Hugging Face's logo

YYYYYYibo
/
two_agent_rdpo_iter_2

Text Generation
Transformers
Safetensors
mistral
alignment-handbook
trl
dpo
Generated from Trainer
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
two_agent_rdpo_iter_2
File size: 11,020 Bytes 
8097c02
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
 
{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9945,
  "eval_steps": 500,
  "global_step": 153,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "eta": 0.0010000000474974513,
      "grad_norm": 18.06976070111927,
      "learning_rate": 3.125e-08,
      "logits/chosen": -2.1194543838500977,
      "logits/rejected": -2.2610020637512207,
      "logps/chosen": -254.6973419189453,
      "logps/pi_response": -318.5512390136719,
      "logps/ref_response": -318.5512390136719,
      "logps/rejected": -224.19918823242188,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.07,
      "eta": 0.0010000000474974513,
      "grad_norm": 16.01968550655723,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -2.385725975036621,
      "logits/rejected": -2.3076765537261963,
      "logps/chosen": -218.1858673095703,
      "logps/pi_response": -268.26239013671875,
      "logps/ref_response": -266.3502197265625,
      "logps/rejected": -224.54605102539062,
      "loss": 0.6927,
      "rewards/accuracies": 0.44871795177459717,
      "rewards/chosen": -0.006226478144526482,
      "rewards/margins": 0.0014922022819519043,
      "rewards/rejected": -0.007718680426478386,
      "step": 10
    },
    {
      "epoch": 0.13,
      "eta": 0.0010000000474974513,
      "grad_norm": 16.02312687332099,
      "learning_rate": 4.989490450759331e-07,
      "logits/chosen": -2.397501230239868,
      "logits/rejected": -2.3145182132720947,
      "logps/chosen": -228.54135131835938,
      "logps/pi_response": -300.3511047363281,
      "logps/ref_response": -260.5257873535156,
      "logps/rejected": -246.08518981933594,
      "loss": 0.6874,
      "rewards/accuracies": 0.5230769515037537,
      "rewards/chosen": -0.22692929208278656,
      "rewards/margins": 0.027445880696177483,
      "rewards/rejected": -0.254375159740448,
      "step": 20
    },
    {
      "epoch": 0.2,
      "eta": 0.0010000000474974513,
      "grad_norm": 18.496071851642174,
      "learning_rate": 4.872270441827174e-07,
      "logits/chosen": -2.2651801109313965,
      "logits/rejected": -2.206923007965088,
      "logps/chosen": -264.01336669921875,
      "logps/pi_response": -342.78717041015625,
      "logps/ref_response": -260.0426940917969,
      "logps/rejected": -266.82000732421875,
      "loss": 0.696,
      "rewards/accuracies": 0.4730769097805023,
      "rewards/chosen": -0.6173264980316162,
      "rewards/margins": 0.023560278117656708,
      "rewards/rejected": -0.6408867835998535,
      "step": 30
    },
    {
      "epoch": 0.26,
      "eta": 0.0010000000474974513,
      "grad_norm": 17.411127458960003,
      "learning_rate": 4.6308512113530063e-07,
      "logits/chosen": -2.4821219444274902,
      "logits/rejected": -2.384908437728882,
      "logps/chosen": -243.66514587402344,
      "logps/pi_response": -308.2298583984375,
      "logps/ref_response": -255.49522399902344,
      "logps/rejected": -256.4287109375,
      "loss": 0.6903,
      "rewards/accuracies": 0.5538461804389954,
      "rewards/chosen": -0.3519066572189331,
      "rewards/margins": 0.04566844180226326,
      "rewards/rejected": -0.3975750505924225,
      "step": 40
    },
    {
      "epoch": 0.33,
      "eta": 0.0010000000474974513,
      "grad_norm": 14.517352316678341,
      "learning_rate": 4.277872161641681e-07,
      "logits/chosen": -2.545145273208618,
      "logits/rejected": -2.5242159366607666,
      "logps/chosen": -247.29806518554688,
      "logps/pi_response": -293.09326171875,
      "logps/ref_response": -275.4074401855469,
      "logps/rejected": -244.6110076904297,
      "loss": 0.6891,
      "rewards/accuracies": 0.557692289352417,
      "rewards/chosen": -0.1450691968202591,
      "rewards/margins": 0.017832614481449127,
      "rewards/rejected": -0.16290180385112762,
      "step": 50
    },
    {
      "epoch": 0.39,
      "eta": 0.0010000000474974513,
      "grad_norm": 17.626546770334993,
      "learning_rate": 3.8318133624280046e-07,
      "logits/chosen": -2.511488676071167,
      "logits/rejected": -2.480903148651123,
      "logps/chosen": -257.10791015625,
      "logps/pi_response": -302.22210693359375,
      "logps/ref_response": -270.9198303222656,
      "logps/rejected": -252.54173278808594,
      "loss": 0.6912,
      "rewards/accuracies": 0.48846152424812317,
      "rewards/chosen": -0.23286302387714386,
      "rewards/margins": 0.02372700721025467,
      "rewards/rejected": -0.2565900385379791,
      "step": 60
    },
    {
      "epoch": 0.46,
      "eta": 0.0010000000474974513,
      "grad_norm": 15.787453453048895,
      "learning_rate": 3.316028034595861e-07,
      "logits/chosen": -2.3820903301239014,
      "logits/rejected": -2.3555104732513428,
      "logps/chosen": -250.5824432373047,
      "logps/pi_response": -322.5419921875,
      "logps/ref_response": -270.929931640625,
      "logps/rejected": -260.3626403808594,
      "loss": 0.695,
      "rewards/accuracies": 0.557692289352417,
      "rewards/chosen": -0.37925985455513,
      "rewards/margins": 0.030029216781258583,
      "rewards/rejected": -0.40928906202316284,
      "step": 70
    },
    {
      "epoch": 0.52,
      "eta": 0.0010000000474974513,
      "grad_norm": 14.876878099596093,
      "learning_rate": 2.7575199021178855e-07,
      "logits/chosen": -2.463655948638916,
      "logits/rejected": -2.3770523071289062,
      "logps/chosen": -270.34197998046875,
      "logps/pi_response": -317.0186767578125,
      "logps/ref_response": -278.0060119628906,
      "logps/rejected": -269.89398193359375,
      "loss": 0.682,
      "rewards/accuracies": 0.5538461804389954,
      "rewards/chosen": -0.3352661728858948,
      "rewards/margins": 0.037235379219055176,
      "rewards/rejected": -0.37250155210494995,
      "step": 80
    },
    {
      "epoch": 0.58,
      "eta": 0.0010000000474974513,
      "grad_norm": 18.475251229694116,
      "learning_rate": 2.1855294234408068e-07,
      "logits/chosen": -2.37835693359375,
      "logits/rejected": -2.3291523456573486,
      "logps/chosen": -244.36837768554688,
      "logps/pi_response": -331.8450012207031,
      "logps/ref_response": -267.61846923828125,
      "logps/rejected": -280.1258544921875,
      "loss": 0.6786,
      "rewards/accuracies": 0.5692307949066162,
      "rewards/chosen": -0.39472243189811707,
      "rewards/margins": 0.06756081432104111,
      "rewards/rejected": -0.46228325366973877,
      "step": 90
    },
    {
      "epoch": 0.65,
      "eta": 0.0010000000474974513,
      "grad_norm": 21.467183952290178,
      "learning_rate": 1.6300029195778453e-07,
      "logits/chosen": -2.353746175765991,
      "logits/rejected": -2.2994236946105957,
      "logps/chosen": -259.2685546875,
      "logps/pi_response": -356.6775817871094,
      "logps/ref_response": -270.3107604980469,
      "logps/rejected": -271.2638854980469,
      "loss": 0.6687,
      "rewards/accuracies": 0.5769230723381042,
      "rewards/chosen": -0.4701143503189087,
      "rewards/margins": 0.08206918090581894,
      "rewards/rejected": -0.5521835088729858,
      "step": 100
    },
    {
      "epoch": 0.71,
      "eta": 0.0010000000474974513,
      "grad_norm": 22.52617676998604,
      "learning_rate": 1.1200247470632392e-07,
      "logits/chosen": -2.292710542678833,
      "logits/rejected": -2.357100486755371,
      "logps/chosen": -274.7545471191406,
      "logps/pi_response": -387.603515625,
      "logps/ref_response": -285.7787780761719,
      "logps/rejected": -272.741943359375,
      "loss": 0.6847,
      "rewards/accuracies": 0.4923076927661896,
      "rewards/chosen": -0.5492157936096191,
      "rewards/margins": 0.018456529825925827,
      "rewards/rejected": -0.5676723718643188,
      "step": 110
    },
    {
      "epoch": 0.78,
      "eta": 0.0010000000474974513,
      "grad_norm": 23.08131985031319,
      "learning_rate": 6.822945986946385e-08,
      "logits/chosen": -2.2610812187194824,
      "logits/rejected": -2.1860787868499756,
      "logps/chosen": -282.33782958984375,
      "logps/pi_response": -372.6535949707031,
      "logps/ref_response": -265.4132080078125,
      "logps/rejected": -290.5935363769531,
      "loss": 0.6808,
      "rewards/accuracies": 0.5461538434028625,
      "rewards/chosen": -0.5871608257293701,
      "rewards/margins": 0.05175128951668739,
      "rewards/rejected": -0.6389120817184448,
      "step": 120
    },
    {
      "epoch": 0.84,
      "eta": 0.0010000000474974513,
      "grad_norm": 22.961865819374246,
      "learning_rate": 3.397296523427806e-08,
      "logits/chosen": -2.2791786193847656,
      "logits/rejected": -2.31884503364563,
      "logps/chosen": -277.06085205078125,
      "logps/pi_response": -365.2495422363281,
      "logps/ref_response": -260.3804626464844,
      "logps/rejected": -295.7681884765625,
      "loss": 0.672,
      "rewards/accuracies": 0.5653846263885498,
      "rewards/chosen": -0.6707223057746887,
      "rewards/margins": 0.08594530820846558,
      "rewards/rejected": -0.7566676139831543,
      "step": 130
    },
    {
      "epoch": 0.91,
      "eta": 0.0010000000474974513,
      "grad_norm": 22.470744853394315,
      "learning_rate": 1.1026475173977978e-08,
      "logits/chosen": -2.3486456871032715,
      "logits/rejected": -2.201983690261841,
      "logps/chosen": -294.6012268066406,
      "logps/pi_response": -374.5638732910156,
      "logps/ref_response": -269.6314697265625,
      "logps/rejected": -284.3193359375,
      "loss": 0.6824,
      "rewards/accuracies": 0.5653846263885498,
      "rewards/chosen": -0.683001697063446,
      "rewards/margins": 0.06894499808549881,
      "rewards/rejected": -0.751946747303009,
      "step": 140
    },
    {
      "epoch": 0.97,
      "eta": 0.0010000000474974513,
      "grad_norm": 26.50699233707175,
      "learning_rate": 5.913435276374834e-10,
      "logits/chosen": -2.3918538093566895,
      "logits/rejected": -2.34391450881958,
      "logps/chosen": -279.7143859863281,
      "logps/pi_response": -387.8644104003906,
      "logps/ref_response": -273.84423828125,
      "logps/rejected": -304.1023254394531,
      "loss": 0.6713,
      "rewards/accuracies": 0.6153846383094788,
      "rewards/chosen": -0.6463515758514404,
      "rewards/margins": 0.1319323182106018,
      "rewards/rejected": -0.7782838940620422,
      "step": 150
    },
    {
      "epoch": 0.99,
      "step": 153,
      "total_flos": 0.0,
      "train_loss": 0.6840069897813734,
      "train_runtime": 41065.6381,
      "train_samples_per_second": 0.487,
      "train_steps_per_second": 0.004
    }
  ],
  "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": 1,
  "trial_name": null,
  "trial_params": null
}