Hugging Face's logo

QYWH
/
DeepSeek-R1-Distill-Qwen-1.5B-sft

Text Generation
Transformers
Safetensors
qwen2
Generated from Trainer
trl
sft
conversational
text-generation-inference
Model card Files
xet
Community
QYWH commited on
Commit
f97f2e3
·
verified ·
1 Parent(s): ecd37d6

Model save

Browse files
Files changed (6) hide show
  1. README.md +1 -1
  2. all_results.json +5 -5
  3. model.safetensors +1 -1
  4. train_results.json +5 -5
  5. trainer_state.json +320 -32
  6. training_args.bin +1 -1
README.md CHANGED
@@ -27,7 +27,7 @@ print(output["generated_text"])
27
 
28
  ## Training procedure
29
 
30
- [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/2495479412-/huggingface/runs/37gehpt9)
31
 
32
 
33
  This model was trained with SFT.
 
27
 
28
  ## Training procedure
29
 
30
+ [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/2495479412-/huggingface/runs/w57ufw1m)
31
 
32
 
33
  This model was trained with SFT.
all_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "total_flos": 6751587926016.0,
3
- "train_loss": 2.2961658954620363,
4
- "train_runtime": 310.7862,
5
  "train_samples": 5000,
6
- "train_samples_per_second": 6.114,
7
- "train_steps_per_second": 0.064
8
  }
 
1
  {
2
+ "total_flos": 34146710650880.0,
3
+ "train_loss": 0.40527356846258045,
4
+ "train_runtime": 2669.3514,
5
  "train_samples": 5000,
6
+ "train_samples_per_second": 3.559,
7
+ "train_steps_per_second": 0.075
8
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3473c8536d7531db94efa7206920af1d4d45278f4ed618ff19d1c4af83aef81c
3
  size 3554214752
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c7d828df414f3d2c511c95cd455105df737fd4d4951e9d75f2288dab1326cda
3
  size 3554214752
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "total_flos": 6751587926016.0,
3
- "train_loss": 2.2961658954620363,
4
- "train_runtime": 310.7862,
5
  "train_samples": 5000,
6
- "train_samples_per_second": 6.114,
7
- "train_steps_per_second": 0.064
8
  }
 
1
  {
2
+ "total_flos": 34146710650880.0,
3
+ "train_loss": 0.40527356846258045,
4
+ "train_runtime": 2669.3514,
5
  "train_samples": 5000,
6
+ "train_samples_per_second": 3.559,
7
+ "train_steps_per_second": 0.075
8
  }
trainer_state.json CHANGED
@@ -1,59 +1,347 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 20.0,
5
  "eval_steps": 1000,
6
- "global_step": 20,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 5.0,
13
- "grad_norm": 25.052859468781215,
14
- "learning_rate": 1.866025403784439e-05,
15
- "loss": 2.8554,
16
- "mean_token_accuracy": 0.6353429824113845,
17
  "step": 5
18
  },
19
  {
20
- "epoch": 10.0,
21
- "grad_norm": 15.2646976196116,
22
- "learning_rate": 1.1736481776669307e-05,
23
- "loss": 2.3331,
24
- "mean_token_accuracy": 0.6901387840509414,
25
  "step": 10
26
  },
27
  {
28
- "epoch": 15.0,
29
- "grad_norm": 6.031402058395469,
30
- "learning_rate": 3.5721239031346067e-06,
31
- "loss": 2.0407,
32
- "mean_token_accuracy": 0.7198028057813645,
33
  "step": 15
34
  },
35
  {
36
- "epoch": 20.0,
37
- "grad_norm": 4.660242359142969,
38
- "learning_rate": 0.0,
39
- "loss": 1.9556,
40
- "mean_token_accuracy": 0.7245746940374375,
41
  "step": 20
42
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
43
  {
44
  "epoch": 20.0,
45
- "step": 20,
46
- "total_flos": 6751587926016.0,
47
- "train_loss": 2.2961658954620363,
48
- "train_runtime": 310.7862,
49
- "train_samples_per_second": 6.114,
50
- "train_steps_per_second": 0.064
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
51
  }
52
  ],
53
  "logging_steps": 5,
54
- "max_steps": 20,
55
  "num_input_tokens_seen": 0,
56
- "num_train_epochs": 20,
57
  "save_steps": 500,
58
  "stateful_callbacks": {
59
  "TrainerControl": {
@@ -67,8 +355,8 @@
67
  "attributes": {}
68
  }
69
  },
70
- "total_flos": 6751587926016.0,
71
- "train_batch_size": 4,
72
  "trial_name": null,
73
  "trial_params": null
74
  }
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 100.0,
5
  "eval_steps": 1000,
6
+ "global_step": 200,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 2.5,
13
+ "grad_norm": 13.470948224931323,
14
+ "learning_rate": 5e-06,
15
+ "loss": 2.9758,
16
+ "mean_token_accuracy": 0.6230694115161896,
17
  "step": 5
18
  },
19
  {
20
+ "epoch": 5.0,
21
+ "grad_norm": 13.750445042685815,
22
+ "learning_rate": 1e-05,
23
+ "loss": 2.659,
24
+ "mean_token_accuracy": 0.6543959975242615,
25
  "step": 10
26
  },
27
  {
28
+ "epoch": 7.5,
29
+ "grad_norm": 2.698895165960937,
30
+ "learning_rate": 1.5000000000000002e-05,
31
+ "loss": 2.1537,
32
+ "mean_token_accuracy": 0.7085180535912514,
33
  "step": 15
34
  },
35
  {
36
+ "epoch": 10.0,
37
+ "grad_norm": 2.021193946571878,
38
+ "learning_rate": 2e-05,
39
+ "loss": 1.7625,
40
+ "mean_token_accuracy": 0.7400283828377724,
41
  "step": 20
42
  },
43
+ {
44
+ "epoch": 12.5,
45
+ "grad_norm": 1.8876805703856514,
46
+ "learning_rate": 1.9961946980917457e-05,
47
+ "loss": 1.4332,
48
+ "mean_token_accuracy": 0.7722540438175202,
49
+ "step": 25
50
+ },
51
+ {
52
+ "epoch": 15.0,
53
+ "grad_norm": 1.9137993288001238,
54
+ "learning_rate": 1.9848077530122083e-05,
55
+ "loss": 1.1868,
56
+ "mean_token_accuracy": 0.7999878883361816,
57
+ "step": 30
58
+ },
59
+ {
60
+ "epoch": 17.5,
61
+ "grad_norm": 1.864430389242646,
62
+ "learning_rate": 1.9659258262890683e-05,
63
+ "loss": 0.9777,
64
+ "mean_token_accuracy": 0.8262525200843811,
65
+ "step": 35
66
+ },
67
  {
68
  "epoch": 20.0,
69
+ "grad_norm": 1.8647110420709154,
70
+ "learning_rate": 1.9396926207859085e-05,
71
+ "loss": 0.7848,
72
+ "mean_token_accuracy": 0.8553504809737206,
73
+ "step": 40
74
+ },
75
+ {
76
+ "epoch": 22.5,
77
+ "grad_norm": 1.8688016725450207,
78
+ "learning_rate": 1.9063077870366504e-05,
79
+ "loss": 0.5982,
80
+ "mean_token_accuracy": 0.8966465726494789,
81
+ "step": 45
82
+ },
83
+ {
84
+ "epoch": 25.0,
85
+ "grad_norm": 1.9284394076191518,
86
+ "learning_rate": 1.866025403784439e-05,
87
+ "loss": 0.4319,
88
+ "mean_token_accuracy": 0.9422784462571144,
89
+ "step": 50
90
+ },
91
+ {
92
+ "epoch": 27.5,
93
+ "grad_norm": 1.9014044414549942,
94
+ "learning_rate": 1.819152044288992e-05,
95
+ "loss": 0.3002,
96
+ "mean_token_accuracy": 0.9657423257827759,
97
+ "step": 55
98
+ },
99
+ {
100
+ "epoch": 30.0,
101
+ "grad_norm": 1.837847610581908,
102
+ "learning_rate": 1.766044443118978e-05,
103
+ "loss": 0.2033,
104
+ "mean_token_accuracy": 0.9815968632698059,
105
+ "step": 60
106
+ },
107
+ {
108
+ "epoch": 32.5,
109
+ "grad_norm": 1.7961955203879951,
110
+ "learning_rate": 1.7071067811865477e-05,
111
+ "loss": 0.1403,
112
+ "mean_token_accuracy": 0.9899852931499481,
113
+ "step": 65
114
+ },
115
+ {
116
+ "epoch": 35.0,
117
+ "grad_norm": 1.7767287140174859,
118
+ "learning_rate": 1.6427876096865394e-05,
119
+ "loss": 0.1046,
120
+ "mean_token_accuracy": 0.9929877564311027,
121
+ "step": 70
122
+ },
123
+ {
124
+ "epoch": 37.5,
125
+ "grad_norm": 1.4092119906646094,
126
+ "learning_rate": 1.573576436351046e-05,
127
+ "loss": 0.0739,
128
+ "mean_token_accuracy": 0.9973135724663734,
129
+ "step": 75
130
+ },
131
+ {
132
+ "epoch": 40.0,
133
+ "grad_norm": 1.1487581792063775,
134
+ "learning_rate": 1.5000000000000002e-05,
135
+ "loss": 0.0542,
136
+ "mean_token_accuracy": 0.9987073212862014,
137
+ "step": 80
138
+ },
139
+ {
140
+ "epoch": 42.5,
141
+ "grad_norm": 0.9136680911828254,
142
+ "learning_rate": 1.4226182617406996e-05,
143
+ "loss": 0.0423,
144
+ "mean_token_accuracy": 0.9989680975675583,
145
+ "step": 85
146
+ },
147
+ {
148
+ "epoch": 45.0,
149
+ "grad_norm": 0.714276120050603,
150
+ "learning_rate": 1.342020143325669e-05,
151
+ "loss": 0.0342,
152
+ "mean_token_accuracy": 0.9991405799984932,
153
+ "step": 90
154
+ },
155
+ {
156
+ "epoch": 47.5,
157
+ "grad_norm": 0.601809233243359,
158
+ "learning_rate": 1.2588190451025209e-05,
159
+ "loss": 0.0285,
160
+ "mean_token_accuracy": 0.9993251755833625,
161
+ "step": 95
162
+ },
163
+ {
164
+ "epoch": 50.0,
165
+ "grad_norm": 0.5529472873720432,
166
+ "learning_rate": 1.1736481776669307e-05,
167
+ "loss": 0.0244,
168
+ "mean_token_accuracy": 0.9994255259633065,
169
+ "step": 100
170
+ },
171
+ {
172
+ "epoch": 52.5,
173
+ "grad_norm": 0.522508977878471,
174
+ "learning_rate": 1.0871557427476585e-05,
175
+ "loss": 0.0213,
176
+ "mean_token_accuracy": 0.9994575038552285,
177
+ "step": 105
178
+ },
179
+ {
180
+ "epoch": 55.0,
181
+ "grad_norm": 0.5010419801089824,
182
+ "learning_rate": 1e-05,
183
+ "loss": 0.0189,
184
+ "mean_token_accuracy": 0.9994710594415664,
185
+ "step": 110
186
+ },
187
+ {
188
+ "epoch": 57.5,
189
+ "grad_norm": 0.4737936422226576,
190
+ "learning_rate": 9.128442572523418e-06,
191
+ "loss": 0.0169,
192
+ "mean_token_accuracy": 0.9995683461427689,
193
+ "step": 115
194
+ },
195
+ {
196
+ "epoch": 60.0,
197
+ "grad_norm": 0.45159464312393605,
198
+ "learning_rate": 8.263518223330698e-06,
199
+ "loss": 0.0155,
200
+ "mean_token_accuracy": 0.9995812460780144,
201
+ "step": 120
202
+ },
203
+ {
204
+ "epoch": 62.5,
205
+ "grad_norm": 0.4315104257213208,
206
+ "learning_rate": 7.411809548974792e-06,
207
+ "loss": 0.0141,
208
+ "mean_token_accuracy": 0.9995992064476014,
209
+ "step": 125
210
+ },
211
+ {
212
+ "epoch": 65.0,
213
+ "grad_norm": 0.41026001713760674,
214
+ "learning_rate": 6.579798566743314e-06,
215
+ "loss": 0.0131,
216
+ "mean_token_accuracy": 0.9995993703603745,
217
+ "step": 130
218
+ },
219
+ {
220
+ "epoch": 67.5,
221
+ "grad_norm": 0.39156974708559894,
222
+ "learning_rate": 5.773817382593008e-06,
223
+ "loss": 0.0122,
224
+ "mean_token_accuracy": 0.9996095344424247,
225
+ "step": 135
226
+ },
227
+ {
228
+ "epoch": 70.0,
229
+ "grad_norm": 0.3758286740687803,
230
+ "learning_rate": 5.000000000000003e-06,
231
+ "loss": 0.0116,
232
+ "mean_token_accuracy": 0.9996033638715744,
233
+ "step": 140
234
+ },
235
+ {
236
+ "epoch": 72.5,
237
+ "grad_norm": 0.36650688039325785,
238
+ "learning_rate": 4.264235636489542e-06,
239
+ "loss": 0.011,
240
+ "mean_token_accuracy": 0.9996233299374581,
241
+ "step": 145
242
+ },
243
+ {
244
+ "epoch": 75.0,
245
+ "grad_norm": 0.355312570554161,
246
+ "learning_rate": 3.5721239031346067e-06,
247
+ "loss": 0.0107,
248
+ "mean_token_accuracy": 0.9996237218379974,
249
+ "step": 150
250
+ },
251
+ {
252
+ "epoch": 77.5,
253
+ "grad_norm": 0.3470766956761054,
254
+ "learning_rate": 2.9289321881345257e-06,
255
+ "loss": 0.0102,
256
+ "mean_token_accuracy": 0.9996343672275543,
257
+ "step": 155
258
+ },
259
+ {
260
+ "epoch": 80.0,
261
+ "grad_norm": 0.3405501982761911,
262
+ "learning_rate": 2.339555568810221e-06,
263
+ "loss": 0.0101,
264
+ "mean_token_accuracy": 0.9996294215321541,
265
+ "step": 160
266
+ },
267
+ {
268
+ "epoch": 82.5,
269
+ "grad_norm": 0.3355965449228851,
270
+ "learning_rate": 1.808479557110081e-06,
271
+ "loss": 0.0098,
272
+ "mean_token_accuracy": 0.9996438756585121,
273
+ "step": 165
274
+ },
275
+ {
276
+ "epoch": 85.0,
277
+ "grad_norm": 0.3310960741356627,
278
+ "learning_rate": 1.339745962155613e-06,
279
+ "loss": 0.0096,
280
+ "mean_token_accuracy": 0.9996403336524964,
281
+ "step": 170
282
+ },
283
+ {
284
+ "epoch": 87.5,
285
+ "grad_norm": 0.3296565871751418,
286
+ "learning_rate": 9.369221296335007e-07,
287
+ "loss": 0.0096,
288
+ "mean_token_accuracy": 0.9996464192867279,
289
+ "step": 175
290
+ },
291
+ {
292
+ "epoch": 90.0,
293
+ "grad_norm": 0.32579571171147914,
294
+ "learning_rate": 6.030737921409169e-07,
295
+ "loss": 0.0095,
296
+ "mean_token_accuracy": 0.9996390745043755,
297
+ "step": 180
298
+ },
299
+ {
300
+ "epoch": 92.5,
301
+ "grad_norm": 0.3268532655575783,
302
+ "learning_rate": 3.4074173710931804e-07,
303
+ "loss": 0.0095,
304
+ "mean_token_accuracy": 0.9996397033333778,
305
+ "step": 185
306
+ },
307
+ {
308
+ "epoch": 95.0,
309
+ "grad_norm": 0.32376162339876563,
310
+ "learning_rate": 1.519224698779198e-07,
311
+ "loss": 0.0093,
312
+ "mean_token_accuracy": 0.9996532008051873,
313
+ "step": 190
314
+ },
315
+ {
316
+ "epoch": 97.5,
317
+ "grad_norm": 0.3250121331654675,
318
+ "learning_rate": 3.805301908254455e-08,
319
+ "loss": 0.0094,
320
+ "mean_token_accuracy": 0.9996466800570488,
321
+ "step": 195
322
+ },
323
+ {
324
+ "epoch": 100.0,
325
+ "grad_norm": 0.3261372439108348,
326
+ "learning_rate": 0.0,
327
+ "loss": 0.0093,
328
+ "mean_token_accuracy": 0.999645346403122,
329
+ "step": 200
330
+ },
331
+ {
332
+ "epoch": 100.0,
333
+ "step": 200,
334
+ "total_flos": 34146710650880.0,
335
+ "train_loss": 0.40527356846258045,
336
+ "train_runtime": 2669.3514,
337
+ "train_samples_per_second": 3.559,
338
+ "train_steps_per_second": 0.075
339
  }
340
  ],
341
  "logging_steps": 5,
342
+ "max_steps": 200,
343
  "num_input_tokens_seen": 0,
344
+ "num_train_epochs": 100,
345
  "save_steps": 500,
346
  "stateful_callbacks": {
347
  "TrainerControl": {
 
355
  "attributes": {}
356
  }
357
  },
358
+ "total_flos": 34146710650880.0,
359
+ "train_batch_size": 2,
360
  "trial_name": null,
361
  "trial_params": null
362
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:94f6e6670bad9721691cc99d080ded4936a55451572e0f24bf129b20fa34ad03
3
  size 7416
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:da6d972b48dec4af739ca0b32b6b53e1b5cfc25964f97c90e6e8517181f3f4c0
3
  size 7416