_rate": 2.722772277227723e-05,
85
+ "loss": 0.5858,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.59,
90
+ "grad_norm": 4.189158916473389,
91
+ "learning_rate": 2.9702970297029702e-05,
92
+ "loss": 0.5849,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.64,
97
+ "grad_norm": 6.533910274505615,
98
+ "learning_rate": 3.217821782178218e-05,
99
+ "loss": 0.5811,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.69,
104
+ "grad_norm": 15.806050300598145,
105
+ "learning_rate": 3.465346534653465e-05,
106
+ "loss": 0.6041,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.74,
111
+ "grad_norm": 3.087372064590454,
112
+ "learning_rate": 3.712871287128713e-05,
113
+ "loss": 0.5402,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.79,
118
+ "grad_norm": 6.852709770202637,
119
+ "learning_rate": 3.9603960396039605e-05,
120
+ "loss": 0.5435,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.84,
125
+ "grad_norm": 3.9751667976379395,
126
+ "learning_rate": 4.207920792079208e-05,
127
+ "loss": 0.5826,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.89,
132
+ "grad_norm": 8.571554183959961,
133
+ "learning_rate": 4.455445544554456e-05,
134
+ "loss": 0.5399,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.94,
139
+ "grad_norm": 2.956425905227661,
140
+ "learning_rate": 4.702970297029703e-05,
141
+ "loss": 0.5627,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.99,
146
+ "grad_norm": 10.44449520111084,
147
+ "learning_rate": 4.950495049504951e-05,
148
+ "loss": 0.5345,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 1.0,
153
+ "eval_accuracy": 0.7770811563915012,
154
+ "eval_loss": 0.4689331650733948,
155
+ "eval_runtime": 5.627,
156
+ "eval_samples_per_second": 510.22,
157
+ "eval_steps_per_second": 15.994,
158
+ "step": 202
159
+ },
160
+ {
161
+ "epoch": 1.04,
162
+ "grad_norm": 5.914970397949219,
163
+ "learning_rate": 4.977997799779978e-05,
164
+ "loss": 0.533,
165
+ "step": 210
166
+ },
167
+ {
168
+ "epoch": 1.09,
169
+ "grad_norm": 7.073808193206787,
170
+ "learning_rate": 4.950495049504951e-05,
171
+ "loss": 0.517,
172
+ "step": 220
173
+ },
174
+ {
175
+ "epoch": 1.14,
176
+ "grad_norm": 3.3154642581939697,
177
+ "learning_rate": 4.9229922992299234e-05,
178
+ "loss": 0.5385,
179
+ "step": 230
180
+ },
181
+ {
182
+ "epoch": 1.19,
183
+ "grad_norm": 3.500061273574829,
184
+ "learning_rate": 4.895489548954896e-05,
185
+ "loss": 0.5345,
186
+ "step": 240
187
+ },
188
+ {
189
+ "epoch": 1.24,
190
+ "grad_norm": 8.62484359741211,
191
+ "learning_rate": 4.867986798679868e-05,
192
+ "loss": 0.5368,
193
+ "step": 250
194
+ },
195
+ {
196
+ "epoch": 1.29,
197
+ "grad_norm": 9.027387619018555,
198
+ "learning_rate": 4.8404840484048406e-05,
199
+ "loss": 0.5247,
200
+ "step": 260
201
+ },
202
+ {
203
+ "epoch": 1.34,
204
+ "grad_norm": 7.457286357879639,
205
+ "learning_rate": 4.812981298129813e-05,
206
+ "loss": 0.5044,
207
+ "step": 270
208
+ },
209
+ {
210
+ "epoch": 1.39,
211
+ "grad_norm": 7.2032790184021,
212
+ "learning_rate": 4.785478547854786e-05,
213
+ "loss": 0.5156,
214
+ "step": 280
215
+ },
216
+ {
217
+ "epoch": 1.44,
218
+ "grad_norm": 5.909176826477051,
219
+ "learning_rate": 4.7579757975797585e-05,
220
+ "loss": 0.5191,
221
+ "step": 290
222
+ },
223
+ {
224
+ "epoch": 1.49,
225
+ "grad_norm": 4.56859827041626,
226
+ "learning_rate": 4.730473047304731e-05,
227
+ "loss": 0.4891,
228
+ "step": 300
229
+ },
230
+ {
231
+ "epoch": 1.53,
232
+ "grad_norm": 5.181271553039551,
233
+ "learning_rate": 4.702970297029703e-05,
234
+ "loss": 0.5016,
235
+ "step": 310
236
+ },
237
+ {
238
+ "epoch": 1.58,
239
+ "grad_norm": 6.948479175567627,
240
+ "learning_rate": 4.675467546754676e-05,
241
+ "loss": 0.5008,
242
+ "step": 320
243
+ },
244
+ {
245
+ "epoch": 1.63,
246
+ "grad_norm": 7.155640125274658,
247
+ "learning_rate": 4.647964796479648e-05,
248
+ "loss": 0.5451,
249
+ "step": 330
250
+ },
251
+ {
252
+ "epoch": 1.68,
253
+ "grad_norm": 3.5051815509796143,
254
+ "learning_rate": 4.62046204620462e-05,
255
+ "loss": 0.4998,
256
+ "step": 340
257
+ },
258
+ {
259
+ "epoch": 1.73,
260
+ "grad_norm": 12.015359878540039,
261
+ "learning_rate": 4.592959295929593e-05,
262
+ "loss": 0.5302,
263
+ "step": 350
264
+ },
265
+ {
266
+ "epoch": 1.78,
267
+ "grad_norm": 5.719058990478516,
268
+ "learning_rate": 4.5654565456545655e-05,
269
+ "loss": 0.5052,
270
+ "step": 360
271
+ },
272
+ {
273
+ "epoch": 1.83,
274
+ "grad_norm": 6.292553424835205,
275
+ "learning_rate": 4.537953795379538e-05,
276
+ "loss": 0.5192,
277
+ "step": 370
278
+ },
279
+ {
280
+ "epoch": 1.88,
281
+ "grad_norm": 2.8755393028259277,
282
+ "learning_rate": 4.510451045104511e-05,
283
+ "loss": 0.4966,
284
+ "step": 380
285
+ },
286
+ {
287
+ "epoch": 1.93,
288
+ "grad_norm": 3.096843719482422,
289
+ "learning_rate": 4.4829482948294834e-05,
290
+ "loss": 0.5032,
291
+ "step": 390
292
+ },
293
+ {
294
+ "epoch": 1.98,
295
+ "grad_norm": 3.9766151905059814,
296
+ "learning_rate": 4.455445544554456e-05,
297
+ "loss": 0.4936,
298
+ "step": 400
299
+ },
300
+ {
301
+ "epoch": 2.0,
302
+ "eval_accuracy": 0.7485196795541623,
303
+ "eval_loss": 0.5022404789924622,
304
+ "eval_runtime": 5.6861,
305
+ "eval_samples_per_second": 504.916,
306
+ "eval_steps_per_second": 15.828,
307
+ "step": 404
308
+ },
309
+ {
310
+ "epoch": 2.03,
311
+ "grad_norm": 7.3236870765686035,
312
+ "learning_rate": 4.427942794279428e-05,
313
+ "loss": 0.4828,
314
+ "step": 410
315
+ },
316
+ {
317
+ "epoch": 2.08,
318
+ "grad_norm": 4.203598976135254,
319
+ "learning_rate": 4.4004400440044006e-05,
320
+ "loss": 0.5108,
321
+ "step": 420
322
+ },
323
+ {
324
+ "epoch": 2.13,
325
+ "grad_norm": 4.851642608642578,
326
+ "learning_rate": 4.372937293729373e-05,
327
+ "loss": 0.474,
328
+ "step": 430
329
+ },
330
+ {
331
+ "epoch": 2.18,
332
+ "grad_norm": 8.77710247039795,
333
+ "learning_rate": 4.345434543454346e-05,
334
+ "loss": 0.4902,
335
+ "step": 440
336
+ },
337
+ {
338
+ "epoch": 2.23,
339
+ "grad_norm": 6.436154365539551,
340
+ "learning_rate": 4.3179317931793185e-05,
341
+ "loss": 0.4838,
342
+ "step": 450
343
+ },
344
+ {
345
+ "epoch": 2.28,
346
+ "grad_norm": 4.809104919433594,
347
+ "learning_rate": 4.2904290429042904e-05,
348
+ "loss": 0.5064,
349
+ "step": 460
350
+ },
351
+ {
352
+ "epoch": 2.33,
353
+ "grad_norm": 4.125964641571045,
354
+ "learning_rate": 4.262926292629263e-05,
355
+ "loss": 0.4819,
356
+ "step": 470
357
+ },
358
+ {
359
+ "epoch": 2.38,
360
+ "grad_norm": 6.8852763175964355,
361
+ "learning_rate": 4.2354235423542356e-05,
362
+ "loss": 0.5011,
363
+ "step": 480
364
+ },
365
+ {
366
+ "epoch": 2.43,
367
+ "grad_norm": 3.3697924613952637,
368
+ "learning_rate": 4.207920792079208e-05,
369
+ "loss": 0.4676,
370
+ "step": 490
371
+ },
372
+ {
373
+ "epoch": 2.48,
374
+ "grad_norm": 5.354856967926025,
375
+ "learning_rate": 4.18041804180418e-05,
376
+ "loss": 0.507,
377
+ "step": 500
378
+ },
379
+ {
380
+ "epoch": 2.52,
381
+ "grad_norm": 7.638093948364258,
382
+ "learning_rate": 4.152915291529153e-05,
383
+ "loss": 0.5046,
384
+ "step": 510
385
+ },
386
+ {
387
+ "epoch": 2.57,
388
+ "grad_norm": 4.444268226623535,
389
+ "learning_rate": 4.1254125412541255e-05,
390
+ "loss": 0.4635,
391
+ "step": 520
392
+ },
393
+ {
394
+ "epoch": 2.62,
395
+ "grad_norm": 8.037334442138672,
396
+ "learning_rate": 4.097909790979098e-05,
397
+ "loss": 0.4796,
398
+ "step": 530
399
+ },
400
+ {
401
+ "epoch": 2.67,
402
+ "grad_norm": 6.52307653427124,
403
+ "learning_rate": 4.070407040704071e-05,
404
+ "loss": 0.4668,
405
+ "step": 540
406
+ },
407
+ {
408
+ "epoch": 2.72,
409
+ "grad_norm": 7.490472316741943,
410
+ "learning_rate": 4.042904290429043e-05,
411
+ "loss": 0.4747,
412
+ "step": 550
413
+ },
414
+ {
415
+ "epoch": 2.77,
416
+ "grad_norm": 3.393075942993164,
417
+ "learning_rate": 4.015401540154016e-05,
418
+ "loss": 0.4667,
419
+ "step": 560
420
+ },
421
+ {
422
+ "epoch": 2.82,
423
+ "grad_norm": 4.454715251922607,
424
+ "learning_rate": 3.987898789878988e-05,
425
+ "loss": 0.4795,
426
+ "step": 570
427
+ },
428
+ {
429
+ "epoch": 2.87,
430
+ "grad_norm": 5.316873550415039,
431
+ "learning_rate": 3.9603960396039605e-05,
432
+ "loss": 0.499,
433
+ "step": 580
434
+ },
435
+ {
436
+ "epoch": 2.92,
437
+ "grad_norm": 2.8893849849700928,
438
+ "learning_rate": 3.932893289328933e-05,
439
+ "loss": 0.4913,
440
+ "step": 590
441
+ },
442
+ {
443
+ "epoch": 2.97,
444
+ "grad_norm": 2.6538729667663574,
445
+ "learning_rate": 3.905390539053906e-05,
446
+ "loss": 0.4911,
447
+ "step": 600
448
+ },
449
+ {
450
+ "epoch": 3.0,
451
+ "eval_accuracy": 0.82793451758969,
452
+ "eval_loss": 0.3886525630950928,
453
+ "eval_runtime": 5.5731,
454
+ "eval_samples_per_second": 515.151,
455
+ "eval_steps_per_second": 16.149,
456
+ "step": 606
457
+ },
458
+ {
459
+ "epoch": 3.02,
460
+ "grad_norm": 9.823522567749023,
461
+ "learning_rate": 3.877887788778878e-05,
462
+ "loss": 0.4731,
463
+ "step": 610
464
+ },
465
+ {
466
+ "epoch": 3.07,
467
+ "grad_norm": 5.899537086486816,
468
+ "learning_rate": 3.8503850385038503e-05,
469
+ "loss": 0.4621,
470
+ "step": 620
471
+ },
472
+ {
473
+ "epoch": 3.12,
474
+ "grad_norm": 4.964333534240723,
475
+ "learning_rate": 3.822882288228823e-05,
476
+ "loss": 0.4762,
477
+ "step": 630
478
+ },
479
+ {
480
+ "epoch": 3.17,
481
+ "grad_norm": 4.469970703125,
482
+ "learning_rate": 3.7953795379537956e-05,
483
+ "loss": 0.4774,
484
+ "step": 640
485
+ },
486
+ {
487
+ "epoch": 3.22,
488
+ "grad_norm": 3.393183469772339,
489
+ "learning_rate": 3.767876787678768e-05,
490
+ "loss": 0.4501,
491
+ "step": 650
492
+ },
493
+ {
494
+ "epoch": 3.27,
495
+ "grad_norm": 11.101873397827148,
496
+ "learning_rate": 3.74037403740374e-05,
497
+ "loss": 0.4596,
498
+ "step": 660
499
+ },
500
+ {
501
+ "epoch": 3.32,
502
+ "grad_norm": 4.9833831787109375,
503
+ "learning_rate": 3.712871287128713e-05,
504
+ "loss": 0.4516,
505
+ "step": 670
506
+ },
507
+ {
508
+ "epoch": 3.37,
509
+ "grad_norm": 5.423743724822998,
510
+ "learning_rate": 3.6853685368536854e-05,
511
+ "loss": 0.4887,
512
+ "step": 680
513
+ },
514
+ {
515
+ "epoch": 3.42,
516
+ "grad_norm": 4.94615364074707,
517
+ "learning_rate": 3.657865786578658e-05,
518
+ "loss": 0.4538,
519
+ "step": 690
520
+ },
521
+ {
522
+ "epoch": 3.47,
523
+ "grad_norm": 5.113562107086182,
524
+ "learning_rate": 3.6303630363036307e-05,
525
+ "loss": 0.4645,
526
+ "step": 700
527
+ },
528
+ {
529
+ "epoch": 3.51,
530
+ "grad_norm": 7.3234357833862305,
531
+ "learning_rate": 3.602860286028603e-05,
532
+ "loss": 0.4645,
533
+ "step": 710
534
+ },
535
+ {
536
+ "epoch": 3.56,
537
+ "grad_norm": 4.696035385131836,
538
+ "learning_rate": 3.575357535753576e-05,
539
+ "loss": 0.4584,
540
+ "step": 720
541
+ },
542
+ {
543
+ "epoch": 3.61,
544
+ "grad_norm": 3.724952459335327,
545
+ "learning_rate": 3.5478547854785485e-05,
546
+ "loss": 0.427,
547
+ "step": 730
548
+ },
549
+ {
550
+ "epoch": 3.66,
551
+ "grad_norm": 3.8439409732818604,
552
+ "learning_rate": 3.5203520352035205e-05,
553
+ "loss": 0.4554,
554
+ "step": 740
555
+ },
556
+ {
557
+ "epoch": 3.71,
558
+ "grad_norm": 3.4024925231933594,
559
+ "learning_rate": 3.492849284928493e-05,
560
+ "loss": 0.4413,
561
+ "step": 750
562
+ },
563
+ {
564
+ "epoch": 3.76,
565
+ "grad_norm": 8.046459197998047,
566
+ "learning_rate": 3.465346534653465e-05,
567
+ "loss": 0.4638,
568
+ "step": 760
569
+ },
570
+ {
571
+ "epoch": 3.81,
572
+ "grad_norm": 5.1486358642578125,
573
+ "learning_rate": 3.4378437843784377e-05,
574
+ "loss": 0.4852,
575
+ "step": 770
576
+ },
577
+ {
578
+ "epoch": 3.86,
579
+ "grad_norm": 3.408259868621826,
580
+ "learning_rate": 3.41034103410341e-05,
581
+ "loss": 0.4347,
582
+ "step": 780
583
+ },
584
+ {
585
+ "epoch": 3.91,
586
+ "grad_norm": 5.650546073913574,
587
+ "learning_rate": 3.382838283828383e-05,
588
+ "loss": 0.4294,
589
+ "step": 790
590
+ },
591
+ {
592
+ "epoch": 3.96,
593
+ "grad_norm": 3.8805394172668457,
594
+ "learning_rate": 3.3553355335533555e-05,
595
+ "loss": 0.4191,
596
+ "step": 800
597
+ },
598
+ {
599
+ "epoch": 4.0,
600
+ "eval_accuracy": 0.8098223615464994,
601
+ "eval_loss": 0.4121237099170685,
602
+ "eval_runtime": 5.4396,
603
+ "eval_samples_per_second": 527.797,
604
+ "eval_steps_per_second": 16.545,
605
+ "step": 808
606
+ },
607
+ {
608
+ "epoch": 4.01,
609
+ "grad_norm": 4.736340522766113,
610
+ "learning_rate": 3.327832783278328e-05,
611
+ "loss": 0.4824,
612
+ "step": 810
613
+ },
614
+ {
615
+ "epoch": 4.06,
616
+ "grad_norm": 4.468814373016357,
617
+ "learning_rate": 3.300330033003301e-05,
618
+ "loss": 0.468,
619
+ "step": 820
620
+ },
621
+ {
622
+ "epoch": 4.11,
623
+ "grad_norm": 3.311174154281616,
624
+ "learning_rate": 3.272827282728273e-05,
625
+ "loss": 0.442,
626
+ "step": 830
627
+ },
628
+ {
629
+ "epoch": 4.16,
630
+ "grad_norm": 6.016156196594238,
631
+ "learning_rate": 3.2453245324532453e-05,
632
+ "loss": 0.4748,
633
+ "step": 840
634
+ },
635
+ {
636
+ "epoch": 4.21,
637
+ "grad_norm": 6.195199966430664,
638
+ "learning_rate": 3.217821782178218e-05,
639
+ "loss": 0.4488,
640
+ "step": 850
641
+ },
642
+ {
643
+ "epoch": 4.26,
644
+ "grad_norm": 3.6519343852996826,
645
+ "learning_rate": 3.1903190319031906e-05,
646
+ "loss": 0.4285,
647
+ "step": 860
648
+ },
649
+ {
650
+ "epoch": 4.31,
651
+ "grad_norm": 5.461367130279541,
652
+ "learning_rate": 3.162816281628163e-05,
653
+ "loss": 0.4124,
654
+ "step": 870
655
+ },
656
+ {
657
+ "epoch": 4.36,
658
+ "grad_norm": 3.7743630409240723,
659
+ "learning_rate": 3.135313531353136e-05,
660
+ "loss": 0.4361,
661
+ "step": 880
662
+ },
663
+ {
664
+ "epoch": 4.41,
665
+ "grad_norm": 3.741748571395874,
666
+ "learning_rate": 3.1078107810781085e-05,
667
+ "loss": 0.4196,
668
+ "step": 890
669
+ },
670
+ {
671
+ "epoch": 4.46,
672
+ "grad_norm": 5.458765983581543,
673
+ "learning_rate": 3.0803080308030804e-05,
674
+ "loss": 0.4814,
675
+ "step": 900
676
+ },
677
+ {
678
+ "epoch": 4.5,
679
+ "grad_norm": 5.118380546569824,
680
+ "learning_rate": 3.052805280528053e-05,
681
+ "loss": 0.4441,
682
+ "step": 910
683
+ },
684
+ {
685
+ "epoch": 4.55,
686
+ "grad_norm": 4.428480625152588,
687
+ "learning_rate": 3.0253025302530253e-05,
688
+ "loss": 0.4217,
689
+ "step": 920
690
+ },
691
+ {
692
+ "epoch": 4.6,
693
+ "grad_norm": 3.2278711795806885,
694
+ "learning_rate": 2.9977997799779976e-05,
695
+ "loss": 0.4259,
696
+ "step": 930
697
+ },
698
+ {
699
+ "epoch": 4.65,
700
+ "grad_norm": 3.5108141899108887,
701
+ "learning_rate": 2.9702970297029702e-05,
702
+ "loss": 0.4507,
703
+ "step": 940
704
+ },
705
+ {
706
+ "epoch": 4.7,
707
+ "grad_norm": 3.02825665473938,
708
+ "learning_rate": 2.942794279427943e-05,
709
+ "loss": 0.4095,
710
+ "step": 950
711
+ },
712
+ {
713
+ "epoch": 4.75,
714
+ "grad_norm": 3.8650879859924316,
715
+ "learning_rate": 2.9152915291529155e-05,
716
+ "loss": 0.4251,
717
+ "step": 960
718
+ },
719
+ {
720
+ "epoch": 4.8,
721
+ "grad_norm": 5.335313320159912,
722
+ "learning_rate": 2.8877887788778878e-05,
723
+ "loss": 0.428,
724
+ "step": 970
725
+ },
726
+ {
727
+ "epoch": 4.85,
728
+ "grad_norm": 5.564798831939697,
729
+ "learning_rate": 2.8602860286028604e-05,
730
+ "loss": 0.4713,
731
+ "step": 980
732
+ },
733
+ {
734
+ "epoch": 4.9,
735
+ "grad_norm": 4.036501407623291,
736
+ "learning_rate": 2.832783278327833e-05,
737
+ "loss": 0.4152,
738
+ "step": 990
739
+ },
740
+ {
741
+ "epoch": 4.95,
742
+ "grad_norm": 4.295027732849121,
743
+ "learning_rate": 2.8052805280528056e-05,
744
+ "loss": 0.4209,
745
+ "step": 1000
746
+ },
747
+ {
748
+ "epoch": 5.0,
749
+ "grad_norm": 2.8911187648773193,
750
+ "learning_rate": 2.777777777777778e-05,
751
+ "loss": 0.4408,
752
+ "step": 1010
753
+ },
754
+ {
755
+ "epoch": 5.0,
756
+ "eval_accuracy": 0.8254963427377221,
757
+ "eval_loss": 0.38967224955558777,
758
+ "eval_runtime": 5.6692,
759
+ "eval_samples_per_second": 506.425,
760
+ "eval_steps_per_second": 15.875,
761
+ "step": 1010
762
+ },
763
+ {
764
+ "epoch": 5.05,
765
+ "grad_norm": 7.270760536193848,
766
+ "learning_rate": 2.7502750275027505e-05,
767
+ "loss": 0.4415,
768
+ "step": 1020
769
+ },
770
+ {
771
+ "epoch": 5.1,
772
+ "grad_norm": 2.75962495803833,
773
+ "learning_rate": 2.722772277227723e-05,
774
+ "loss": 0.4282,
775
+ "step": 1030
776
+ },
777
+ {
778
+ "epoch": 5.15,
779
+ "grad_norm": 3.6654067039489746,
780
+ "learning_rate": 2.6952695269526958e-05,
781
+ "loss": 0.4265,
782
+ "step": 1040
783
+ },
784
+ {
785
+ "epoch": 5.2,
786
+ "grad_norm": 5.461548805236816,
787
+ "learning_rate": 2.667766776677668e-05,
788
+ "loss": 0.4188,
789
+ "step": 1050
790
+ },
791
+ {
792
+ "epoch": 5.25,
793
+ "grad_norm": 3.629088878631592,
794
+ "learning_rate": 2.64026402640264e-05,
795
+ "loss": 0.438,
796
+ "step": 1060
797
+ },
798
+ {
799
+ "epoch": 5.3,
800
+ "grad_norm": 5.7036614418029785,
801
+ "learning_rate": 2.6127612761276126e-05,
802
+ "loss": 0.4073,
803
+ "step": 1070
804
+ },
805
+ {
806
+ "epoch": 5.35,
807
+ "grad_norm": 7.140379905700684,
808
+ "learning_rate": 2.5852585258525853e-05,
809
+ "loss": 0.4287,
810
+ "step": 1080
811
+ },
812
+ {
813
+ "epoch": 5.4,
814
+ "grad_norm": 3.2449722290039062,
815
+ "learning_rate": 2.557755775577558e-05,
816
+ "loss": 0.425,
817
+ "step": 1090
818
+ },
819
+ {
820
+ "epoch": 5.45,
821
+ "grad_norm": 4.488211631774902,
822
+ "learning_rate": 2.53025302530253e-05,
823
+ "loss": 0.4203,
824
+ "step": 1100
825
+ },
826
+ {
827
+ "epoch": 5.5,
828
+ "grad_norm": 3.7270421981811523,
829
+ "learning_rate": 2.5027502750275028e-05,
830
+ "loss": 0.4174,
831
+ "step": 1110
832
+ },
833
+ {
834
+ "epoch": 5.54,
835
+ "grad_norm": 4.254336357116699,
836
+ "learning_rate": 2.4752475247524754e-05,
837
+ "loss": 0.4533,
838
+ "step": 1120
839
+ },
840
+ {
841
+ "epoch": 5.59,
842
+ "grad_norm": 3.1208856105804443,
843
+ "learning_rate": 2.447744774477448e-05,
844
+ "loss": 0.4221,
845
+ "step": 1130
846
+ },
847
+ {
848
+ "epoch": 5.64,
849
+ "grad_norm": 6.634130001068115,
850
+ "learning_rate": 2.4202420242024203e-05,
851
+ "loss": 0.4377,
852
+ "step": 1140
853
+ },
854
+ {
855
+ "epoch": 5.69,
856
+ "grad_norm": 5.457045078277588,
857
+ "learning_rate": 2.392739273927393e-05,
858
+ "loss": 0.4459,
859
+ "step": 1150
860
+ },
861
+ {
862
+ "epoch": 5.74,
863
+ "grad_norm": 3.537189483642578,
864
+ "learning_rate": 2.3652365236523656e-05,
865
+ "loss": 0.4423,
866
+ "step": 1160
867
+ },
868
+ {
869
+ "epoch": 5.79,
870
+ "grad_norm": 5.883282661437988,
871
+ "learning_rate": 2.337733773377338e-05,
872
+ "loss": 0.4121,
873
+ "step": 1170
874
+ },
875
+ {
876
+ "epoch": 5.84,
877
+ "grad_norm": 4.0562744140625,
878
+ "learning_rate": 2.31023102310231e-05,
879
+ "loss": 0.4114,
880
+ "step": 1180
881
+ },
882
+ {
883
+ "epoch": 5.89,
884
+ "grad_norm": 5.0152974128723145,
885
+ "learning_rate": 2.2827282728272828e-05,
886
+ "loss": 0.4287,
887
+ "step": 1190
888
+ },
889
+ {
890
+ "epoch": 5.94,
891
+ "grad_norm": 4.672534465789795,
892
+ "learning_rate": 2.2552255225522554e-05,
893
+ "loss": 0.4124,
894
+ "step": 1200
895
+ },
896
+ {
897
+ "epoch": 5.99,
898
+ "grad_norm": 5.874300479888916,
899
+ "learning_rate": 2.227722772277228e-05,
900
+ "loss": 0.4134,
901
+ "step": 1210
902
+ },
903
+ {
904
+ "epoch": 6.0,
905
+ "eval_accuracy": 0.8331591779867642,
906
+ "eval_loss": 0.3714359402656555,
907
+ "eval_runtime": 5.6334,
908
+ "eval_samples_per_second": 509.636,
909
+ "eval_steps_per_second": 15.976,
910
+ "step": 1212
911
+ },
912
+ {
913
+ "epoch": 6.04,
914
+ "grad_norm": 3.3733067512512207,
915
+ "learning_rate": 2.2002200220022003e-05,
916
+ "loss": 0.404,
917
+ "step": 1220
918
+ },
919
+ {
920
+ "epoch": 6.09,
921
+ "grad_norm": 3.479614734649658,
922
+ "learning_rate": 2.172717271727173e-05,
923
+ "loss": 0.4102,
924
+ "step": 1230
925
+ },
926
+ {
927
+ "epoch": 6.14,
928
+ "grad_norm": 6.636502742767334,
929
+ "learning_rate": 2.1452145214521452e-05,
930
+ "loss": 0.4261,
931
+ "step": 1240
932
+ },
933
+ {
934
+ "epoch": 6.19,
935
+ "grad_norm": 3.6467645168304443,
936
+ "learning_rate": 2.1177117711771178e-05,
937
+ "loss": 0.4061,
938
+ "step": 1250
939
+ },
940
+ {
941
+ "epoch": 6.24,
942
+ "grad_norm": 3.435811996459961,
943
+ "learning_rate": 2.09020902090209e-05,
944
+ "loss": 0.4198,
945
+ "step": 1260
946
+ },
947
+ {
948
+ "epoch": 6.29,
949
+ "grad_norm": 5.085671424865723,
950
+ "learning_rate": 2.0627062706270627e-05,
951
+ "loss": 0.3963,
952
+ "step": 1270
953
+ },
954
+ {
955
+ "epoch": 6.34,
956
+ "grad_norm": 9.153255462646484,
957
+ "learning_rate": 2.0352035203520354e-05,
958
+ "loss": 0.4239,
959
+ "step": 1280
960
+ },
961
+ {
962
+ "epoch": 6.39,
963
+ "grad_norm": 5.1853251457214355,
964
+ "learning_rate": 2.007700770077008e-05,
965
+ "loss": 0.3908,
966
+ "step": 1290
967
+ },
968
+ {
969
+ "epoch": 6.44,
970
+ "grad_norm": 7.626086235046387,
971
+ "learning_rate": 1.9801980198019803e-05,
972
+ "loss": 0.4132,
973
+ "step": 1300
974
+ },
975
+ {
976
+ "epoch": 6.49,
977
+ "grad_norm": 5.670557022094727,
978
+ "learning_rate": 1.952695269526953e-05,
979
+ "loss": 0.4228,
980
+ "step": 1310
981
+ },
982
+ {
983
+ "epoch": 6.53,
984
+ "grad_norm": 5.062951564788818,
985
+ "learning_rate": 1.9251925192519252e-05,
986
+ "loss": 0.4385,
987
+ "step": 1320
988
+ },
989
+ {
990
+ "epoch": 6.58,
991
+ "grad_norm": 3.8000316619873047,
992
+ "learning_rate": 1.8976897689768978e-05,
993
+ "loss": 0.414,
994
+ "step": 1330
995
+ },
996
+ {
997
+ "epoch": 6.63,
998
+ "grad_norm": 3.444866895675659,
999
+ "learning_rate": 1.87018701870187e-05,
1000
+ "loss": 0.4246,
1001
+ "step": 1340
1002
+ },
1003
+ {
1004
+ "epoch": 6.68,
1005
+ "grad_norm": 3.809499740600586,
1006
+ "learning_rate": 1.8426842684268427e-05,
1007
+ "loss": 0.4198,
1008
+ "step": 1350
1009
+ },
1010
+ {
1011
+ "epoch": 6.73,
1012
+ "grad_norm": 4.515871047973633,
1013
+ "learning_rate": 1.8151815181518153e-05,
1014
+ "loss": 0.4051,
1015
+ "step": 1360
1016
+ },
1017
+ {
1018
+ "epoch": 6.78,
1019
+ "grad_norm": 4.38671350479126,
1020
+ "learning_rate": 1.787678767876788e-05,
1021
+ "loss": 0.4153,
1022
+ "step": 1370
1023
+ },
1024
+ {
1025
+ "epoch": 6.83,
1026
+ "grad_norm": 3.238205909729004,
1027
+ "learning_rate": 1.7601760176017602e-05,
1028
+ "loss": 0.3947,
1029
+ "step": 1380
1030
+ },
1031
+ {
1032
+ "epoch": 6.88,
1033
+ "grad_norm": 3.455275058746338,
1034
+ "learning_rate": 1.7326732673267325e-05,
1035
+ "loss": 0.4018,
1036
+ "step": 1390
1037
+ },
1038
+ {
1039
+ "epoch": 6.93,
1040
+ "grad_norm": 3.587259531021118,
1041
+ "learning_rate": 1.705170517051705e-05,
1042
+ "loss": 0.3983,
1043
+ "step": 1400
1044
+ },
1045
+ {
1046
+ "epoch": 6.98,
1047
+ "grad_norm": 4.972535610198975,
1048
+ "learning_rate": 1.6776677667766778e-05,
1049
+ "loss": 0.4117,
1050
+ "step": 1410
1051
+ },
1052
+ {
1053
+ "epoch": 7.0,
1054
+ "eval_accuracy": 0.8376872169975619,
1055
+ "eval_loss": 0.36851075291633606,
1056
+ "eval_runtime": 5.6001,
1057
+ "eval_samples_per_second": 512.666,
1058
+ "eval_steps_per_second": 16.071,
1059
+ "step": 1414
1060
+ },
1061
+ {
1062
+ "epoch": 7.03,
1063
+ "grad_norm": 4.680665493011475,
1064
+ "learning_rate": 1.6501650165016504e-05,
1065
+ "loss": 0.3791,
1066
+ "step": 1420
1067
+ },
1068
+ {
1069
+ "epoch": 7.08,
1070
+ "grad_norm": 4.995221138000488,
1071
+ "learning_rate": 1.6226622662266227e-05,
1072
+ "loss": 0.4119,
1073
+ "step": 1430
1074
+ },
1075
+ {
1076
+ "epoch": 7.13,
1077
+ "grad_norm": 3.969187021255493,
1078
+ "learning_rate": 1.5951595159515953e-05,
1079
+ "loss": 0.3864,
1080
+ "step": 1440
1081
+ },
1082
+ {
1083
+ "epoch": 7.18,
1084
+ "grad_norm": 3.4930386543273926,
1085
+ "learning_rate": 1.567656765676568e-05,
1086
+ "loss": 0.3914,
1087
+ "step": 1450
1088
+ },
1089
+ {
1090
+ "epoch": 7.23,
1091
+ "grad_norm": 4.228696823120117,
1092
+ "learning_rate": 1.5401540154015402e-05,
1093
+ "loss": 0.4072,
1094
+ "step": 1460
1095
+ },
1096
+ {
1097
+ "epoch": 7.28,
1098
+ "grad_norm": 3.5527050495147705,
1099
+ "learning_rate": 1.5126512651265127e-05,
1100
+ "loss": 0.3864,
1101
+ "step": 1470
1102
+ },
1103
+ {
1104
+ "epoch": 7.33,
1105
+ "grad_norm": 3.4274041652679443,
1106
+ "learning_rate": 1.4851485148514851e-05,
1107
+ "loss": 0.3957,
1108
+ "step": 1480
1109
+ },
1110
+ {
1111
+ "epoch": 7.38,
1112
+ "grad_norm": 4.775713920593262,
1113
+ "learning_rate": 1.4576457645764577e-05,
1114
+ "loss": 0.4034,
1115
+ "step": 1490
1116
+ },
1117
+ {
1118
+ "epoch": 7.43,
1119
+ "grad_norm": 4.741795539855957,
1120
+ "learning_rate": 1.4301430143014302e-05,
1121
+ "loss": 0.4042,
1122
+ "step": 1500
1123
+ },
1124
+ {
1125
+ "epoch": 7.48,
1126
+ "grad_norm": 6.42737340927124,
1127
+ "learning_rate": 1.4026402640264028e-05,
1128
+ "loss": 0.4197,
1129
+ "step": 1510
1130
+ },
1131
+ {
1132
+ "epoch": 7.52,
1133
+ "grad_norm": 5.1894121170043945,
1134
+ "learning_rate": 1.3751375137513753e-05,
1135
+ "loss": 0.3956,
1136
+ "step": 1520
1137
+ },
1138
+ {
1139
+ "epoch": 7.57,
1140
+ "grad_norm": 5.648491859436035,
1141
+ "learning_rate": 1.3476347634763479e-05,
1142
+ "loss": 0.4005,
1143
+ "step": 1530
1144
+ },
1145
+ {
1146
+ "epoch": 7.62,
1147
+ "grad_norm": 4.315338611602783,
1148
+ "learning_rate": 1.32013201320132e-05,
1149
+ "loss": 0.3784,
1150
+ "step": 1540
1151
+ },
1152
+ {
1153
+ "epoch": 7.67,
1154
+ "grad_norm": 4.007127285003662,
1155
+ "learning_rate": 1.2926292629262926e-05,
1156
+ "loss": 0.3948,
1157
+ "step": 1550
1158
+ },
1159
+ {
1160
+ "epoch": 7.72,
1161
+ "grad_norm": 4.685734748840332,
1162
+ "learning_rate": 1.265126512651265e-05,
1163
+ "loss": 0.4165,
1164
+ "step": 1560
1165
+ },
1166
+ {
1167
+ "epoch": 7.77,
1168
+ "grad_norm": 5.51500940322876,
1169
+ "learning_rate": 1.2376237623762377e-05,
1170
+ "loss": 0.3872,
1171
+ "step": 1570
1172
+ },
1173
+ {
1174
+ "epoch": 7.82,
1175
+ "grad_norm": 3.208813428878784,
1176
+ "learning_rate": 1.2101210121012102e-05,
1177
+ "loss": 0.3886,
1178
+ "step": 1580
1179
+ },
1180
+ {
1181
+ "epoch": 7.87,
1182
+ "grad_norm": 4.27180814743042,
1183
+ "learning_rate": 1.1826182618261828e-05,
1184
+ "loss": 0.3837,
1185
+ "step": 1590
1186
+ },
1187
+ {
1188
+ "epoch": 7.92,
1189
+ "grad_norm": 9.008807182312012,
1190
+ "learning_rate": 1.155115511551155e-05,
1191
+ "loss": 0.4174,
1192
+ "step": 1600
1193
+ },
1194
+ {
1195
+ "epoch": 7.97,
1196
+ "grad_norm": 3.7245564460754395,
1197
+ "learning_rate": 1.1276127612761277e-05,
1198
+ "loss": 0.3991,
1199
+ "step": 1610
1200
+ },
1201
+ {
1202
+ "epoch": 8.0,
1203
+ "eval_accuracy": 0.8411703239289446,
1204
+ "eval_loss": 0.3601776659488678,
1205
+ "eval_runtime": 5.6002,
1206
+ "eval_samples_per_second": 512.663,
1207
+ "eval_steps_per_second": 16.071,
1208
+ "step": 1616
1209
+ },
1210
+ {
1211
+ "epoch": 8.02,
1212
+ "grad_norm": 4.604074954986572,
1213
+ "learning_rate": 1.1001100110011001e-05,
1214
+ "loss": 0.4121,
1215
+ "step": 1620
1216
+ },
1217
+ {
1218
+ "epoch": 8.07,
1219
+ "grad_norm": 6.1971611976623535,
1220
+ "learning_rate": 1.0726072607260726e-05,
1221
+ "loss": 0.3925,
1222
+ "step": 1630
1223
+ },
1224
+ {
1225
+ "epoch": 8.12,
1226
+ "grad_norm": 3.589857816696167,
1227
+ "learning_rate": 1.045104510451045e-05,
1228
+ "loss": 0.4339,
1229
+ "step": 1640
1230
+ },
1231
+ {
1232
+ "epoch": 8.17,
1233
+ "grad_norm": 3.007169246673584,
1234
+ "learning_rate": 1.0176017601760177e-05,
1235
+ "loss": 0.3841,
1236
+ "step": 1650
1237
+ },
1238
+ {
1239
+ "epoch": 8.22,
1240
+ "grad_norm": 8.007019996643066,
1241
+ "learning_rate": 9.900990099009901e-06,
1242
+ "loss": 0.4005,
1243
+ "step": 1660
1244
+ },
1245
+ {
1246
+ "epoch": 8.27,
1247
+ "grad_norm": 4.514450550079346,
1248
+ "learning_rate": 9.625962596259626e-06,
1249
+ "loss": 0.4072,
1250
+ "step": 1670
1251
+ },
1252
+ {
1253
+ "epoch": 8.32,
1254
+ "grad_norm": 6.048130989074707,
1255
+ "learning_rate": 9.35093509350935e-06,
1256
+ "loss": 0.3756,
1257
+ "step": 1680
1258
+ },
1259
+ {
1260
+ "epoch": 8.37,
1261
+ "grad_norm": 6.211109638214111,
1262
+ "learning_rate": 9.075907590759077e-06,
1263
+ "loss": 0.3953,
1264
+ "step": 1690
1265
+ },
1266
+ {
1267
+ "epoch": 8.42,
1268
+ "grad_norm": 3.7817318439483643,
1269
+ "learning_rate": 8.800880088008801e-06,
1270
+ "loss": 0.3685,
1271
+ "step": 1700
1272
+ },
1273
+ {
1274
+ "epoch": 8.47,
1275
+ "grad_norm": 5.379549980163574,
1276
+ "learning_rate": 8.525852585258526e-06,
1277
+ "loss": 0.3713,
1278
+ "step": 1710
1279
+ },
1280
+ {
1281
+ "epoch": 8.51,
1282
+ "grad_norm": 4.123336315155029,
1283
+ "learning_rate": 8.250825082508252e-06,
1284
+ "loss": 0.3629,
1285
+ "step": 1720
1286
+ },
1287
+ {
1288
+ "epoch": 8.56,
1289
+ "grad_norm": 4.12966251373291,
1290
+ "learning_rate": 7.975797579757976e-06,
1291
+ "loss": 0.3884,
1292
+ "step": 1730
1293
+ },
1294
+ {
1295
+ "epoch": 8.61,
1296
+ "grad_norm": 4.9768967628479,
1297
+ "learning_rate": 7.700770077007701e-06,
1298
+ "loss": 0.4092,
1299
+ "step": 1740
1300
+ },
1301
+ {
1302
+ "epoch": 8.66,
1303
+ "grad_norm": 5.166417598724365,
1304
+ "learning_rate": 7.4257425742574256e-06,
1305
+ "loss": 0.3891,
1306
+ "step": 1750
1307
+ },
1308
+ {
1309
+ "epoch": 8.71,
1310
+ "grad_norm": 6.434201240539551,
1311
+ "learning_rate": 7.150715071507151e-06,
1312
+ "loss": 0.3592,
1313
+ "step": 1760
1314
+ },
1315
+ {
1316
+ "epoch": 8.76,
1317
+ "grad_norm": 7.199638843536377,
1318
+ "learning_rate": 6.875687568756876e-06,
1319
+ "loss": 0.3909,
1320
+ "step": 1770
1321
+ },
1322
+ {
1323
+ "epoch": 8.81,
1324
+ "grad_norm": 7.480076313018799,
1325
+ "learning_rate": 6.6006600660066e-06,
1326
+ "loss": 0.3895,
1327
+ "step": 1780
1328
+ },
1329
+ {
1330
+ "epoch": 8.86,
1331
+ "grad_norm": 5.124411582946777,
1332
+ "learning_rate": 6.325632563256325e-06,
1333
+ "loss": 0.391,
1334
+ "step": 1790
1335
+ },
1336
+ {
1337
+ "epoch": 8.91,
1338
+ "grad_norm": 9.751836776733398,
1339
+ "learning_rate": 6.050605060506051e-06,
1340
+ "loss": 0.3908,
1341
+ "step": 1800
1342
+ },
1343
+ {
1344
+ "epoch": 8.96,
1345
+ "grad_norm": 4.501916408538818,
1346
+ "learning_rate": 5.775577557755775e-06,
1347
+ "loss": 0.3936,
1348
+ "step": 1810
1349
+ },
1350
+ {
1351
+ "epoch": 9.0,
1352
+ "eval_accuracy": 0.842911877394636,
1353
+ "eval_loss": 0.3542243540287018,
1354
+ "eval_runtime": 5.623,
1355
+ "eval_samples_per_second": 510.58,
1356
+ "eval_steps_per_second": 16.006,
1357
+ "step": 1818
1358
+ },
1359
+ {
1360
+ "epoch": 9.01,
1361
+ "grad_norm": 4.195034027099609,
1362
+ "learning_rate": 5.500550055005501e-06,
1363
+ "loss": 0.4147,
1364
+ "step": 1820
1365
+ },
1366
+ {
1367
+ "epoch": 9.06,
1368
+ "grad_norm": 4.904818058013916,
1369
+ "learning_rate": 5.225522552255225e-06,
1370
+ "loss": 0.3822,
1371
+ "step": 1830
1372
+ },
1373
+ {
1374
+ "epoch": 9.11,
1375
+ "grad_norm": 4.105266094207764,
1376
+ "learning_rate": 4.950495049504951e-06,
1377
+ "loss": 0.401,
1378
+ "step": 1840
1379
+ },
1380
+ {
1381
+ "epoch": 9.16,
1382
+ "grad_norm": 4.364089488983154,
1383
+ "learning_rate": 4.675467546754675e-06,
1384
+ "loss": 0.398,
1385
+ "step": 1850
1386
+ },
1387
+ {
1388
+ "epoch": 9.21,
1389
+ "grad_norm": 5.043737411499023,
1390
+ "learning_rate": 4.400440044004401e-06,
1391
+ "loss": 0.4218,
1392
+ "step": 1860
1393
+ },
1394
+ {
1395
+ "epoch": 9.26,
1396
+ "grad_norm": 4.617894172668457,
1397
+ "learning_rate": 4.125412541254126e-06,
1398
+ "loss": 0.4002,
1399
+ "step": 1870
1400
+ },
1401
+ {
1402
+ "epoch": 9.31,
1403
+ "grad_norm": 3.750427007675171,
1404
+ "learning_rate": 3.8503850385038505e-06,
1405
+ "loss": 0.4005,
1406
+ "step": 1880
1407
+ },
1408
+ {
1409
+ "epoch": 9.36,
1410
+ "grad_norm": 4.708156585693359,
1411
+ "learning_rate": 3.5753575357535755e-06,
1412
+ "loss": 0.4067,
1413
+ "step": 1890
1414
+ },
1415
+ {
1416
+ "epoch": 9.41,
1417
+ "grad_norm": 4.125495910644531,
1418
+ "learning_rate": 3.3003300330033e-06,
1419
+ "loss": 0.3745,
1420
+ "step": 1900
1421
+ },
1422
+ {
1423
+ "epoch": 9.46,
1424
+ "grad_norm": 3.2616422176361084,
1425
+ "learning_rate": 3.0253025302530254e-06,
1426
+ "loss": 0.3914,
1427
+ "step": 1910
1428
+ },
1429
+ {
1430
+ "epoch": 9.5,
1431
+ "grad_norm": 3.8317136764526367,
1432
+ "learning_rate": 2.7502750275027504e-06,
1433
+ "loss": 0.383,
1434
+ "step": 1920
1435
+ },
1436
+ {
1437
+ "epoch": 9.55,
1438
+ "grad_norm": 4.044767379760742,
1439
+ "learning_rate": 2.4752475247524753e-06,
1440
+ "loss": 0.3713,
1441
+ "step": 1930
1442
+ },
1443
+ {
1444
+ "epoch": 9.6,
1445
+ "grad_norm": 5.957603931427002,
1446
+ "learning_rate": 2.2002200220022003e-06,
1447
+ "loss": 0.3524,
1448
+ "step": 1940
1449
+ },
1450
+ {
1451
+ "epoch": 9.65,
1452
+ "grad_norm": 3.8878390789031982,
1453
+ "learning_rate": 1.9251925192519253e-06,
1454
+ "loss": 0.3876,
1455
+ "step": 1950
1456
+ },
1457
+ {
1458
+ "epoch": 9.7,
1459
+ "grad_norm": 4.049957752227783,
1460
+ "learning_rate": 1.65016501650165e-06,
1461
+ "loss": 0.3726,
1462
+ "step": 1960
1463
+ },
1464
+ {
1465
+ "epoch": 9.75,
1466
+ "grad_norm": 3.9568910598754883,
1467
+ "learning_rate": 1.3751375137513752e-06,
1468
+ "loss": 0.397,
1469
+ "step": 1970
1470
+ },
1471
+ {
1472
+ "epoch": 9.8,
1473
+ "grad_norm": 5.049424171447754,
1474
+ "learning_rate": 1.1001100110011001e-06,
1475
+ "loss": 0.3771,
1476
+ "step": 1980
1477
+ },
1478
+ {
1479
+ "epoch": 9.85,
1480
+ "grad_norm": 3.9506354331970215,
1481
+ "learning_rate": 8.25082508250825e-07,
1482
+ "loss": 0.3816,
1483
+ "step": 1990
1484
+ },
1485
+ {
1486
+ "epoch": 9.9,
1487
+ "grad_norm": 4.543933391571045,
1488
+ "learning_rate": 5.500550055005501e-07,
1489
+ "loss": 0.3961,
1490
+ "step": 2000
1491
+ },
1492
+ {
1493
+ "epoch": 9.95,
1494
+ "grad_norm": 4.267475605010986,
1495
+ "learning_rate": 2.7502750275027504e-07,
1496
+ "loss": 0.3612,
1497
+ "step": 2010
1498
+ },
1499
+ {
1500
+ "epoch": 10.0,
1501
+ "grad_norm": 7.206649303436279,
1502
+ "learning_rate": 0.0,
1503
+ "loss": 0.3422,
1504
+ "step": 2020
1505
+ },
1506
+ {
1507
+ "epoch": 10.0,
1508
+ "eval_accuracy": 0.8397770811563915,
1509
+ "eval_loss": 0.35397040843963623,
1510
+ "eval_runtime": 5.6519,
1511
+ "eval_samples_per_second": 507.973,
1512
+ "eval_steps_per_second": 15.924,
1513
+ "step": 2020
1514
+ },
1515
+ {
1516
+ "epoch": 10.0,
1517
+ "step": 2020,
1518
+ "total_flos": 1.2892209231838003e+18,
1519
+ "train_loss": 0.45288851414576614,
1520
+ "train_runtime": 810.9556,
1521
+ "train_samples_per_second": 318.599,
1522
+ "train_steps_per_second": 2.491
1523
+ }
1524
+ ],
1525
+ "logging_steps": 10,
1526
+ "max_steps": 2020,
1527
+ "num_input_tokens_seen": 0,
1528
+ "num_train_epochs": 10,
1529
+ "save_steps": 500,
1530
+ "total_flos": 1.2892209231838003e+18,
1531
+ "train_batch_size": 32,
1532
+ "trial_name": null,
1533
+ "trial_params": null
1534
+ }