File size: 7,270 Bytes
9e1e448
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9966024915062288,
  "eval_steps": 1000,
  "global_step": 110,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009060022650056626,
      "grad_norm": 2.142748189776569,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": -2.2157700061798096,
      "logits/rejected": -2.1868345737457275,
      "logps/chosen": -314.38787841796875,
      "logps/rejected": -291.1216735839844,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.09060022650056625,
      "grad_norm": 2.150777513425362,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.2591588497161865,
      "logits/rejected": -2.233074188232422,
      "logps/chosen": -323.4332275390625,
      "logps/rejected": -301.56719970703125,
      "loss": 0.6929,
      "rewards/accuracies": 0.4713541567325592,
      "rewards/chosen": 0.0009152439888566732,
      "rewards/margins": 0.000497353496029973,
      "rewards/rejected": 0.00041789052193053067,
      "step": 10
    },
    {
      "epoch": 0.1812004530011325,
      "grad_norm": 2.066434141751538,
      "learning_rate": 4.898732434036243e-07,
      "logits/chosen": -2.2525153160095215,
      "logits/rejected": -2.2378909587860107,
      "logps/chosen": -325.8227233886719,
      "logps/rejected": -305.6023254394531,
      "loss": 0.6861,
      "rewards/accuracies": 0.7789062261581421,
      "rewards/chosen": 0.0187881700694561,
      "rewards/margins": 0.014950500801205635,
      "rewards/rejected": 0.0038376704324036837,
      "step": 20
    },
    {
      "epoch": 0.2718006795016987,
      "grad_norm": 1.935984789640625,
      "learning_rate": 4.5591914535745817e-07,
      "logits/chosen": -2.1835825443267822,
      "logits/rejected": -2.173578977584839,
      "logps/chosen": -313.19586181640625,
      "logps/rejected": -300.5938720703125,
      "loss": 0.6661,
      "rewards/accuracies": 0.8046875,
      "rewards/chosen": 0.04926218464970589,
      "rewards/margins": 0.05489668250083923,
      "rewards/rejected": -0.005634505767375231,
      "step": 30
    },
    {
      "epoch": 0.362400906002265,
      "grad_norm": 2.2867184944223293,
      "learning_rate": 4.0140242178441665e-07,
      "logits/chosen": -2.1189799308776855,
      "logits/rejected": -2.114716053009033,
      "logps/chosen": -317.4134826660156,
      "logps/rejected": -306.8409423828125,
      "loss": 0.6458,
      "rewards/accuracies": 0.8046875,
      "rewards/chosen": 0.03598688170313835,
      "rewards/margins": 0.09652377665042877,
      "rewards/rejected": -0.06053689122200012,
      "step": 40
    },
    {
      "epoch": 0.45300113250283125,
      "grad_norm": 2.40453710231171,
      "learning_rate": 3.317669908293554e-07,
      "logits/chosen": -1.9152988195419312,
      "logits/rejected": -1.9253225326538086,
      "logps/chosen": -321.01800537109375,
      "logps/rejected": -334.33349609375,
      "loss": 0.5959,
      "rewards/accuracies": 0.8109375238418579,
      "rewards/chosen": -0.13218708336353302,
      "rewards/margins": 0.22598442435264587,
      "rewards/rejected": -0.3581715524196625,
      "step": 50
    },
    {
      "epoch": 0.5436013590033975,
      "grad_norm": 2.548257476263302,
      "learning_rate": 2.53966490958702e-07,
      "logits/chosen": -1.8517974615097046,
      "logits/rejected": -1.851008653640747,
      "logps/chosen": -349.90252685546875,
      "logps/rejected": -362.4208984375,
      "loss": 0.5612,
      "rewards/accuracies": 0.815625011920929,
      "rewards/chosen": -0.2689761817455292,
      "rewards/margins": 0.33339887857437134,
      "rewards/rejected": -0.6023750305175781,
      "step": 60
    },
    {
      "epoch": 0.6342015855039638,
      "grad_norm": 2.564454636474787,
      "learning_rate": 1.7576990616793137e-07,
      "logits/chosen": -1.8310279846191406,
      "logits/rejected": -1.8546864986419678,
      "logps/chosen": -362.08538818359375,
      "logps/rejected": -394.1742858886719,
      "loss": 0.5353,
      "rewards/accuracies": 0.803906261920929,
      "rewards/chosen": -0.4547205865383148,
      "rewards/margins": 0.4052005410194397,
      "rewards/rejected": -0.8599211573600769,
      "step": 70
    },
    {
      "epoch": 0.72480181200453,
      "grad_norm": 2.5983243283209183,
      "learning_rate": 1.0498577260720048e-07,
      "logits/chosen": -1.8076627254486084,
      "logits/rejected": -1.8311001062393188,
      "logps/chosen": -375.9088134765625,
      "logps/rejected": -410.9624938964844,
      "loss": 0.5223,
      "rewards/accuracies": 0.782031238079071,
      "rewards/chosen": -0.5610671639442444,
      "rewards/margins": 0.4681544303894043,
      "rewards/rejected": -1.029221534729004,
      "step": 80
    },
    {
      "epoch": 0.8154020385050963,
      "grad_norm": 2.742604490623868,
      "learning_rate": 4.868243561723534e-08,
      "logits/chosen": -1.809565544128418,
      "logits/rejected": -1.8266630172729492,
      "logps/chosen": -381.6236267089844,
      "logps/rejected": -415.415771484375,
      "loss": 0.513,
      "rewards/accuracies": 0.792187511920929,
      "rewards/chosen": -0.6585050225257874,
      "rewards/margins": 0.4940672814846039,
      "rewards/rejected": -1.1525723934173584,
      "step": 90
    },
    {
      "epoch": 0.9060022650056625,
      "grad_norm": 2.4312610291944887,
      "learning_rate": 1.2482220564763667e-08,
      "logits/chosen": -1.8194091320037842,
      "logits/rejected": -1.811342477798462,
      "logps/chosen": -386.8002624511719,
      "logps/rejected": -417.9185485839844,
      "loss": 0.508,
      "rewards/accuracies": 0.817187488079071,
      "rewards/chosen": -0.6888748407363892,
      "rewards/margins": 0.544217050075531,
      "rewards/rejected": -1.2330917119979858,
      "step": 100
    },
    {
      "epoch": 0.9966024915062288,
      "grad_norm": 2.609201337420324,
      "learning_rate": 0.0,
      "logits/chosen": -1.8002452850341797,
      "logits/rejected": -1.796565294265747,
      "logps/chosen": -391.4776611328125,
      "logps/rejected": -427.072265625,
      "loss": 0.506,
      "rewards/accuracies": 0.8046875,
      "rewards/chosen": -0.7028344869613647,
      "rewards/margins": 0.5477779507637024,
      "rewards/rejected": -1.250612497329712,
      "step": 110
    },
    {
      "epoch": 0.9966024915062288,
      "step": 110,
      "total_flos": 0.0,
      "train_loss": 0.5847668994556774,
      "train_runtime": 2901.4735,
      "train_samples_per_second": 38.945,
      "train_steps_per_second": 0.038
    }
  ],
  "logging_steps": 10,
  "max_steps": 110,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}