File size: 9,465 Bytes
a4205b5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.996742671009772,
  "eval_steps": 100,
  "global_step": 153,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 20.72440115890601,
      "learning_rate": 3.125e-08,
      "logits/chosen": -2.8101868629455566,
      "logits/rejected": -2.7995030879974365,
      "logps/chosen": -574.7534790039062,
      "logps/rejected": -1069.037109375,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.07,
      "grad_norm": 18.465682889641315,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -2.828524589538574,
      "logits/rejected": -2.7697534561157227,
      "logps/chosen": -477.94866943359375,
      "logps/rejected": -1011.936767578125,
      "loss": 0.6922,
      "rewards/accuracies": 0.5277777910232544,
      "rewards/chosen": 0.0030513808596879244,
      "rewards/margins": 0.0022961744107306004,
      "rewards/rejected": 0.000755206448957324,
      "step": 10
    },
    {
      "epoch": 0.13,
      "grad_norm": 20.63587915637583,
      "learning_rate": 4.989490450759331e-07,
      "logits/chosen": -2.914271593093872,
      "logits/rejected": -2.8110787868499756,
      "logps/chosen": -559.07470703125,
      "logps/rejected": -1107.041748046875,
      "loss": 0.6681,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.043809644877910614,
      "rewards/margins": 0.053940802812576294,
      "rewards/rejected": -0.010131158865988255,
      "step": 20
    },
    {
      "epoch": 0.2,
      "grad_norm": 18.63790840633573,
      "learning_rate": 4.872270441827174e-07,
      "logits/chosen": -2.95158314704895,
      "logits/rejected": -2.8900146484375,
      "logps/chosen": -586.1901245117188,
      "logps/rejected": -1104.927978515625,
      "loss": 0.6191,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.07399635016918182,
      "rewards/margins": 0.1986941397190094,
      "rewards/rejected": -0.12469778954982758,
      "step": 30
    },
    {
      "epoch": 0.26,
      "grad_norm": 27.298745582804646,
      "learning_rate": 4.6308512113530063e-07,
      "logits/chosen": -3.0119404792785645,
      "logits/rejected": -3.0303287506103516,
      "logps/chosen": -568.1453247070312,
      "logps/rejected": -1168.633544921875,
      "loss": 0.4834,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.1426580548286438,
      "rewards/margins": 0.7169780731201172,
      "rewards/rejected": -0.8596361875534058,
      "step": 40
    },
    {
      "epoch": 0.33,
      "grad_norm": 28.357781831320167,
      "learning_rate": 4.277872161641681e-07,
      "logits/chosen": -3.2949161529541016,
      "logits/rejected": -3.241774320602417,
      "logps/chosen": -628.916015625,
      "logps/rejected": -1339.3349609375,
      "loss": 0.2917,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -0.8088345527648926,
      "rewards/margins": 2.2371134757995605,
      "rewards/rejected": -3.0459485054016113,
      "step": 50
    },
    {
      "epoch": 0.39,
      "grad_norm": 21.80399650152746,
      "learning_rate": 3.8318133624280046e-07,
      "logits/chosen": -3.2456772327423096,
      "logits/rejected": -3.213705539703369,
      "logps/chosen": -700.1171264648438,
      "logps/rejected": -1630.8173828125,
      "loss": 0.2407,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -1.35238516330719,
      "rewards/margins": 4.4795427322387695,
      "rewards/rejected": -5.831927299499512,
      "step": 60
    },
    {
      "epoch": 0.46,
      "grad_norm": 18.258479162120135,
      "learning_rate": 3.316028034595861e-07,
      "logits/chosen": -3.1275057792663574,
      "logits/rejected": -3.216618061065674,
      "logps/chosen": -673.2752075195312,
      "logps/rejected": -1530.3189697265625,
      "loss": 0.2029,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.7634323835372925,
      "rewards/margins": 3.932297468185425,
      "rewards/rejected": -4.695730686187744,
      "step": 70
    },
    {
      "epoch": 0.52,
      "grad_norm": 40.88867782206525,
      "learning_rate": 2.7575199021178855e-07,
      "logits/chosen": -3.0596213340759277,
      "logits/rejected": -3.1614341735839844,
      "logps/chosen": -607.4539184570312,
      "logps/rejected": -1750.110107421875,
      "loss": 0.1651,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -1.0371493101119995,
      "rewards/margins": 5.035472869873047,
      "rewards/rejected": -6.072621822357178,
      "step": 80
    },
    {
      "epoch": 0.59,
      "grad_norm": 20.720968605328384,
      "learning_rate": 2.1855294234408068e-07,
      "logits/chosen": -3.212242841720581,
      "logits/rejected": -3.208061695098877,
      "logps/chosen": -692.4600830078125,
      "logps/rejected": -1919.945068359375,
      "loss": 0.1388,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.2357938289642334,
      "rewards/margins": 6.9475884437561035,
      "rewards/rejected": -8.183381080627441,
      "step": 90
    },
    {
      "epoch": 0.65,
      "grad_norm": 21.014024444843912,
      "learning_rate": 1.6300029195778453e-07,
      "logits/chosen": -3.062844753265381,
      "logits/rejected": -3.173572063446045,
      "logps/chosen": -761.4752807617188,
      "logps/rejected": -1871.939208984375,
      "loss": 0.1362,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": -1.5748692750930786,
      "rewards/margins": 6.684431552886963,
      "rewards/rejected": -8.259300231933594,
      "step": 100
    },
    {
      "epoch": 0.65,
      "eval_logits/chosen": -3.279001474380493,
      "eval_logits/rejected": -3.112272024154663,
      "eval_logps/chosen": -721.5318603515625,
      "eval_logps/rejected": -1477.787353515625,
      "eval_loss": 0.32064372301101685,
      "eval_rewards/accuracies": 0.8707386255264282,
      "eval_rewards/chosen": -1.9160078763961792,
      "eval_rewards/margins": 3.549539089202881,
      "eval_rewards/rejected": -5.465547561645508,
      "eval_runtime": 295.1481,
      "eval_samples_per_second": 9.5,
      "eval_steps_per_second": 0.298,
      "step": 100
    },
    {
      "epoch": 0.72,
      "grad_norm": 18.750300976193962,
      "learning_rate": 1.1200247470632392e-07,
      "logits/chosen": -3.0732262134552,
      "logits/rejected": -3.194598436355591,
      "logps/chosen": -676.952880859375,
      "logps/rejected": -1844.5322265625,
      "loss": 0.1452,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -1.3141238689422607,
      "rewards/margins": 6.629952907562256,
      "rewards/rejected": -7.944077491760254,
      "step": 110
    },
    {
      "epoch": 0.78,
      "grad_norm": 26.345958267827655,
      "learning_rate": 6.822945986946385e-08,
      "logits/chosen": -3.1366074085235596,
      "logits/rejected": -3.1714870929718018,
      "logps/chosen": -760.309814453125,
      "logps/rejected": -1834.424560546875,
      "loss": 0.1493,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": -1.501333475112915,
      "rewards/margins": 6.5416059494018555,
      "rewards/rejected": -8.042940139770508,
      "step": 120
    },
    {
      "epoch": 0.85,
      "grad_norm": 28.267952965151345,
      "learning_rate": 3.397296523427806e-08,
      "logits/chosen": -3.0828897953033447,
      "logits/rejected": -3.2096328735351562,
      "logps/chosen": -751.352783203125,
      "logps/rejected": -1817.760986328125,
      "loss": 0.1283,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -1.5563374757766724,
      "rewards/margins": 6.55484676361084,
      "rewards/rejected": -8.111185073852539,
      "step": 130
    },
    {
      "epoch": 0.91,
      "grad_norm": 33.23618703081401,
      "learning_rate": 1.1026475173977978e-08,
      "logits/chosen": -3.1680922508239746,
      "logits/rejected": -3.1776795387268066,
      "logps/chosen": -692.7706298828125,
      "logps/rejected": -1928.9267578125,
      "loss": 0.1068,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": -1.3918243646621704,
      "rewards/margins": 6.6903181076049805,
      "rewards/rejected": -8.08214282989502,
      "step": 140
    },
    {
      "epoch": 0.98,
      "grad_norm": 35.04978007474125,
      "learning_rate": 5.913435276374834e-10,
      "logits/chosen": -3.089778423309326,
      "logits/rejected": -3.155916452407837,
      "logps/chosen": -713.7938842773438,
      "logps/rejected": -1777.8834228515625,
      "loss": 0.1383,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -1.6856982707977295,
      "rewards/margins": 5.726717472076416,
      "rewards/rejected": -7.412415504455566,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 153,
      "total_flos": 0.0,
      "train_loss": 0.04516718044779659,
      "train_runtime": 721.6494,
      "train_samples_per_second": 13.591,
      "train_steps_per_second": 0.212
    }
  ],
  "logging_steps": 10,
  "max_steps": 153,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}