Blancy commited on
Commit
1977f3d
·
verified ·
1 Parent(s): 9308579

Model save

Browse files
Files changed (4) hide show
  1. README.md +1 -3
  2. all_results.json +5 -5
  3. train_results.json +5 -5
  4. trainer_state.json +860 -69
README.md CHANGED
@@ -1,10 +1,8 @@
1
  ---
2
- datasets: Blancy/verifiable-coding-problems-SFT
3
  library_name: transformers
4
  model_name: Qwen3-0.6B-Open-R1-Distill
5
  tags:
6
  - generated_from_trainer
7
- - open-r1
8
  - trl
9
  - sft
10
  licence: license
@@ -12,7 +10,7 @@ licence: license
12
 
13
  # Model Card for Qwen3-0.6B-Open-R1-Distill
14
 
15
- This model is a fine-tuned version of [None](https://huggingface.co/None) on the [Blancy/verifiable-coding-problems-SFT](https://huggingface.co/datasets/Blancy/verifiable-coding-problems-SFT) dataset.
16
  It has been trained using [TRL](https://github.com/huggingface/trl).
17
 
18
  ## Quick start
 
1
  ---
 
2
  library_name: transformers
3
  model_name: Qwen3-0.6B-Open-R1-Distill
4
  tags:
5
  - generated_from_trainer
 
6
  - trl
7
  - sft
8
  licence: license
 
10
 
11
  # Model Card for Qwen3-0.6B-Open-R1-Distill
12
 
13
+ This model is a fine-tuned version of [None](https://huggingface.co/None).
14
  It has been trained using [TRL](https://github.com/huggingface/trl).
15
 
16
  ## Quick start
all_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "total_flos": 1.0840722057776333e+17,
3
- "train_loss": 0.09704786779904606,
4
- "train_runtime": 144.1296,
5
  "train_samples": 1086,
6
- "train_samples_per_second": 86.859,
7
- "train_steps_per_second": 0.687
8
  }
 
1
  {
2
+ "total_flos": 7.227148038517555e+17,
3
+ "train_loss": 0.08479578431808588,
4
+ "train_runtime": 842.3759,
5
  "train_samples": 1086,
6
+ "train_samples_per_second": 99.077,
7
+ "train_steps_per_second": 0.783
8
  }
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "total_flos": 1.0840722057776333e+17,
3
- "train_loss": 0.09704786779904606,
4
- "train_runtime": 144.1296,
5
  "train_samples": 1086,
6
- "train_samples_per_second": 86.859,
7
- "train_steps_per_second": 0.687
8
  }
 
1
  {
2
+ "total_flos": 7.227148038517555e+17,
3
+ "train_loss": 0.08479578431808588,
4
+ "train_runtime": 842.3759,
5
  "train_samples": 1086,
6
+ "train_samples_per_second": 99.077,
7
+ "train_steps_per_second": 0.783
8
  }
trainer_state.json CHANGED
@@ -2,160 +2,951 @@
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
- "epoch": 3.0,
6
  "eval_steps": 500,
7
- "global_step": 99,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
11
  "log_history": [
12
  {
13
  "epoch": 0.15151515151515152,
14
- "grad_norm": 0.77734375,
15
- "learning_rate": 2.4e-05,
16
- "loss": 0.1367,
17
  "step": 5
18
  },
19
  {
20
  "epoch": 0.30303030303030304,
21
- "grad_norm": 0.44140625,
22
- "learning_rate": 2.9879546090089533e-05,
23
- "loss": 0.1117,
24
  "step": 10
25
  },
26
  {
27
  "epoch": 0.45454545454545453,
28
- "grad_norm": 0.28515625,
29
- "learning_rate": 2.9393883712293316e-05,
30
- "loss": 0.1015,
31
  "step": 15
32
  },
33
  {
34
  "epoch": 0.6060606060606061,
35
- "grad_norm": 0.244140625,
36
- "learning_rate": 2.8549004284108398e-05,
37
- "loss": 0.099,
38
  "step": 20
39
  },
40
  {
41
  "epoch": 0.7575757575757576,
42
- "grad_norm": 0.228515625,
43
- "learning_rate": 2.7368445717222102e-05,
44
- "loss": 0.0965,
45
  "step": 25
46
  },
47
  {
48
  "epoch": 0.9090909090909091,
49
- "grad_norm": 0.220703125,
50
- "learning_rate": 2.5885097773607675e-05,
51
- "loss": 0.0973,
52
  "step": 30
53
  },
54
  {
55
  "epoch": 1.0606060606060606,
56
- "grad_norm": 0.2255859375,
57
- "learning_rate": 2.4140285773463036e-05,
58
- "loss": 0.0936,
59
  "step": 35
60
  },
61
  {
62
  "epoch": 1.2121212121212122,
63
- "grad_norm": 0.2138671875,
64
- "learning_rate": 2.2182619292782524e-05,
65
- "loss": 0.0961,
66
  "step": 40
67
  },
68
  {
69
  "epoch": 1.3636363636363638,
70
- "grad_norm": 0.2041015625,
71
- "learning_rate": 2.0066637925262362e-05,
72
- "loss": 0.0931,
73
  "step": 45
74
  },
75
  {
76
  "epoch": 1.5151515151515151,
77
- "grad_norm": 0.2177734375,
78
- "learning_rate": 1.7851291836925332e-05,
79
- "loss": 0.0936,
80
  "step": 50
81
  },
82
  {
83
  "epoch": 1.6666666666666665,
84
- "grad_norm": 0.2119140625,
85
- "learning_rate": 1.559829944444086e-05,
86
- "loss": 0.0932,
87
  "step": 55
88
  },
89
  {
90
  "epoch": 1.8181818181818183,
91
- "grad_norm": 0.2099609375,
92
- "learning_rate": 1.3370427971388369e-05,
93
- "loss": 0.0933,
94
  "step": 60
95
  },
96
  {
97
  "epoch": 1.9696969696969697,
98
- "grad_norm": 0.2275390625,
99
- "learning_rate": 1.1229744785292821e-05,
100
- "loss": 0.0924,
101
  "step": 65
102
  },
103
  {
104
  "epoch": 2.121212121212121,
105
- "grad_norm": 0.197265625,
106
- "learning_rate": 9.235888232294472e-06,
107
- "loss": 0.092,
108
  "step": 70
109
  },
110
  {
111
  "epoch": 2.2727272727272725,
112
- "grad_norm": 0.20703125,
113
- "learning_rate": 7.444406143120487e-06,
114
- "loss": 0.0904,
115
  "step": 75
116
  },
117
  {
118
  "epoch": 2.4242424242424243,
119
- "grad_norm": 0.220703125,
120
- "learning_rate": 5.9052082987380775e-06,
121
- "loss": 0.0926,
122
  "step": 80
123
  },
124
  {
125
  "epoch": 2.5757575757575757,
126
- "grad_norm": 0.2080078125,
127
- "learning_rate": 4.6611759692099345e-06,
128
- "loss": 0.0925,
129
  "step": 85
130
  },
131
  {
132
  "epoch": 2.7272727272727275,
133
- "grad_norm": 0.2099609375,
134
- "learning_rate": 3.746967263293098e-06,
135
- "loss": 0.0932,
136
  "step": 90
137
  },
138
  {
139
  "epoch": 2.878787878787879,
140
- "grad_norm": 0.2109375,
141
- "learning_rate": 3.188051571134615e-06,
142
- "loss": 0.0916,
143
  "step": 95
144
  },
145
  {
146
- "epoch": 3.0,
147
- "step": 99,
148
- "total_flos": 1.0840722057776333e+17,
149
- "train_loss": 0.09704786779904606,
150
- "train_runtime": 144.1296,
151
- "train_samples_per_second": 86.859,
152
- "train_steps_per_second": 0.687
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
153
  }
154
  ],
155
  "logging_steps": 5,
156
- "max_steps": 99,
157
  "num_input_tokens_seen": 0,
158
- "num_train_epochs": 3,
159
  "save_steps": 500,
160
  "stateful_callbacks": {
161
  "TrainerControl": {
@@ -169,7 +960,7 @@
169
  "attributes": {}
170
  }
171
  },
172
- "total_flos": 1.0840722057776333e+17,
173
  "train_batch_size": 128,
174
  "trial_name": null,
175
  "trial_params": null
 
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
+ "epoch": 20.0,
6
  "eval_steps": 500,
7
+ "global_step": 660,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
11
  "log_history": [
12
  {
13
  "epoch": 0.15151515151515152,
14
+ "grad_norm": 1.984375,
15
+ "learning_rate": 3.6363636363636366e-06,
16
+ "loss": 0.1441,
17
  "step": 5
18
  },
19
  {
20
  "epoch": 0.30303030303030304,
21
+ "grad_norm": 1.40625,
22
+ "learning_rate": 8.181818181818181e-06,
23
+ "loss": 0.1383,
24
  "step": 10
25
  },
26
  {
27
  "epoch": 0.45454545454545453,
28
+ "grad_norm": 0.6484375,
29
+ "learning_rate": 1.2727272727272728e-05,
30
+ "loss": 0.1213,
31
  "step": 15
32
  },
33
  {
34
  "epoch": 0.6060606060606061,
35
+ "grad_norm": 0.423828125,
36
+ "learning_rate": 1.7272727272727274e-05,
37
+ "loss": 0.1133,
38
  "step": 20
39
  },
40
  {
41
  "epoch": 0.7575757575757576,
42
+ "grad_norm": 0.392578125,
43
+ "learning_rate": 2.1818181818181818e-05,
44
+ "loss": 0.1085,
45
  "step": 25
46
  },
47
  {
48
  "epoch": 0.9090909090909091,
49
+ "grad_norm": 0.28125,
50
+ "learning_rate": 2.6363636363636365e-05,
51
+ "loss": 0.1055,
52
  "step": 30
53
  },
54
  {
55
  "epoch": 1.0606060606060606,
56
+ "grad_norm": 0.2353515625,
57
+ "learning_rate": 2.9999830539872836e-05,
58
+ "loss": 0.0987,
59
  "step": 35
60
  },
61
  {
62
  "epoch": 1.2121212121212122,
63
+ "grad_norm": 0.2412109375,
64
+ "learning_rate": 2.9993899882114902e-05,
65
+ "loss": 0.1005,
66
  "step": 40
67
  },
68
  {
69
  "epoch": 1.3636363636363638,
70
+ "grad_norm": 0.21484375,
71
+ "learning_rate": 2.997950047184977e-05,
72
+ "loss": 0.0954,
73
  "step": 45
74
  },
75
  {
76
  "epoch": 1.5151515151515151,
77
+ "grad_norm": 0.2333984375,
78
+ "learning_rate": 2.9956641346126986e-05,
79
+ "loss": 0.0955,
80
  "step": 50
81
  },
82
  {
83
  "epoch": 1.6666666666666665,
84
+ "grad_norm": 0.224609375,
85
+ "learning_rate": 2.9925336851301575e-05,
86
+ "loss": 0.0946,
87
  "step": 55
88
  },
89
  {
90
  "epoch": 1.8181818181818183,
91
+ "grad_norm": 0.224609375,
92
+ "learning_rate": 2.9885606634030267e-05,
93
+ "loss": 0.0941,
94
  "step": 60
95
  },
96
  {
97
  "epoch": 1.9696969696969697,
98
+ "grad_norm": 0.2265625,
99
+ "learning_rate": 2.98374756289413e-05,
100
+ "loss": 0.0926,
101
  "step": 65
102
  },
103
  {
104
  "epoch": 2.121212121212121,
105
+ "grad_norm": 0.20703125,
106
+ "learning_rate": 2.9780974042985506e-05,
107
+ "loss": 0.0913,
108
  "step": 70
109
  },
110
  {
111
  "epoch": 2.2727272727272725,
112
+ "grad_norm": 0.2119140625,
113
+ "learning_rate": 2.971613733647841e-05,
114
+ "loss": 0.0898,
115
  "step": 75
116
  },
117
  {
118
  "epoch": 2.4242424242424243,
119
+ "grad_norm": 0.2236328125,
120
+ "learning_rate": 2.9643006200845458e-05,
121
+ "loss": 0.0914,
122
  "step": 80
123
  },
124
  {
125
  "epoch": 2.5757575757575757,
126
+ "grad_norm": 0.2177734375,
127
+ "learning_rate": 2.9561626533084068e-05,
128
+ "loss": 0.0912,
129
  "step": 85
130
  },
131
  {
132
  "epoch": 2.7272727272727275,
133
+ "grad_norm": 0.2138671875,
134
+ "learning_rate": 2.9472049406958788e-05,
135
+ "loss": 0.0906,
136
  "step": 90
137
  },
138
  {
139
  "epoch": 2.878787878787879,
140
+ "grad_norm": 0.2197265625,
141
+ "learning_rate": 2.937433104094746e-05,
142
+ "loss": 0.09,
143
  "step": 95
144
  },
145
  {
146
+ "epoch": 3.0303030303030303,
147
+ "grad_norm": 0.2197265625,
148
+ "learning_rate": 2.9268532762958568e-05,
149
+ "loss": 0.0873,
150
+ "step": 100
151
+ },
152
+ {
153
+ "epoch": 3.1818181818181817,
154
+ "grad_norm": 0.2216796875,
155
+ "learning_rate": 2.915472097184196e-05,
156
+ "loss": 0.0893,
157
+ "step": 105
158
+ },
159
+ {
160
+ "epoch": 3.3333333333333335,
161
+ "grad_norm": 0.212890625,
162
+ "learning_rate": 2.903296709571698e-05,
163
+ "loss": 0.0877,
164
+ "step": 110
165
+ },
166
+ {
167
+ "epoch": 3.484848484848485,
168
+ "grad_norm": 0.20703125,
169
+ "learning_rate": 2.8903347547144327e-05,
170
+ "loss": 0.0881,
171
+ "step": 115
172
+ },
173
+ {
174
+ "epoch": 3.6363636363636362,
175
+ "grad_norm": 0.21875,
176
+ "learning_rate": 2.876594367516961e-05,
177
+ "loss": 0.0885,
178
+ "step": 120
179
+ },
180
+ {
181
+ "epoch": 3.787878787878788,
182
+ "grad_norm": 0.21875,
183
+ "learning_rate": 2.8620841714268804e-05,
184
+ "loss": 0.0886,
185
+ "step": 125
186
+ },
187
+ {
188
+ "epoch": 3.9393939393939394,
189
+ "grad_norm": 0.2158203125,
190
+ "learning_rate": 2.846813273022764e-05,
191
+ "loss": 0.0867,
192
+ "step": 130
193
+ },
194
+ {
195
+ "epoch": 4.090909090909091,
196
+ "grad_norm": 0.2041015625,
197
+ "learning_rate": 2.83079125629888e-05,
198
+ "loss": 0.0877,
199
+ "step": 135
200
+ },
201
+ {
202
+ "epoch": 4.242424242424242,
203
+ "grad_norm": 0.2294921875,
204
+ "learning_rate": 2.8140281766502957e-05,
205
+ "loss": 0.0877,
206
+ "step": 140
207
+ },
208
+ {
209
+ "epoch": 4.393939393939394,
210
+ "grad_norm": 0.2216796875,
211
+ "learning_rate": 2.7965345545621217e-05,
212
+ "loss": 0.0875,
213
+ "step": 145
214
+ },
215
+ {
216
+ "epoch": 4.545454545454545,
217
+ "grad_norm": 0.228515625,
218
+ "learning_rate": 2.7783213690068737e-05,
219
+ "loss": 0.0861,
220
+ "step": 150
221
+ },
222
+ {
223
+ "epoch": 4.696969696969697,
224
+ "grad_norm": 0.2158203125,
225
+ "learning_rate": 2.7594000505540807e-05,
226
+ "loss": 0.0896,
227
+ "step": 155
228
+ },
229
+ {
230
+ "epoch": 4.848484848484849,
231
+ "grad_norm": 0.224609375,
232
+ "learning_rate": 2.7397824741964805e-05,
233
+ "loss": 0.0834,
234
+ "step": 160
235
+ },
236
+ {
237
+ "epoch": 5.0,
238
+ "grad_norm": 0.2890625,
239
+ "learning_rate": 2.7194809518972856e-05,
240
+ "loss": 0.0865,
241
+ "step": 165
242
+ },
243
+ {
244
+ "epoch": 5.151515151515151,
245
+ "grad_norm": 0.2177734375,
246
+ "learning_rate": 2.6985082248632174e-05,
247
+ "loss": 0.0828,
248
+ "step": 170
249
+ },
250
+ {
251
+ "epoch": 5.303030303030303,
252
+ "grad_norm": 0.2197265625,
253
+ "learning_rate": 2.676877455548141e-05,
254
+ "loss": 0.0852,
255
+ "step": 175
256
+ },
257
+ {
258
+ "epoch": 5.454545454545454,
259
+ "grad_norm": 0.2109375,
260
+ "learning_rate": 2.6546022193923274e-05,
261
+ "loss": 0.086,
262
+ "step": 180
263
+ },
264
+ {
265
+ "epoch": 5.606060606060606,
266
+ "grad_norm": 0.22265625,
267
+ "learning_rate": 2.631696496302526e-05,
268
+ "loss": 0.0869,
269
+ "step": 185
270
+ },
271
+ {
272
+ "epoch": 5.757575757575758,
273
+ "grad_norm": 0.2158203125,
274
+ "learning_rate": 2.6081746618781953e-05,
275
+ "loss": 0.0862,
276
+ "step": 190
277
+ },
278
+ {
279
+ "epoch": 5.909090909090909,
280
+ "grad_norm": 0.23046875,
281
+ "learning_rate": 2.584051478389399e-05,
282
+ "loss": 0.0847,
283
+ "step": 195
284
+ },
285
+ {
286
+ "epoch": 6.0606060606060606,
287
+ "grad_norm": 0.212890625,
288
+ "learning_rate": 2.559342085512022e-05,
289
+ "loss": 0.0853,
290
+ "step": 200
291
+ },
292
+ {
293
+ "epoch": 6.212121212121212,
294
+ "grad_norm": 0.2265625,
295
+ "learning_rate": 2.5340619908261352e-05,
296
+ "loss": 0.0877,
297
+ "step": 205
298
+ },
299
+ {
300
+ "epoch": 6.363636363636363,
301
+ "grad_norm": 0.2197265625,
302
+ "learning_rate": 2.508227060083457e-05,
303
+ "loss": 0.0807,
304
+ "step": 210
305
+ },
306
+ {
307
+ "epoch": 6.515151515151516,
308
+ "grad_norm": 0.2236328125,
309
+ "learning_rate": 2.4818535072500327e-05,
310
+ "loss": 0.0793,
311
+ "step": 215
312
+ },
313
+ {
314
+ "epoch": 6.666666666666667,
315
+ "grad_norm": 0.224609375,
316
+ "learning_rate": 2.4549578843303708e-05,
317
+ "loss": 0.0821,
318
+ "step": 220
319
+ },
320
+ {
321
+ "epoch": 6.818181818181818,
322
+ "grad_norm": 0.2412109375,
323
+ "learning_rate": 2.427557070979427e-05,
324
+ "loss": 0.0832,
325
+ "step": 225
326
+ },
327
+ {
328
+ "epoch": 6.96969696969697,
329
+ "grad_norm": 0.220703125,
330
+ "learning_rate": 2.399668263908961e-05,
331
+ "loss": 0.0844,
332
+ "step": 230
333
+ },
334
+ {
335
+ "epoch": 7.121212121212121,
336
+ "grad_norm": 0.2236328125,
337
+ "learning_rate": 2.3713089660948985e-05,
338
+ "loss": 0.0846,
339
+ "step": 235
340
+ },
341
+ {
342
+ "epoch": 7.2727272727272725,
343
+ "grad_norm": 0.208984375,
344
+ "learning_rate": 2.342496975792494e-05,
345
+ "loss": 0.0824,
346
+ "step": 240
347
+ },
348
+ {
349
+ "epoch": 7.424242424242424,
350
+ "grad_norm": 0.1953125,
351
+ "learning_rate": 2.313250375366167e-05,
352
+ "loss": 0.0796,
353
+ "step": 245
354
+ },
355
+ {
356
+ "epoch": 7.575757575757576,
357
+ "grad_norm": 0.21484375,
358
+ "learning_rate": 2.283587519941036e-05,
359
+ "loss": 0.0846,
360
+ "step": 250
361
+ },
362
+ {
363
+ "epoch": 7.7272727272727275,
364
+ "grad_norm": 0.2275390625,
365
+ "learning_rate": 2.253527025883271e-05,
366
+ "loss": 0.0814,
367
+ "step": 255
368
+ },
369
+ {
370
+ "epoch": 7.878787878787879,
371
+ "grad_norm": 0.228515625,
372
+ "learning_rate": 2.2230877591164858e-05,
373
+ "loss": 0.0846,
374
+ "step": 260
375
+ },
376
+ {
377
+ "epoch": 8.030303030303031,
378
+ "grad_norm": 0.21875,
379
+ "learning_rate": 2.192288823281509e-05,
380
+ "loss": 0.0809,
381
+ "step": 265
382
+ },
383
+ {
384
+ "epoch": 8.181818181818182,
385
+ "grad_norm": 0.2373046875,
386
+ "learning_rate": 2.1611495477469712e-05,
387
+ "loss": 0.0819,
388
+ "step": 270
389
+ },
390
+ {
391
+ "epoch": 8.333333333333334,
392
+ "grad_norm": 0.2197265625,
393
+ "learning_rate": 2.1296894754782155e-05,
394
+ "loss": 0.0851,
395
+ "step": 275
396
+ },
397
+ {
398
+ "epoch": 8.484848484848484,
399
+ "grad_norm": 0.22265625,
400
+ "learning_rate": 2.0979283507721653e-05,
401
+ "loss": 0.0821,
402
+ "step": 280
403
+ },
404
+ {
405
+ "epoch": 8.636363636363637,
406
+ "grad_norm": 0.2119140625,
407
+ "learning_rate": 2.0658861068658254e-05,
408
+ "loss": 0.0788,
409
+ "step": 285
410
+ },
411
+ {
412
+ "epoch": 8.787878787878787,
413
+ "grad_norm": 0.2216796875,
414
+ "learning_rate": 2.0335828534262148e-05,
415
+ "loss": 0.0819,
416
+ "step": 290
417
+ },
418
+ {
419
+ "epoch": 8.93939393939394,
420
+ "grad_norm": 0.2353515625,
421
+ "learning_rate": 2.001038863929568e-05,
422
+ "loss": 0.0839,
423
+ "step": 295
424
+ },
425
+ {
426
+ "epoch": 9.090909090909092,
427
+ "grad_norm": 0.2412109375,
428
+ "learning_rate": 1.9682745629377267e-05,
429
+ "loss": 0.0826,
430
+ "step": 300
431
+ },
432
+ {
433
+ "epoch": 9.242424242424242,
434
+ "grad_norm": 0.220703125,
435
+ "learning_rate": 1.9353105132797175e-05,
436
+ "loss": 0.0796,
437
+ "step": 305
438
+ },
439
+ {
440
+ "epoch": 9.393939393939394,
441
+ "grad_norm": 0.216796875,
442
+ "learning_rate": 1.902167403146548e-05,
443
+ "loss": 0.0803,
444
+ "step": 310
445
+ },
446
+ {
447
+ "epoch": 9.545454545454545,
448
+ "grad_norm": 0.2314453125,
449
+ "learning_rate": 1.8688660331073253e-05,
450
+ "loss": 0.0812,
451
+ "step": 315
452
+ },
453
+ {
454
+ "epoch": 9.696969696969697,
455
+ "grad_norm": 0.2158203125,
456
+ "learning_rate": 1.8354273030548512e-05,
457
+ "loss": 0.0792,
458
+ "step": 320
459
+ },
460
+ {
461
+ "epoch": 9.848484848484848,
462
+ "grad_norm": 0.22265625,
463
+ "learning_rate": 1.801872199088878e-05,
464
+ "loss": 0.0783,
465
+ "step": 325
466
+ },
467
+ {
468
+ "epoch": 10.0,
469
+ "grad_norm": 0.271484375,
470
+ "learning_rate": 1.7682217803452616e-05,
471
+ "loss": 0.0825,
472
+ "step": 330
473
+ },
474
+ {
475
+ "epoch": 10.151515151515152,
476
+ "grad_norm": 0.216796875,
477
+ "learning_rate": 1.7344971657792768e-05,
478
+ "loss": 0.0846,
479
+ "step": 335
480
+ },
481
+ {
482
+ "epoch": 10.303030303030303,
483
+ "grad_norm": 0.21484375,
484
+ "learning_rate": 1.7007195209113934e-05,
485
+ "loss": 0.0809,
486
+ "step": 340
487
+ },
488
+ {
489
+ "epoch": 10.454545454545455,
490
+ "grad_norm": 0.2236328125,
491
+ "learning_rate": 1.666910044543822e-05,
492
+ "loss": 0.0823,
493
+ "step": 345
494
+ },
495
+ {
496
+ "epoch": 10.606060606060606,
497
+ "grad_norm": 0.21484375,
498
+ "learning_rate": 1.6330899554561785e-05,
499
+ "loss": 0.0814,
500
+ "step": 350
501
+ },
502
+ {
503
+ "epoch": 10.757575757575758,
504
+ "grad_norm": 0.2197265625,
505
+ "learning_rate": 1.5992804790886075e-05,
506
+ "loss": 0.0795,
507
+ "step": 355
508
+ },
509
+ {
510
+ "epoch": 10.909090909090908,
511
+ "grad_norm": 0.2177734375,
512
+ "learning_rate": 1.5655028342207235e-05,
513
+ "loss": 0.0818,
514
+ "step": 360
515
+ },
516
+ {
517
+ "epoch": 11.06060606060606,
518
+ "grad_norm": 0.228515625,
519
+ "learning_rate": 1.5317782196547387e-05,
520
+ "loss": 0.0817,
521
+ "step": 365
522
+ },
523
+ {
524
+ "epoch": 11.212121212121213,
525
+ "grad_norm": 0.2373046875,
526
+ "learning_rate": 1.4981278009111222e-05,
527
+ "loss": 0.0819,
528
+ "step": 370
529
+ },
530
+ {
531
+ "epoch": 11.363636363636363,
532
+ "grad_norm": 0.2294921875,
533
+ "learning_rate": 1.4645726969451489e-05,
534
+ "loss": 0.0778,
535
+ "step": 375
536
+ },
537
+ {
538
+ "epoch": 11.515151515151516,
539
+ "grad_norm": 0.2197265625,
540
+ "learning_rate": 1.4311339668926748e-05,
541
+ "loss": 0.0796,
542
+ "step": 380
543
+ },
544
+ {
545
+ "epoch": 11.666666666666666,
546
+ "grad_norm": 0.23828125,
547
+ "learning_rate": 1.397832596853452e-05,
548
+ "loss": 0.0823,
549
+ "step": 385
550
+ },
551
+ {
552
+ "epoch": 11.818181818181818,
553
+ "grad_norm": 0.228515625,
554
+ "learning_rate": 1.3646894867202821e-05,
555
+ "loss": 0.0794,
556
+ "step": 390
557
+ },
558
+ {
559
+ "epoch": 11.969696969696969,
560
+ "grad_norm": 0.2314453125,
561
+ "learning_rate": 1.3317254370622732e-05,
562
+ "loss": 0.0834,
563
+ "step": 395
564
+ },
565
+ {
566
+ "epoch": 12.121212121212121,
567
+ "grad_norm": 0.22265625,
568
+ "learning_rate": 1.298961136070432e-05,
569
+ "loss": 0.0811,
570
+ "step": 400
571
+ },
572
+ {
573
+ "epoch": 12.272727272727273,
574
+ "grad_norm": 0.2275390625,
575
+ "learning_rate": 1.266417146573785e-05,
576
+ "loss": 0.078,
577
+ "step": 405
578
+ },
579
+ {
580
+ "epoch": 12.424242424242424,
581
+ "grad_norm": 0.2216796875,
582
+ "learning_rate": 1.2341138931341752e-05,
583
+ "loss": 0.0814,
584
+ "step": 410
585
+ },
586
+ {
587
+ "epoch": 12.575757575757576,
588
+ "grad_norm": 0.2373046875,
589
+ "learning_rate": 1.2020716492278353e-05,
590
+ "loss": 0.0811,
591
+ "step": 415
592
+ },
593
+ {
594
+ "epoch": 12.727272727272727,
595
+ "grad_norm": 0.212890625,
596
+ "learning_rate": 1.1703105245217848e-05,
597
+ "loss": 0.0787,
598
+ "step": 420
599
+ },
600
+ {
601
+ "epoch": 12.878787878787879,
602
+ "grad_norm": 0.224609375,
603
+ "learning_rate": 1.1388504522530296e-05,
604
+ "loss": 0.0792,
605
+ "step": 425
606
+ },
607
+ {
608
+ "epoch": 13.030303030303031,
609
+ "grad_norm": 0.2119140625,
610
+ "learning_rate": 1.1077111767184916e-05,
611
+ "loss": 0.0816,
612
+ "step": 430
613
+ },
614
+ {
615
+ "epoch": 13.181818181818182,
616
+ "grad_norm": 0.234375,
617
+ "learning_rate": 1.0769122408835148e-05,
618
+ "loss": 0.0786,
619
+ "step": 435
620
+ },
621
+ {
622
+ "epoch": 13.333333333333334,
623
+ "grad_norm": 0.216796875,
624
+ "learning_rate": 1.0464729741167291e-05,
625
+ "loss": 0.0825,
626
+ "step": 440
627
+ },
628
+ {
629
+ "epoch": 13.484848484848484,
630
+ "grad_norm": 0.2119140625,
631
+ "learning_rate": 1.016412480058964e-05,
632
+ "loss": 0.0807,
633
+ "step": 445
634
+ },
635
+ {
636
+ "epoch": 13.636363636363637,
637
+ "grad_norm": 0.2255859375,
638
+ "learning_rate": 9.86749624633833e-06,
639
+ "loss": 0.0832,
640
+ "step": 450
641
+ },
642
+ {
643
+ "epoch": 13.787878787878787,
644
+ "grad_norm": 0.2294921875,
645
+ "learning_rate": 9.575030242075062e-06,
646
+ "loss": 0.0778,
647
+ "step": 455
648
+ },
649
+ {
650
+ "epoch": 13.93939393939394,
651
+ "grad_norm": 0.4296875,
652
+ "learning_rate": 9.286910339051015e-06,
653
+ "loss": 0.0846,
654
+ "step": 460
655
+ },
656
+ {
657
+ "epoch": 14.090909090909092,
658
+ "grad_norm": 0.216796875,
659
+ "learning_rate": 9.003317360910392e-06,
660
+ "loss": 0.0806,
661
+ "step": 465
662
+ },
663
+ {
664
+ "epoch": 14.242424242424242,
665
+ "grad_norm": 0.220703125,
666
+ "learning_rate": 8.724429290205732e-06,
667
+ "loss": 0.0787,
668
+ "step": 470
669
+ },
670
+ {
671
+ "epoch": 14.393939393939394,
672
+ "grad_norm": 0.2197265625,
673
+ "learning_rate": 8.450421156696298e-06,
674
+ "loss": 0.0817,
675
+ "step": 475
676
+ },
677
+ {
678
+ "epoch": 14.545454545454545,
679
+ "grad_norm": 0.2158203125,
680
+ "learning_rate": 8.181464927499674e-06,
681
+ "loss": 0.0776,
682
+ "step": 480
683
+ },
684
+ {
685
+ "epoch": 14.696969696969697,
686
+ "grad_norm": 0.228515625,
687
+ "learning_rate": 7.917729399165435e-06,
688
+ "loss": 0.0801,
689
+ "step": 485
690
+ },
691
+ {
692
+ "epoch": 14.848484848484848,
693
+ "grad_norm": 0.2314453125,
694
+ "learning_rate": 7.659380091738652e-06,
695
+ "loss": 0.0781,
696
+ "step": 490
697
+ },
698
+ {
699
+ "epoch": 15.0,
700
+ "grad_norm": 0.31640625,
701
+ "learning_rate": 7.406579144879779e-06,
702
+ "loss": 0.0786,
703
+ "step": 495
704
+ },
705
+ {
706
+ "epoch": 15.151515151515152,
707
+ "grad_norm": 0.2275390625,
708
+ "learning_rate": 7.159485216106013e-06,
709
+ "loss": 0.0807,
710
+ "step": 500
711
+ },
712
+ {
713
+ "epoch": 15.303030303030303,
714
+ "grad_norm": 0.232421875,
715
+ "learning_rate": 6.918253381218046e-06,
716
+ "loss": 0.0767,
717
+ "step": 505
718
+ },
719
+ {
720
+ "epoch": 15.454545454545455,
721
+ "grad_norm": 0.234375,
722
+ "learning_rate": 6.683035036974742e-06,
723
+ "loss": 0.0787,
724
+ "step": 510
725
+ },
726
+ {
727
+ "epoch": 15.606060606060606,
728
+ "grad_norm": 0.2373046875,
729
+ "learning_rate": 6.45397780607673e-06,
730
+ "loss": 0.0763,
731
+ "step": 515
732
+ },
733
+ {
734
+ "epoch": 15.757575757575758,
735
+ "grad_norm": 0.2216796875,
736
+ "learning_rate": 6.23122544451859e-06,
737
+ "loss": 0.081,
738
+ "step": 520
739
+ },
740
+ {
741
+ "epoch": 15.909090909090908,
742
+ "grad_norm": 0.234375,
743
+ "learning_rate": 6.014917751367825e-06,
744
+ "loss": 0.0794,
745
+ "step": 525
746
+ },
747
+ {
748
+ "epoch": 16.060606060606062,
749
+ "grad_norm": 0.2236328125,
750
+ "learning_rate": 5.80519048102715e-06,
751
+ "loss": 0.0787,
752
+ "step": 530
753
+ },
754
+ {
755
+ "epoch": 16.21212121212121,
756
+ "grad_norm": 0.2314453125,
757
+ "learning_rate": 5.602175258035204e-06,
758
+ "loss": 0.077,
759
+ "step": 535
760
+ },
761
+ {
762
+ "epoch": 16.363636363636363,
763
+ "grad_norm": 0.2412109375,
764
+ "learning_rate": 5.4059994944591914e-06,
765
+ "loss": 0.0798,
766
+ "step": 540
767
+ },
768
+ {
769
+ "epoch": 16.515151515151516,
770
+ "grad_norm": 0.21875,
771
+ "learning_rate": 5.2167863099312636e-06,
772
+ "loss": 0.0794,
773
+ "step": 545
774
+ },
775
+ {
776
+ "epoch": 16.666666666666668,
777
+ "grad_norm": 0.2373046875,
778
+ "learning_rate": 5.034654454378783e-06,
779
+ "loss": 0.0793,
780
+ "step": 550
781
+ },
782
+ {
783
+ "epoch": 16.818181818181817,
784
+ "grad_norm": 0.2353515625,
785
+ "learning_rate": 4.859718233497048e-06,
786
+ "loss": 0.0801,
787
+ "step": 555
788
+ },
789
+ {
790
+ "epoch": 16.96969696969697,
791
+ "grad_norm": 0.2177734375,
792
+ "learning_rate": 4.692087437011203e-06,
793
+ "loss": 0.0791,
794
+ "step": 560
795
+ },
796
+ {
797
+ "epoch": 17.12121212121212,
798
+ "grad_norm": 0.2255859375,
799
+ "learning_rate": 4.5318672697723665e-06,
800
+ "loss": 0.081,
801
+ "step": 565
802
+ },
803
+ {
804
+ "epoch": 17.272727272727273,
805
+ "grad_norm": 0.2265625,
806
+ "learning_rate": 4.3791582857311975e-06,
807
+ "loss": 0.0792,
808
+ "step": 570
809
+ },
810
+ {
811
+ "epoch": 17.424242424242426,
812
+ "grad_norm": 0.2119140625,
813
+ "learning_rate": 4.2340563248303915e-06,
814
+ "loss": 0.0805,
815
+ "step": 575
816
+ },
817
+ {
818
+ "epoch": 17.575757575757574,
819
+ "grad_norm": 0.2197265625,
820
+ "learning_rate": 4.096652452855675e-06,
821
+ "loss": 0.0797,
822
+ "step": 580
823
+ },
824
+ {
825
+ "epoch": 17.727272727272727,
826
+ "grad_norm": 0.212890625,
827
+ "learning_rate": 3.967032904283021e-06,
828
+ "loss": 0.0809,
829
+ "step": 585
830
+ },
831
+ {
832
+ "epoch": 17.87878787878788,
833
+ "grad_norm": 0.236328125,
834
+ "learning_rate": 3.8452790281580445e-06,
835
+ "loss": 0.0803,
836
+ "step": 590
837
+ },
838
+ {
839
+ "epoch": 18.03030303030303,
840
+ "grad_norm": 0.23828125,
841
+ "learning_rate": 3.731467237041433e-06,
842
+ "loss": 0.0804,
843
+ "step": 595
844
+ },
845
+ {
846
+ "epoch": 18.181818181818183,
847
+ "grad_norm": 0.2236328125,
848
+ "learning_rate": 3.6256689590525444e-06,
849
+ "loss": 0.0824,
850
+ "step": 600
851
+ },
852
+ {
853
+ "epoch": 18.333333333333332,
854
+ "grad_norm": 0.224609375,
855
+ "learning_rate": 3.5279505930412164e-06,
856
+ "loss": 0.0809,
857
+ "step": 605
858
+ },
859
+ {
860
+ "epoch": 18.484848484848484,
861
+ "grad_norm": 0.22265625,
862
+ "learning_rate": 3.4383734669159366e-06,
863
+ "loss": 0.0811,
864
+ "step": 610
865
+ },
866
+ {
867
+ "epoch": 18.636363636363637,
868
+ "grad_norm": 0.2265625,
869
+ "learning_rate": 3.356993799154545e-06,
870
+ "loss": 0.0786,
871
+ "step": 615
872
+ },
873
+ {
874
+ "epoch": 18.78787878787879,
875
+ "grad_norm": 0.2158203125,
876
+ "learning_rate": 3.2838626635215874e-06,
877
+ "loss": 0.0811,
878
+ "step": 620
879
+ },
880
+ {
881
+ "epoch": 18.939393939393938,
882
+ "grad_norm": 0.2216796875,
883
+ "learning_rate": 3.2190259570144957e-06,
884
+ "loss": 0.0834,
885
+ "step": 625
886
+ },
887
+ {
888
+ "epoch": 19.09090909090909,
889
+ "grad_norm": 0.2255859375,
890
+ "learning_rate": 3.162524371058697e-06,
891
+ "loss": 0.0804,
892
+ "step": 630
893
+ },
894
+ {
895
+ "epoch": 19.242424242424242,
896
+ "grad_norm": 0.220703125,
897
+ "learning_rate": 3.1143933659697377e-06,
898
+ "loss": 0.0777,
899
+ "step": 635
900
+ },
901
+ {
902
+ "epoch": 19.393939393939394,
903
+ "grad_norm": 0.2333984375,
904
+ "learning_rate": 3.0746631486984266e-06,
905
+ "loss": 0.081,
906
+ "step": 640
907
+ },
908
+ {
909
+ "epoch": 19.545454545454547,
910
+ "grad_norm": 0.2294921875,
911
+ "learning_rate": 3.043358653873013e-06,
912
+ "loss": 0.0776,
913
+ "step": 645
914
+ },
915
+ {
916
+ "epoch": 19.696969696969695,
917
+ "grad_norm": 0.22265625,
918
+ "learning_rate": 3.020499528150232e-06,
919
+ "loss": 0.077,
920
+ "step": 650
921
+ },
922
+ {
923
+ "epoch": 19.848484848484848,
924
+ "grad_norm": 0.220703125,
925
+ "learning_rate": 3.006100117885101e-06,
926
+ "loss": 0.0814,
927
+ "step": 655
928
+ },
929
+ {
930
+ "epoch": 20.0,
931
+ "grad_norm": 0.298828125,
932
+ "learning_rate": 3.000169460127164e-06,
933
+ "loss": 0.0805,
934
+ "step": 660
935
+ },
936
+ {
937
+ "epoch": 20.0,
938
+ "step": 660,
939
+ "total_flos": 7.227148038517555e+17,
940
+ "train_loss": 0.08479578431808588,
941
+ "train_runtime": 842.3759,
942
+ "train_samples_per_second": 99.077,
943
+ "train_steps_per_second": 0.783
944
  }
945
  ],
946
  "logging_steps": 5,
947
+ "max_steps": 660,
948
  "num_input_tokens_seen": 0,
949
+ "num_train_epochs": 20,
950
  "save_steps": 500,
951
  "stateful_callbacks": {
952
  "TrainerControl": {
 
960
  "attributes": {}
961
  }
962
  },
963
+ "total_flos": 7.227148038517555e+17,
964
  "train_batch_size": 128,
965
  "trial_name": null,
966
  "trial_params": null