File size: 9,779 Bytes
fcec091
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 0.9968652037617555,
  "eval_steps": 500,
  "global_step": 159,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 3.125e-08,
      "logits/chosen": -1.995415210723877,
      "logits/rejected": -2.0361223220825195,
      "logps/chosen": -475.42401123046875,
      "logps/pi_response": -236.53262329101562,
      "logps/ref_response": -236.53262329101562,
      "logps/rejected": -571.512451171875,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -2.0094504356384277,
      "logits/rejected": -1.9493383169174194,
      "logps/chosen": -398.70062255859375,
      "logps/pi_response": -160.86041259765625,
      "logps/ref_response": -161.05593872070312,
      "logps/rejected": -436.8709411621094,
      "loss": 0.6906,
      "rewards/accuracies": 0.4513888955116272,
      "rewards/chosen": -0.010171439498662949,
      "rewards/margins": 0.0004877760075032711,
      "rewards/rejected": -0.010659217834472656,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.990353313429303e-07,
      "logits/chosen": -1.9780817031860352,
      "logits/rejected": -1.9359722137451172,
      "logps/chosen": -396.9286193847656,
      "logps/pi_response": -165.99008178710938,
      "logps/ref_response": -170.16830444335938,
      "logps/rejected": -542.5230712890625,
      "loss": 0.6473,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.28707781434059143,
      "rewards/margins": 0.1905716061592102,
      "rewards/rejected": -0.4776495099067688,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.882681251368548e-07,
      "logits/chosen": -1.929377794265747,
      "logits/rejected": -1.9014297723770142,
      "logps/chosen": -477.72247314453125,
      "logps/pi_response": -175.51304626464844,
      "logps/ref_response": -176.73623657226562,
      "logps/rejected": -558.6651000976562,
      "loss": 0.6794,
      "rewards/accuracies": 0.5687500238418579,
      "rewards/chosen": -0.7204787135124207,
      "rewards/margins": 0.3288179636001587,
      "rewards/rejected": -1.0492966175079346,
      "step": 30
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.6604720940421207e-07,
      "logits/chosen": -1.9873077869415283,
      "logits/rejected": -1.9341981410980225,
      "logps/chosen": -473.9336853027344,
      "logps/pi_response": -187.4553680419922,
      "logps/ref_response": -175.1200408935547,
      "logps/rejected": -595.444091796875,
      "loss": 0.6401,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.7007579803466797,
      "rewards/margins": 0.272124320268631,
      "rewards/rejected": -0.9728823900222778,
      "step": 40
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.3344075855595097e-07,
      "logits/chosen": -1.8471710681915283,
      "logits/rejected": -1.641099214553833,
      "logps/chosen": -511.49871826171875,
      "logps/pi_response": -183.08493041992188,
      "logps/ref_response": -184.63198852539062,
      "logps/rejected": -632.9375,
      "loss": 0.6017,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.8219456672668457,
      "rewards/margins": 0.4130307137966156,
      "rewards/rejected": -1.2349765300750732,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.920161866827889e-07,
      "logits/chosen": -0.9633913040161133,
      "logits/rejected": -0.7572700381278992,
      "logps/chosen": -479.69775390625,
      "logps/pi_response": -182.07357788085938,
      "logps/ref_response": -175.98574829101562,
      "logps/rejected": -607.6907958984375,
      "loss": 0.5892,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.6976214051246643,
      "rewards/margins": 0.5892900228500366,
      "rewards/rejected": -1.2869113683700562,
      "step": 60
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4376480090239047e-07,
      "logits/chosen": -0.3013351559638977,
      "logits/rejected": 0.12189887464046478,
      "logps/chosen": -462.14190673828125,
      "logps/pi_response": -178.45272827148438,
      "logps/ref_response": -161.18075561523438,
      "logps/rejected": -615.686279296875,
      "loss": 0.5997,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.698739767074585,
      "rewards/margins": 0.6993860006332397,
      "rewards/rejected": -1.3981258869171143,
      "step": 70
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.910060778827554e-07,
      "logits/chosen": -0.2273164540529251,
      "logits/rejected": 0.14264746010303497,
      "logps/chosen": -505.594970703125,
      "logps/pi_response": -187.5181121826172,
      "logps/ref_response": -175.85252380371094,
      "logps/rejected": -639.3907470703125,
      "loss": 0.5449,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.9091317057609558,
      "rewards/margins": 0.6041504144668579,
      "rewards/rejected": -1.5132819414138794,
      "step": 80
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.3627616503391812e-07,
      "logits/chosen": 0.14409923553466797,
      "logits/rejected": 0.62093585729599,
      "logps/chosen": -543.806640625,
      "logps/pi_response": -203.8004608154297,
      "logps/ref_response": -184.4833526611328,
      "logps/rejected": -670.08984375,
      "loss": 0.5898,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -1.1904714107513428,
      "rewards/margins": 0.5562640428543091,
      "rewards/rejected": -1.7467355728149414,
      "step": 90
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8220596619089573e-07,
      "logits/chosen": 0.1400291472673416,
      "logits/rejected": 0.6031023263931274,
      "logps/chosen": -544.8568115234375,
      "logps/pi_response": -222.86630249023438,
      "logps/ref_response": -191.95008850097656,
      "logps/rejected": -661.2197265625,
      "loss": 0.566,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -1.1847233772277832,
      "rewards/margins": 0.5735085010528564,
      "rewards/rejected": -1.7582318782806396,
      "step": 100
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.3139467229135998e-07,
      "logits/chosen": 0.13480663299560547,
      "logits/rejected": 0.6124667525291443,
      "logps/chosen": -551.712158203125,
      "logps/pi_response": -201.6997833251953,
      "logps/ref_response": -172.76661682128906,
      "logps/rejected": -627.7642211914062,
      "loss": 0.5799,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -1.042403221130371,
      "rewards/margins": 0.4938434660434723,
      "rewards/rejected": -1.536246657371521,
      "step": 110
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.628481651367875e-08,
      "logits/chosen": 0.29530060291290283,
      "logits/rejected": 0.561973512172699,
      "logps/chosen": -507.9276428222656,
      "logps/pi_response": -198.93344116210938,
      "logps/ref_response": -166.47335815429688,
      "logps/rejected": -658.5081787109375,
      "loss": 0.5442,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -1.0466843843460083,
      "rewards/margins": 0.6271711587905884,
      "rewards/rejected": -1.6738557815551758,
      "step": 120
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.904486005914027e-08,
      "logits/chosen": 0.4397956430912018,
      "logits/rejected": 0.7937263250350952,
      "logps/chosen": -523.13525390625,
      "logps/pi_response": -213.9718017578125,
      "logps/ref_response": -178.47569274902344,
      "logps/rejected": -653.7128295898438,
      "loss": 0.5582,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -1.0310872793197632,
      "rewards/margins": 0.6268772482872009,
      "rewards/rejected": -1.6579644680023193,
      "step": 130
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.1464952759020856e-08,
      "logits/chosen": 0.6194955706596375,
      "logits/rejected": 1.0747044086456299,
      "logps/chosen": -514.37744140625,
      "logps/pi_response": -208.5031280517578,
      "logps/ref_response": -176.0931396484375,
      "logps/rejected": -671.9991455078125,
      "loss": 0.5577,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -1.111130952835083,
      "rewards/margins": 0.6132029294967651,
      "rewards/rejected": -1.7243340015411377,
      "step": 140
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.8708793644441086e-09,
      "logits/chosen": 0.08109824359416962,
      "logits/rejected": 0.7494246959686279,
      "logps/chosen": -524.765625,
      "logps/pi_response": -212.7974090576172,
      "logps/ref_response": -180.5318145751953,
      "logps/rejected": -664.6768798828125,
      "loss": 0.5556,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.9752191305160522,
      "rewards/margins": 0.7606547474861145,
      "rewards/rejected": -1.735873818397522,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 159,
      "total_flos": 0.0,
      "train_loss": 0.5931253643155848,
      "train_runtime": 4224.9045,
      "train_samples_per_second": 4.823,
      "train_steps_per_second": 0.038
    }
  ],
  "logging_steps": 10,
  "max_steps": 159,
  "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
}