presencesw commited on
Commit
d09a996
·
verified ·
1 Parent(s): b5c218a

End of training

Browse files
README.md CHANGED
@@ -4,6 +4,7 @@ license: gemma
4
  base_model: google/gemma-3-1b-it
5
  tags:
6
  - llama-factory
 
7
  - generated_from_trainer
8
  model-index:
9
  - name: gemma-3-1b-it_MED_NLI
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # gemma-3-1b-it_MED_NLI
17
 
18
- This model is a fine-tuned version of [google/gemma-3-1b-it](https://huggingface.co/google/gemma-3-1b-it) on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
  - Loss: 0.0177
21
 
 
4
  base_model: google/gemma-3-1b-it
5
  tags:
6
  - llama-factory
7
+ - full
8
  - generated_from_trainer
9
  model-index:
10
  - name: gemma-3-1b-it_MED_NLI
 
16
 
17
  # gemma-3-1b-it_MED_NLI
18
 
19
+ This model is a fine-tuned version of [google/gemma-3-1b-it](https://huggingface.co/google/gemma-3-1b-it) on the zero_shot dataset.
20
  It achieves the following results on the evaluation set:
21
  - Loss: 0.0177
22
 
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_loss": 0.017735935747623444,
4
+ "eval_runtime": 610.4181,
5
+ "eval_samples_per_second": 47.757,
6
+ "eval_steps_per_second": 11.939,
7
+ "total_flos": 1.8987963054248755e+17,
8
+ "train_loss": 0.0597800020030789,
9
+ "train_runtime": 12067.3267,
10
+ "train_samples_per_second": 5.637,
11
+ "train_steps_per_second": 0.705
12
+ }
eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_loss": 0.017735935747623444,
4
+ "eval_runtime": 610.4181,
5
+ "eval_samples_per_second": 47.757,
6
+ "eval_steps_per_second": 11.939
7
+ }
runs/Apr13_02-06-24_9e72a503a884/events.out.tfevents.1744522924.9e72a503a884.3411.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c7b8e33f644892977a9ebf391ac276b2110262b5e1db0b023d03c6d24e237f7
3
+ size 359
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "total_flos": 1.8987963054248755e+17,
4
+ "train_loss": 0.0597800020030789,
5
+ "train_runtime": 12067.3267,
6
+ "train_samples_per_second": 5.637,
7
+ "train_steps_per_second": 0.705
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1297 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.0,
6
+ "eval_steps": 1000,
7
+ "global_step": 8503,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.005880277549100318,
14
+ "grad_norm": 387.39825439453125,
15
+ "learning_rate": 5.875440658049355e-07,
16
+ "loss": 5.0951,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.011760555098200636,
21
+ "grad_norm": 66.9789047241211,
22
+ "learning_rate": 1.175088131609871e-06,
23
+ "loss": 0.8587,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.01764083264730095,
28
+ "grad_norm": 11.23346996307373,
29
+ "learning_rate": 1.762632197414806e-06,
30
+ "loss": 0.084,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.02352111019640127,
35
+ "grad_norm": 11.451313018798828,
36
+ "learning_rate": 2.350176263219742e-06,
37
+ "loss": 0.051,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.02940138774550159,
42
+ "grad_norm": 12.508331298828125,
43
+ "learning_rate": 2.937720329024677e-06,
44
+ "loss": 0.0592,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.0352816652946019,
49
+ "grad_norm": 42.7663688659668,
50
+ "learning_rate": 3.525264394829612e-06,
51
+ "loss": 0.0519,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.041161942843702226,
56
+ "grad_norm": 9.149981498718262,
57
+ "learning_rate": 4.112808460634548e-06,
58
+ "loss": 0.0418,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 0.04704222039280254,
63
+ "grad_norm": 16.077157974243164,
64
+ "learning_rate": 4.700352526439484e-06,
65
+ "loss": 0.0665,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 0.05292249794190286,
70
+ "grad_norm": 8.89875316619873,
71
+ "learning_rate": 5.287896592244419e-06,
72
+ "loss": 0.052,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 0.05880277549100318,
77
+ "grad_norm": 10.934527397155762,
78
+ "learning_rate": 5.875440658049354e-06,
79
+ "loss": 0.0405,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 0.0646830530401035,
84
+ "grad_norm": 5.134288787841797,
85
+ "learning_rate": 6.46298472385429e-06,
86
+ "loss": 0.0499,
87
+ "step": 550
88
+ },
89
+ {
90
+ "epoch": 0.0705633305892038,
91
+ "grad_norm": 2.750032663345337,
92
+ "learning_rate": 7.050528789659224e-06,
93
+ "loss": 0.0444,
94
+ "step": 600
95
+ },
96
+ {
97
+ "epoch": 0.07644360813830413,
98
+ "grad_norm": 3.613896131515503,
99
+ "learning_rate": 7.63807285546416e-06,
100
+ "loss": 0.0417,
101
+ "step": 650
102
+ },
103
+ {
104
+ "epoch": 0.08232388568740445,
105
+ "grad_norm": 5.510785102844238,
106
+ "learning_rate": 8.225616921269097e-06,
107
+ "loss": 0.0337,
108
+ "step": 700
109
+ },
110
+ {
111
+ "epoch": 0.08820416323650476,
112
+ "grad_norm": 21.227962493896484,
113
+ "learning_rate": 8.81316098707403e-06,
114
+ "loss": 0.0314,
115
+ "step": 750
116
+ },
117
+ {
118
+ "epoch": 0.09408444078560509,
119
+ "grad_norm": 11.67798137664795,
120
+ "learning_rate": 9.400705052878968e-06,
121
+ "loss": 0.0386,
122
+ "step": 800
123
+ },
124
+ {
125
+ "epoch": 0.0999647183347054,
126
+ "grad_norm": 0.9770076870918274,
127
+ "learning_rate": 9.988249118683903e-06,
128
+ "loss": 0.0322,
129
+ "step": 850
130
+ },
131
+ {
132
+ "epoch": 0.10584499588380572,
133
+ "grad_norm": 2.259112596511841,
134
+ "learning_rate": 9.998988263671598e-06,
135
+ "loss": 0.037,
136
+ "step": 900
137
+ },
138
+ {
139
+ "epoch": 0.11172527343290603,
140
+ "grad_norm": 5.629507064819336,
141
+ "learning_rate": 9.995870471854679e-06,
142
+ "loss": 0.0328,
143
+ "step": 950
144
+ },
145
+ {
146
+ "epoch": 0.11760555098200635,
147
+ "grad_norm": 2.3545737266540527,
148
+ "learning_rate": 9.990647516930925e-06,
149
+ "loss": 0.0331,
150
+ "step": 1000
151
+ },
152
+ {
153
+ "epoch": 0.11760555098200635,
154
+ "eval_loss": 0.03788302466273308,
155
+ "eval_runtime": 619.4182,
156
+ "eval_samples_per_second": 47.064,
157
+ "eval_steps_per_second": 11.766,
158
+ "step": 1000
159
+ },
160
+ {
161
+ "epoch": 0.12348582853110666,
162
+ "grad_norm": 1.4001481533050537,
163
+ "learning_rate": 9.983321599752438e-06,
164
+ "loss": 0.0359,
165
+ "step": 1050
166
+ },
167
+ {
168
+ "epoch": 0.129366106080207,
169
+ "grad_norm": 10.03685188293457,
170
+ "learning_rate": 9.9738958073189e-06,
171
+ "loss": 0.0306,
172
+ "step": 1100
173
+ },
174
+ {
175
+ "epoch": 0.1352463836293073,
176
+ "grad_norm": 8.22728443145752,
177
+ "learning_rate": 9.962374111476778e-06,
178
+ "loss": 0.031,
179
+ "step": 1150
180
+ },
181
+ {
182
+ "epoch": 0.1411266611784076,
183
+ "grad_norm": 4.021151065826416,
184
+ "learning_rate": 9.948761367245665e-06,
185
+ "loss": 0.0333,
186
+ "step": 1200
187
+ },
188
+ {
189
+ "epoch": 0.14700693872750795,
190
+ "grad_norm": 4.213313102722168,
191
+ "learning_rate": 9.933063310772463e-06,
192
+ "loss": 0.0347,
193
+ "step": 1250
194
+ },
195
+ {
196
+ "epoch": 0.15288721627660826,
197
+ "grad_norm": 1.7054405212402344,
198
+ "learning_rate": 9.915286556914286e-06,
199
+ "loss": 0.0326,
200
+ "step": 1300
201
+ },
202
+ {
203
+ "epoch": 0.15876749382570857,
204
+ "grad_norm": 3.276946783065796,
205
+ "learning_rate": 9.89543859645109e-06,
206
+ "loss": 0.0323,
207
+ "step": 1350
208
+ },
209
+ {
210
+ "epoch": 0.1646477713748089,
211
+ "grad_norm": 5.019853591918945,
212
+ "learning_rate": 9.873527792929196e-06,
213
+ "loss": 0.0274,
214
+ "step": 1400
215
+ },
216
+ {
217
+ "epoch": 0.17052804892390921,
218
+ "grad_norm": 6.062413692474365,
219
+ "learning_rate": 9.84956337913706e-06,
220
+ "loss": 0.0294,
221
+ "step": 1450
222
+ },
223
+ {
224
+ "epoch": 0.17640832647300952,
225
+ "grad_norm": 4.063215255737305,
226
+ "learning_rate": 9.82355545321475e-06,
227
+ "loss": 0.0336,
228
+ "step": 1500
229
+ },
230
+ {
231
+ "epoch": 0.18228860402210983,
232
+ "grad_norm": 3.2010011672973633,
233
+ "learning_rate": 9.795514974398789e-06,
234
+ "loss": 0.0295,
235
+ "step": 1550
236
+ },
237
+ {
238
+ "epoch": 0.18816888157121017,
239
+ "grad_norm": 4.517515182495117,
240
+ "learning_rate": 9.765453758404144e-06,
241
+ "loss": 0.0301,
242
+ "step": 1600
243
+ },
244
+ {
245
+ "epoch": 0.19404915912031048,
246
+ "grad_norm": 4.797945022583008,
247
+ "learning_rate": 9.733384472445308e-06,
248
+ "loss": 0.0309,
249
+ "step": 1650
250
+ },
251
+ {
252
+ "epoch": 0.1999294366694108,
253
+ "grad_norm": 2.705932378768921,
254
+ "learning_rate": 9.699320629898589e-06,
255
+ "loss": 0.026,
256
+ "step": 1700
257
+ },
258
+ {
259
+ "epoch": 0.2058097142185111,
260
+ "grad_norm": 20.035690307617188,
261
+ "learning_rate": 9.663276584607831e-06,
262
+ "loss": 0.0264,
263
+ "step": 1750
264
+ },
265
+ {
266
+ "epoch": 0.21168999176761144,
267
+ "grad_norm": 1.2530241012573242,
268
+ "learning_rate": 9.625267524835974e-06,
269
+ "loss": 0.0277,
270
+ "step": 1800
271
+ },
272
+ {
273
+ "epoch": 0.21757026931671175,
274
+ "grad_norm": 3.373883008956909,
275
+ "learning_rate": 9.585309466865029e-06,
276
+ "loss": 0.0276,
277
+ "step": 1850
278
+ },
279
+ {
280
+ "epoch": 0.22345054686581206,
281
+ "grad_norm": 5.847758769989014,
282
+ "learning_rate": 9.54341924824712e-06,
283
+ "loss": 0.0289,
284
+ "step": 1900
285
+ },
286
+ {
287
+ "epoch": 0.22933082441491237,
288
+ "grad_norm": 1.0913113355636597,
289
+ "learning_rate": 9.499614520709457e-06,
290
+ "loss": 0.0256,
291
+ "step": 1950
292
+ },
293
+ {
294
+ "epoch": 0.2352111019640127,
295
+ "grad_norm": 4.371528148651123,
296
+ "learning_rate": 9.453913742716256e-06,
297
+ "loss": 0.0299,
298
+ "step": 2000
299
+ },
300
+ {
301
+ "epoch": 0.2352111019640127,
302
+ "eval_loss": 0.026240274310112,
303
+ "eval_runtime": 619.3143,
304
+ "eval_samples_per_second": 47.071,
305
+ "eval_steps_per_second": 11.768,
306
+ "step": 2000
307
+ },
308
+ {
309
+ "epoch": 0.24109137951311302,
310
+ "grad_norm": 4.8858208656311035,
311
+ "learning_rate": 9.40633617169069e-06,
312
+ "loss": 0.0273,
313
+ "step": 2050
314
+ },
315
+ {
316
+ "epoch": 0.24697165706221333,
317
+ "grad_norm": 1.7557135820388794,
318
+ "learning_rate": 9.35690185590018e-06,
319
+ "loss": 0.0265,
320
+ "step": 2100
321
+ },
322
+ {
323
+ "epoch": 0.25285193461131367,
324
+ "grad_norm": 2.1983985900878906,
325
+ "learning_rate": 9.305631626008454e-06,
326
+ "loss": 0.0249,
327
+ "step": 2150
328
+ },
329
+ {
330
+ "epoch": 0.258732212160414,
331
+ "grad_norm": 4.288529872894287,
332
+ "learning_rate": 9.252547086297895e-06,
333
+ "loss": 0.0243,
334
+ "step": 2200
335
+ },
336
+ {
337
+ "epoch": 0.2646124897095143,
338
+ "grad_norm": 1.228149652481079,
339
+ "learning_rate": 9.197670605565932e-06,
340
+ "loss": 0.0271,
341
+ "step": 2250
342
+ },
343
+ {
344
+ "epoch": 0.2704927672586146,
345
+ "grad_norm": 4.099296569824219,
346
+ "learning_rate": 9.141025307699246e-06,
347
+ "loss": 0.0265,
348
+ "step": 2300
349
+ },
350
+ {
351
+ "epoch": 0.2763730448077149,
352
+ "grad_norm": 3.512291431427002,
353
+ "learning_rate": 9.082635061929817e-06,
354
+ "loss": 0.0286,
355
+ "step": 2350
356
+ },
357
+ {
358
+ "epoch": 0.2822533223568152,
359
+ "grad_norm": 1.2961852550506592,
360
+ "learning_rate": 9.022524472776897e-06,
361
+ "loss": 0.0288,
362
+ "step": 2400
363
+ },
364
+ {
365
+ "epoch": 0.2881335999059156,
366
+ "grad_norm": 4.3626322746276855,
367
+ "learning_rate": 8.960718869679132e-06,
368
+ "loss": 0.0252,
369
+ "step": 2450
370
+ },
371
+ {
372
+ "epoch": 0.2940138774550159,
373
+ "grad_norm": 1.3140321969985962,
374
+ "learning_rate": 8.89724429632124e-06,
375
+ "loss": 0.0243,
376
+ "step": 2500
377
+ },
378
+ {
379
+ "epoch": 0.2998941550041162,
380
+ "grad_norm": 4.2938432693481445,
381
+ "learning_rate": 8.832127499659687e-06,
382
+ "loss": 0.0216,
383
+ "step": 2550
384
+ },
385
+ {
386
+ "epoch": 0.3057744325532165,
387
+ "grad_norm": 2.0306389331817627,
388
+ "learning_rate": 8.765395918652062e-06,
389
+ "loss": 0.0245,
390
+ "step": 2600
391
+ },
392
+ {
393
+ "epoch": 0.3116547101023168,
394
+ "grad_norm": 3.160212993621826,
395
+ "learning_rate": 8.697077672694809e-06,
396
+ "loss": 0.0292,
397
+ "step": 2650
398
+ },
399
+ {
400
+ "epoch": 0.31753498765141713,
401
+ "grad_norm": 8.705582618713379,
402
+ "learning_rate": 8.627201549774273e-06,
403
+ "loss": 0.0243,
404
+ "step": 2700
405
+ },
406
+ {
407
+ "epoch": 0.32341526520051744,
408
+ "grad_norm": 5.516971588134766,
409
+ "learning_rate": 8.55579699433599e-06,
410
+ "loss": 0.0272,
411
+ "step": 2750
412
+ },
413
+ {
414
+ "epoch": 0.3292955427496178,
415
+ "grad_norm": 3.2388885021209717,
416
+ "learning_rate": 8.482894094877372e-06,
417
+ "loss": 0.0268,
418
+ "step": 2800
419
+ },
420
+ {
421
+ "epoch": 0.3351758202987181,
422
+ "grad_norm": 1.1581066846847534,
423
+ "learning_rate": 8.408523571269e-06,
424
+ "loss": 0.0251,
425
+ "step": 2850
426
+ },
427
+ {
428
+ "epoch": 0.34105609784781843,
429
+ "grad_norm": 1.0647870302200317,
430
+ "learning_rate": 8.332716761809857e-06,
431
+ "loss": 0.0304,
432
+ "step": 2900
433
+ },
434
+ {
435
+ "epoch": 0.34693637539691874,
436
+ "grad_norm": 1.6037498712539673,
437
+ "learning_rate": 8.255505610021981e-06,
438
+ "loss": 0.0241,
439
+ "step": 2950
440
+ },
441
+ {
442
+ "epoch": 0.35281665294601905,
443
+ "grad_norm": 3.1624505519866943,
444
+ "learning_rate": 8.176922651190085e-06,
445
+ "loss": 0.0251,
446
+ "step": 3000
447
+ },
448
+ {
449
+ "epoch": 0.35281665294601905,
450
+ "eval_loss": 0.028408875688910484,
451
+ "eval_runtime": 619.7012,
452
+ "eval_samples_per_second": 47.042,
453
+ "eval_steps_per_second": 11.761,
454
+ "step": 3000
455
+ },
456
+ {
457
+ "epoch": 0.35869693049511936,
458
+ "grad_norm": 4.156674861907959,
459
+ "learning_rate": 8.097000998651812e-06,
460
+ "loss": 0.0316,
461
+ "step": 3050
462
+ },
463
+ {
464
+ "epoch": 0.36457720804421967,
465
+ "grad_norm": 1.2711337804794312,
466
+ "learning_rate": 8.015774329844417e-06,
467
+ "loss": 0.0239,
468
+ "step": 3100
469
+ },
470
+ {
471
+ "epoch": 0.37045748559332,
472
+ "grad_norm": 1.4552150964736938,
473
+ "learning_rate": 7.933276872113754e-06,
474
+ "loss": 0.0226,
475
+ "step": 3150
476
+ },
477
+ {
478
+ "epoch": 0.37633776314242035,
479
+ "grad_norm": 6.432965278625488,
480
+ "learning_rate": 7.849543388291524e-06,
481
+ "loss": 0.0256,
482
+ "step": 3200
483
+ },
484
+ {
485
+ "epoch": 0.38221804069152066,
486
+ "grad_norm": 3.6660749912261963,
487
+ "learning_rate": 7.764609162046894e-06,
488
+ "loss": 0.0265,
489
+ "step": 3250
490
+ },
491
+ {
492
+ "epoch": 0.38809831824062097,
493
+ "grad_norm": 3.6166951656341553,
494
+ "learning_rate": 7.678509983018656e-06,
495
+ "loss": 0.0256,
496
+ "step": 3300
497
+ },
498
+ {
499
+ "epoch": 0.3939785957897213,
500
+ "grad_norm": 8.923437118530273,
501
+ "learning_rate": 7.591282131734139e-06,
502
+ "loss": 0.024,
503
+ "step": 3350
504
+ },
505
+ {
506
+ "epoch": 0.3998588733388216,
507
+ "grad_norm": 3.535114049911499,
508
+ "learning_rate": 7.50296236432132e-06,
509
+ "loss": 0.0208,
510
+ "step": 3400
511
+ },
512
+ {
513
+ "epoch": 0.4057391508879219,
514
+ "grad_norm": 3.5117030143737793,
515
+ "learning_rate": 7.413587897020496e-06,
516
+ "loss": 0.0243,
517
+ "step": 3450
518
+ },
519
+ {
520
+ "epoch": 0.4116194284370222,
521
+ "grad_norm": 1.3753975629806519,
522
+ "learning_rate": 7.323196390502074e-06,
523
+ "loss": 0.0213,
524
+ "step": 3500
525
+ },
526
+ {
527
+ "epoch": 0.41749970598612257,
528
+ "grad_norm": 1.682301640510559,
529
+ "learning_rate": 7.231825933997105e-06,
530
+ "loss": 0.0261,
531
+ "step": 3550
532
+ },
533
+ {
534
+ "epoch": 0.4233799835352229,
535
+ "grad_norm": 4.877748489379883,
536
+ "learning_rate": 7.139515029247213e-06,
537
+ "loss": 0.0286,
538
+ "step": 3600
539
+ },
540
+ {
541
+ "epoch": 0.4292602610843232,
542
+ "grad_norm": 1.995045781135559,
543
+ "learning_rate": 7.046302574280703e-06,
544
+ "loss": 0.0244,
545
+ "step": 3650
546
+ },
547
+ {
548
+ "epoch": 0.4351405386334235,
549
+ "grad_norm": 2.8061089515686035,
550
+ "learning_rate": 6.952227847021697e-06,
551
+ "loss": 0.023,
552
+ "step": 3700
553
+ },
554
+ {
555
+ "epoch": 0.4410208161825238,
556
+ "grad_norm": 1.9934951066970825,
557
+ "learning_rate": 6.857330488739159e-06,
558
+ "loss": 0.0224,
559
+ "step": 3750
560
+ },
561
+ {
562
+ "epoch": 0.4469010937316241,
563
+ "grad_norm": 4.9554290771484375,
564
+ "learning_rate": 6.76165048734285e-06,
565
+ "loss": 0.0235,
566
+ "step": 3800
567
+ },
568
+ {
569
+ "epoch": 0.45278137128072443,
570
+ "grad_norm": 2.459325075149536,
571
+ "learning_rate": 6.665228160533186e-06,
572
+ "loss": 0.0251,
573
+ "step": 3850
574
+ },
575
+ {
576
+ "epoch": 0.45866164882982474,
577
+ "grad_norm": 3.0527613162994385,
578
+ "learning_rate": 6.568104138812141e-06,
579
+ "loss": 0.0208,
580
+ "step": 3900
581
+ },
582
+ {
583
+ "epoch": 0.4645419263789251,
584
+ "grad_norm": 2.7639405727386475,
585
+ "learning_rate": 6.470319348362344e-06,
586
+ "loss": 0.0242,
587
+ "step": 3950
588
+ },
589
+ {
590
+ "epoch": 0.4704222039280254,
591
+ "grad_norm": 1.0516586303710938,
592
+ "learning_rate": 6.371914993801573e-06,
593
+ "loss": 0.0213,
594
+ "step": 4000
595
+ },
596
+ {
597
+ "epoch": 0.4704222039280254,
598
+ "eval_loss": 0.02518468163907528,
599
+ "eval_runtime": 612.2227,
600
+ "eval_samples_per_second": 47.617,
601
+ "eval_steps_per_second": 11.904,
602
+ "step": 4000
603
+ },
604
+ {
605
+ "epoch": 0.4763024814771257,
606
+ "grad_norm": 2.6117072105407715,
607
+ "learning_rate": 6.272932540819929e-06,
608
+ "loss": 0.0201,
609
+ "step": 4050
610
+ },
611
+ {
612
+ "epoch": 0.48218275902622604,
613
+ "grad_norm": 3.2772061824798584,
614
+ "learning_rate": 6.173413698706999e-06,
615
+ "loss": 0.0245,
616
+ "step": 4100
617
+ },
618
+ {
619
+ "epoch": 0.48806303657532635,
620
+ "grad_norm": 9.428531646728516,
621
+ "learning_rate": 6.073400402776364e-06,
622
+ "loss": 0.0229,
623
+ "step": 4150
624
+ },
625
+ {
626
+ "epoch": 0.49394331412442666,
627
+ "grad_norm": 3.652819871902466,
628
+ "learning_rate": 5.972934796694871e-06,
629
+ "loss": 0.0203,
630
+ "step": 4200
631
+ },
632
+ {
633
+ "epoch": 0.49982359167352697,
634
+ "grad_norm": 3.571290969848633,
635
+ "learning_rate": 5.872059214724112e-06,
636
+ "loss": 0.0224,
637
+ "step": 4250
638
+ },
639
+ {
640
+ "epoch": 0.5057038692226273,
641
+ "grad_norm": 7.393253326416016,
642
+ "learning_rate": 5.770816163881581e-06,
643
+ "loss": 0.0233,
644
+ "step": 4300
645
+ },
646
+ {
647
+ "epoch": 0.5115841467717276,
648
+ "grad_norm": 3.7029826641082764,
649
+ "learning_rate": 5.669248306029042e-06,
650
+ "loss": 0.0211,
651
+ "step": 4350
652
+ },
653
+ {
654
+ "epoch": 0.517464424320828,
655
+ "grad_norm": 1.5849096775054932,
656
+ "learning_rate": 5.567398439895643e-06,
657
+ "loss": 0.0228,
658
+ "step": 4400
659
+ },
660
+ {
661
+ "epoch": 0.5233447018699282,
662
+ "grad_norm": 4.666280269622803,
663
+ "learning_rate": 5.465309483043364e-06,
664
+ "loss": 0.019,
665
+ "step": 4450
666
+ },
667
+ {
668
+ "epoch": 0.5292249794190286,
669
+ "grad_norm": 1.970779299736023,
670
+ "learning_rate": 5.363024453782388e-06,
671
+ "loss": 0.0262,
672
+ "step": 4500
673
+ },
674
+ {
675
+ "epoch": 0.5351052569681289,
676
+ "grad_norm": 3.114126682281494,
677
+ "learning_rate": 5.260586453044011e-06,
678
+ "loss": 0.0208,
679
+ "step": 4550
680
+ },
681
+ {
682
+ "epoch": 0.5409855345172292,
683
+ "grad_norm": 2.33022141456604,
684
+ "learning_rate": 5.158038646218749e-06,
685
+ "loss": 0.0248,
686
+ "step": 4600
687
+ },
688
+ {
689
+ "epoch": 0.5468658120663296,
690
+ "grad_norm": 1.2761564254760742,
691
+ "learning_rate": 5.055424244967284e-06,
692
+ "loss": 0.0206,
693
+ "step": 4650
694
+ },
695
+ {
696
+ "epoch": 0.5527460896154298,
697
+ "grad_norm": 1.679206132888794,
698
+ "learning_rate": 4.95278648901189e-06,
699
+ "loss": 0.0202,
700
+ "step": 4700
701
+ },
702
+ {
703
+ "epoch": 0.5586263671645302,
704
+ "grad_norm": 3.895859479904175,
705
+ "learning_rate": 4.850168627916068e-06,
706
+ "loss": 0.023,
707
+ "step": 4750
708
+ },
709
+ {
710
+ "epoch": 0.5645066447136304,
711
+ "grad_norm": 1.8703380823135376,
712
+ "learning_rate": 4.7476139028600085e-06,
713
+ "loss": 0.0243,
714
+ "step": 4800
715
+ },
716
+ {
717
+ "epoch": 0.5703869222627308,
718
+ "grad_norm": 1.986175298690796,
719
+ "learning_rate": 4.645165528419598e-06,
720
+ "loss": 0.0199,
721
+ "step": 4850
722
+ },
723
+ {
724
+ "epoch": 0.5762671998118312,
725
+ "grad_norm": 1.6648274660110474,
726
+ "learning_rate": 4.542866674356627e-06,
727
+ "loss": 0.0224,
728
+ "step": 4900
729
+ },
730
+ {
731
+ "epoch": 0.5821474773609314,
732
+ "grad_norm": 1.7521028518676758,
733
+ "learning_rate": 4.440760447427899e-06,
734
+ "loss": 0.0206,
735
+ "step": 4950
736
+ },
737
+ {
738
+ "epoch": 0.5880277549100318,
739
+ "grad_norm": 3.9067554473876953,
740
+ "learning_rate": 4.338889873220875e-06,
741
+ "loss": 0.0264,
742
+ "step": 5000
743
+ },
744
+ {
745
+ "epoch": 0.5880277549100318,
746
+ "eval_loss": 0.022188851609826088,
747
+ "eval_runtime": 615.8241,
748
+ "eval_samples_per_second": 47.338,
749
+ "eval_steps_per_second": 11.835,
750
+ "step": 5000
751
+ },
752
+ {
753
+ "epoch": 0.593908032459132,
754
+ "grad_norm": 2.8073413372039795,
755
+ "learning_rate": 4.237297878023512e-06,
756
+ "loss": 0.0238,
757
+ "step": 5050
758
+ },
759
+ {
760
+ "epoch": 0.5997883100082324,
761
+ "grad_norm": 2.105537176132202,
762
+ "learning_rate": 4.136027270735971e-06,
763
+ "loss": 0.0207,
764
+ "step": 5100
765
+ },
766
+ {
767
+ "epoch": 0.6056685875573327,
768
+ "grad_norm": 3.5026049613952637,
769
+ "learning_rate": 4.035120724831766e-06,
770
+ "loss": 0.0208,
771
+ "step": 5150
772
+ },
773
+ {
774
+ "epoch": 0.611548865106433,
775
+ "grad_norm": 3.762202262878418,
776
+ "learning_rate": 3.9346207603759966e-06,
777
+ "loss": 0.0214,
778
+ "step": 5200
779
+ },
780
+ {
781
+ "epoch": 0.6174291426555334,
782
+ "grad_norm": 2.426508903503418,
783
+ "learning_rate": 3.834569726108201e-06,
784
+ "loss": 0.0195,
785
+ "step": 5250
786
+ },
787
+ {
788
+ "epoch": 0.6233094202046336,
789
+ "grad_norm": 2.668912410736084,
790
+ "learning_rate": 3.7350097815974395e-06,
791
+ "loss": 0.0229,
792
+ "step": 5300
793
+ },
794
+ {
795
+ "epoch": 0.629189697753734,
796
+ "grad_norm": 3.8856732845306396,
797
+ "learning_rate": 3.6359828794770467e-06,
798
+ "loss": 0.0253,
799
+ "step": 5350
800
+ },
801
+ {
802
+ "epoch": 0.6350699753028343,
803
+ "grad_norm": 0.85421222448349,
804
+ "learning_rate": 3.5375307477666134e-06,
805
+ "loss": 0.0197,
806
+ "step": 5400
807
+ },
808
+ {
809
+ "epoch": 0.6409502528519346,
810
+ "grad_norm": 6.5461225509643555,
811
+ "learning_rate": 3.4396948722886065e-06,
812
+ "loss": 0.018,
813
+ "step": 5450
814
+ },
815
+ {
816
+ "epoch": 0.6468305304010349,
817
+ "grad_norm": 4.219273567199707,
818
+ "learning_rate": 3.342516479187047e-06,
819
+ "loss": 0.019,
820
+ "step": 5500
821
+ },
822
+ {
823
+ "epoch": 0.6527108079501353,
824
+ "grad_norm": 2.8508059978485107,
825
+ "learning_rate": 3.246036517555611e-06,
826
+ "loss": 0.0218,
827
+ "step": 5550
828
+ },
829
+ {
830
+ "epoch": 0.6585910854992356,
831
+ "grad_norm": 0.25915786623954773,
832
+ "learning_rate": 3.1502956421824714e-06,
833
+ "loss": 0.0172,
834
+ "step": 5600
835
+ },
836
+ {
837
+ "epoch": 0.6644713630483359,
838
+ "grad_norm": 3.543858766555786,
839
+ "learning_rate": 3.0553341964191587e-06,
840
+ "loss": 0.0213,
841
+ "step": 5650
842
+ },
843
+ {
844
+ "epoch": 0.6703516405974362,
845
+ "grad_norm": 4.268918514251709,
846
+ "learning_rate": 2.961192195180657e-06,
847
+ "loss": 0.0175,
848
+ "step": 5700
849
+ },
850
+ {
851
+ "epoch": 0.6762319181465365,
852
+ "grad_norm": 3.376847982406616,
853
+ "learning_rate": 2.867909308083885e-06,
854
+ "loss": 0.0167,
855
+ "step": 5750
856
+ },
857
+ {
858
+ "epoch": 0.6821121956956369,
859
+ "grad_norm": 3.2170565128326416,
860
+ "learning_rate": 2.7755248427316976e-06,
861
+ "loss": 0.0192,
862
+ "step": 5800
863
+ },
864
+ {
865
+ "epoch": 0.6879924732447371,
866
+ "grad_norm": 4.281452178955078,
867
+ "learning_rate": 2.68407772814942e-06,
868
+ "loss": 0.0176,
869
+ "step": 5850
870
+ },
871
+ {
872
+ "epoch": 0.6938727507938375,
873
+ "grad_norm": 2.3371615409851074,
874
+ "learning_rate": 2.5936064983808994e-06,
875
+ "loss": 0.0212,
876
+ "step": 5900
877
+ },
878
+ {
879
+ "epoch": 0.6997530283429377,
880
+ "grad_norm": 3.5211522579193115,
881
+ "learning_rate": 2.5041492762510245e-06,
882
+ "loss": 0.0206,
883
+ "step": 5950
884
+ },
885
+ {
886
+ "epoch": 0.7056333058920381,
887
+ "grad_norm": 3.2257282733917236,
888
+ "learning_rate": 2.415743757301486e-06,
889
+ "loss": 0.0183,
890
+ "step": 6000
891
+ },
892
+ {
893
+ "epoch": 0.7056333058920381,
894
+ "eval_loss": 0.019069144502282143,
895
+ "eval_runtime": 622.7936,
896
+ "eval_samples_per_second": 46.808,
897
+ "eval_steps_per_second": 11.702,
898
+ "step": 6000
899
+ },
900
+ {
901
+ "epoch": 0.7115135834411385,
902
+ "grad_norm": 1.6855459213256836,
903
+ "learning_rate": 2.3284271939066127e-06,
904
+ "loss": 0.0175,
905
+ "step": 6050
906
+ },
907
+ {
908
+ "epoch": 0.7173938609902387,
909
+ "grad_norm": 1.7634873390197754,
910
+ "learning_rate": 2.2422363795759534e-06,
911
+ "loss": 0.0212,
912
+ "step": 6100
913
+ },
914
+ {
915
+ "epoch": 0.7232741385393391,
916
+ "grad_norm": 4.85552978515625,
917
+ "learning_rate": 2.157207633450183e-06,
918
+ "loss": 0.0192,
919
+ "step": 6150
920
+ },
921
+ {
922
+ "epoch": 0.7291544160884393,
923
+ "grad_norm": 1.8883658647537231,
924
+ "learning_rate": 2.073376784996931e-06,
925
+ "loss": 0.0205,
926
+ "step": 6200
927
+ },
928
+ {
929
+ "epoch": 0.7350346936375397,
930
+ "grad_norm": 2.118055820465088,
931
+ "learning_rate": 1.990779158912943e-06,
932
+ "loss": 0.0179,
933
+ "step": 6250
934
+ },
935
+ {
936
+ "epoch": 0.74091497118664,
937
+ "grad_norm": 0.919160783290863,
938
+ "learning_rate": 1.9094495602389235e-06,
939
+ "loss": 0.0172,
940
+ "step": 6300
941
+ },
942
+ {
943
+ "epoch": 0.7467952487357403,
944
+ "grad_norm": 1.7510900497436523,
945
+ "learning_rate": 1.829422259693377e-06,
946
+ "loss": 0.0186,
947
+ "step": 6350
948
+ },
949
+ {
950
+ "epoch": 0.7526755262848407,
951
+ "grad_norm": 3.255988597869873,
952
+ "learning_rate": 1.750730979231588e-06,
953
+ "loss": 0.0209,
954
+ "step": 6400
955
+ },
956
+ {
957
+ "epoch": 0.758555803833941,
958
+ "grad_norm": 4.849563121795654,
959
+ "learning_rate": 1.6734088778358371e-06,
960
+ "loss": 0.0217,
961
+ "step": 6450
962
+ },
963
+ {
964
+ "epoch": 0.7644360813830413,
965
+ "grad_norm": 2.414062261581421,
966
+ "learning_rate": 1.5974885375428494e-06,
967
+ "loss": 0.0169,
968
+ "step": 6500
969
+ },
970
+ {
971
+ "epoch": 0.7703163589321416,
972
+ "grad_norm": 2.1367931365966797,
973
+ "learning_rate": 1.5230019497143633e-06,
974
+ "loss": 0.0207,
975
+ "step": 6550
976
+ },
977
+ {
978
+ "epoch": 0.7761966364812419,
979
+ "grad_norm": 5.911067008972168,
980
+ "learning_rate": 1.4499805015565754e-06,
981
+ "loss": 0.0158,
982
+ "step": 6600
983
+ },
984
+ {
985
+ "epoch": 0.7820769140303422,
986
+ "grad_norm": 2.751194953918457,
987
+ "learning_rate": 1.378454962894193e-06,
988
+ "loss": 0.0189,
989
+ "step": 6650
990
+ },
991
+ {
992
+ "epoch": 0.7879571915794426,
993
+ "grad_norm": 2.061358690261841,
994
+ "learning_rate": 1.308455473204619e-06,
995
+ "loss": 0.0171,
996
+ "step": 6700
997
+ },
998
+ {
999
+ "epoch": 0.7938374691285429,
1000
+ "grad_norm": 2.2449493408203125,
1001
+ "learning_rate": 1.240011528917756e-06,
1002
+ "loss": 0.019,
1003
+ "step": 6750
1004
+ },
1005
+ {
1006
+ "epoch": 0.7997177466776432,
1007
+ "grad_norm": 4.053592205047607,
1008
+ "learning_rate": 1.1731519709867933e-06,
1009
+ "loss": 0.0167,
1010
+ "step": 6800
1011
+ },
1012
+ {
1013
+ "epoch": 0.8055980242267435,
1014
+ "grad_norm": 0.6525304317474365,
1015
+ "learning_rate": 1.1079049727351726e-06,
1016
+ "loss": 0.0184,
1017
+ "step": 6850
1018
+ },
1019
+ {
1020
+ "epoch": 0.8114783017758438,
1021
+ "grad_norm": 1.0814751386642456,
1022
+ "learning_rate": 1.0442980279849086e-06,
1023
+ "loss": 0.0193,
1024
+ "step": 6900
1025
+ },
1026
+ {
1027
+ "epoch": 0.8173585793249442,
1028
+ "grad_norm": 4.694520473480225,
1029
+ "learning_rate": 9.823579394712175e-07,
1030
+ "loss": 0.0195,
1031
+ "step": 6950
1032
+ },
1033
+ {
1034
+ "epoch": 0.8232388568740444,
1035
+ "grad_norm": 3.2695419788360596,
1036
+ "learning_rate": 9.221108075483615e-07,
1037
+ "loss": 0.0171,
1038
+ "step": 7000
1039
+ },
1040
+ {
1041
+ "epoch": 0.8232388568740444,
1042
+ "eval_loss": 0.017895469442009926,
1043
+ "eval_runtime": 618.1531,
1044
+ "eval_samples_per_second": 47.16,
1045
+ "eval_steps_per_second": 11.79,
1046
+ "step": 7000
1047
+ },
1048
+ {
1049
+ "epoch": 0.8291191344231448,
1050
+ "grad_norm": 2.8549704551696777,
1051
+ "learning_rate": 8.63582019191469e-07,
1052
+ "loss": 0.0191,
1053
+ "step": 7050
1054
+ },
1055
+ {
1056
+ "epoch": 0.8349994119722451,
1057
+ "grad_norm": 1.8346871137619019,
1058
+ "learning_rate": 8.067962372989563e-07,
1059
+ "loss": 0.0168,
1060
+ "step": 7100
1061
+ },
1062
+ {
1063
+ "epoch": 0.8408796895213454,
1064
+ "grad_norm": 1.3449316024780273,
1065
+ "learning_rate": 7.517773903000519e-07,
1066
+ "loss": 0.0199,
1067
+ "step": 7150
1068
+ },
1069
+ {
1070
+ "epoch": 0.8467599670704458,
1071
+ "grad_norm": 1.1152535676956177,
1072
+ "learning_rate": 6.98548662071828e-07,
1073
+ "loss": 0.0166,
1074
+ "step": 7200
1075
+ },
1076
+ {
1077
+ "epoch": 0.852640244619546,
1078
+ "grad_norm": 2.3504021167755127,
1079
+ "learning_rate": 6.471324821699603e-07,
1080
+ "loss": 0.0156,
1081
+ "step": 7250
1082
+ },
1083
+ {
1084
+ "epoch": 0.8585205221686464,
1085
+ "grad_norm": 0.3554767072200775,
1086
+ "learning_rate": 5.975505163773437e-07,
1087
+ "loss": 0.0215,
1088
+ "step": 7300
1089
+ },
1090
+ {
1091
+ "epoch": 0.8644007997177466,
1092
+ "grad_norm": 0.13760164380073547,
1093
+ "learning_rate": 5.498236575745564e-07,
1094
+ "loss": 0.0196,
1095
+ "step": 7350
1096
+ },
1097
+ {
1098
+ "epoch": 0.870281077266847,
1099
+ "grad_norm": 1.4082413911819458,
1100
+ "learning_rate": 5.039720169360013e-07,
1101
+ "loss": 0.0159,
1102
+ "step": 7400
1103
+ },
1104
+ {
1105
+ "epoch": 0.8761613548159473,
1106
+ "grad_norm": 3.8390660285949707,
1107
+ "learning_rate": 4.600149154554501e-07,
1108
+ "loss": 0.0214,
1109
+ "step": 7450
1110
+ },
1111
+ {
1112
+ "epoch": 0.8820416323650476,
1113
+ "grad_norm": 3.1669695377349854,
1114
+ "learning_rate": 4.179708758045431e-07,
1115
+ "loss": 0.0181,
1116
+ "step": 7500
1117
+ },
1118
+ {
1119
+ "epoch": 0.887921909914148,
1120
+ "grad_norm": 6.687104225158691,
1121
+ "learning_rate": 3.7785761452770295e-07,
1122
+ "loss": 0.02,
1123
+ "step": 7550
1124
+ },
1125
+ {
1126
+ "epoch": 0.8938021874632482,
1127
+ "grad_norm": 4.694364070892334,
1128
+ "learning_rate": 3.396920345767185e-07,
1129
+ "loss": 0.0203,
1130
+ "step": 7600
1131
+ },
1132
+ {
1133
+ "epoch": 0.8996824650123486,
1134
+ "grad_norm": 2.0906476974487305,
1135
+ "learning_rate": 3.0349021818817326e-07,
1136
+ "loss": 0.0176,
1137
+ "step": 7650
1138
+ },
1139
+ {
1140
+ "epoch": 0.9055627425614489,
1141
+ "grad_norm": 3.2009048461914062,
1142
+ "learning_rate": 2.692674201066975e-07,
1143
+ "loss": 0.0189,
1144
+ "step": 7700
1145
+ },
1146
+ {
1147
+ "epoch": 0.9114430201105492,
1148
+ "grad_norm": 4.380706310272217,
1149
+ "learning_rate": 2.3703806115691951e-07,
1150
+ "loss": 0.0171,
1151
+ "step": 7750
1152
+ },
1153
+ {
1154
+ "epoch": 0.9173232976596495,
1155
+ "grad_norm": 0.9385294318199158,
1156
+ "learning_rate": 2.068157221668049e-07,
1157
+ "loss": 0.0161,
1158
+ "step": 7800
1159
+ },
1160
+ {
1161
+ "epoch": 0.9232035752087498,
1162
+ "grad_norm": 2.8777430057525635,
1163
+ "learning_rate": 1.786131382449602e-07,
1164
+ "loss": 0.0232,
1165
+ "step": 7850
1166
+ },
1167
+ {
1168
+ "epoch": 0.9290838527578502,
1169
+ "grad_norm": 0.5679008960723877,
1170
+ "learning_rate": 1.5244219341430443e-07,
1171
+ "loss": 0.0197,
1172
+ "step": 7900
1173
+ },
1174
+ {
1175
+ "epoch": 0.9349641303069505,
1176
+ "grad_norm": 2.9544992446899414,
1177
+ "learning_rate": 1.2831391560437278e-07,
1178
+ "loss": 0.0187,
1179
+ "step": 7950
1180
+ },
1181
+ {
1182
+ "epoch": 0.9408444078560508,
1183
+ "grad_norm": 2.460313081741333,
1184
+ "learning_rate": 1.0623847200435966e-07,
1185
+ "loss": 0.0185,
1186
+ "step": 8000
1187
+ },
1188
+ {
1189
+ "epoch": 0.9408444078560508,
1190
+ "eval_loss": 0.01769772544503212,
1191
+ "eval_runtime": 613.4256,
1192
+ "eval_samples_per_second": 47.523,
1193
+ "eval_steps_per_second": 11.881,
1194
+ "step": 8000
1195
+ },
1196
+ {
1197
+ "epoch": 0.9467246854051511,
1198
+ "grad_norm": 3.39328670501709,
1199
+ "learning_rate": 8.62251647788609e-08,
1200
+ "loss": 0.019,
1201
+ "step": 8050
1202
+ },
1203
+ {
1204
+ "epoch": 0.9526049629542515,
1205
+ "grad_norm": 4.006555080413818,
1206
+ "learning_rate": 6.828242714812527e-08,
1207
+ "loss": 0.0183,
1208
+ "step": 8100
1209
+ },
1210
+ {
1211
+ "epoch": 0.9584852405033517,
1212
+ "grad_norm": 2.6548595428466797,
1213
+ "learning_rate": 5.2417819834454374e-08,
1214
+ "loss": 0.0156,
1215
+ "step": 8150
1216
+ },
1217
+ {
1218
+ "epoch": 0.9643655180524521,
1219
+ "grad_norm": 1.8154182434082031,
1220
+ "learning_rate": 3.863802787626325e-08,
1221
+ "loss": 0.0175,
1222
+ "step": 8200
1223
+ },
1224
+ {
1225
+ "epoch": 0.9702457956015524,
1226
+ "grad_norm": 4.321004867553711,
1227
+ "learning_rate": 2.694885781113432e-08,
1228
+ "loss": 0.0195,
1229
+ "step": 8250
1230
+ },
1231
+ {
1232
+ "epoch": 0.9761260731506527,
1233
+ "grad_norm": 2.2515978813171387,
1234
+ "learning_rate": 1.735523522905347e-08,
1235
+ "loss": 0.0182,
1236
+ "step": 8300
1237
+ },
1238
+ {
1239
+ "epoch": 0.9820063506997531,
1240
+ "grad_norm": 2.5015530586242676,
1241
+ "learning_rate": 9.861202696864191e-09,
1242
+ "loss": 0.0197,
1243
+ "step": 8350
1244
+ },
1245
+ {
1246
+ "epoch": 0.9878866282488533,
1247
+ "grad_norm": 2.121727705001831,
1248
+ "learning_rate": 4.469918054806344e-09,
1249
+ "loss": 0.0156,
1250
+ "step": 8400
1251
+ },
1252
+ {
1253
+ "epoch": 0.9937669057979537,
1254
+ "grad_norm": 3.144368886947632,
1255
+ "learning_rate": 1.1836530858633234e-09,
1256
+ "loss": 0.0197,
1257
+ "step": 8450
1258
+ },
1259
+ {
1260
+ "epoch": 0.9996471833470539,
1261
+ "grad_norm": 2.387354850769043,
1262
+ "learning_rate": 3.792558477266894e-12,
1263
+ "loss": 0.0191,
1264
+ "step": 8500
1265
+ },
1266
+ {
1267
+ "epoch": 1.0,
1268
+ "step": 8503,
1269
+ "total_flos": 1.8987963054248755e+17,
1270
+ "train_loss": 0.0597800020030789,
1271
+ "train_runtime": 12067.3267,
1272
+ "train_samples_per_second": 5.637,
1273
+ "train_steps_per_second": 0.705
1274
+ }
1275
+ ],
1276
+ "logging_steps": 50,
1277
+ "max_steps": 8503,
1278
+ "num_input_tokens_seen": 0,
1279
+ "num_train_epochs": 1,
1280
+ "save_steps": 1000,
1281
+ "stateful_callbacks": {
1282
+ "TrainerControl": {
1283
+ "args": {
1284
+ "should_epoch_stop": false,
1285
+ "should_evaluate": false,
1286
+ "should_log": false,
1287
+ "should_save": true,
1288
+ "should_training_stop": true
1289
+ },
1290
+ "attributes": {}
1291
+ }
1292
+ },
1293
+ "total_flos": 1.8987963054248755e+17,
1294
+ "train_batch_size": 4,
1295
+ "trial_name": null,
1296
+ "trial_params": null
1297
+ }
training_eval_loss.png ADDED
training_loss.png ADDED