Hugging Face's logo

YYYYYYibo
/
imp_sam_iter_1

PEFT
Safetensors
mistral
alignment-handbook
Generated from Trainer
trl
dpo
Model card Files Community
YYYYYYibo commited on
Commit
974e2fd
·
verified ·
1 Parent(s): 91bd3f0

Model save

Browse files
Files changed (4) hide show
  1. README.md +75 -0
  2. all_results.json +8 -0
  3. train_results.json +8 -0
  4. trainer_state.json +270 -0
README.md ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: alignment-handbook/zephyr-7b-sft-full
3
+ library_name: peft
4
+ license: apache-2.0
5
+ tags:
6
+ - trl
7
+ - dpo
8
+ - generated_from_trainer
9
+ model-index:
10
+ - name: imp_sam_iter_1
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # imp_sam_iter_1
18
+
19
+ This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the None dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 0.6285
22
+ - Rewards/chosen: -0.1131
23
+ - Rewards/rejected: -0.2857
24
+ - Rewards/accuracies: 0.7000
25
+ - Rewards/margins: 0.1725
26
+ - Logps/rejected: -286.0817
27
+ - Logps/chosen: -295.3544
28
+ - Logits/rejected: -2.5416
29
+ - Logits/chosen: -2.6261
30
+
31
+ ## Model description
32
+
33
+ More information needed
34
+
35
+ ## Intended uses & limitations
36
+
37
+ More information needed
38
+
39
+ ## Training and evaluation data
40
+
41
+ More information needed
42
+
43
+ ## Training procedure
44
+
45
+ ### Training hyperparameters
46
+
47
+ The following hyperparameters were used during training:
48
+ - learning_rate: 5e-06
49
+ - train_batch_size: 2
50
+ - eval_batch_size: 2
51
+ - seed: 42
52
+ - distributed_type: multi-GPU
53
+ - num_devices: 4
54
+ - gradient_accumulation_steps: 16
55
+ - total_train_batch_size: 128
56
+ - total_eval_batch_size: 8
57
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
58
+ - lr_scheduler_type: cosine
59
+ - lr_scheduler_warmup_ratio: 0.1
60
+ - num_epochs: 1
61
+
62
+ ### Training results
63
+
64
+ | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
65
+ |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
66
+ | 0.6297 | 0.64 | 100 | 0.6285 | -0.1131 | -0.2857 | 0.7000 | 0.1725 | -286.0817 | -295.3544 | -2.5416 | -2.6261 |
67
+
68
+
69
+ ### Framework versions
70
+
71
+ - PEFT 0.7.1
72
+ - Transformers 4.36.2
73
+ - Pytorch 2.1.2+cu121
74
+ - Datasets 2.14.6
75
+ - Tokenizers 0.15.2
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.6446946973984058,
4
+ "train_runtime": 7153.8565,
5
+ "train_samples": 20000,
6
+ "train_samples_per_second": 2.796,
7
+ "train_steps_per_second": 0.022
8
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.6446946973984058,
4
+ "train_runtime": 7153.8565,
5
+ "train_samples": 20000,
6
+ "train_samples_per_second": 2.796,
7
+ "train_steps_per_second": 0.022
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,270 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9984,
5
+ "eval_steps": 100,
6
+ "global_step": 156,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 3.125e-07,
14
+ "logits/chosen": -2.89351749420166,
15
+ "logits/rejected": -2.7752203941345215,
16
+ "logps/chosen": -345.7324523925781,
17
+ "logps/rejected": -319.42047119140625,
18
+ "loss": 0.6931,
19
+ "rewards/accuracies": 0.0,
20
+ "rewards/chosen": 0.0,
21
+ "rewards/margins": 0.0,
22
+ "rewards/rejected": 0.0,
23
+ "step": 1
24
+ },
25
+ {
26
+ "epoch": 0.06,
27
+ "learning_rate": 3.125e-06,
28
+ "logits/chosen": -2.8028833866119385,
29
+ "logits/rejected": -2.7471988201141357,
30
+ "logps/chosen": -255.036865234375,
31
+ "logps/rejected": -252.82679748535156,
32
+ "loss": 0.6928,
33
+ "rewards/accuracies": 0.4583333432674408,
34
+ "rewards/chosen": 0.0003066221543122083,
35
+ "rewards/margins": 0.0006635435856878757,
36
+ "rewards/rejected": -0.0003569214604794979,
37
+ "step": 10
38
+ },
39
+ {
40
+ "epoch": 0.13,
41
+ "learning_rate": 4.989935734988098e-06,
42
+ "logits/chosen": -2.7710824012756348,
43
+ "logits/rejected": -2.7166454792022705,
44
+ "logps/chosen": -277.1798400878906,
45
+ "logps/rejected": -256.997802734375,
46
+ "loss": 0.6889,
47
+ "rewards/accuracies": 0.65625,
48
+ "rewards/chosen": 0.012738336808979511,
49
+ "rewards/margins": 0.008854442276060581,
50
+ "rewards/rejected": 0.00388389453291893,
51
+ "step": 20
52
+ },
53
+ {
54
+ "epoch": 0.19,
55
+ "learning_rate": 4.8776412907378845e-06,
56
+ "logits/chosen": -2.7218027114868164,
57
+ "logits/rejected": -2.658277750015259,
58
+ "logps/chosen": -274.3503112792969,
59
+ "logps/rejected": -246.41128540039062,
60
+ "loss": 0.6795,
61
+ "rewards/accuracies": 0.6875,
62
+ "rewards/chosen": 0.03419749438762665,
63
+ "rewards/margins": 0.027278240770101547,
64
+ "rewards/rejected": 0.006919251289218664,
65
+ "step": 30
66
+ },
67
+ {
68
+ "epoch": 0.26,
69
+ "learning_rate": 4.646121984004666e-06,
70
+ "logits/chosen": -2.7631821632385254,
71
+ "logits/rejected": -2.6550638675689697,
72
+ "logps/chosen": -268.45220947265625,
73
+ "logps/rejected": -251.58743286132812,
74
+ "loss": 0.6693,
75
+ "rewards/accuracies": 0.715624988079071,
76
+ "rewards/chosen": 0.03486743941903114,
77
+ "rewards/margins": 0.05773182958364487,
78
+ "rewards/rejected": -0.022864393889904022,
79
+ "step": 40
80
+ },
81
+ {
82
+ "epoch": 0.32,
83
+ "learning_rate": 4.3069871595684795e-06,
84
+ "logits/chosen": -2.676445960998535,
85
+ "logits/rejected": -2.62298321723938,
86
+ "logps/chosen": -274.287841796875,
87
+ "logps/rejected": -281.6821594238281,
88
+ "loss": 0.6605,
89
+ "rewards/accuracies": 0.671875,
90
+ "rewards/chosen": -0.003023784141987562,
91
+ "rewards/margins": 0.07856948673725128,
92
+ "rewards/rejected": -0.08159326761960983,
93
+ "step": 50
94
+ },
95
+ {
96
+ "epoch": 0.38,
97
+ "learning_rate": 3.8772424536302565e-06,
98
+ "logits/chosen": -2.687849521636963,
99
+ "logits/rejected": -2.6154909133911133,
100
+ "logps/chosen": -287.0888366699219,
101
+ "logps/rejected": -274.48822021484375,
102
+ "loss": 0.6507,
103
+ "rewards/accuracies": 0.715624988079071,
104
+ "rewards/chosen": -0.03279999643564224,
105
+ "rewards/margins": 0.11418493092060089,
106
+ "rewards/rejected": -0.14698493480682373,
107
+ "step": 60
108
+ },
109
+ {
110
+ "epoch": 0.45,
111
+ "learning_rate": 3.3784370602033572e-06,
112
+ "logits/chosen": -2.6265475749969482,
113
+ "logits/rejected": -2.570312976837158,
114
+ "logps/chosen": -304.10125732421875,
115
+ "logps/rejected": -292.9505310058594,
116
+ "loss": 0.6431,
117
+ "rewards/accuracies": 0.7250000238418579,
118
+ "rewards/chosen": -0.033603884279727936,
119
+ "rewards/margins": 0.12847770750522614,
120
+ "rewards/rejected": -0.16208159923553467,
121
+ "step": 70
122
+ },
123
+ {
124
+ "epoch": 0.51,
125
+ "learning_rate": 2.835583164544139e-06,
126
+ "logits/chosen": -2.6333694458007812,
127
+ "logits/rejected": -2.5370612144470215,
128
+ "logps/chosen": -302.35736083984375,
129
+ "logps/rejected": -276.11175537109375,
130
+ "loss": 0.634,
131
+ "rewards/accuracies": 0.6812499761581421,
132
+ "rewards/chosen": -0.04339645802974701,
133
+ "rewards/margins": 0.14248773455619812,
134
+ "rewards/rejected": -0.18588420748710632,
135
+ "step": 80
136
+ },
137
+ {
138
+ "epoch": 0.58,
139
+ "learning_rate": 2.2759017277414165e-06,
140
+ "logits/chosen": -2.584850311279297,
141
+ "logits/rejected": -2.5956850051879883,
142
+ "logps/chosen": -292.6184997558594,
143
+ "logps/rejected": -311.091796875,
144
+ "loss": 0.6249,
145
+ "rewards/accuracies": 0.699999988079071,
146
+ "rewards/chosen": -0.09801232814788818,
147
+ "rewards/margins": 0.17943526804447174,
148
+ "rewards/rejected": -0.27744758129119873,
149
+ "step": 90
150
+ },
151
+ {
152
+ "epoch": 0.64,
153
+ "learning_rate": 1.7274575140626318e-06,
154
+ "logits/chosen": -2.6319050788879395,
155
+ "logits/rejected": -2.5922951698303223,
156
+ "logps/chosen": -290.19976806640625,
157
+ "logps/rejected": -273.79815673828125,
158
+ "loss": 0.6297,
159
+ "rewards/accuracies": 0.643750011920929,
160
+ "rewards/chosen": -0.1279720962047577,
161
+ "rewards/margins": 0.16270975768566132,
162
+ "rewards/rejected": -0.2906818985939026,
163
+ "step": 100
164
+ },
165
+ {
166
+ "epoch": 0.64,
167
+ "eval_logits/chosen": -2.6261305809020996,
168
+ "eval_logits/rejected": -2.541619062423706,
169
+ "eval_logps/chosen": -295.3544006347656,
170
+ "eval_logps/rejected": -286.08172607421875,
171
+ "eval_loss": 0.6284892559051514,
172
+ "eval_rewards/accuracies": 0.699999988079071,
173
+ "eval_rewards/chosen": -0.11314628273248672,
174
+ "eval_rewards/margins": 0.1725120097398758,
175
+ "eval_rewards/rejected": -0.2856582701206207,
176
+ "eval_runtime": 382.1442,
177
+ "eval_samples_per_second": 5.234,
178
+ "eval_steps_per_second": 0.654,
179
+ "step": 100
180
+ },
181
+ {
182
+ "epoch": 0.7,
183
+ "learning_rate": 1.217751806485235e-06,
184
+ "logits/chosen": -2.6338038444519043,
185
+ "logits/rejected": -2.5288445949554443,
186
+ "logps/chosen": -289.0359802246094,
187
+ "logps/rejected": -275.2894287109375,
188
+ "loss": 0.6177,
189
+ "rewards/accuracies": 0.684374988079071,
190
+ "rewards/chosen": -0.06765095144510269,
191
+ "rewards/margins": 0.20255950093269348,
192
+ "rewards/rejected": -0.2702104449272156,
193
+ "step": 110
194
+ },
195
+ {
196
+ "epoch": 0.77,
197
+ "learning_rate": 7.723433775328385e-07,
198
+ "logits/chosen": -2.552917003631592,
199
+ "logits/rejected": -2.550574779510498,
200
+ "logps/chosen": -274.2412109375,
201
+ "logps/rejected": -288.6865539550781,
202
+ "loss": 0.6251,
203
+ "rewards/accuracies": 0.6812499761581421,
204
+ "rewards/chosen": -0.10379727929830551,
205
+ "rewards/margins": 0.1611437350511551,
206
+ "rewards/rejected": -0.2649410367012024,
207
+ "step": 120
208
+ },
209
+ {
210
+ "epoch": 0.83,
211
+ "learning_rate": 4.1356686569674344e-07,
212
+ "logits/chosen": -2.611356735229492,
213
+ "logits/rejected": -2.5477137565612793,
214
+ "logps/chosen": -291.66290283203125,
215
+ "logps/rejected": -301.75537109375,
216
+ "loss": 0.6233,
217
+ "rewards/accuracies": 0.690625011920929,
218
+ "rewards/chosen": -0.11176357418298721,
219
+ "rewards/margins": 0.20609335601329803,
220
+ "rewards/rejected": -0.31785690784454346,
221
+ "step": 130
222
+ },
223
+ {
224
+ "epoch": 0.9,
225
+ "learning_rate": 1.59412823400657e-07,
226
+ "logits/chosen": -2.6064445972442627,
227
+ "logits/rejected": -2.479429244995117,
228
+ "logps/chosen": -313.6204833984375,
229
+ "logps/rejected": -287.5814514160156,
230
+ "loss": 0.6146,
231
+ "rewards/accuracies": 0.684374988079071,
232
+ "rewards/chosen": -0.09765791893005371,
233
+ "rewards/margins": 0.21785131096839905,
234
+ "rewards/rejected": -0.31550922989845276,
235
+ "step": 140
236
+ },
237
+ {
238
+ "epoch": 0.96,
239
+ "learning_rate": 2.262559558016325e-08,
240
+ "logits/chosen": -2.58495831489563,
241
+ "logits/rejected": -2.4780099391937256,
242
+ "logps/chosen": -300.6700134277344,
243
+ "logps/rejected": -281.14935302734375,
244
+ "loss": 0.6193,
245
+ "rewards/accuracies": 0.746874988079071,
246
+ "rewards/chosen": -0.11531106382608414,
247
+ "rewards/margins": 0.22770515084266663,
248
+ "rewards/rejected": -0.34301620721817017,
249
+ "step": 150
250
+ },
251
+ {
252
+ "epoch": 1.0,
253
+ "step": 156,
254
+ "total_flos": 0.0,
255
+ "train_loss": 0.6446946973984058,
256
+ "train_runtime": 7153.8565,
257
+ "train_samples_per_second": 2.796,
258
+ "train_steps_per_second": 0.022
259
+ }
260
+ ],
261
+ "logging_steps": 10,
262
+ "max_steps": 156,
263
+ "num_input_tokens_seen": 0,
264
+ "num_train_epochs": 1,
265
+ "save_steps": 100,
266
+ "total_flos": 0.0,
267
+ "train_batch_size": 2,
268
+ "trial_name": null,
269
+ "trial_params": null
270
+ }