gsmyrnis commited on
Commit
de2a417
·
verified ·
1 Parent(s): 1581f04

End of training

Browse files
Files changed (5) hide show
  1. README.md +2 -1
  2. all_results.json +5 -9
  3. train_results.json +5 -5
  4. trainer_state.json +758 -26
  5. training_loss.png +0 -0
README.md CHANGED
@@ -4,6 +4,7 @@ license: apache-2.0
4
  base_model: Qwen/Qwen2.5-7B-Instruct
5
  tags:
6
  - llama-factory
 
7
  - generated_from_trainer
8
  model-index:
9
  - name: llama3-1_8b_4o_annotated_aime
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # llama3-1_8b_4o_annotated_aime
17
 
18
- This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on an unknown dataset.
19
 
20
  ## Model description
21
 
 
4
  base_model: Qwen/Qwen2.5-7B-Instruct
5
  tags:
6
  - llama-factory
7
+ - full
8
  - generated_from_trainer
9
  model-index:
10
  - name: llama3-1_8b_4o_annotated_aime
 
16
 
17
  # llama3-1_8b_4o_annotated_aime
18
 
19
+ This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the mlfoundations-dev/4o_annotated_aime dataset.
20
 
21
  ## Model description
22
 
all_results.json CHANGED
@@ -1,12 +1,8 @@
1
  {
2
  "epoch": 3.0,
3
- "eval_loss": 0.6393457651138306,
4
- "eval_runtime": 0.9691,
5
- "eval_samples_per_second": 75.325,
6
- "eval_steps_per_second": 1.032,
7
- "total_flos": 15075335208960.0,
8
- "train_loss": 0.7744514147440592,
9
- "train_runtime": 468.4661,
10
- "train_samples_per_second": 8.837,
11
- "train_steps_per_second": 0.019
12
  }
 
1
  {
2
  "epoch": 3.0,
3
+ "total_flos": 25167957131264.0,
4
+ "train_loss": 0.4632317288606255,
5
+ "train_runtime": 439.9384,
6
+ "train_samples_per_second": 23.206,
7
+ "train_steps_per_second": 0.245
 
 
 
 
8
  }
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "epoch": 3.0,
3
- "total_flos": 15075335208960.0,
4
- "train_loss": 0.7744514147440592,
5
- "train_runtime": 468.4661,
6
- "train_samples_per_second": 8.837,
7
- "train_steps_per_second": 0.019
8
  }
 
1
  {
2
  "epoch": 3.0,
3
+ "total_flos": 25167957131264.0,
4
+ "train_loss": 0.4632317288606255,
5
+ "train_runtime": 439.9384,
6
+ "train_samples_per_second": 23.206,
7
+ "train_steps_per_second": 0.245
8
  }
trainer_state.json CHANGED
@@ -3,47 +3,779 @@
3
  "best_model_checkpoint": null,
4
  "epoch": 3.0,
5
  "eval_steps": 500,
6
- "global_step": 9,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 1.0,
13
- "eval_loss": 0.7094656229019165,
14
- "eval_runtime": 0.9868,
15
- "eval_samples_per_second": 73.973,
16
- "eval_steps_per_second": 1.013,
 
 
 
 
 
 
 
 
 
 
 
 
 
17
  "step": 3
18
  },
19
  {
20
- "epoch": 2.0,
21
- "eval_loss": 0.6780223846435547,
22
- "eval_runtime": 0.9143,
23
- "eval_samples_per_second": 79.839,
24
- "eval_steps_per_second": 1.094,
 
 
 
 
 
 
 
 
 
 
 
 
 
25
  "step": 6
26
  },
27
  {
28
- "epoch": 3.0,
29
- "eval_loss": 0.6393457651138306,
30
- "eval_runtime": 0.9548,
31
- "eval_samples_per_second": 76.453,
32
- "eval_steps_per_second": 1.047,
 
 
 
 
 
 
 
 
 
 
 
 
 
33
  "step": 9
34
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
35
  {
36
  "epoch": 3.0,
37
- "step": 9,
38
- "total_flos": 15075335208960.0,
39
- "train_loss": 0.7744514147440592,
40
- "train_runtime": 468.4661,
41
- "train_samples_per_second": 8.837,
42
- "train_steps_per_second": 0.019
43
  }
44
  ],
45
- "logging_steps": 10,
46
- "max_steps": 9,
47
  "num_input_tokens_seen": 0,
48
  "num_train_epochs": 3,
49
  "save_steps": 500,
@@ -59,8 +791,8 @@
59
  "attributes": {}
60
  }
61
  },
62
- "total_flos": 15075335208960.0,
63
- "train_batch_size": 16,
64
  "trial_name": null,
65
  "trial_params": null
66
  }
 
3
  "best_model_checkpoint": null,
4
  "epoch": 3.0,
5
  "eval_steps": 500,
6
+ "global_step": 108,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 0.027777777777777776,
13
+ "grad_norm": 5.499531247448967,
14
+ "learning_rate": 9.090909090909091e-07,
15
+ "loss": 0.6855,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.05555555555555555,
20
+ "grad_norm": 5.557838812807806,
21
+ "learning_rate": 1.8181818181818183e-06,
22
+ "loss": 0.6554,
23
+ "step": 2
24
+ },
25
+ {
26
+ "epoch": 0.08333333333333333,
27
+ "grad_norm": 5.5626937869370305,
28
+ "learning_rate": 2.7272727272727272e-06,
29
+ "loss": 0.631,
30
  "step": 3
31
  },
32
  {
33
+ "epoch": 0.1111111111111111,
34
+ "grad_norm": 5.202877857676261,
35
+ "learning_rate": 3.6363636363636366e-06,
36
+ "loss": 0.6684,
37
+ "step": 4
38
+ },
39
+ {
40
+ "epoch": 0.1388888888888889,
41
+ "grad_norm": 3.392348666163089,
42
+ "learning_rate": 4.5454545454545455e-06,
43
+ "loss": 0.551,
44
+ "step": 5
45
+ },
46
+ {
47
+ "epoch": 0.16666666666666666,
48
+ "grad_norm": 2.213517617001493,
49
+ "learning_rate": 5.4545454545454545e-06,
50
+ "loss": 0.5989,
51
  "step": 6
52
  },
53
  {
54
+ "epoch": 0.19444444444444445,
55
+ "grad_norm": 2.001366109944542,
56
+ "learning_rate": 6.363636363636364e-06,
57
+ "loss": 0.5333,
58
+ "step": 7
59
+ },
60
+ {
61
+ "epoch": 0.2222222222222222,
62
+ "grad_norm": 1.9419958988527681,
63
+ "learning_rate": 7.272727272727273e-06,
64
+ "loss": 0.5054,
65
+ "step": 8
66
+ },
67
+ {
68
+ "epoch": 0.25,
69
+ "grad_norm": 1.5833591254215535,
70
+ "learning_rate": 8.181818181818183e-06,
71
+ "loss": 0.5322,
72
  "step": 9
73
  },
74
+ {
75
+ "epoch": 0.2777777777777778,
76
+ "grad_norm": 1.436746398759113,
77
+ "learning_rate": 9.090909090909091e-06,
78
+ "loss": 0.5433,
79
+ "step": 10
80
+ },
81
+ {
82
+ "epoch": 0.3055555555555556,
83
+ "grad_norm": 1.7345453054514894,
84
+ "learning_rate": 1e-05,
85
+ "loss": 0.5709,
86
+ "step": 11
87
+ },
88
+ {
89
+ "epoch": 0.3333333333333333,
90
+ "grad_norm": 1.597109242203359,
91
+ "learning_rate": 9.997377845227577e-06,
92
+ "loss": 0.4906,
93
+ "step": 12
94
+ },
95
+ {
96
+ "epoch": 0.3611111111111111,
97
+ "grad_norm": 1.2794918602228373,
98
+ "learning_rate": 9.98951413118856e-06,
99
+ "loss": 0.6254,
100
+ "step": 13
101
+ },
102
+ {
103
+ "epoch": 0.3888888888888889,
104
+ "grad_norm": 1.2457098159062354,
105
+ "learning_rate": 9.97641710583307e-06,
106
+ "loss": 0.5566,
107
+ "step": 14
108
+ },
109
+ {
110
+ "epoch": 0.4166666666666667,
111
+ "grad_norm": 1.0051622325160827,
112
+ "learning_rate": 9.958100506132127e-06,
113
+ "loss": 0.5086,
114
+ "step": 15
115
+ },
116
+ {
117
+ "epoch": 0.4444444444444444,
118
+ "grad_norm": 1.0232838508074662,
119
+ "learning_rate": 9.934583543669454e-06,
120
+ "loss": 0.5798,
121
+ "step": 16
122
+ },
123
+ {
124
+ "epoch": 0.4722222222222222,
125
+ "grad_norm": 1.0687266662109796,
126
+ "learning_rate": 9.905890884491196e-06,
127
+ "loss": 0.547,
128
+ "step": 17
129
+ },
130
+ {
131
+ "epoch": 0.5,
132
+ "grad_norm": 0.8078554986405523,
133
+ "learning_rate": 9.872052623234632e-06,
134
+ "loss": 0.4698,
135
+ "step": 18
136
+ },
137
+ {
138
+ "epoch": 0.5277777777777778,
139
+ "grad_norm": 0.9700266562907995,
140
+ "learning_rate": 9.833104251563058e-06,
141
+ "loss": 0.6414,
142
+ "step": 19
143
+ },
144
+ {
145
+ "epoch": 0.5555555555555556,
146
+ "grad_norm": 0.7994292138406759,
147
+ "learning_rate": 9.789086620939936e-06,
148
+ "loss": 0.4365,
149
+ "step": 20
150
+ },
151
+ {
152
+ "epoch": 0.5833333333333334,
153
+ "grad_norm": 0.8745740021635219,
154
+ "learning_rate": 9.740045899781353e-06,
155
+ "loss": 0.4913,
156
+ "step": 21
157
+ },
158
+ {
159
+ "epoch": 0.6111111111111112,
160
+ "grad_norm": 0.8854836350036159,
161
+ "learning_rate": 9.68603352503172e-06,
162
+ "loss": 0.5518,
163
+ "step": 22
164
+ },
165
+ {
166
+ "epoch": 0.6388888888888888,
167
+ "grad_norm": 0.847456579465944,
168
+ "learning_rate": 9.627106148213521e-06,
169
+ "loss": 0.5354,
170
+ "step": 23
171
+ },
172
+ {
173
+ "epoch": 0.6666666666666666,
174
+ "grad_norm": 0.821018000951143,
175
+ "learning_rate": 9.563325576007702e-06,
176
+ "loss": 0.5213,
177
+ "step": 24
178
+ },
179
+ {
180
+ "epoch": 0.6944444444444444,
181
+ "grad_norm": 0.9295894881714469,
182
+ "learning_rate": 9.494758705426978e-06,
183
+ "loss": 0.598,
184
+ "step": 25
185
+ },
186
+ {
187
+ "epoch": 0.7222222222222222,
188
+ "grad_norm": 0.7448130004829931,
189
+ "learning_rate": 9.421477453650118e-06,
190
+ "loss": 0.4658,
191
+ "step": 26
192
+ },
193
+ {
194
+ "epoch": 0.75,
195
+ "grad_norm": 0.9080917046249632,
196
+ "learning_rate": 9.343558682590757e-06,
197
+ "loss": 0.6228,
198
+ "step": 27
199
+ },
200
+ {
201
+ "epoch": 0.7777777777777778,
202
+ "grad_norm": 0.8163873567576216,
203
+ "learning_rate": 9.261084118279846e-06,
204
+ "loss": 0.5333,
205
+ "step": 28
206
+ },
207
+ {
208
+ "epoch": 0.8055555555555556,
209
+ "grad_norm": 0.7764439372547906,
210
+ "learning_rate": 9.174140265146356e-06,
211
+ "loss": 0.4741,
212
+ "step": 29
213
+ },
214
+ {
215
+ "epoch": 0.8333333333333334,
216
+ "grad_norm": 1.1961962564927606,
217
+ "learning_rate": 9.082818315286054e-06,
218
+ "loss": 0.5714,
219
+ "step": 30
220
+ },
221
+ {
222
+ "epoch": 0.8611111111111112,
223
+ "grad_norm": 0.7267955393856498,
224
+ "learning_rate": 8.987214052813605e-06,
225
+ "loss": 0.5379,
226
+ "step": 31
227
+ },
228
+ {
229
+ "epoch": 0.8888888888888888,
230
+ "grad_norm": 0.769564835206009,
231
+ "learning_rate": 8.887427753398249e-06,
232
+ "loss": 0.5221,
233
+ "step": 32
234
+ },
235
+ {
236
+ "epoch": 0.9166666666666666,
237
+ "grad_norm": 0.7557825851349546,
238
+ "learning_rate": 8.783564079088478e-06,
239
+ "loss": 0.5237,
240
+ "step": 33
241
+ },
242
+ {
243
+ "epoch": 0.9444444444444444,
244
+ "grad_norm": 0.8013832351115299,
245
+ "learning_rate": 8.675731968536004e-06,
246
+ "loss": 0.5525,
247
+ "step": 34
248
+ },
249
+ {
250
+ "epoch": 0.9722222222222222,
251
+ "grad_norm": 0.6888521920999732,
252
+ "learning_rate": 8.564044522734147e-06,
253
+ "loss": 0.4881,
254
+ "step": 35
255
+ },
256
+ {
257
+ "epoch": 1.0,
258
+ "grad_norm": 0.6587728956884636,
259
+ "learning_rate": 8.448618886390523e-06,
260
+ "loss": 0.4985,
261
+ "step": 36
262
+ },
263
+ {
264
+ "epoch": 1.0277777777777777,
265
+ "grad_norm": 0.795241641899969,
266
+ "learning_rate": 8.329576125058406e-06,
267
+ "loss": 0.4365,
268
+ "step": 37
269
+ },
270
+ {
271
+ "epoch": 1.0555555555555556,
272
+ "grad_norm": 2.210851457630018,
273
+ "learning_rate": 8.207041098155701e-06,
274
+ "loss": 0.5556,
275
+ "step": 38
276
+ },
277
+ {
278
+ "epoch": 1.0833333333333333,
279
+ "grad_norm": 0.7613786102792487,
280
+ "learning_rate": 8.081142328004638e-06,
281
+ "loss": 0.4846,
282
+ "step": 39
283
+ },
284
+ {
285
+ "epoch": 1.1111111111111112,
286
+ "grad_norm": 0.723429529329392,
287
+ "learning_rate": 7.952011865029614e-06,
288
+ "loss": 0.4603,
289
+ "step": 40
290
+ },
291
+ {
292
+ "epoch": 1.1388888888888888,
293
+ "grad_norm": 0.7836107269741927,
294
+ "learning_rate": 7.819785149254534e-06,
295
+ "loss": 0.4483,
296
+ "step": 41
297
+ },
298
+ {
299
+ "epoch": 1.1666666666666667,
300
+ "grad_norm": 0.8412482780955819,
301
+ "learning_rate": 7.68460086824492e-06,
302
+ "loss": 0.5091,
303
+ "step": 42
304
+ },
305
+ {
306
+ "epoch": 1.1944444444444444,
307
+ "grad_norm": 0.7560956617656049,
308
+ "learning_rate": 7.546600811643816e-06,
309
+ "loss": 0.4768,
310
+ "step": 43
311
+ },
312
+ {
313
+ "epoch": 1.2222222222222223,
314
+ "grad_norm": 0.7532433430539396,
315
+ "learning_rate": 7.405929722454026e-06,
316
+ "loss": 0.4879,
317
+ "step": 44
318
+ },
319
+ {
320
+ "epoch": 1.25,
321
+ "grad_norm": 0.7250342590564604,
322
+ "learning_rate": 7.262735145222696e-06,
323
+ "loss": 0.4424,
324
+ "step": 45
325
+ },
326
+ {
327
+ "epoch": 1.2777777777777777,
328
+ "grad_norm": 0.7424652764504065,
329
+ "learning_rate": 7.117167271287453e-06,
330
+ "loss": 0.4268,
331
+ "step": 46
332
+ },
333
+ {
334
+ "epoch": 1.3055555555555556,
335
+ "grad_norm": 0.5613548384089758,
336
+ "learning_rate": 6.969378781246436e-06,
337
+ "loss": 0.3586,
338
+ "step": 47
339
+ },
340
+ {
341
+ "epoch": 1.3333333333333333,
342
+ "grad_norm": 0.7217076640093948,
343
+ "learning_rate": 6.819524684817439e-06,
344
+ "loss": 0.4861,
345
+ "step": 48
346
+ },
347
+ {
348
+ "epoch": 1.3611111111111112,
349
+ "grad_norm": 0.6561868334638776,
350
+ "learning_rate": 6.667762158254104e-06,
351
+ "loss": 0.4543,
352
+ "step": 49
353
+ },
354
+ {
355
+ "epoch": 1.3888888888888888,
356
+ "grad_norm": 0.7110755871975613,
357
+ "learning_rate": 6.514250379489754e-06,
358
+ "loss": 0.4277,
359
+ "step": 50
360
+ },
361
+ {
362
+ "epoch": 1.4166666666666667,
363
+ "grad_norm": 0.7340306300343789,
364
+ "learning_rate": 6.3591503611817155e-06,
365
+ "loss": 0.4245,
366
+ "step": 51
367
+ },
368
+ {
369
+ "epoch": 1.4444444444444444,
370
+ "grad_norm": 0.7373696398545522,
371
+ "learning_rate": 6.202624781831269e-06,
372
+ "loss": 0.4865,
373
+ "step": 52
374
+ },
375
+ {
376
+ "epoch": 1.4722222222222223,
377
+ "grad_norm": 0.7694890579716093,
378
+ "learning_rate": 6.044837815156377e-06,
379
+ "loss": 0.467,
380
+ "step": 53
381
+ },
382
+ {
383
+ "epoch": 1.5,
384
+ "grad_norm": 0.7185745702090169,
385
+ "learning_rate": 5.885954957896115e-06,
386
+ "loss": 0.4313,
387
+ "step": 54
388
+ },
389
+ {
390
+ "epoch": 1.5277777777777777,
391
+ "grad_norm": 0.6305358582918202,
392
+ "learning_rate": 5.726142856227453e-06,
393
+ "loss": 0.4161,
394
+ "step": 55
395
+ },
396
+ {
397
+ "epoch": 1.5555555555555556,
398
+ "grad_norm": 0.7209314959601887,
399
+ "learning_rate": 5.5655691309764225e-06,
400
+ "loss": 0.4562,
401
+ "step": 56
402
+ },
403
+ {
404
+ "epoch": 1.5833333333333335,
405
+ "grad_norm": 0.64260058204023,
406
+ "learning_rate": 5.404402201807022e-06,
407
+ "loss": 0.4636,
408
+ "step": 57
409
+ },
410
+ {
411
+ "epoch": 1.6111111111111112,
412
+ "grad_norm": 0.612522134619282,
413
+ "learning_rate": 5.242811110572243e-06,
414
+ "loss": 0.3946,
415
+ "step": 58
416
+ },
417
+ {
418
+ "epoch": 1.6388888888888888,
419
+ "grad_norm": 0.640209879297769,
420
+ "learning_rate": 5.080965344012509e-06,
421
+ "loss": 0.4234,
422
+ "step": 59
423
+ },
424
+ {
425
+ "epoch": 1.6666666666666665,
426
+ "grad_norm": 0.684267323330591,
427
+ "learning_rate": 4.919034655987493e-06,
428
+ "loss": 0.4956,
429
+ "step": 60
430
+ },
431
+ {
432
+ "epoch": 1.6944444444444444,
433
+ "grad_norm": 0.6560173146533167,
434
+ "learning_rate": 4.757188889427761e-06,
435
+ "loss": 0.4513,
436
+ "step": 61
437
+ },
438
+ {
439
+ "epoch": 1.7222222222222223,
440
+ "grad_norm": 0.6024082771930704,
441
+ "learning_rate": 4.59559779819298e-06,
442
+ "loss": 0.3554,
443
+ "step": 62
444
+ },
445
+ {
446
+ "epoch": 1.75,
447
+ "grad_norm": 0.5988731751099766,
448
+ "learning_rate": 4.434430869023579e-06,
449
+ "loss": 0.423,
450
+ "step": 63
451
+ },
452
+ {
453
+ "epoch": 1.7777777777777777,
454
+ "grad_norm": 0.7714096061714903,
455
+ "learning_rate": 4.27385714377255e-06,
456
+ "loss": 0.3804,
457
+ "step": 64
458
+ },
459
+ {
460
+ "epoch": 1.8055555555555556,
461
+ "grad_norm": 0.6870558807208743,
462
+ "learning_rate": 4.1140450421038865e-06,
463
+ "loss": 0.46,
464
+ "step": 65
465
+ },
466
+ {
467
+ "epoch": 1.8333333333333335,
468
+ "grad_norm": 0.6342926359159472,
469
+ "learning_rate": 3.955162184843625e-06,
470
+ "loss": 0.5136,
471
+ "step": 66
472
+ },
473
+ {
474
+ "epoch": 1.8611111111111112,
475
+ "grad_norm": 0.6155115323488202,
476
+ "learning_rate": 3.7973752181687336e-06,
477
+ "loss": 0.4108,
478
+ "step": 67
479
+ },
480
+ {
481
+ "epoch": 1.8888888888888888,
482
+ "grad_norm": 0.7006350597692972,
483
+ "learning_rate": 3.6408496388182857e-06,
484
+ "loss": 0.5141,
485
+ "step": 68
486
+ },
487
+ {
488
+ "epoch": 1.9166666666666665,
489
+ "grad_norm": 0.6853399554911791,
490
+ "learning_rate": 3.4857496205102475e-06,
491
+ "loss": 0.4591,
492
+ "step": 69
493
+ },
494
+ {
495
+ "epoch": 1.9444444444444444,
496
+ "grad_norm": 0.6807578527187025,
497
+ "learning_rate": 3.3322378417458985e-06,
498
+ "loss": 0.4447,
499
+ "step": 70
500
+ },
501
+ {
502
+ "epoch": 1.9722222222222223,
503
+ "grad_norm": 0.6002571912158542,
504
+ "learning_rate": 3.180475315182563e-06,
505
+ "loss": 0.3775,
506
+ "step": 71
507
+ },
508
+ {
509
+ "epoch": 2.0,
510
+ "grad_norm": 0.6176679391981694,
511
+ "learning_rate": 3.0306212187535653e-06,
512
+ "loss": 0.362,
513
+ "step": 72
514
+ },
515
+ {
516
+ "epoch": 2.0277777777777777,
517
+ "grad_norm": 0.6348434037827244,
518
+ "learning_rate": 2.882832728712551e-06,
519
+ "loss": 0.4395,
520
+ "step": 73
521
+ },
522
+ {
523
+ "epoch": 2.0555555555555554,
524
+ "grad_norm": 0.6194546514321693,
525
+ "learning_rate": 2.7372648547773063e-06,
526
+ "loss": 0.385,
527
+ "step": 74
528
+ },
529
+ {
530
+ "epoch": 2.0833333333333335,
531
+ "grad_norm": 0.5947575965373516,
532
+ "learning_rate": 2.594070277545975e-06,
533
+ "loss": 0.3634,
534
+ "step": 75
535
+ },
536
+ {
537
+ "epoch": 2.111111111111111,
538
+ "grad_norm": 0.6412237789531587,
539
+ "learning_rate": 2.4533991883561868e-06,
540
+ "loss": 0.3352,
541
+ "step": 76
542
+ },
543
+ {
544
+ "epoch": 2.138888888888889,
545
+ "grad_norm": 0.5993096167073967,
546
+ "learning_rate": 2.315399131755081e-06,
547
+ "loss": 0.3654,
548
+ "step": 77
549
+ },
550
+ {
551
+ "epoch": 2.1666666666666665,
552
+ "grad_norm": 0.5863285405996705,
553
+ "learning_rate": 2.1802148507454675e-06,
554
+ "loss": 0.3648,
555
+ "step": 78
556
+ },
557
+ {
558
+ "epoch": 2.1944444444444446,
559
+ "grad_norm": 0.6176169061591418,
560
+ "learning_rate": 2.0479881349703885e-06,
561
+ "loss": 0.3567,
562
+ "step": 79
563
+ },
564
+ {
565
+ "epoch": 2.2222222222222223,
566
+ "grad_norm": 2.036643091775969,
567
+ "learning_rate": 1.9188576719953635e-06,
568
+ "loss": 0.5043,
569
+ "step": 80
570
+ },
571
+ {
572
+ "epoch": 2.25,
573
+ "grad_norm": 0.5707024837734276,
574
+ "learning_rate": 1.7929589018443016e-06,
575
+ "loss": 0.3835,
576
+ "step": 81
577
+ },
578
+ {
579
+ "epoch": 2.2777777777777777,
580
+ "grad_norm": 0.6703596535434576,
581
+ "learning_rate": 1.6704238749415958e-06,
582
+ "loss": 0.4195,
583
+ "step": 82
584
+ },
585
+ {
586
+ "epoch": 2.3055555555555554,
587
+ "grad_norm": 0.6161351940524724,
588
+ "learning_rate": 1.5513811136094786e-06,
589
+ "loss": 0.3808,
590
+ "step": 83
591
+ },
592
+ {
593
+ "epoch": 2.3333333333333335,
594
+ "grad_norm": 0.567006481121317,
595
+ "learning_rate": 1.4359554772658551e-06,
596
+ "loss": 0.3484,
597
+ "step": 84
598
+ },
599
+ {
600
+ "epoch": 2.361111111111111,
601
+ "grad_norm": 0.5645679908744803,
602
+ "learning_rate": 1.3242680314639995e-06,
603
+ "loss": 0.3442,
604
+ "step": 85
605
+ },
606
+ {
607
+ "epoch": 2.388888888888889,
608
+ "grad_norm": 0.6134174690452912,
609
+ "learning_rate": 1.2164359209115235e-06,
610
+ "loss": 0.3701,
611
+ "step": 86
612
+ },
613
+ {
614
+ "epoch": 2.4166666666666665,
615
+ "grad_norm": 0.6351405467839962,
616
+ "learning_rate": 1.1125722466017547e-06,
617
+ "loss": 0.4531,
618
+ "step": 87
619
+ },
620
+ {
621
+ "epoch": 2.4444444444444446,
622
+ "grad_norm": 0.6153449983061651,
623
+ "learning_rate": 1.012785947186397e-06,
624
+ "loss": 0.3512,
625
+ "step": 88
626
+ },
627
+ {
628
+ "epoch": 2.4722222222222223,
629
+ "grad_norm": 0.5815107547799697,
630
+ "learning_rate": 9.171816847139447e-07,
631
+ "loss": 0.3224,
632
+ "step": 89
633
+ },
634
+ {
635
+ "epoch": 2.5,
636
+ "grad_norm": 0.5972403522847891,
637
+ "learning_rate": 8.258597348536452e-07,
638
+ "loss": 0.3708,
639
+ "step": 90
640
+ },
641
+ {
642
+ "epoch": 2.5277777777777777,
643
+ "grad_norm": 0.6224293908552574,
644
+ "learning_rate": 7.389158817201541e-07,
645
+ "loss": 0.4313,
646
+ "step": 91
647
+ },
648
+ {
649
+ "epoch": 2.5555555555555554,
650
+ "grad_norm": 0.6024938044516791,
651
+ "learning_rate": 6.564413174092443e-07,
652
+ "loss": 0.3964,
653
+ "step": 92
654
+ },
655
+ {
656
+ "epoch": 2.5833333333333335,
657
+ "grad_norm": 0.6108443482475023,
658
+ "learning_rate": 5.785225463498828e-07,
659
+ "loss": 0.3827,
660
+ "step": 93
661
+ },
662
+ {
663
+ "epoch": 2.611111111111111,
664
+ "grad_norm": 0.6722881160180866,
665
+ "learning_rate": 5.05241294573024e-07,
666
+ "loss": 0.4383,
667
+ "step": 94
668
+ },
669
+ {
670
+ "epoch": 2.638888888888889,
671
+ "grad_norm": 0.6149268282937064,
672
+ "learning_rate": 4.3667442399229985e-07,
673
+ "loss": 0.4371,
674
+ "step": 95
675
+ },
676
+ {
677
+ "epoch": 2.6666666666666665,
678
+ "grad_norm": 0.6113966401547459,
679
+ "learning_rate": 3.728938517864794e-07,
680
+ "loss": 0.3829,
681
+ "step": 96
682
+ },
683
+ {
684
+ "epoch": 2.6944444444444446,
685
+ "grad_norm": 0.6671411413349442,
686
+ "learning_rate": 3.1396647496828245e-07,
687
+ "loss": 0.3965,
688
+ "step": 97
689
+ },
690
+ {
691
+ "epoch": 2.7222222222222223,
692
+ "grad_norm": 0.547101720665063,
693
+ "learning_rate": 2.599541002186479e-07,
694
+ "loss": 0.3122,
695
+ "step": 98
696
+ },
697
+ {
698
+ "epoch": 2.75,
699
+ "grad_norm": 0.6044035422441768,
700
+ "learning_rate": 2.109133790600648e-07,
701
+ "loss": 0.4733,
702
+ "step": 99
703
+ },
704
+ {
705
+ "epoch": 2.7777777777777777,
706
+ "grad_norm": 0.6069536296371835,
707
+ "learning_rate": 1.6689574843694433e-07,
708
+ "loss": 0.4024,
709
+ "step": 100
710
+ },
711
+ {
712
+ "epoch": 2.8055555555555554,
713
+ "grad_norm": 0.5982249811836392,
714
+ "learning_rate": 1.2794737676536993e-07,
715
+ "loss": 0.4336,
716
+ "step": 101
717
+ },
718
+ {
719
+ "epoch": 2.8333333333333335,
720
+ "grad_norm": 0.6884925904147986,
721
+ "learning_rate": 9.410911550880474e-08,
722
+ "loss": 0.3595,
723
+ "step": 102
724
+ },
725
+ {
726
+ "epoch": 2.861111111111111,
727
+ "grad_norm": 0.7589234561889401,
728
+ "learning_rate": 6.54164563305465e-08,
729
+ "loss": 0.4163,
730
+ "step": 103
731
+ },
732
+ {
733
+ "epoch": 2.888888888888889,
734
+ "grad_norm": 0.6469041727623386,
735
+ "learning_rate": 4.189949386787462e-08,
736
+ "loss": 0.4264,
737
+ "step": 104
738
+ },
739
+ {
740
+ "epoch": 2.9166666666666665,
741
+ "grad_norm": 0.6755069676275332,
742
+ "learning_rate": 2.358289416693027e-08,
743
+ "loss": 0.4073,
744
+ "step": 105
745
+ },
746
+ {
747
+ "epoch": 2.9444444444444446,
748
+ "grad_norm": 0.6272555116057572,
749
+ "learning_rate": 1.0485868811441757e-08,
750
+ "loss": 0.4084,
751
+ "step": 106
752
+ },
753
+ {
754
+ "epoch": 2.9722222222222223,
755
+ "grad_norm": 0.6263081374423879,
756
+ "learning_rate": 2.6221547724253337e-09,
757
+ "loss": 0.3967,
758
+ "step": 107
759
+ },
760
+ {
761
+ "epoch": 3.0,
762
+ "grad_norm": 0.5852999921009103,
763
+ "learning_rate": 0.0,
764
+ "loss": 0.3847,
765
+ "step": 108
766
+ },
767
  {
768
  "epoch": 3.0,
769
+ "step": 108,
770
+ "total_flos": 25167957131264.0,
771
+ "train_loss": 0.4632317288606255,
772
+ "train_runtime": 439.9384,
773
+ "train_samples_per_second": 23.206,
774
+ "train_steps_per_second": 0.245
775
  }
776
  ],
777
+ "logging_steps": 1,
778
+ "max_steps": 108,
779
  "num_input_tokens_seen": 0,
780
  "num_train_epochs": 3,
781
  "save_steps": 500,
 
791
  "attributes": {}
792
  }
793
  },
794
+ "total_flos": 25167957131264.0,
795
+ "train_batch_size": 3,
796
  "trial_name": null,
797
  "trial_params": null
798
  }
training_loss.png ADDED