File size: 3,928 Bytes
bfeab83
 
 
 
 
 
 
 
 
 
 
 
d4fbdad
bfeab83
 
d4fbdad
 
 
 
 
 
 
 
 
bfeab83
 
 
 
d4fbdad
bfeab83
d4fbdad
 
 
 
 
 
 
 
 
bfeab83
 
 
 
d4fbdad
bfeab83
 
d4fbdad
 
 
 
 
 
 
 
 
bfeab83
 
 
 
d4fbdad
bfeab83
 
d4fbdad
 
 
 
 
 
 
 
 
bfeab83
 
 
 
d4fbdad
bfeab83
d4fbdad
 
 
 
 
 
 
 
 
bfeab83
 
 
 
d4fbdad
bfeab83
 
d4fbdad
 
 
 
 
 
 
 
 
bfeab83
 
 
 
 
 
 
 
 
 
 
 
 
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
{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8064,
  "eval_steps": 63,
  "global_step": 252,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2016,
      "eval_logps/chosen": -70.4515609741211,
      "eval_logps/ref_chosen": -70.265625,
      "eval_logps/ref_rejected": -63.51186752319336,
      "eval_logps/rejected": -70.4515609741211,
      "eval_loss": 0.700020968914032,
      "eval_rewards/accuracies": 0.5043316388630343,
      "eval_rewards/chosen": -0.009296582080423832,
      "eval_rewards/margins": -0.0003258216893300414,
      "eval_rewards/rejected": -0.00897076167166233,
      "eval_runtime": 2598.2788,
      "eval_samples_per_second": 32.253,
      "eval_steps_per_second": 4.032,
      "step": 63
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.734374523162842,
      "learning_rate": 2.3016987391917016e-05,
      "logps/chosen": -164.50128173828125,
      "logps/ref_chosen": -159.2122802734375,
      "logps/ref_rejected": -181.2625732421875,
      "logps/rejected": -164.50128173828125,
      "loss": 0.6682,
      "rewards/accuracies": 0.5696874856948853,
      "rewards/chosen": -0.26445069909095764,
      "rewards/margins": 0.10393363237380981,
      "rewards/rejected": -0.36838433146476746,
      "step": 100
    },
    {
      "epoch": 0.4032,
      "eval_logps/chosen": -70.71819305419922,
      "eval_logps/ref_chosen": -70.265625,
      "eval_logps/ref_rejected": -63.51186752319336,
      "eval_logps/rejected": -70.71819305419922,
      "eval_loss": 0.6713600158691406,
      "eval_rewards/accuracies": 0.5957017732273693,
      "eval_rewards/chosen": -0.022628214210271835,
      "eval_rewards/margins": 0.06350255757570267,
      "eval_rewards/rejected": -0.0861307755112648,
      "eval_runtime": 2598.0044,
      "eval_samples_per_second": 32.256,
      "eval_steps_per_second": 4.032,
      "step": 126
    },
    {
      "epoch": 0.6048,
      "eval_logps/chosen": -71.90546417236328,
      "eval_logps/ref_chosen": -70.265625,
      "eval_logps/ref_rejected": -63.51186752319336,
      "eval_logps/rejected": -71.90546417236328,
      "eval_loss": 0.6622863411903381,
      "eval_rewards/accuracies": 0.6141738860647717,
      "eval_rewards/chosen": -0.08199150860309601,
      "eval_rewards/margins": 0.08873386681079865,
      "eval_rewards/rejected": -0.17072536051273346,
      "eval_runtime": 2596.18,
      "eval_samples_per_second": 32.279,
      "eval_steps_per_second": 4.035,
      "step": 189
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.594386100769043,
      "learning_rate": 8.569611578954186e-06,
      "logps/chosen": -161.8042755126953,
      "logps/ref_chosen": -158.50271606445312,
      "logps/ref_rejected": -179.81715393066406,
      "logps/rejected": -161.8042755126953,
      "loss": 0.6423,
      "rewards/accuracies": 0.6114062666893005,
      "rewards/chosen": -0.16507746279239655,
      "rewards/margins": 0.20611368119716644,
      "rewards/rejected": -0.371191143989563,
      "step": 200
    },
    {
      "epoch": 0.8064,
      "eval_logps/chosen": -72.42017364501953,
      "eval_logps/ref_chosen": -70.265625,
      "eval_logps/ref_rejected": -63.51186752319336,
      "eval_logps/rejected": -72.42017364501953,
      "eval_loss": 0.6550462245941162,
      "eval_rewards/accuracies": 0.6268585475286986,
      "eval_rewards/chosen": -0.10772793740034103,
      "eval_rewards/margins": 0.10725179314613342,
      "eval_rewards/rejected": -0.21497976779937744,
      "eval_runtime": 2596.7824,
      "eval_samples_per_second": 32.271,
      "eval_steps_per_second": 4.034,
      "step": 252
    }
  ],
  "logging_steps": 100,
  "max_steps": 312,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 63,
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}