Hugging Face's logo

AdversarialRLHF
/
rloo_pythia410m_tldr6.9b_rm410mdata_allprefixsft_prefix

Safetensors
gpt_neox
Model card Files Community
Muqeeth commited on
Commit
4625eb7
·
verified ·
1 Parent(s): 3e68818

Training in progress, step 32, checkpoint

Browse files
Files changed (11) hide show
  1. checkpoint-32/config.json +31 -0
  2. checkpoint-32/generation_config.json +5 -0
  3. checkpoint-32/model.safetensors +3 -0
  4. checkpoint-32/optimizer.pt +3 -0
  5. checkpoint-32/rng_state.pth +3 -0
  6. checkpoint-32/scheduler.pt +3 -0
  7. checkpoint-32/special_tokens_map.json +30 -0
  8. checkpoint-32/tokenizer.json +0 -0
  9. checkpoint-32/tokenizer_config.json +216 -0
  10. checkpoint-32/trainer_state.json +675 -0
  11. checkpoint-32/training_args.bin +3 -0
checkpoint-32/config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "GPTNeoXForCausalLM"
4
+ ],
5
+ "attention_bias": true,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": 0.1,
9
+ "eos_token_id": 0,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout": 0.0,
12
+ "hidden_size": 1024,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 4096,
15
+ "layer_norm_eps": 1e-05,
16
+ "max_position_embeddings": 2048,
17
+ "model_type": "gpt_neox",
18
+ "num_attention_heads": 16,
19
+ "num_hidden_layers": 24,
20
+ "partial_rotary_factor": 0.25,
21
+ "rope_scaling": null,
22
+ "rope_theta": 10000,
23
+ "rotary_emb_base": 10000,
24
+ "rotary_pct": 0.25,
25
+ "tie_word_embeddings": false,
26
+ "torch_dtype": "float32",
27
+ "transformers_version": "4.50.3",
28
+ "use_cache": true,
29
+ "use_parallel_residual": true,
30
+ "vocab_size": 50304
31
+ }
checkpoint-32/generation_config.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "transformers_version": "4.50.3"
5
+ }
checkpoint-32/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4b5d97f7a8184c07c61f6b8065b218db44cadbceb17a29429ff689bcbb9f654
3
+ size 1621370224
checkpoint-32/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c54a7f47337bc22e12d44d9179b70887912e34ee7c9a39da3609bf3bc3f43b44
3
+ size 3242916026
checkpoint-32/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb93e8888f3bd7bfc7d97ee5a3432f068ecedb32e813f7162074a941b6bf0609
3
+ size 14308
checkpoint-32/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:def327218d564c242c83a06a2a6ec6d22e2e906119a059f4313f04b378d73de0
3
+ size 1064
checkpoint-32/special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|endoftext|>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|endoftext|>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<|padding|>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "<|endoftext|>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ }
30
+ }
checkpoint-32/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-32/tokenizer_config.json ADDED
@@ -0,0 +1,216 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_eos_token": false,
4
+ "add_prefix_space": false,
5
+ "added_tokens_decoder": {
6
+ "0": {
7
+ "content": "<|endoftext|>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false,
12
+ "special": true
13
+ },
14
+ "1": {
15
+ "content": "<|padding|>",
16
+ "lstrip": false,
17
+ "normalized": false,
18
+ "rstrip": false,
19
+ "single_word": false,
20
+ "special": true
21
+ },
22
+ "50254": {
23
+ "content": " ",
24
+ "lstrip": false,
25
+ "normalized": true,
26
+ "rstrip": false,
27
+ "single_word": false,
28
+ "special": false
29
+ },
30
+ "50255": {
31
+ "content": " ",
32
+ "lstrip": false,
33
+ "normalized": true,
34
+ "rstrip": false,
35
+ "single_word": false,
36
+ "special": false
37
+ },
38
+ "50256": {
39
+ "content": " ",
40
+ "lstrip": false,
41
+ "normalized": true,
42
+ "rstrip": false,
43
+ "single_word": false,
44
+ "special": false
45
+ },
46
+ "50257": {
47
+ "content": " ",
48
+ "lstrip": false,
49
+ "normalized": true,
50
+ "rstrip": false,
51
+ "single_word": false,
52
+ "special": false
53
+ },
54
+ "50258": {
55
+ "content": " ",
56
+ "lstrip": false,
57
+ "normalized": true,
58
+ "rstrip": false,
59
+ "single_word": false,
60
+ "special": false
61
+ },
62
+ "50259": {
63
+ "content": " ",
64
+ "lstrip": false,
65
+ "normalized": true,
66
+ "rstrip": false,
67
+ "single_word": false,
68
+ "special": false
69
+ },
70
+ "50260": {
71
+ "content": " ",
72
+ "lstrip": false,
73
+ "normalized": true,
74
+ "rstrip": false,
75
+ "single_word": false,
76
+ "special": false
77
+ },
78
+ "50261": {
79
+ "content": " ",
80
+ "lstrip": false,
81
+ "normalized": true,
82
+ "rstrip": false,
83
+ "single_word": false,
84
+ "special": false
85
+ },
86
+ "50262": {
87
+ "content": " ",
88
+ "lstrip": false,
89
+ "normalized": true,
90
+ "rstrip": false,
91
+ "single_word": false,
92
+ "special": false
93
+ },
94
+ "50263": {
95
+ "content": " ",
96
+ "lstrip": false,
97
+ "normalized": true,
98
+ "rstrip": false,
99
+ "single_word": false,
100
+ "special": false
101
+ },
102
+ "50264": {
103
+ "content": " ",
104
+ "lstrip": false,
105
+ "normalized": true,
106
+ "rstrip": false,
107
+ "single_word": false,
108
+ "special": false
109
+ },
110
+ "50265": {
111
+ "content": " ",
112
+ "lstrip": false,
113
+ "normalized": true,
114
+ "rstrip": false,
115
+ "single_word": false,
116
+ "special": false
117
+ },
118
+ "50266": {
119
+ "content": " ",
120
+ "lstrip": false,
121
+ "normalized": true,
122
+ "rstrip": false,
123
+ "single_word": false,
124
+ "special": false
125
+ },
126
+ "50267": {
127
+ "content": " ",
128
+ "lstrip": false,
129
+ "normalized": true,
130
+ "rstrip": false,
131
+ "single_word": false,
132
+ "special": false
133
+ },
134
+ "50268": {
135
+ "content": " ",
136
+ "lstrip": false,
137
+ "normalized": true,
138
+ "rstrip": false,
139
+ "single_word": false,
140
+ "special": false
141
+ },
142
+ "50269": {
143
+ "content": " ",
144
+ "lstrip": false,
145
+ "normalized": true,
146
+ "rstrip": false,
147
+ "single_word": false,
148
+ "special": false
149
+ },
150
+ "50270": {
151
+ "content": " ",
152
+ "lstrip": false,
153
+ "normalized": true,
154
+ "rstrip": false,
155
+ "single_word": false,
156
+ "special": false
157
+ },
158
+ "50271": {
159
+ "content": " ",
160
+ "lstrip": false,
161
+ "normalized": true,
162
+ "rstrip": false,
163
+ "single_word": false,
164
+ "special": false
165
+ },
166
+ "50272": {
167
+ "content": " ",
168
+ "lstrip": false,
169
+ "normalized": true,
170
+ "rstrip": false,
171
+ "single_word": false,
172
+ "special": false
173
+ },
174
+ "50273": {
175
+ "content": " ",
176
+ "lstrip": false,
177
+ "normalized": true,
178
+ "rstrip": false,
179
+ "single_word": false,
180
+ "special": false
181
+ },
182
+ "50274": {
183
+ "content": " ",
184
+ "lstrip": false,
185
+ "normalized": true,
186
+ "rstrip": false,
187
+ "single_word": false,
188
+ "special": false
189
+ },
190
+ "50275": {
191
+ "content": " ",
192
+ "lstrip": false,
193
+ "normalized": true,
194
+ "rstrip": false,
195
+ "single_word": false,
196
+ "special": false
197
+ },
198
+ "50276": {
199
+ "content": " ",
200
+ "lstrip": false,
201
+ "normalized": true,
202
+ "rstrip": false,
203
+ "single_word": false,
204
+ "special": false
205
+ }
206
+ },
207
+ "bos_token": "<|endoftext|>",
208
+ "clean_up_tokenization_spaces": false,
209
+ "eos_token": "<|endoftext|>",
210
+ "extra_special_tokens": {},
211
+ "model_max_length": 1000000000000000019884624838656,
212
+ "pad_token": "<|padding|>",
213
+ "padding_side": "left",
214
+ "tokenizer_class": "GPTNeoXTokenizer",
215
+ "unk_token": "<|endoftext|>"
216
+ }
checkpoint-32/trainer_state.json ADDED
@@ -0,0 +1,675 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "episode": 16384,
6
+ "epoch": 0.07018505825908156,
7
+ "eval_steps": 500,
8
+ "global_step": 32,
9
+ "is_hyper_param_search": false,
10
+ "is_local_process_zero": true,
11
+ "is_world_process_zero": true,
12
+ "log_history": [
13
+ {
14
+ "episode": 512,
15
+ "epoch": 0.0021932830705962986,
16
+ "eps": 6,
17
+ "loss/policy_avg": 0.18060211837291718,
18
+ "lr": 3e-06,
19
+ "objective/entropy": -44.39540100097656,
20
+ "objective/kl": 0.23749515414237976,
21
+ "objective/non_score_reward": 0.0,
22
+ "objective/rlhf_reward": 4.621584892272949,
23
+ "objective/scores": 4.621584892272949,
24
+ "policy/approxkl_avg": 0.09178640693426132,
25
+ "policy/clipfrac_avg": 0.271484375,
26
+ "policy/entropy_avg": 0.7151485681533813,
27
+ "step": 0,
28
+ "val/clipfrac_avg": 0.0,
29
+ "val/num_eos_tokens": 5024,
30
+ "val/ratio": 1.0001094341278076,
31
+ "val/ratio_var": 1.1996758075838443e-06
32
+ },
33
+ {
34
+ "episode": 1024,
35
+ "epoch": 0.004386566141192597,
36
+ "eps": 6,
37
+ "loss/policy_avg": 0.10823587328195572,
38
+ "lr": 2.9882812500000002e-06,
39
+ "objective/entropy": -40.33261489868164,
40
+ "objective/kl": 0.8562099933624268,
41
+ "objective/non_score_reward": 0.0,
42
+ "objective/rlhf_reward": 4.949611186981201,
43
+ "objective/scores": 4.949611186981201,
44
+ "policy/approxkl_avg": 0.09872549772262573,
45
+ "policy/clipfrac_avg": 0.29296875,
46
+ "policy/entropy_avg": 0.6909540891647339,
47
+ "step": 1,
48
+ "val/clipfrac_avg": 0.0,
49
+ "val/num_eos_tokens": 3988,
50
+ "val/ratio": 1.0000126361846924,
51
+ "val/ratio_var": 1.5850138197492925e-06
52
+ },
53
+ {
54
+ "episode": 1536,
55
+ "epoch": 0.006579849211788897,
56
+ "eps": 6,
57
+ "loss/policy_avg": 0.24130496382713318,
58
+ "lr": 2.9765625e-06,
59
+ "objective/entropy": -37.140586853027344,
60
+ "objective/kl": 2.1461260318756104,
61
+ "objective/non_score_reward": 0.0,
62
+ "objective/rlhf_reward": 5.433126449584961,
63
+ "objective/scores": 5.433126449584961,
64
+ "policy/approxkl_avg": 0.1151532381772995,
65
+ "policy/clipfrac_avg": 0.30078125,
66
+ "policy/entropy_avg": 0.6479897499084473,
67
+ "step": 2,
68
+ "val/clipfrac_avg": 0.0,
69
+ "val/num_eos_tokens": 2924,
70
+ "val/ratio": 0.999896764755249,
71
+ "val/ratio_var": 1.4944761232982273e-06
72
+ },
73
+ {
74
+ "episode": 2048,
75
+ "epoch": 0.008773132282385195,
76
+ "eps": 6,
77
+ "loss/policy_avg": 0.11201904714107513,
78
+ "lr": 2.96484375e-06,
79
+ "objective/entropy": -23.162431716918945,
80
+ "objective/kl": 4.706851005554199,
81
+ "objective/non_score_reward": 0.0,
82
+ "objective/rlhf_reward": 5.616954803466797,
83
+ "objective/scores": 5.616954803466797,
84
+ "policy/approxkl_avg": 0.13956432044506073,
85
+ "policy/clipfrac_avg": 0.365234375,
86
+ "policy/entropy_avg": 0.6664378643035889,
87
+ "step": 3,
88
+ "val/clipfrac_avg": 0.0,
89
+ "val/num_eos_tokens": 2146,
90
+ "val/ratio": 0.9996535778045654,
91
+ "val/ratio_var": 2.1856692455912707e-06
92
+ },
93
+ {
94
+ "episode": 2560,
95
+ "epoch": 0.010966415352981495,
96
+ "eps": 6,
97
+ "loss/policy_avg": 0.21055178344249725,
98
+ "lr": 2.953125e-06,
99
+ "objective/entropy": -17.498517990112305,
100
+ "objective/kl": 6.893625259399414,
101
+ "objective/non_score_reward": 0.0,
102
+ "objective/rlhf_reward": 5.975703716278076,
103
+ "objective/scores": 5.975703716278076,
104
+ "policy/approxkl_avg": 0.14695778489112854,
105
+ "policy/clipfrac_avg": 0.33203125,
106
+ "policy/entropy_avg": 0.6391204595565796,
107
+ "step": 4,
108
+ "val/clipfrac_avg": 0.0,
109
+ "val/num_eos_tokens": 1715,
110
+ "val/ratio": 0.9999969005584717,
111
+ "val/ratio_var": 2.4044477413553977e-06
112
+ },
113
+ {
114
+ "episode": 3072,
115
+ "epoch": 0.013159698423577794,
116
+ "eps": 6,
117
+ "loss/policy_avg": 0.28335219621658325,
118
+ "lr": 2.94140625e-06,
119
+ "objective/entropy": -16.186603546142578,
120
+ "objective/kl": 8.215164184570312,
121
+ "objective/non_score_reward": 0.0,
122
+ "objective/rlhf_reward": 6.314523696899414,
123
+ "objective/scores": 6.314523696899414,
124
+ "policy/approxkl_avg": 0.15406616032123566,
125
+ "policy/clipfrac_avg": 0.34375,
126
+ "policy/entropy_avg": 0.6234536170959473,
127
+ "step": 5,
128
+ "val/clipfrac_avg": 0.0,
129
+ "val/num_eos_tokens": 1743,
130
+ "val/ratio": 1.0000813007354736,
131
+ "val/ratio_var": 2.2491783511213725e-06
132
+ },
133
+ {
134
+ "episode": 3584,
135
+ "epoch": 0.015352981494174092,
136
+ "eps": 6,
137
+ "loss/policy_avg": 0.2581750154495239,
138
+ "lr": 2.9296875e-06,
139
+ "objective/entropy": -16.82525634765625,
140
+ "objective/kl": 10.012128829956055,
141
+ "objective/non_score_reward": 0.0,
142
+ "objective/rlhf_reward": 6.68141508102417,
143
+ "objective/scores": 6.68141508102417,
144
+ "policy/approxkl_avg": 0.15727216005325317,
145
+ "policy/clipfrac_avg": 0.306640625,
146
+ "policy/entropy_avg": 0.6279575824737549,
147
+ "step": 6,
148
+ "val/clipfrac_avg": 0.0,
149
+ "val/num_eos_tokens": 2024,
150
+ "val/ratio": 1.0000149011611938,
151
+ "val/ratio_var": 3.910876330337487e-06
152
+ },
153
+ {
154
+ "episode": 4096,
155
+ "epoch": 0.01754626456477039,
156
+ "eps": 6,
157
+ "loss/policy_avg": 0.4508611857891083,
158
+ "lr": 2.91796875e-06,
159
+ "objective/entropy": -13.531968116760254,
160
+ "objective/kl": 11.674339294433594,
161
+ "objective/non_score_reward": 0.0,
162
+ "objective/rlhf_reward": 6.797825813293457,
163
+ "objective/scores": 6.797825813293457,
164
+ "policy/approxkl_avg": 0.17787906527519226,
165
+ "policy/clipfrac_avg": 0.333984375,
166
+ "policy/entropy_avg": 0.6140714287757874,
167
+ "step": 7,
168
+ "val/clipfrac_avg": 0.0,
169
+ "val/num_eos_tokens": 2248,
170
+ "val/ratio": 0.9998712539672852,
171
+ "val/ratio_var": 2.3673501345911063e-06
172
+ },
173
+ {
174
+ "episode": 4608,
175
+ "epoch": 0.01973954763536669,
176
+ "eps": 6,
177
+ "loss/policy_avg": 0.25109419226646423,
178
+ "lr": 2.90625e-06,
179
+ "objective/entropy": -15.348489761352539,
180
+ "objective/kl": 12.88166332244873,
181
+ "objective/non_score_reward": 0.0,
182
+ "objective/rlhf_reward": 7.097723484039307,
183
+ "objective/scores": 7.097723484039307,
184
+ "policy/approxkl_avg": 0.18050235509872437,
185
+ "policy/clipfrac_avg": 0.302734375,
186
+ "policy/entropy_avg": 0.6190727949142456,
187
+ "step": 8,
188
+ "val/clipfrac_avg": 0.0,
189
+ "val/num_eos_tokens": 2698,
190
+ "val/ratio": 1.000298023223877,
191
+ "val/ratio_var": 2.175341933252639e-06
192
+ },
193
+ {
194
+ "episode": 5120,
195
+ "epoch": 0.02193283070596299,
196
+ "eps": 6,
197
+ "loss/policy_avg": 0.2478555142879486,
198
+ "lr": 2.89453125e-06,
199
+ "objective/entropy": -20.807220458984375,
200
+ "objective/kl": 12.126293182373047,
201
+ "objective/non_score_reward": 0.0,
202
+ "objective/rlhf_reward": 7.294116020202637,
203
+ "objective/scores": 7.294116020202637,
204
+ "policy/approxkl_avg": 0.16978076100349426,
205
+ "policy/clipfrac_avg": 0.31640625,
206
+ "policy/entropy_avg": 0.5916793346405029,
207
+ "step": 9,
208
+ "val/clipfrac_avg": 0.0,
209
+ "val/num_eos_tokens": 3166,
210
+ "val/ratio": 0.999906063079834,
211
+ "val/ratio_var": 2.3764121124258963e-06
212
+ },
213
+ {
214
+ "episode": 5632,
215
+ "epoch": 0.024126113776559287,
216
+ "eps": 6,
217
+ "loss/policy_avg": 0.2194754183292389,
218
+ "lr": 2.8828125e-06,
219
+ "objective/entropy": -23.919815063476562,
220
+ "objective/kl": 13.048431396484375,
221
+ "objective/non_score_reward": 0.0,
222
+ "objective/rlhf_reward": 7.783618927001953,
223
+ "objective/scores": 7.783618927001953,
224
+ "policy/approxkl_avg": 0.1719370186328888,
225
+ "policy/clipfrac_avg": 0.3359375,
226
+ "policy/entropy_avg": 0.6023513078689575,
227
+ "step": 10,
228
+ "val/clipfrac_avg": 0.0,
229
+ "val/num_eos_tokens": 3395,
230
+ "val/ratio": 0.9998449683189392,
231
+ "val/ratio_var": 1.784112214409106e-06
232
+ },
233
+ {
234
+ "episode": 6144,
235
+ "epoch": 0.026319396847155587,
236
+ "eps": 6,
237
+ "loss/policy_avg": 0.13436169922351837,
238
+ "lr": 2.87109375e-06,
239
+ "objective/entropy": -21.395191192626953,
240
+ "objective/kl": 15.173453330993652,
241
+ "objective/non_score_reward": 0.0,
242
+ "objective/rlhf_reward": 7.682435512542725,
243
+ "objective/scores": 7.682435512542725,
244
+ "policy/approxkl_avg": 0.19314789772033691,
245
+ "policy/clipfrac_avg": 0.369140625,
246
+ "policy/entropy_avg": 0.6016405820846558,
247
+ "step": 11,
248
+ "val/clipfrac_avg": 0.0,
249
+ "val/num_eos_tokens": 3629,
250
+ "val/ratio": 0.9996920824050903,
251
+ "val/ratio_var": 1.8341333998250775e-06
252
+ },
253
+ {
254
+ "episode": 6656,
255
+ "epoch": 0.028512679917751884,
256
+ "eps": 6,
257
+ "loss/policy_avg": 0.27384719252586365,
258
+ "lr": 2.859375e-06,
259
+ "objective/entropy": -23.660961151123047,
260
+ "objective/kl": 15.459722518920898,
261
+ "objective/non_score_reward": 0.0,
262
+ "objective/rlhf_reward": 8.303886413574219,
263
+ "objective/scores": 8.303886413574219,
264
+ "policy/approxkl_avg": 0.1933055967092514,
265
+ "policy/clipfrac_avg": 0.359375,
266
+ "policy/entropy_avg": 0.5889922976493835,
267
+ "step": 12,
268
+ "val/clipfrac_avg": 0.0,
269
+ "val/num_eos_tokens": 3434,
270
+ "val/ratio": 1.0000567436218262,
271
+ "val/ratio_var": 2.589923496998381e-06
272
+ },
273
+ {
274
+ "episode": 7168,
275
+ "epoch": 0.030705962988348184,
276
+ "eps": 6,
277
+ "loss/policy_avg": 0.20046143233776093,
278
+ "lr": 2.84765625e-06,
279
+ "objective/entropy": -24.642715454101562,
280
+ "objective/kl": 17.541698455810547,
281
+ "objective/non_score_reward": 0.0,
282
+ "objective/rlhf_reward": 8.798042297363281,
283
+ "objective/scores": 8.798042297363281,
284
+ "policy/approxkl_avg": 0.1510610282421112,
285
+ "policy/clipfrac_avg": 0.337890625,
286
+ "policy/entropy_avg": 0.552618145942688,
287
+ "step": 13,
288
+ "val/clipfrac_avg": 0.0,
289
+ "val/num_eos_tokens": 3045,
290
+ "val/ratio": 1.0000343322753906,
291
+ "val/ratio_var": 2.4570517780375667e-06
292
+ },
293
+ {
294
+ "episode": 7680,
295
+ "epoch": 0.03289924605894448,
296
+ "eps": 6,
297
+ "loss/policy_avg": 0.22966830432415009,
298
+ "lr": 2.8359375e-06,
299
+ "objective/entropy": -19.635862350463867,
300
+ "objective/kl": 18.844629287719727,
301
+ "objective/non_score_reward": 0.0,
302
+ "objective/rlhf_reward": 8.394830703735352,
303
+ "objective/scores": 8.394830703735352,
304
+ "policy/approxkl_avg": 0.1538720726966858,
305
+ "policy/clipfrac_avg": 0.3359375,
306
+ "policy/entropy_avg": 0.5296885371208191,
307
+ "step": 14,
308
+ "val/clipfrac_avg": 0.0,
309
+ "val/num_eos_tokens": 2402,
310
+ "val/ratio": 1.0000078678131104,
311
+ "val/ratio_var": 2.192526608268963e-06
312
+ },
313
+ {
314
+ "episode": 8192,
315
+ "epoch": 0.03509252912954078,
316
+ "eps": 6,
317
+ "loss/policy_avg": 0.2523574233055115,
318
+ "lr": 2.82421875e-06,
319
+ "objective/entropy": -19.664382934570312,
320
+ "objective/kl": 19.9348201751709,
321
+ "objective/non_score_reward": 0.0,
322
+ "objective/rlhf_reward": 8.683082580566406,
323
+ "objective/scores": 8.683082580566406,
324
+ "policy/approxkl_avg": 0.216833233833313,
325
+ "policy/clipfrac_avg": 0.34375,
326
+ "policy/entropy_avg": 0.5138433575630188,
327
+ "step": 15,
328
+ "val/clipfrac_avg": 0.0,
329
+ "val/num_eos_tokens": 2232,
330
+ "val/ratio": 0.9998021721839905,
331
+ "val/ratio_var": 1.7273125649808208e-06
332
+ },
333
+ {
334
+ "episode": 8704,
335
+ "epoch": 0.03728581220013708,
336
+ "eps": 6,
337
+ "loss/policy_avg": 0.1616867184638977,
338
+ "lr": 2.8125e-06,
339
+ "objective/entropy": -19.297401428222656,
340
+ "objective/kl": 20.52892303466797,
341
+ "objective/non_score_reward": 0.0,
342
+ "objective/rlhf_reward": 8.915764808654785,
343
+ "objective/scores": 8.915764808654785,
344
+ "policy/approxkl_avg": 0.1600520759820938,
345
+ "policy/clipfrac_avg": 0.318359375,
346
+ "policy/entropy_avg": 0.47873735427856445,
347
+ "step": 16,
348
+ "val/clipfrac_avg": 0.0,
349
+ "val/num_eos_tokens": 1931,
350
+ "val/ratio": 0.9997690916061401,
351
+ "val/ratio_var": 2.2947442630538717e-06
352
+ },
353
+ {
354
+ "episode": 9216,
355
+ "epoch": 0.03947909527073338,
356
+ "eps": 6,
357
+ "loss/policy_avg": 0.163500115275383,
358
+ "lr": 2.80078125e-06,
359
+ "objective/entropy": -16.193389892578125,
360
+ "objective/kl": 22.477802276611328,
361
+ "objective/non_score_reward": 0.0,
362
+ "objective/rlhf_reward": 8.87047290802002,
363
+ "objective/scores": 8.87047290802002,
364
+ "policy/approxkl_avg": 0.17684966325759888,
365
+ "policy/clipfrac_avg": 0.33984375,
366
+ "policy/entropy_avg": 0.4807279706001282,
367
+ "step": 17,
368
+ "val/clipfrac_avg": 0.0,
369
+ "val/num_eos_tokens": 1798,
370
+ "val/ratio": 0.9998293519020081,
371
+ "val/ratio_var": 2.4208854938478908e-06
372
+ },
373
+ {
374
+ "episode": 9728,
375
+ "epoch": 0.041672378341329676,
376
+ "eps": 6,
377
+ "loss/policy_avg": 0.19272606074810028,
378
+ "lr": 2.7890625e-06,
379
+ "objective/entropy": -12.483476638793945,
380
+ "objective/kl": 24.110105514526367,
381
+ "objective/non_score_reward": 0.0,
382
+ "objective/rlhf_reward": 8.77183723449707,
383
+ "objective/scores": 8.77183723449707,
384
+ "policy/approxkl_avg": 0.1895884871482849,
385
+ "policy/clipfrac_avg": 0.36328125,
386
+ "policy/entropy_avg": 0.4732131361961365,
387
+ "step": 18,
388
+ "val/clipfrac_avg": 0.0,
389
+ "val/num_eos_tokens": 1580,
390
+ "val/ratio": 1.0000896453857422,
391
+ "val/ratio_var": 2.6988111585524166e-06
392
+ },
393
+ {
394
+ "episode": 10240,
395
+ "epoch": 0.04386566141192598,
396
+ "eps": 6,
397
+ "loss/policy_avg": 0.31339162588119507,
398
+ "lr": 2.77734375e-06,
399
+ "objective/entropy": -14.506279945373535,
400
+ "objective/kl": 23.571977615356445,
401
+ "objective/non_score_reward": 0.0,
402
+ "objective/rlhf_reward": 8.617586135864258,
403
+ "objective/scores": 8.617586135864258,
404
+ "policy/approxkl_avg": 0.18291257321834564,
405
+ "policy/clipfrac_avg": 0.349609375,
406
+ "policy/entropy_avg": 0.4734336733818054,
407
+ "step": 19,
408
+ "val/clipfrac_avg": 0.0,
409
+ "val/num_eos_tokens": 1565,
410
+ "val/ratio": 0.9999023675918579,
411
+ "val/ratio_var": 2.692267571546836e-06
412
+ },
413
+ {
414
+ "episode": 10752,
415
+ "epoch": 0.04605894448252228,
416
+ "eps": 6,
417
+ "loss/policy_avg": 0.4666588306427002,
418
+ "lr": 2.765625e-06,
419
+ "objective/entropy": -12.953289985656738,
420
+ "objective/kl": 24.32935333251953,
421
+ "objective/non_score_reward": 0.0,
422
+ "objective/rlhf_reward": 8.778571128845215,
423
+ "objective/scores": 8.778571128845215,
424
+ "policy/approxkl_avg": 0.20027068257331848,
425
+ "policy/clipfrac_avg": 0.30859375,
426
+ "policy/entropy_avg": 0.4659819006919861,
427
+ "step": 20,
428
+ "val/clipfrac_avg": 0.0,
429
+ "val/num_eos_tokens": 1506,
430
+ "val/ratio": 1.0001583099365234,
431
+ "val/ratio_var": 3.233054258089396e-06
432
+ },
433
+ {
434
+ "episode": 11264,
435
+ "epoch": 0.048252227553118573,
436
+ "eps": 6,
437
+ "loss/policy_avg": 0.27946847677230835,
438
+ "lr": 2.75390625e-06,
439
+ "objective/entropy": -11.48613166809082,
440
+ "objective/kl": 24.395483016967773,
441
+ "objective/non_score_reward": 0.0,
442
+ "objective/rlhf_reward": 9.02904224395752,
443
+ "objective/scores": 9.02904224395752,
444
+ "policy/approxkl_avg": 0.18909631669521332,
445
+ "policy/clipfrac_avg": 0.333984375,
446
+ "policy/entropy_avg": 0.48570871353149414,
447
+ "step": 21,
448
+ "val/clipfrac_avg": 0.0,
449
+ "val/num_eos_tokens": 1560,
450
+ "val/ratio": 1.0002957582473755,
451
+ "val/ratio_var": 2.959979838124127e-06
452
+ },
453
+ {
454
+ "episode": 11776,
455
+ "epoch": 0.05044551062371487,
456
+ "eps": 6,
457
+ "loss/policy_avg": 0.2717439830303192,
458
+ "lr": 2.7421875e-06,
459
+ "objective/entropy": -9.29305362701416,
460
+ "objective/kl": 26.03940200805664,
461
+ "objective/non_score_reward": 0.0,
462
+ "objective/rlhf_reward": 9.028656005859375,
463
+ "objective/scores": 9.028656005859375,
464
+ "policy/approxkl_avg": 0.1726618856191635,
465
+ "policy/clipfrac_avg": 0.34765625,
466
+ "policy/entropy_avg": 0.5084494352340698,
467
+ "step": 22,
468
+ "val/clipfrac_avg": 0.0,
469
+ "val/num_eos_tokens": 1547,
470
+ "val/ratio": 1.0000401735305786,
471
+ "val/ratio_var": 3.0132259780657478e-06
472
+ },
473
+ {
474
+ "episode": 12288,
475
+ "epoch": 0.052638793694311174,
476
+ "eps": 6,
477
+ "loss/policy_avg": 0.19815856218338013,
478
+ "lr": 2.73046875e-06,
479
+ "objective/entropy": -13.52733039855957,
480
+ "objective/kl": 25.79766273498535,
481
+ "objective/non_score_reward": 0.0,
482
+ "objective/rlhf_reward": 9.250341415405273,
483
+ "objective/scores": 9.250341415405273,
484
+ "policy/approxkl_avg": 0.1738530993461609,
485
+ "policy/clipfrac_avg": 0.3125,
486
+ "policy/entropy_avg": 0.4977012574672699,
487
+ "step": 23,
488
+ "val/clipfrac_avg": 0.0,
489
+ "val/num_eos_tokens": 1648,
490
+ "val/ratio": 0.9999476671218872,
491
+ "val/ratio_var": 2.497435616533039e-06
492
+ },
493
+ {
494
+ "episode": 12800,
495
+ "epoch": 0.05483207676490747,
496
+ "eps": 6,
497
+ "loss/policy_avg": 0.2345161885023117,
498
+ "lr": 2.71875e-06,
499
+ "objective/entropy": -9.971322059631348,
500
+ "objective/kl": 26.595800399780273,
501
+ "objective/non_score_reward": 0.0,
502
+ "objective/rlhf_reward": 9.097506523132324,
503
+ "objective/scores": 9.097506523132324,
504
+ "policy/approxkl_avg": 0.18735162913799286,
505
+ "policy/clipfrac_avg": 0.33203125,
506
+ "policy/entropy_avg": 0.4894530177116394,
507
+ "step": 24,
508
+ "val/clipfrac_avg": 0.0,
509
+ "val/num_eos_tokens": 1580,
510
+ "val/ratio": 1.000110149383545,
511
+ "val/ratio_var": 2.1261726033117156e-06
512
+ },
513
+ {
514
+ "episode": 13312,
515
+ "epoch": 0.05702535983550377,
516
+ "eps": 6,
517
+ "loss/policy_avg": 0.2765505909919739,
518
+ "lr": 2.70703125e-06,
519
+ "objective/entropy": -8.664510726928711,
520
+ "objective/kl": 27.207202911376953,
521
+ "objective/non_score_reward": 0.0,
522
+ "objective/rlhf_reward": 9.248998641967773,
523
+ "objective/scores": 9.248998641967773,
524
+ "policy/approxkl_avg": 0.21922221779823303,
525
+ "policy/clipfrac_avg": 0.3671875,
526
+ "policy/entropy_avg": 0.4936632513999939,
527
+ "step": 25,
528
+ "val/clipfrac_avg": 0.0,
529
+ "val/num_eos_tokens": 1689,
530
+ "val/ratio": 0.9997528195381165,
531
+ "val/ratio_var": 2.2092231120041106e-06
532
+ },
533
+ {
534
+ "episode": 13824,
535
+ "epoch": 0.05921864290610007,
536
+ "eps": 6,
537
+ "loss/policy_avg": 0.20894497632980347,
538
+ "lr": 2.6953125e-06,
539
+ "objective/entropy": -13.794546127319336,
540
+ "objective/kl": 27.585330963134766,
541
+ "objective/non_score_reward": 0.0,
542
+ "objective/rlhf_reward": 9.134411811828613,
543
+ "objective/scores": 9.134411811828613,
544
+ "policy/approxkl_avg": 0.19758901000022888,
545
+ "policy/clipfrac_avg": 0.337890625,
546
+ "policy/entropy_avg": 0.4570607841014862,
547
+ "step": 26,
548
+ "val/clipfrac_avg": 0.0,
549
+ "val/num_eos_tokens": 1664,
550
+ "val/ratio": 0.999937117099762,
551
+ "val/ratio_var": 3.2459099656989565e-06
552
+ },
553
+ {
554
+ "episode": 14336,
555
+ "epoch": 0.06141192597669637,
556
+ "eps": 6,
557
+ "loss/policy_avg": 0.187041237950325,
558
+ "lr": 2.68359375e-06,
559
+ "objective/entropy": -18.817930221557617,
560
+ "objective/kl": 26.306095123291016,
561
+ "objective/non_score_reward": 0.0,
562
+ "objective/rlhf_reward": 9.336585998535156,
563
+ "objective/scores": 9.336585998535156,
564
+ "policy/approxkl_avg": 0.20577886700630188,
565
+ "policy/clipfrac_avg": 0.279296875,
566
+ "policy/entropy_avg": 0.4518057107925415,
567
+ "step": 27,
568
+ "val/clipfrac_avg": 0.0,
569
+ "val/num_eos_tokens": 1731,
570
+ "val/ratio": 1.0001513957977295,
571
+ "val/ratio_var": 2.708766032810672e-06
572
+ },
573
+ {
574
+ "episode": 14848,
575
+ "epoch": 0.06360520904729267,
576
+ "eps": 6,
577
+ "loss/policy_avg": 0.15964296460151672,
578
+ "lr": 2.671875e-06,
579
+ "objective/entropy": -19.800785064697266,
580
+ "objective/kl": 27.51376724243164,
581
+ "objective/non_score_reward": 0.0,
582
+ "objective/rlhf_reward": 9.47825813293457,
583
+ "objective/scores": 9.47825813293457,
584
+ "policy/approxkl_avg": 0.19460758566856384,
585
+ "policy/clipfrac_avg": 0.34765625,
586
+ "policy/entropy_avg": 0.4188844859600067,
587
+ "step": 28,
588
+ "val/clipfrac_avg": 0.0,
589
+ "val/num_eos_tokens": 1687,
590
+ "val/ratio": 0.9997619390487671,
591
+ "val/ratio_var": 2.6442203306942247e-06
592
+ },
593
+ {
594
+ "episode": 15360,
595
+ "epoch": 0.06579849211788896,
596
+ "eps": 6,
597
+ "loss/policy_avg": 0.15885861217975616,
598
+ "lr": 2.66015625e-06,
599
+ "objective/entropy": -17.79833221435547,
600
+ "objective/kl": 27.949443817138672,
601
+ "objective/non_score_reward": 0.0,
602
+ "objective/rlhf_reward": 9.496625900268555,
603
+ "objective/scores": 9.496625900268555,
604
+ "policy/approxkl_avg": 0.16567930579185486,
605
+ "policy/clipfrac_avg": 0.34765625,
606
+ "policy/entropy_avg": 0.4135560691356659,
607
+ "step": 29,
608
+ "val/clipfrac_avg": 0.0,
609
+ "val/num_eos_tokens": 1689,
610
+ "val/ratio": 1.0000743865966797,
611
+ "val/ratio_var": 3.2084180929814465e-06
612
+ },
613
+ {
614
+ "episode": 15872,
615
+ "epoch": 0.06799177518848526,
616
+ "eps": 6,
617
+ "loss/policy_avg": 0.16605864465236664,
618
+ "lr": 2.6484375e-06,
619
+ "objective/entropy": -18.510845184326172,
620
+ "objective/kl": 29.722782135009766,
621
+ "objective/non_score_reward": 0.0,
622
+ "objective/rlhf_reward": 9.869793891906738,
623
+ "objective/scores": 9.869793891906738,
624
+ "policy/approxkl_avg": 0.1928042322397232,
625
+ "policy/clipfrac_avg": 0.421875,
626
+ "policy/entropy_avg": 0.3933394253253937,
627
+ "step": 30,
628
+ "val/clipfrac_avg": 0.0,
629
+ "val/num_eos_tokens": 1600,
630
+ "val/ratio": 1.0000554323196411,
631
+ "val/ratio_var": 2.9572597668447997e-06
632
+ },
633
+ {
634
+ "episode": 16384,
635
+ "epoch": 0.07018505825908156,
636
+ "eps": 6,
637
+ "loss/policy_avg": 0.19288870692253113,
638
+ "lr": 2.63671875e-06,
639
+ "objective/entropy": -17.214672088623047,
640
+ "objective/kl": 30.913219451904297,
641
+ "objective/non_score_reward": 0.0,
642
+ "objective/rlhf_reward": 9.56223201751709,
643
+ "objective/scores": 9.56223201751709,
644
+ "policy/approxkl_avg": 0.1885090172290802,
645
+ "policy/clipfrac_avg": 0.345703125,
646
+ "policy/entropy_avg": 0.3990550935268402,
647
+ "step": 31,
648
+ "val/clipfrac_avg": 0.0,
649
+ "val/num_eos_tokens": 1621,
650
+ "val/ratio": 0.9998235106468201,
651
+ "val/ratio_var": 2.50669563683914e-06
652
+ }
653
+ ],
654
+ "logging_steps": 1,
655
+ "max_steps": 128,
656
+ "num_input_tokens_seen": 0,
657
+ "num_train_epochs": 1.122960932145305,
658
+ "save_steps": 2,
659
+ "stateful_callbacks": {
660
+ "TrainerControl": {
661
+ "args": {
662
+ "should_epoch_stop": false,
663
+ "should_evaluate": false,
664
+ "should_log": true,
665
+ "should_save": true,
666
+ "should_training_stop": false
667
+ },
668
+ "attributes": {}
669
+ }
670
+ },
671
+ "total_flos": 0,
672
+ "train_batch_size": null,
673
+ "trial_name": null,
674
+ "trial_params": null
675
+ }
checkpoint-32/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d466b2207c4065e08d4717869aa8511625bf2d5a24b4c4af0a8221284315ed03
3
+ size 6584