Hugging Face's logo

AmberYifan
/
Llama-3.1-8B-Mistral-7B-v0.3-mix

Text Generation
Transformers
Safetensors
llama
Generated from Trainer
trl
dpo
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
AmberYifan commited on
Commit
5b7fba2
·
verified ·
1 Parent(s): 595d5a7

Training in progress, epoch 1, checkpoint

Browse files
Files changed (17) hide show
  1. last-checkpoint/global_step313/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
  2. last-checkpoint/global_step313/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
  3. last-checkpoint/global_step313/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +3 -0
  4. last-checkpoint/global_step313/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +3 -0
  5. last-checkpoint/global_step313/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
  6. last-checkpoint/global_step313/zero_pp_rank_1_mp_rank_00_model_states.pt +3 -0
  7. last-checkpoint/global_step313/zero_pp_rank_2_mp_rank_00_model_states.pt +3 -0
  8. last-checkpoint/global_step313/zero_pp_rank_3_mp_rank_00_model_states.pt +3 -0
  9. last-checkpoint/latest +1 -1
  10. last-checkpoint/model-00001-of-00004.safetensors +1 -1
  11. last-checkpoint/model-00002-of-00004.safetensors +1 -1
  12. last-checkpoint/model-00003-of-00004.safetensors +1 -1
  13. last-checkpoint/model-00004-of-00004.safetensors +1 -1
  14. last-checkpoint/scheduler.pt +1 -1
  15. last-checkpoint/tokenizer_config.json +5 -1
  16. last-checkpoint/trainer_state.json +469 -94
  17. last-checkpoint/training_args.bin +1 -1
last-checkpoint/global_step313/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f1d9689f2d0ae9f68a8f6539d51c2d7e670c16188c74bbe0992a6bfc87b4c8f
3
+ size 16060527788
last-checkpoint/global_step313/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:df7d85ada27417feaf42c89d7c7fba88263f2b69b78d9cfc726b5f4a3dabed2e
3
+ size 16060527788
last-checkpoint/global_step313/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1ddae1c9df2055d0d15e51a52cc3d19cb8c5c187859490959e5687dddbbb0f2
3
+ size 16060527788
last-checkpoint/global_step313/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3497debe6f45b7aef09f7dba1e0aace30f7e64b49d2d15793db679b009c5ddb5
3
+ size 16060527788
last-checkpoint/global_step313/zero_pp_rank_0_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17eed26ba27472ce711147ef53e0c1021013065243d4cdfd5b3dbaf0908918d0
3
+ size 150693
last-checkpoint/global_step313/zero_pp_rank_1_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:75fba867bcf326c87113723c73497ab82dc56e55ff12f4121ac20ef4a86775d7
3
+ size 150693
last-checkpoint/global_step313/zero_pp_rank_2_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb9431acd6ed41df752d6ef9c54ed239402ddee455ae1f7fe51bc63e56f6ae3a
3
+ size 150693
last-checkpoint/global_step313/zero_pp_rank_3_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a90127d04eed8748d360b3c545c66a8557d9845b7acaac61b591ad660515d071
3
+ size 150693
last-checkpoint/latest CHANGED
@@ -1 +1 @@
1
- global_step63
 
1
+ global_step313
last-checkpoint/model-00001-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ade34906b4cbdc902daa864eccbadf2a8cd52b90a0c82c00e5ea078e8ec76711
3
  size 4976698672
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e43ccfbf09c4ed8af08feda5bcae4e3997daf4d9e193ea0455aea2851ea65f1
3
  size 4976698672
last-checkpoint/model-00002-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:412ac5e7b41b49f159f1ab0e79fec711d071236d0c039b9213a6e0e970bac99c
3
  size 4999802720
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:091f8e1bd9cba29d3dd77cf70431df45558323bcb81475f0cbe3939b90192125
3
  size 4999802720
last-checkpoint/model-00003-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cf0acadd41f37958f2d32c0e85d94b5f67f43915a39704444b1b9040fa49ec51
3
  size 4915916176
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e54fac96c6aa2e0495935eac8ca6ffff6770cdc037c904f9cb5bd493f787da8
3
  size 4915916176
last-checkpoint/model-00004-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:75cb0755a39def127f63de3fa5df3890ca0820f19b74340bb4adcb75c6f7b9f9
3
  size 1168138808
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a2262b675f2d413afb23fc20133126e1ef385833b70462ddbd7a17b8eb9cd4d0
3
  size 1168138808
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:845f8eea760ff90e22dadf5fb7f80c9079f8c2298a8770c5d4c3b3309b38a6ef
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b9aad3a748e15719e20120f0a760f22686338e176ee3f1c011b6dab2f3ab19b
3
  size 1064
last-checkpoint/tokenizer_config.json CHANGED
@@ -2053,11 +2053,15 @@
2053
  "chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
2054
  "clean_up_tokenization_spaces": true,
2055
  "eos_token": "<|end_of_text|>",
 
2056
  "model_input_names": [
2057
  "input_ids",
2058
  "attention_mask"
2059
  ],
2060
  "model_max_length": 2048,
2061
  "pad_token": "<|end_of_text|>",
2062
- "tokenizer_class": "PreTrainedTokenizerFast"
 
 
 
2063
  }
 
2053
  "chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
2054
  "clean_up_tokenization_spaces": true,
2055
  "eos_token": "<|end_of_text|>",
2056
+ "max_length": 1024,
2057
  "model_input_names": [
2058
  "input_ids",
2059
  "attention_mask"
2060
  ],
2061
  "model_max_length": 2048,
2062
  "pad_token": "<|end_of_text|>",
2063
+ "stride": 0,
2064
+ "tokenizer_class": "PreTrainedTokenizerFast",
2065
+ "truncation_side": "left",
2066
+ "truncation_strategy": "longest_first"
2067
  }
last-checkpoint/trainer_state.json CHANGED
@@ -3,19 +3,19 @@
3
  "best_model_checkpoint": null,
4
  "epoch": 1.0,
5
  "eval_steps": 500,
6
- "global_step": 63,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 0.015873015873015872,
13
- "grad_norm": 2.956210802841563,
14
- "learning_rate": 2.6315789473684208e-08,
15
- "logits/chosen": -0.84765625,
16
- "logits/rejected": -0.84765625,
17
- "logps/chosen": -14.875,
18
- "logps/rejected": -14.875,
19
  "loss": 0.6914,
20
  "rewards/accuracies": 0.0,
21
  "rewards/chosen": 0.0,
@@ -24,114 +24,489 @@
24
  "step": 1
25
  },
26
  {
27
- "epoch": 0.15873015873015872,
28
- "grad_norm": 0.0026515697903897284,
29
- "learning_rate": 2.631578947368421e-07,
30
- "logits/chosen": -0.8828125,
31
- "logits/rejected": -0.88671875,
32
- "logps/chosen": -15.9375,
33
- "logps/rejected": -15.8125,
34
- "loss": 0.6914,
35
- "rewards/accuracies": 0.02777777798473835,
36
- "rewards/chosen": 0.0008697509765625,
37
- "rewards/margins": -0.00043487548828125,
38
- "rewards/rejected": 0.00130462646484375,
39
  "step": 10
40
  },
41
  {
42
- "epoch": 0.31746031746031744,
43
- "grad_norm": 2.201768075236467,
44
- "learning_rate": 4.970588235294118e-07,
45
- "logits/chosen": -0.890625,
46
- "logits/rejected": -0.88671875,
47
- "logps/chosen": -15.625,
48
- "logps/rejected": -15.5625,
49
- "loss": 0.6894,
50
- "rewards/accuracies": 0.07500000298023224,
51
- "rewards/chosen": 0.0228271484375,
52
- "rewards/margins": 0.00860595703125,
53
- "rewards/rejected": 0.01422119140625,
54
  "step": 20
55
  },
56
  {
57
- "epoch": 0.47619047619047616,
58
- "grad_norm": 2.7819271307964892,
59
- "learning_rate": 4.676470588235294e-07,
60
- "logits/chosen": -0.890625,
61
- "logits/rejected": -0.875,
62
- "logps/chosen": -14.9375,
63
- "logps/rejected": -14.75,
64
- "loss": 0.689,
65
- "rewards/accuracies": 0.0625,
66
- "rewards/chosen": 0.0830078125,
67
- "rewards/margins": 0.0111083984375,
68
- "rewards/rejected": 0.07177734375,
69
  "step": 30
70
  },
71
  {
72
- "epoch": 0.6349206349206349,
73
- "grad_norm": 2.468434193598462,
74
- "learning_rate": 4.38235294117647e-07,
75
- "logits/chosen": -0.8984375,
76
- "logits/rejected": -0.91015625,
77
- "logps/chosen": -14.375,
78
- "logps/rejected": -14.8125,
79
- "loss": 0.6849,
80
- "rewards/accuracies": 0.11249999701976776,
81
- "rewards/chosen": 0.1552734375,
82
- "rewards/margins": 0.038330078125,
83
- "rewards/rejected": 0.11669921875,
84
  "step": 40
85
  },
86
  {
87
- "epoch": 0.7936507936507936,
88
- "grad_norm": 3.217396320056085,
89
- "learning_rate": 4.0882352941176465e-07,
90
- "logits/chosen": -0.87109375,
91
- "logits/rejected": -0.8828125,
92
- "logps/chosen": -13.125,
93
- "logps/rejected": -12.625,
94
- "loss": 0.6828,
95
- "rewards/accuracies": 0.03750000149011612,
96
- "rewards/chosen": 0.267578125,
97
- "rewards/margins": 0.01129150390625,
98
- "rewards/rejected": 0.255859375,
99
  "step": 50
100
  },
101
  {
102
- "epoch": 0.9523809523809523,
103
- "grad_norm": 3.5576289077982315,
104
- "learning_rate": 3.7941176470588235e-07,
105
- "logits/chosen": -0.85546875,
106
- "logits/rejected": -0.8671875,
107
- "logps/chosen": -12.125,
108
- "logps/rejected": -12.625,
109
- "loss": 0.6772,
110
- "rewards/accuracies": 0.02500000037252903,
111
- "rewards/chosen": 0.3359375,
112
- "rewards/margins": 0.0196533203125,
113
- "rewards/rejected": 0.31640625,
114
  "step": 60
115
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
116
  {
117
  "epoch": 1.0,
118
- "eval_logits/chosen": -0.875,
119
- "eval_logits/rejected": -0.87109375,
120
- "eval_logps/chosen": -10.25,
121
- "eval_logps/rejected": -10.625,
122
- "eval_loss": 0.6818749904632568,
123
- "eval_rewards/accuracies": 0.0357142873108387,
124
- "eval_rewards/chosen": 0.458984375,
125
- "eval_rewards/margins": -0.0004177093505859375,
126
- "eval_rewards/rejected": 0.4609375,
127
- "eval_runtime": 12.3547,
128
- "eval_samples_per_second": 16.188,
129
- "eval_steps_per_second": 0.567,
130
- "step": 63
131
  }
132
  ],
133
  "logging_steps": 10,
134
- "max_steps": 189,
135
  "num_input_tokens_seen": 0,
136
  "num_train_epochs": 3,
137
  "save_steps": 500,
 
3
  "best_model_checkpoint": null,
4
  "epoch": 1.0,
5
  "eval_steps": 500,
6
+ "global_step": 313,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 0.003194888178913738,
13
+ "grad_norm": 34.47548268439031,
14
+ "learning_rate": 5.3191489361702125e-09,
15
+ "logits/chosen": -1.6953125,
16
+ "logits/rejected": -1.640625,
17
+ "logps/chosen": -235.0,
18
+ "logps/rejected": -262.0,
19
  "loss": 0.6914,
20
  "rewards/accuracies": 0.0,
21
  "rewards/chosen": 0.0,
 
24
  "step": 1
25
  },
26
  {
27
+ "epoch": 0.03194888178913738,
28
+ "grad_norm": 33.08858843822245,
29
+ "learning_rate": 5.3191489361702123e-08,
30
+ "logits/chosen": -1.7890625,
31
+ "logits/rejected": -1.7734375,
32
+ "logps/chosen": -245.0,
33
+ "logps/rejected": -272.0,
34
+ "loss": 0.6967,
35
+ "rewards/accuracies": 0.1805555522441864,
36
+ "rewards/chosen": -0.004180908203125,
37
+ "rewards/margins": -0.006988525390625,
38
+ "rewards/rejected": 0.002777099609375,
39
  "step": 10
40
  },
41
  {
42
+ "epoch": 0.06389776357827476,
43
+ "grad_norm": 35.17419870896503,
44
+ "learning_rate": 1.0638297872340425e-07,
45
+ "logits/chosen": -1.7578125,
46
+ "logits/rejected": -1.765625,
47
+ "logps/chosen": -247.0,
48
+ "logps/rejected": -256.0,
49
+ "loss": 0.6896,
50
+ "rewards/accuracies": 0.22499999403953552,
51
+ "rewards/chosen": 0.003753662109375,
52
+ "rewards/margins": 0.01129150390625,
53
+ "rewards/rejected": -0.00750732421875,
54
  "step": 20
55
  },
56
  {
57
+ "epoch": 0.09584664536741214,
58
+ "grad_norm": 36.64639475774933,
59
+ "learning_rate": 1.5957446808510638e-07,
60
+ "logits/chosen": -1.796875,
61
+ "logits/rejected": -1.8046875,
62
+ "logps/chosen": -242.0,
63
+ "logps/rejected": -264.0,
64
+ "loss": 0.6837,
65
+ "rewards/accuracies": 0.3125,
66
+ "rewards/chosen": 0.0012664794921875,
67
+ "rewards/margins": -0.0012054443359375,
68
+ "rewards/rejected": 0.00250244140625,
69
  "step": 30
70
  },
71
  {
72
+ "epoch": 0.12779552715654952,
73
+ "grad_norm": 35.2711796102512,
74
+ "learning_rate": 2.127659574468085e-07,
75
+ "logits/chosen": -1.7421875,
76
+ "logits/rejected": -1.7421875,
77
+ "logps/chosen": -246.0,
78
+ "logps/rejected": -264.0,
79
+ "loss": 0.663,
80
+ "rewards/accuracies": 0.5375000238418579,
81
+ "rewards/chosen": 0.038818359375,
82
+ "rewards/margins": 0.08154296875,
83
+ "rewards/rejected": -0.04248046875,
84
  "step": 40
85
  },
86
  {
87
+ "epoch": 0.1597444089456869,
88
+ "grad_norm": 31.56733318399803,
89
+ "learning_rate": 2.659574468085106e-07,
90
+ "logits/chosen": -1.75,
91
+ "logits/rejected": -1.71875,
92
+ "logps/chosen": -238.0,
93
+ "logps/rejected": -266.0,
94
+ "loss": 0.65,
95
+ "rewards/accuracies": 0.675000011920929,
96
+ "rewards/chosen": 0.04248046875,
97
+ "rewards/margins": 0.1826171875,
98
+ "rewards/rejected": -0.140625,
99
  "step": 50
100
  },
101
  {
102
+ "epoch": 0.19169329073482427,
103
+ "grad_norm": 32.09814766896921,
104
+ "learning_rate": 3.1914893617021275e-07,
105
+ "logits/chosen": -1.75,
106
+ "logits/rejected": -1.7265625,
107
+ "logps/chosen": -236.0,
108
+ "logps/rejected": -268.0,
109
+ "loss": 0.619,
110
+ "rewards/accuracies": 0.675000011920929,
111
+ "rewards/chosen": 0.053955078125,
112
+ "rewards/margins": 0.2099609375,
113
+ "rewards/rejected": -0.15625,
114
  "step": 60
115
  },
116
+ {
117
+ "epoch": 0.22364217252396165,
118
+ "grad_norm": 32.03938813135862,
119
+ "learning_rate": 3.7234042553191484e-07,
120
+ "logits/chosen": -1.765625,
121
+ "logits/rejected": -1.7578125,
122
+ "logps/chosen": -242.0,
123
+ "logps/rejected": -264.0,
124
+ "loss": 0.611,
125
+ "rewards/accuracies": 0.625,
126
+ "rewards/chosen": 0.0277099609375,
127
+ "rewards/margins": 0.255859375,
128
+ "rewards/rejected": -0.2275390625,
129
+ "step": 70
130
+ },
131
+ {
132
+ "epoch": 0.25559105431309903,
133
+ "grad_norm": 43.96835624603879,
134
+ "learning_rate": 4.25531914893617e-07,
135
+ "logits/chosen": -1.7109375,
136
+ "logits/rejected": -1.75,
137
+ "logps/chosen": -243.0,
138
+ "logps/rejected": -262.0,
139
+ "loss": 0.6006,
140
+ "rewards/accuracies": 0.6625000238418579,
141
+ "rewards/chosen": -0.103515625,
142
+ "rewards/margins": 0.392578125,
143
+ "rewards/rejected": -0.49609375,
144
+ "step": 80
145
+ },
146
+ {
147
+ "epoch": 0.28753993610223644,
148
+ "grad_norm": 28.877852885313988,
149
+ "learning_rate": 4.787234042553192e-07,
150
+ "logits/chosen": -1.7421875,
151
+ "logits/rejected": -1.71875,
152
+ "logps/chosen": -245.0,
153
+ "logps/rejected": -272.0,
154
+ "loss": 0.5725,
155
+ "rewards/accuracies": 0.5874999761581421,
156
+ "rewards/chosen": -0.03369140625,
157
+ "rewards/margins": 0.34765625,
158
+ "rewards/rejected": -0.380859375,
159
+ "step": 90
160
+ },
161
+ {
162
+ "epoch": 0.3194888178913738,
163
+ "grad_norm": 33.04150137611911,
164
+ "learning_rate": 4.964497041420119e-07,
165
+ "logits/chosen": -1.75,
166
+ "logits/rejected": -1.734375,
167
+ "logps/chosen": -240.0,
168
+ "logps/rejected": -268.0,
169
+ "loss": 0.6132,
170
+ "rewards/accuracies": 0.762499988079071,
171
+ "rewards/chosen": 0.08544921875,
172
+ "rewards/margins": 0.5078125,
173
+ "rewards/rejected": -0.423828125,
174
+ "step": 100
175
+ },
176
+ {
177
+ "epoch": 0.3514376996805112,
178
+ "grad_norm": 31.292790431153033,
179
+ "learning_rate": 4.905325443786982e-07,
180
+ "logits/chosen": -1.71875,
181
+ "logits/rejected": -1.7421875,
182
+ "logps/chosen": -241.0,
183
+ "logps/rejected": -260.0,
184
+ "loss": 0.5318,
185
+ "rewards/accuracies": 0.7250000238418579,
186
+ "rewards/chosen": 0.07421875,
187
+ "rewards/margins": 0.51953125,
188
+ "rewards/rejected": -0.443359375,
189
+ "step": 110
190
+ },
191
+ {
192
+ "epoch": 0.38338658146964855,
193
+ "grad_norm": 28.540124264789696,
194
+ "learning_rate": 4.846153846153846e-07,
195
+ "logits/chosen": -1.734375,
196
+ "logits/rejected": -1.703125,
197
+ "logps/chosen": -238.0,
198
+ "logps/rejected": -268.0,
199
+ "loss": 0.5829,
200
+ "rewards/accuracies": 0.6875,
201
+ "rewards/chosen": 0.0263671875,
202
+ "rewards/margins": 0.57421875,
203
+ "rewards/rejected": -0.546875,
204
+ "step": 120
205
+ },
206
+ {
207
+ "epoch": 0.41533546325878595,
208
+ "grad_norm": 29.08477368681588,
209
+ "learning_rate": 4.78698224852071e-07,
210
+ "logits/chosen": -1.765625,
211
+ "logits/rejected": -1.7265625,
212
+ "logps/chosen": -245.0,
213
+ "logps/rejected": -274.0,
214
+ "loss": 0.5456,
215
+ "rewards/accuracies": 0.7250000238418579,
216
+ "rewards/chosen": -0.0712890625,
217
+ "rewards/margins": 0.52734375,
218
+ "rewards/rejected": -0.59765625,
219
+ "step": 130
220
+ },
221
+ {
222
+ "epoch": 0.4472843450479233,
223
+ "grad_norm": 24.895122773023758,
224
+ "learning_rate": 4.727810650887574e-07,
225
+ "logits/chosen": -1.6796875,
226
+ "logits/rejected": -1.7109375,
227
+ "logps/chosen": -240.0,
228
+ "logps/rejected": -268.0,
229
+ "loss": 0.5276,
230
+ "rewards/accuracies": 0.6875,
231
+ "rewards/chosen": -0.0252685546875,
232
+ "rewards/margins": 0.67578125,
233
+ "rewards/rejected": -0.69921875,
234
+ "step": 140
235
+ },
236
+ {
237
+ "epoch": 0.4792332268370607,
238
+ "grad_norm": 25.086960468247135,
239
+ "learning_rate": 4.668639053254438e-07,
240
+ "logits/chosen": -1.71875,
241
+ "logits/rejected": -1.7265625,
242
+ "logps/chosen": -236.0,
243
+ "logps/rejected": -264.0,
244
+ "loss": 0.555,
245
+ "rewards/accuracies": 0.762499988079071,
246
+ "rewards/chosen": 0.0751953125,
247
+ "rewards/margins": 0.66796875,
248
+ "rewards/rejected": -0.59375,
249
+ "step": 150
250
+ },
251
+ {
252
+ "epoch": 0.5111821086261981,
253
+ "grad_norm": 28.17100030607696,
254
+ "learning_rate": 4.6094674556213014e-07,
255
+ "logits/chosen": -1.7109375,
256
+ "logits/rejected": -1.75,
257
+ "logps/chosen": -244.0,
258
+ "logps/rejected": -264.0,
259
+ "loss": 0.5135,
260
+ "rewards/accuracies": 0.7124999761581421,
261
+ "rewards/chosen": -0.0162353515625,
262
+ "rewards/margins": 0.609375,
263
+ "rewards/rejected": -0.625,
264
+ "step": 160
265
+ },
266
+ {
267
+ "epoch": 0.5431309904153354,
268
+ "grad_norm": 26.942036465103737,
269
+ "learning_rate": 4.5502958579881655e-07,
270
+ "logits/chosen": -1.75,
271
+ "logits/rejected": -1.7265625,
272
+ "logps/chosen": -238.0,
273
+ "logps/rejected": -268.0,
274
+ "loss": 0.5144,
275
+ "rewards/accuracies": 0.800000011920929,
276
+ "rewards/chosen": 0.1201171875,
277
+ "rewards/margins": 0.875,
278
+ "rewards/rejected": -0.75390625,
279
+ "step": 170
280
+ },
281
+ {
282
+ "epoch": 0.5750798722044729,
283
+ "grad_norm": 27.057109325150893,
284
+ "learning_rate": 4.491124260355029e-07,
285
+ "logits/chosen": -1.703125,
286
+ "logits/rejected": -1.7265625,
287
+ "logps/chosen": -237.0,
288
+ "logps/rejected": -284.0,
289
+ "loss": 0.4916,
290
+ "rewards/accuracies": 0.7749999761581421,
291
+ "rewards/chosen": 0.09130859375,
292
+ "rewards/margins": 0.94140625,
293
+ "rewards/rejected": -0.8515625,
294
+ "step": 180
295
+ },
296
+ {
297
+ "epoch": 0.6070287539936102,
298
+ "grad_norm": 30.82037140778774,
299
+ "learning_rate": 4.4319526627218936e-07,
300
+ "logits/chosen": -1.7109375,
301
+ "logits/rejected": -1.703125,
302
+ "logps/chosen": -242.0,
303
+ "logps/rejected": -266.0,
304
+ "loss": 0.54,
305
+ "rewards/accuracies": 0.6625000238418579,
306
+ "rewards/chosen": 0.09130859375,
307
+ "rewards/margins": 0.69140625,
308
+ "rewards/rejected": -0.59765625,
309
+ "step": 190
310
+ },
311
+ {
312
+ "epoch": 0.6389776357827476,
313
+ "grad_norm": 24.497963916643677,
314
+ "learning_rate": 4.372781065088757e-07,
315
+ "logits/chosen": -1.78125,
316
+ "logits/rejected": -1.7265625,
317
+ "logps/chosen": -245.0,
318
+ "logps/rejected": -268.0,
319
+ "loss": 0.5101,
320
+ "rewards/accuracies": 0.762499988079071,
321
+ "rewards/chosen": 0.06884765625,
322
+ "rewards/margins": 0.73828125,
323
+ "rewards/rejected": -0.671875,
324
+ "step": 200
325
+ },
326
+ {
327
+ "epoch": 0.670926517571885,
328
+ "grad_norm": 30.284876833786186,
329
+ "learning_rate": 4.313609467455621e-07,
330
+ "logits/chosen": -1.7265625,
331
+ "logits/rejected": -1.75,
332
+ "logps/chosen": -242.0,
333
+ "logps/rejected": -270.0,
334
+ "loss": 0.551,
335
+ "rewards/accuracies": 0.737500011920929,
336
+ "rewards/chosen": -0.107421875,
337
+ "rewards/margins": 0.78515625,
338
+ "rewards/rejected": -0.89453125,
339
+ "step": 210
340
+ },
341
+ {
342
+ "epoch": 0.7028753993610224,
343
+ "grad_norm": 31.01459795666554,
344
+ "learning_rate": 4.2544378698224847e-07,
345
+ "logits/chosen": -1.734375,
346
+ "logits/rejected": -1.6953125,
347
+ "logps/chosen": -237.0,
348
+ "logps/rejected": -266.0,
349
+ "loss": 0.5036,
350
+ "rewards/accuracies": 0.6875,
351
+ "rewards/chosen": -0.10888671875,
352
+ "rewards/margins": 0.68359375,
353
+ "rewards/rejected": -0.7890625,
354
+ "step": 220
355
+ },
356
+ {
357
+ "epoch": 0.7348242811501597,
358
+ "grad_norm": 31.681398259902394,
359
+ "learning_rate": 4.195266272189349e-07,
360
+ "logits/chosen": -1.71875,
361
+ "logits/rejected": -1.703125,
362
+ "logps/chosen": -243.0,
363
+ "logps/rejected": -264.0,
364
+ "loss": 0.5146,
365
+ "rewards/accuracies": 0.7250000238418579,
366
+ "rewards/chosen": -0.0986328125,
367
+ "rewards/margins": 0.7421875,
368
+ "rewards/rejected": -0.83984375,
369
+ "step": 230
370
+ },
371
+ {
372
+ "epoch": 0.7667731629392971,
373
+ "grad_norm": 30.3843807531397,
374
+ "learning_rate": 4.1360946745562133e-07,
375
+ "logits/chosen": -1.7109375,
376
+ "logits/rejected": -1.7265625,
377
+ "logps/chosen": -242.0,
378
+ "logps/rejected": -272.0,
379
+ "loss": 0.4829,
380
+ "rewards/accuracies": 0.7124999761581421,
381
+ "rewards/chosen": -0.0751953125,
382
+ "rewards/margins": 0.75,
383
+ "rewards/rejected": -0.828125,
384
+ "step": 240
385
+ },
386
+ {
387
+ "epoch": 0.7987220447284346,
388
+ "grad_norm": 27.627369180135428,
389
+ "learning_rate": 4.076923076923077e-07,
390
+ "logits/chosen": -1.7421875,
391
+ "logits/rejected": -1.734375,
392
+ "logps/chosen": -240.0,
393
+ "logps/rejected": -266.0,
394
+ "loss": 0.4946,
395
+ "rewards/accuracies": 0.675000011920929,
396
+ "rewards/chosen": -0.26171875,
397
+ "rewards/margins": 0.6328125,
398
+ "rewards/rejected": -0.890625,
399
+ "step": 250
400
+ },
401
+ {
402
+ "epoch": 0.8306709265175719,
403
+ "grad_norm": 28.61228168078618,
404
+ "learning_rate": 4.017751479289941e-07,
405
+ "logits/chosen": -1.7421875,
406
+ "logits/rejected": -1.765625,
407
+ "logps/chosen": -240.0,
408
+ "logps/rejected": -270.0,
409
+ "loss": 0.4946,
410
+ "rewards/accuracies": 0.737500011920929,
411
+ "rewards/chosen": -0.111328125,
412
+ "rewards/margins": 0.8671875,
413
+ "rewards/rejected": -0.98046875,
414
+ "step": 260
415
+ },
416
+ {
417
+ "epoch": 0.8626198083067093,
418
+ "grad_norm": 33.147902615790635,
419
+ "learning_rate": 3.9585798816568044e-07,
420
+ "logits/chosen": -1.765625,
421
+ "logits/rejected": -1.71875,
422
+ "logps/chosen": -237.0,
423
+ "logps/rejected": -276.0,
424
+ "loss": 0.5384,
425
+ "rewards/accuracies": 0.7250000238418579,
426
+ "rewards/chosen": -0.099609375,
427
+ "rewards/margins": 0.65234375,
428
+ "rewards/rejected": -0.75,
429
+ "step": 270
430
+ },
431
+ {
432
+ "epoch": 0.8945686900958466,
433
+ "grad_norm": 27.58369117700387,
434
+ "learning_rate": 3.8994082840236685e-07,
435
+ "logits/chosen": -1.703125,
436
+ "logits/rejected": -1.734375,
437
+ "logps/chosen": -241.0,
438
+ "logps/rejected": -260.0,
439
+ "loss": 0.5174,
440
+ "rewards/accuracies": 0.737500011920929,
441
+ "rewards/chosen": 0.09912109375,
442
+ "rewards/margins": 0.7109375,
443
+ "rewards/rejected": -0.609375,
444
+ "step": 280
445
+ },
446
+ {
447
+ "epoch": 0.9265175718849841,
448
+ "grad_norm": 27.845142754279863,
449
+ "learning_rate": 3.840236686390532e-07,
450
+ "logits/chosen": -1.671875,
451
+ "logits/rejected": -1.71875,
452
+ "logps/chosen": -238.0,
453
+ "logps/rejected": -264.0,
454
+ "loss": 0.4757,
455
+ "rewards/accuracies": 0.7875000238418579,
456
+ "rewards/chosen": 0.0177001953125,
457
+ "rewards/margins": 0.984375,
458
+ "rewards/rejected": -0.96875,
459
+ "step": 290
460
+ },
461
+ {
462
+ "epoch": 0.9584664536741214,
463
+ "grad_norm": 32.815897540130955,
464
+ "learning_rate": 3.7810650887573966e-07,
465
+ "logits/chosen": -1.71875,
466
+ "logits/rejected": -1.7421875,
467
+ "logps/chosen": -238.0,
468
+ "logps/rejected": -276.0,
469
+ "loss": 0.4941,
470
+ "rewards/accuracies": 0.737500011920929,
471
+ "rewards/chosen": -0.228515625,
472
+ "rewards/margins": 1.0,
473
+ "rewards/rejected": -1.2265625,
474
+ "step": 300
475
+ },
476
+ {
477
+ "epoch": 0.9904153354632588,
478
+ "grad_norm": 28.360981127394723,
479
+ "learning_rate": 3.72189349112426e-07,
480
+ "logits/chosen": -1.7421875,
481
+ "logits/rejected": -1.7578125,
482
+ "logps/chosen": -248.0,
483
+ "logps/rejected": -274.0,
484
+ "loss": 0.4374,
485
+ "rewards/accuracies": 0.800000011920929,
486
+ "rewards/chosen": -0.2099609375,
487
+ "rewards/margins": 1.3359375,
488
+ "rewards/rejected": -1.5390625,
489
+ "step": 310
490
+ },
491
  {
492
  "epoch": 1.0,
493
+ "eval_logits/chosen": -1.7109375,
494
+ "eval_logits/rejected": -1.7421875,
495
+ "eval_logps/chosen": -240.0,
496
+ "eval_logps/rejected": -278.0,
497
+ "eval_loss": 0.5235937237739563,
498
+ "eval_rewards/accuracies": 0.7142857313156128,
499
+ "eval_rewards/chosen": -0.326171875,
500
+ "eval_rewards/margins": 0.63671875,
501
+ "eval_rewards/rejected": -0.9609375,
502
+ "eval_runtime": 12.6819,
503
+ "eval_samples_per_second": 15.771,
504
+ "eval_steps_per_second": 0.552,
505
+ "step": 313
506
  }
507
  ],
508
  "logging_steps": 10,
509
+ "max_steps": 939,
510
  "num_input_tokens_seen": 0,
511
  "num_train_epochs": 3,
512
  "save_steps": 500,
last-checkpoint/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3671c42fe1669bb58ad27cf3d2da0ac69f3c5a100754abbfaf825b7a3321a2bd
3
  size 7864
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27657f99ecbe3e90addf13347d0446d96ca2b8f78352849b6ec80d9bc2a67640
3
  size 7864