Hugging Face's logo

miyako6105
/
Qwen3-32B-ORPO-test

Text Generation
Transformers
Safetensors
qwen3
Generated from Trainer
open-r1
trl
orpo
conversational
text-generation-inference
Model card Files
xet
Community
miyako6105 commited on
Commit
3569ad2
·
verified ·
1 Parent(s): 2979848

Model save

Browse files
Files changed (4) hide show
  1. all_results.json +9 -0
  2. generation_config.json +10 -0
  3. train_results.json +9 -0
  4. trainer_state.json +133 -0
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.019436345966958212,
3
+ "total_flos": 0.0,
4
+ "train_loss": 2.112666893005371,
5
+ "train_runtime": 260.1894,
6
+ "train_samples": 49375,
7
+ "train_samples_per_second": 3.69,
8
+ "train_steps_per_second": 0.019
9
+ }
generation_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 151643,
3
+ "do_sample": true,
4
+ "eos_token_id": 151645,
5
+ "pad_token_id": 151643,
6
+ "temperature": 0.6,
7
+ "top_k": 20,
8
+ "top_p": 0.95,
9
+ "transformers_version": "4.54.1"
10
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.019436345966958212,
3
+ "total_flos": 0.0,
4
+ "train_loss": 2.112666893005371,
5
+ "train_runtime": 260.1894,
6
+ "train_samples": 49375,
7
+ "train_samples_per_second": 3.69,
8
+ "train_steps_per_second": 0.019
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,133 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.019436345966958212,
6
+ "eval_steps": 500,
7
+ "global_step": 5,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.003887269193391642,
14
+ "grad_norm": 8.51073697097684,
15
+ "learning_rate": 0.0,
16
+ "log_odds_chosen": 0.4093017578125,
17
+ "log_odds_ratio": -0.6591796875,
18
+ "logits/chosen": 1.3271484375,
19
+ "logits/rejected": -1.109375,
20
+ "logps/chosen": -2.451171875,
21
+ "logps/rejected": -2.826171875,
22
+ "loss": 2.5861,
23
+ "nll_loss": 2.521484375,
24
+ "rewards/accuracies": 0.661458432674408,
25
+ "rewards/chosen": -0.2449951171875,
26
+ "rewards/margins": 0.03772735595703125,
27
+ "rewards/rejected": -0.2825927734375,
28
+ "step": 1
29
+ },
30
+ {
31
+ "epoch": 0.007774538386783284,
32
+ "grad_norm": 8.185303894379656,
33
+ "learning_rate": 2e-05,
34
+ "log_odds_chosen": 0.496826171875,
35
+ "log_odds_ratio": -0.69140625,
36
+ "logits/chosen": 1.3232421875,
37
+ "logits/rejected": -1.0576171875,
38
+ "logps/chosen": -2.431640625,
39
+ "logps/rejected": -2.9140625,
40
+ "loss": 2.5667,
41
+ "nll_loss": 2.49609375,
42
+ "rewards/accuracies": 0.6458333134651184,
43
+ "rewards/chosen": -0.2430419921875,
44
+ "rewards/margins": 0.048309326171875,
45
+ "rewards/rejected": -0.291259765625,
46
+ "step": 2
47
+ },
48
+ {
49
+ "epoch": 0.011661807580174927,
50
+ "grad_norm": 5.074940327572319,
51
+ "learning_rate": 1.736396103067893e-05,
52
+ "log_odds_chosen": 0.789306640625,
53
+ "log_odds_ratio": -0.4697265625,
54
+ "logits/chosen": 0.429443359375,
55
+ "logits/rejected": -1.03173828125,
56
+ "logps/chosen": -1.962890625,
57
+ "logps/rejected": -2.6796875,
58
+ "loss": 2.128,
59
+ "nll_loss": 2.0810546875,
60
+ "rewards/accuracies": 0.7864583730697632,
61
+ "rewards/chosen": -0.1962890625,
62
+ "rewards/margins": 0.071533203125,
63
+ "rewards/rejected": -0.267822265625,
64
+ "step": 3
65
+ },
66
+ {
67
+ "epoch": 0.015549076773566569,
68
+ "grad_norm": 3.76757860778528,
69
+ "learning_rate": 1.1000000000000001e-05,
70
+ "log_odds_chosen": 1.40673828125,
71
+ "log_odds_ratio": -0.316162109375,
72
+ "logits/chosen": 1.06201171875,
73
+ "logits/rejected": -1.263671875,
74
+ "logps/chosen": -1.5263671875,
75
+ "logps/rejected": -2.7578125,
76
+ "loss": 1.6897,
77
+ "nll_loss": 1.6591796875,
78
+ "rewards/accuracies": 0.9166667461395264,
79
+ "rewards/chosen": -0.152587890625,
80
+ "rewards/margins": 0.12322998046875,
81
+ "rewards/rejected": -0.2757568359375,
82
+ "step": 4
83
+ },
84
+ {
85
+ "epoch": 0.019436345966958212,
86
+ "grad_norm": 3.555226780695627,
87
+ "learning_rate": 4.636038969321073e-06,
88
+ "log_odds_chosen": 1.5322265625,
89
+ "log_odds_ratio": -0.2637939453125,
90
+ "logits/chosen": 0.80712890625,
91
+ "logits/rejected": -1.6494140625,
92
+ "logps/chosen": -1.4091796875,
93
+ "logps/rejected": -2.720703125,
94
+ "loss": 1.5929,
95
+ "nll_loss": 1.56640625,
96
+ "rewards/accuracies": 0.96875,
97
+ "rewards/chosen": -0.14105224609375,
98
+ "rewards/margins": 0.131591796875,
99
+ "rewards/rejected": -0.2725830078125,
100
+ "step": 5
101
+ },
102
+ {
103
+ "epoch": 0.019436345966958212,
104
+ "step": 5,
105
+ "total_flos": 0.0,
106
+ "train_loss": 2.112666893005371,
107
+ "train_runtime": 260.1894,
108
+ "train_samples_per_second": 3.69,
109
+ "train_steps_per_second": 0.019
110
+ }
111
+ ],
112
+ "logging_steps": 1,
113
+ "max_steps": 5,
114
+ "num_input_tokens_seen": 0,
115
+ "num_train_epochs": 1,
116
+ "save_steps": 500,
117
+ "stateful_callbacks": {
118
+ "TrainerControl": {
119
+ "args": {
120
+ "should_epoch_stop": false,
121
+ "should_evaluate": false,
122
+ "should_log": false,
123
+ "should_save": true,
124
+ "should_training_stop": true
125
+ },
126
+ "attributes": {}
127
+ }
128
+ },
129
+ "total_flos": 0.0,
130
+ "train_batch_size": 1,
131
+ "trial_name": null,
132
+ "trial_params": null
133
+ }