File size: 10,004 Bytes
544fffb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 165,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 4.652278347218786,
      "learning_rate": 2.941176470588235e-08,
      "logits/chosen": -0.8284896612167358,
      "logits/rejected": -0.9010236263275146,
      "logps/chosen": -1066.3585205078125,
      "logps/rejected": -1448.19970703125,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.746771519966634,
      "learning_rate": 2.941176470588235e-07,
      "logits/chosen": -0.8115520477294922,
      "logits/rejected": -0.8255029320716858,
      "logps/chosen": -1131.291259765625,
      "logps/rejected": -1369.7412109375,
      "loss": 0.6932,
      "rewards/accuracies": 0.4652777910232544,
      "rewards/chosen": 0.0002041943371295929,
      "rewards/margins": -4.850090044783428e-05,
      "rewards/rejected": 0.0002526953467167914,
      "step": 10
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.451745733301935,
      "learning_rate": 4.994932636402031e-07,
      "logits/chosen": -0.7243806719779968,
      "logits/rejected": -0.8158847093582153,
      "logps/chosen": -1020.7599487304688,
      "logps/rejected": -1355.944091796875,
      "loss": 0.6921,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": 0.001858971663750708,
      "rewards/margins": 0.0021505323238670826,
      "rewards/rejected": -0.00029156063101254404,
      "step": 20
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.532594184835871,
      "learning_rate": 4.905416503522123e-07,
      "logits/chosen": -0.7353666424751282,
      "logits/rejected": -0.8100309371948242,
      "logps/chosen": -1033.032470703125,
      "logps/rejected": -1331.6929931640625,
      "loss": 0.688,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 0.008143061771988869,
      "rewards/margins": 0.010795501992106438,
      "rewards/rejected": -0.002652441617101431,
      "step": 30
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.452955387273571,
      "learning_rate": 4.707922373336523e-07,
      "logits/chosen": -0.7547545433044434,
      "logits/rejected": -0.7800291776657104,
      "logps/chosen": -1057.7445068359375,
      "logps/rejected": -1296.575439453125,
      "loss": 0.6825,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.008273174986243248,
      "rewards/margins": 0.016675911843776703,
      "rewards/rejected": -0.008402736857533455,
      "step": 40
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.931805222376995,
      "learning_rate": 4.4113156629677313e-07,
      "logits/chosen": -0.7371411919593811,
      "logits/rejected": -0.6845098733901978,
      "logps/chosen": -1045.1011962890625,
      "logps/rejected": -1151.344970703125,
      "loss": 0.6718,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": 0.017855554819107056,
      "rewards/margins": 0.04448147863149643,
      "rewards/rejected": -0.026625927537679672,
      "step": 50
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.689422676957636,
      "learning_rate": 4.0289109058972283e-07,
      "logits/chosen": -0.7692807912826538,
      "logits/rejected": -0.7662399411201477,
      "logps/chosen": -999.9730224609375,
      "logps/rejected": -1286.947509765625,
      "loss": 0.6595,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.027734506875276566,
      "rewards/margins": 0.07751207053661346,
      "rewards/rejected": -0.0497775673866272,
      "step": 60
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.956697872711078,
      "learning_rate": 3.577874068920446e-07,
      "logits/chosen": -0.7923519611358643,
      "logits/rejected": -0.8132171630859375,
      "logps/chosen": -1077.121337890625,
      "logps/rejected": -1317.41845703125,
      "loss": 0.6474,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 0.009537232108414173,
      "rewards/margins": 0.09025295078754425,
      "rewards/rejected": -0.08071572333574295,
      "step": 70
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.011926472486656,
      "learning_rate": 3.078451980100854e-07,
      "logits/chosen": -0.7588658928871155,
      "logits/rejected": -0.8289008140563965,
      "logps/chosen": -1011.5177612304688,
      "logps/rejected": -1298.1904296875,
      "loss": 0.6262,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.023491863161325455,
      "rewards/margins": 0.1584763377904892,
      "rewards/rejected": -0.13498449325561523,
      "step": 80
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.367792800931542,
      "learning_rate": 2.553063458334059e-07,
      "logits/chosen": -0.7922073006629944,
      "logits/rejected": -0.8237783312797546,
      "logps/chosen": -1067.434326171875,
      "logps/rejected": -1301.37109375,
      "loss": 0.5865,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -0.015568578615784645,
      "rewards/margins": 0.31723320484161377,
      "rewards/rejected": -0.3328017592430115,
      "step": 90
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.346586894544636,
      "learning_rate": 2.0252929432814287e-07,
      "logits/chosen": -0.779016375541687,
      "logits/rejected": -0.9120697975158691,
      "logps/chosen": -1015.7185668945312,
      "logps/rejected": -1394.1680908203125,
      "loss": 0.5615,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": -0.05275765806436539,
      "rewards/margins": 0.5331910848617554,
      "rewards/rejected": -0.5859487056732178,
      "step": 100
    },
    {
      "epoch": 0.61,
      "eval_logits/chosen": -0.6675596237182617,
      "eval_logits/rejected": -0.8939424753189087,
      "eval_logps/chosen": -826.0933837890625,
      "eval_logps/rejected": -1433.1563720703125,
      "eval_loss": 0.6218963861465454,
      "eval_rewards/accuracies": 0.7459239363670349,
      "eval_rewards/chosen": -0.044515106827020645,
      "eval_rewards/margins": 0.19978085160255432,
      "eval_rewards/rejected": -0.24429598450660706,
      "eval_runtime": 353.1381,
      "eval_samples_per_second": 8.289,
      "eval_steps_per_second": 0.261,
      "step": 100
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.41704752041782,
      "learning_rate": 1.5188318011445906e-07,
      "logits/chosen": -0.7974969744682312,
      "logits/rejected": -0.8456804156303406,
      "logps/chosen": -1040.8197021484375,
      "logps/rejected": -1285.4654541015625,
      "loss": 0.5406,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.08253589272499084,
      "rewards/margins": 0.788347601890564,
      "rewards/rejected": -0.870883584022522,
      "step": 110
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.860188078136068,
      "learning_rate": 1.0564148305586295e-07,
      "logits/chosen": -0.7957097291946411,
      "logits/rejected": -0.8646506071090698,
      "logps/chosen": -979.1201171875,
      "logps/rejected": -1402.442138671875,
      "loss": 0.5115,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": -0.07362432777881622,
      "rewards/margins": 0.8031437993049622,
      "rewards/rejected": -0.8767681121826172,
      "step": 120
    },
    {
      "epoch": 0.79,
      "grad_norm": 6.241835191835041,
      "learning_rate": 6.587997083462196e-08,
      "logits/chosen": -0.828117847442627,
      "logits/rejected": -0.8768518567085266,
      "logps/chosen": -1065.3460693359375,
      "logps/rejected": -1385.936767578125,
      "loss": 0.5146,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.0992397740483284,
      "rewards/margins": 1.1248447895050049,
      "rewards/rejected": -1.2240846157073975,
      "step": 130
    },
    {
      "epoch": 0.85,
      "grad_norm": 6.81375264804862,
      "learning_rate": 3.438351873250492e-08,
      "logits/chosen": -0.8043051958084106,
      "logits/rejected": -0.8954287767410278,
      "logps/chosen": -1059.5267333984375,
      "logps/rejected": -1414.9984130859375,
      "loss": 0.5123,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": -0.17837993800640106,
      "rewards/margins": 0.9624137878417969,
      "rewards/rejected": -1.1407936811447144,
      "step": 140
    },
    {
      "epoch": 0.91,
      "grad_norm": 7.068585992863548,
      "learning_rate": 1.256598743236703e-08,
      "logits/chosen": -0.8356617093086243,
      "logits/rejected": -0.8929821252822876,
      "logps/chosen": -1066.185302734375,
      "logps/rejected": -1415.41064453125,
      "loss": 0.5079,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.16673186421394348,
      "rewards/margins": 0.8981729745864868,
      "rewards/rejected": -1.0649049282073975,
      "step": 150
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.546330593670502,
      "learning_rate": 1.406755487774386e-09,
      "logits/chosen": -0.816728949546814,
      "logits/rejected": -0.8898676633834839,
      "logps/chosen": -1021.1590576171875,
      "logps/rejected": -1444.29296875,
      "loss": 0.4992,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -0.196590393781662,
      "rewards/margins": 0.9931901097297668,
      "rewards/rejected": -1.189780592918396,
      "step": 160
    },
    {
      "epoch": 1.0,
      "step": 165,
      "total_flos": 0.0,
      "train_loss": 0.20238888480446554,
      "train_runtime": 919.6394,
      "train_samples_per_second": 11.461,
      "train_steps_per_second": 0.179
    }
  ],
  "logging_steps": 10,
  "max_steps": 165,
  "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
}