Hugging Face's logo

EllieS
/
zephyr-dpo-timedial-selfgen-mix

PEFT
TensorBoard
Safetensors
mistral
alignment-handbook
trl
dpo
Generated from Trainer
Model card Files Metrics Community
EllieS commited on
Commit
79b8525
·
verified ·
1 Parent(s): 60bccda

Model save

Browse files
Files changed (6) hide show
  1. README.md +74 -0
  2. adapter_model.safetensors +1 -1
  3. all_results.json +8 -0
  4. runs/Apr11_18-43-53_612e66badb5c/events.out.tfevents.1712861164.612e66badb5c.7843.0 +2 -2
  5. train_results.json +8 -0
  6. trainer_state.json +497 -0
README.md ADDED
@@ -0,0 +1,74 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ library_name: peft
4
+ tags:
5
+ - trl
6
+ - dpo
7
+ - generated_from_trainer
8
+ base_model: alignment-handbook/zephyr-7b-sft-full
9
+ model-index:
10
+ - name: zephyr-dpo-timedial-selfgen-mix
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
+ # zephyr-dpo-timedial-selfgen-mix
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: 0.0236
22
+ - Rewards/chosen: 0.5600
23
+ - Rewards/rejected: -3.3438
24
+ - Rewards/accuracies: 1.0
25
+ - Rewards/margins: 3.9038
26
+ - Logps/rejected: -542.8434
27
+ - Logps/chosen: -74.4535
28
+ - Logits/rejected: -3.2030
29
+ - Logits/chosen: -2.7110
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
+ - gradient_accumulation_steps: 2
54
+ - total_train_batch_size: 4
55
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
56
+ - lr_scheduler_type: cosine
57
+ - lr_scheduler_warmup_ratio: 0.1
58
+ - num_epochs: 1
59
+
60
+ ### Training results
61
+
62
+ | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
63
+ |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
64
+ | 0.039 | 0.35 | 100 | 0.0525 | 0.5119 | -2.4372 | 1.0 | 2.9492 | -452.1829 | -79.2593 | -3.2065 | -2.7171 |
65
+ | 0.0106 | 0.7 | 200 | 0.0236 | 0.5600 | -3.3438 | 1.0 | 3.9038 | -542.8434 | -74.4535 | -3.2030 | -2.7110 |
66
+
67
+
68
+ ### Framework versions
69
+
70
+ - PEFT 0.7.1
71
+ - Transformers 4.39.0.dev0
72
+ - Pytorch 2.1.2+cu121
73
+ - Datasets 2.14.6
74
+ - Tokenizers 0.15.2
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4a5cdd8ac2cf1a6029846ff6482b39a3e04112a0ab00adcc39d20f367bf2d8fd
3
  size 83946192
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de1aa36b0e544ed8d95bfdcf4e6df5d20a9271ddb38c68705bbaa8129bbef57e
3
  size 83946192
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.12933478820742222,
4
+ "train_runtime": 748.758,
5
+ "train_samples": 1139,
6
+ "train_samples_per_second": 1.521,
7
+ "train_steps_per_second": 0.381
8
+ }
runs/Apr11_18-43-53_612e66badb5c/events.out.tfevents.1712861164.612e66badb5c.7843.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2729513a3279feace152632d59ed48eca876286fc88c7eaf010647305592adea
3
- size 20573
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e8e97251b2f8b42524c507185b9664895948896f20d71ecd50ada2c1b012a0dd
3
+ size 26431
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.12933478820742222,
4
+ "train_runtime": 748.758,
5
+ "train_samples": 1139,
6
+ "train_samples_per_second": 1.521,
7
+ "train_steps_per_second": 0.381
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,497 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 100,
6
+ "global_step": 285,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "grad_norm": 1.875,
14
+ "learning_rate": 1.7241379310344828e-07,
15
+ "logits/chosen": -2.735238790512085,
16
+ "logits/rejected": -3.2783570289611816,
17
+ "logps/chosen": -165.14926147460938,
18
+ "logps/rejected": -228.84988403320312,
19
+ "loss": 0.6931,
20
+ "rewards/accuracies": 0.0,
21
+ "rewards/chosen": 0.0,
22
+ "rewards/margins": 0.0,
23
+ "rewards/rejected": 0.0,
24
+ "step": 1
25
+ },
26
+ {
27
+ "epoch": 0.04,
28
+ "grad_norm": 1.7421875,
29
+ "learning_rate": 1.724137931034483e-06,
30
+ "logits/chosen": -2.7123961448669434,
31
+ "logits/rejected": -3.239219903945923,
32
+ "logps/chosen": -164.08265686035156,
33
+ "logps/rejected": -227.88357543945312,
34
+ "loss": 0.6921,
35
+ "rewards/accuracies": 0.6111111044883728,
36
+ "rewards/chosen": 0.0008442008402198553,
37
+ "rewards/margins": 0.002071016002446413,
38
+ "rewards/rejected": -0.0012268151622265577,
39
+ "step": 10
40
+ },
41
+ {
42
+ "epoch": 0.07,
43
+ "grad_norm": 1.828125,
44
+ "learning_rate": 3.448275862068966e-06,
45
+ "logits/chosen": -2.7431998252868652,
46
+ "logits/rejected": -3.2384581565856934,
47
+ "logps/chosen": -163.16790771484375,
48
+ "logps/rejected": -229.4684600830078,
49
+ "loss": 0.6817,
50
+ "rewards/accuracies": 1.0,
51
+ "rewards/chosen": 0.007237003184854984,
52
+ "rewards/margins": 0.023129161447286606,
53
+ "rewards/rejected": -0.015892159193754196,
54
+ "step": 20
55
+ },
56
+ {
57
+ "epoch": 0.11,
58
+ "grad_norm": 1.6640625,
59
+ "learning_rate": 4.999811754597862e-06,
60
+ "logits/chosen": -2.704376220703125,
61
+ "logits/rejected": -3.244763135910034,
62
+ "logps/chosen": -158.67416381835938,
63
+ "logps/rejected": -237.81668090820312,
64
+ "loss": 0.6284,
65
+ "rewards/accuracies": 1.0,
66
+ "rewards/chosen": 0.03751087933778763,
67
+ "rewards/margins": 0.13478627800941467,
68
+ "rewards/rejected": -0.09727539122104645,
69
+ "step": 30
70
+ },
71
+ {
72
+ "epoch": 0.14,
73
+ "grad_norm": 1.578125,
74
+ "learning_rate": 4.97725658856945e-06,
75
+ "logits/chosen": -2.718313217163086,
76
+ "logits/rejected": -3.2308642864227295,
77
+ "logps/chosen": -150.95199584960938,
78
+ "logps/rejected": -256.7232666015625,
79
+ "loss": 0.5081,
80
+ "rewards/accuracies": 1.0,
81
+ "rewards/chosen": 0.11078281700611115,
82
+ "rewards/margins": 0.415455162525177,
83
+ "rewards/rejected": -0.3046723008155823,
84
+ "step": 40
85
+ },
86
+ {
87
+ "epoch": 0.18,
88
+ "grad_norm": 1.3828125,
89
+ "learning_rate": 4.917441177612131e-06,
90
+ "logits/chosen": -2.699450731277466,
91
+ "logits/rejected": -3.2096571922302246,
92
+ "logps/chosen": -140.04672241210938,
93
+ "logps/rejected": -286.2795104980469,
94
+ "loss": 0.3683,
95
+ "rewards/accuracies": 1.0,
96
+ "rewards/chosen": 0.2172900140285492,
97
+ "rewards/margins": 0.8159033060073853,
98
+ "rewards/rejected": -0.5986132025718689,
99
+ "step": 50
100
+ },
101
+ {
102
+ "epoch": 0.21,
103
+ "grad_norm": 1.0390625,
104
+ "learning_rate": 4.82126520118304e-06,
105
+ "logits/chosen": -2.7382638454437256,
106
+ "logits/rejected": -3.209529399871826,
107
+ "logps/chosen": -127.2552261352539,
108
+ "logps/rejected": -322.1763610839844,
109
+ "loss": 0.2407,
110
+ "rewards/accuracies": 1.0,
111
+ "rewards/chosen": 0.36736661195755005,
112
+ "rewards/margins": 1.3110934495925903,
113
+ "rewards/rejected": -0.9437268972396851,
114
+ "step": 60
115
+ },
116
+ {
117
+ "epoch": 0.25,
118
+ "grad_norm": 0.6875,
119
+ "learning_rate": 4.6901752354885166e-06,
120
+ "logits/chosen": -2.696700096130371,
121
+ "logits/rejected": -3.2043445110321045,
122
+ "logps/chosen": -115.544677734375,
123
+ "logps/rejected": -366.6884765625,
124
+ "loss": 0.1455,
125
+ "rewards/accuracies": 1.0,
126
+ "rewards/chosen": 0.4782423973083496,
127
+ "rewards/margins": 1.8693335056304932,
128
+ "rewards/rejected": -1.391090750694275,
129
+ "step": 70
130
+ },
131
+ {
132
+ "epoch": 0.28,
133
+ "grad_norm": 0.48828125,
134
+ "learning_rate": 4.526142995631488e-06,
135
+ "logits/chosen": -2.7269909381866455,
136
+ "logits/rejected": -3.2105610370635986,
137
+ "logps/chosen": -109.22001647949219,
138
+ "logps/rejected": -415.4657287597656,
139
+ "loss": 0.0858,
140
+ "rewards/accuracies": 1.0,
141
+ "rewards/chosen": 0.5552297234535217,
142
+ "rewards/margins": 2.424295663833618,
143
+ "rewards/rejected": -1.8690656423568726,
144
+ "step": 80
145
+ },
146
+ {
147
+ "epoch": 0.32,
148
+ "grad_norm": 0.392578125,
149
+ "learning_rate": 4.331635679181032e-06,
150
+ "logits/chosen": -2.685180425643921,
151
+ "logits/rejected": -3.2150089740753174,
152
+ "logps/chosen": -104.05949401855469,
153
+ "logps/rejected": -453.306884765625,
154
+ "loss": 0.0559,
155
+ "rewards/accuracies": 1.0,
156
+ "rewards/chosen": 0.5974525213241577,
157
+ "rewards/margins": 2.865081548690796,
158
+ "rewards/rejected": -2.2676291465759277,
159
+ "step": 90
160
+ },
161
+ {
162
+ "epoch": 0.35,
163
+ "grad_norm": 0.302734375,
164
+ "learning_rate": 4.109578857224478e-06,
165
+ "logits/chosen": -2.702922821044922,
166
+ "logits/rejected": -3.216787338256836,
167
+ "logps/chosen": -100.30770111083984,
168
+ "logps/rejected": -488.27081298828125,
169
+ "loss": 0.039,
170
+ "rewards/accuracies": 1.0,
171
+ "rewards/chosen": 0.6299440264701843,
172
+ "rewards/margins": 3.2341396808624268,
173
+ "rewards/rejected": -2.6041955947875977,
174
+ "step": 100
175
+ },
176
+ {
177
+ "epoch": 0.35,
178
+ "eval_logits/chosen": -2.7171199321746826,
179
+ "eval_logits/rejected": -3.2064568996429443,
180
+ "eval_logps/chosen": -79.25934600830078,
181
+ "eval_logps/rejected": -452.1828918457031,
182
+ "eval_loss": 0.052521564066410065,
183
+ "eval_rewards/accuracies": 1.0,
184
+ "eval_rewards/chosen": 0.5119398236274719,
185
+ "eval_rewards/margins": 2.9491662979125977,
186
+ "eval_rewards/rejected": -2.4372265338897705,
187
+ "eval_runtime": 1.2939,
188
+ "eval_samples_per_second": 3.864,
189
+ "eval_steps_per_second": 2.319,
190
+ "step": 100
191
+ },
192
+ {
193
+ "epoch": 0.39,
194
+ "grad_norm": 0.259765625,
195
+ "learning_rate": 3.863312471055116e-06,
196
+ "logits/chosen": -2.7052559852600098,
197
+ "logits/rejected": -3.191067934036255,
198
+ "logps/chosen": -98.0762710571289,
199
+ "logps/rejected": -517.7929077148438,
200
+ "loss": 0.0281,
201
+ "rewards/accuracies": 1.0,
202
+ "rewards/chosen": 0.654212236404419,
203
+ "rewards/margins": 3.561893939971924,
204
+ "rewards/rejected": -2.907681465148926,
205
+ "step": 110
206
+ },
207
+ {
208
+ "epoch": 0.42,
209
+ "grad_norm": 0.212890625,
210
+ "learning_rate": 3.5965405963463197e-06,
211
+ "logits/chosen": -2.7348670959472656,
212
+ "logits/rejected": -3.205155849456787,
213
+ "logps/chosen": -95.99646759033203,
214
+ "logps/rejected": -544.2891845703125,
215
+ "loss": 0.0213,
216
+ "rewards/accuracies": 1.0,
217
+ "rewards/chosen": 0.6714237332344055,
218
+ "rewards/margins": 3.842583417892456,
219
+ "rewards/rejected": -3.1711599826812744,
220
+ "step": 120
221
+ },
222
+ {
223
+ "epoch": 0.46,
224
+ "grad_norm": 0.1845703125,
225
+ "learning_rate": 3.313275730405658e-06,
226
+ "logits/chosen": -2.698134183883667,
227
+ "logits/rejected": -3.215333938598633,
228
+ "logps/chosen": -94.20452880859375,
229
+ "logps/rejected": -564.6442260742188,
230
+ "loss": 0.0173,
231
+ "rewards/accuracies": 1.0,
232
+ "rewards/chosen": 0.6862392425537109,
233
+ "rewards/margins": 4.055664539337158,
234
+ "rewards/rejected": -3.369424819946289,
235
+ "step": 130
236
+ },
237
+ {
238
+ "epoch": 0.49,
239
+ "grad_norm": 0.1591796875,
240
+ "learning_rate": 3.0177784404805466e-06,
241
+ "logits/chosen": -2.728977680206299,
242
+ "logits/rejected": -3.2040677070617676,
243
+ "logps/chosen": -95.00137329101562,
244
+ "logps/rejected": -582.2269287109375,
245
+ "loss": 0.0145,
246
+ "rewards/accuracies": 1.0,
247
+ "rewards/chosen": 0.6958636045455933,
248
+ "rewards/margins": 4.233551502227783,
249
+ "rewards/rejected": -3.5376884937286377,
250
+ "step": 140
251
+ },
252
+ {
253
+ "epoch": 0.53,
254
+ "grad_norm": 0.1640625,
255
+ "learning_rate": 2.7144932808611002e-06,
256
+ "logits/chosen": -2.705024003982544,
257
+ "logits/rejected": -3.1964850425720215,
258
+ "logps/chosen": -94.14689636230469,
259
+ "logps/rejected": -587.1410522460938,
260
+ "loss": 0.0136,
261
+ "rewards/accuracies": 1.0,
262
+ "rewards/chosen": 0.6878983974456787,
263
+ "rewards/margins": 4.294985294342041,
264
+ "rewards/rejected": -3.6070873737335205,
265
+ "step": 150
266
+ },
267
+ {
268
+ "epoch": 0.56,
269
+ "grad_norm": 0.1572265625,
270
+ "learning_rate": 2.407981942646603e-06,
271
+ "logits/chosen": -2.7173984050750732,
272
+ "logits/rejected": -3.229872226715088,
273
+ "logps/chosen": -93.74874877929688,
274
+ "logps/rejected": -597.3945922851562,
275
+ "loss": 0.0121,
276
+ "rewards/accuracies": 1.0,
277
+ "rewards/chosen": 0.6988341212272644,
278
+ "rewards/margins": 4.410033226013184,
279
+ "rewards/rejected": -3.7111988067626953,
280
+ "step": 160
281
+ },
282
+ {
283
+ "epoch": 0.6,
284
+ "grad_norm": 0.1435546875,
285
+ "learning_rate": 2.102854641665347e-06,
286
+ "logits/chosen": -2.711818218231201,
287
+ "logits/rejected": -3.219580888748169,
288
+ "logps/chosen": -92.11972045898438,
289
+ "logps/rejected": -603.1061401367188,
290
+ "loss": 0.0115,
291
+ "rewards/accuracies": 1.0,
292
+ "rewards/chosen": 0.7005870342254639,
293
+ "rewards/margins": 4.465435981750488,
294
+ "rewards/rejected": -3.7648491859436035,
295
+ "step": 170
296
+ },
297
+ {
298
+ "epoch": 0.63,
299
+ "grad_norm": 0.154296875,
300
+ "learning_rate": 1.8037007765373677e-06,
301
+ "logits/chosen": -2.736745834350586,
302
+ "logits/rejected": -3.206763505935669,
303
+ "logps/chosen": -93.66650390625,
304
+ "logps/rejected": -605.5599365234375,
305
+ "loss": 0.0113,
306
+ "rewards/accuracies": 1.0,
307
+ "rewards/chosen": 0.700027346611023,
308
+ "rewards/margins": 4.484536170959473,
309
+ "rewards/rejected": -3.7845091819763184,
310
+ "step": 180
311
+ },
312
+ {
313
+ "epoch": 0.67,
314
+ "grad_norm": 0.1484375,
315
+ "learning_rate": 1.5150198998473802e-06,
316
+ "logits/chosen": -2.6961426734924316,
317
+ "logits/rejected": -3.201364517211914,
318
+ "logps/chosen": -92.75911712646484,
319
+ "logps/rejected": -609.6072998046875,
320
+ "loss": 0.0107,
321
+ "rewards/accuracies": 1.0,
322
+ "rewards/chosen": 0.7066330313682556,
323
+ "rewards/margins": 4.538401126861572,
324
+ "rewards/rejected": -3.8317675590515137,
325
+ "step": 190
326
+ },
327
+ {
328
+ "epoch": 0.7,
329
+ "grad_norm": 0.142578125,
330
+ "learning_rate": 1.2411540406857064e-06,
331
+ "logits/chosen": -2.697049140930176,
332
+ "logits/rejected": -3.2053134441375732,
333
+ "logps/chosen": -92.19438171386719,
334
+ "logps/rejected": -611.137939453125,
335
+ "loss": 0.0106,
336
+ "rewards/accuracies": 1.0,
337
+ "rewards/chosen": 0.7058606147766113,
338
+ "rewards/margins": 4.546854496002197,
339
+ "rewards/rejected": -3.840993881225586,
340
+ "step": 200
341
+ },
342
+ {
343
+ "epoch": 0.7,
344
+ "eval_logits/chosen": -2.7109670639038086,
345
+ "eval_logits/rejected": -3.2030317783355713,
346
+ "eval_logps/chosen": -74.4535140991211,
347
+ "eval_logps/rejected": -542.8434448242188,
348
+ "eval_loss": 0.023632029071450233,
349
+ "eval_rewards/accuracies": 1.0,
350
+ "eval_rewards/chosen": 0.5599982738494873,
351
+ "eval_rewards/margins": 3.903829574584961,
352
+ "eval_rewards/rejected": -3.343831777572632,
353
+ "eval_runtime": 1.2832,
354
+ "eval_samples_per_second": 3.896,
355
+ "eval_steps_per_second": 2.338,
356
+ "step": 200
357
+ },
358
+ {
359
+ "epoch": 0.74,
360
+ "grad_norm": 0.1318359375,
361
+ "learning_rate": 9.862223964891864e-07,
362
+ "logits/chosen": -2.728084087371826,
363
+ "logits/rejected": -3.2059364318847656,
364
+ "logps/chosen": -90.77999114990234,
365
+ "logps/rejected": -605.8444213867188,
366
+ "loss": 0.0124,
367
+ "rewards/accuracies": 1.0,
368
+ "rewards/chosen": 0.694692850112915,
369
+ "rewards/margins": 4.4922895431518555,
370
+ "rewards/rejected": -3.797595977783203,
371
+ "step": 210
372
+ },
373
+ {
374
+ "epoch": 0.77,
375
+ "grad_norm": 0.1279296875,
376
+ "learning_rate": 7.54059376477568e-07,
377
+ "logits/chosen": -2.712827682495117,
378
+ "logits/rejected": -3.2164974212646484,
379
+ "logps/chosen": -93.00579071044922,
380
+ "logps/rejected": -616.4993286132812,
381
+ "loss": 0.0102,
382
+ "rewards/accuracies": 1.0,
383
+ "rewards/chosen": 0.7107797861099243,
384
+ "rewards/margins": 4.593611717224121,
385
+ "rewards/rejected": -3.8828322887420654,
386
+ "step": 220
387
+ },
388
+ {
389
+ "epoch": 0.81,
390
+ "grad_norm": 0.130859375,
391
+ "learning_rate": 5.481569285697638e-07,
392
+ "logits/chosen": -2.7499520778656006,
393
+ "logits/rejected": -3.2049834728240967,
394
+ "logps/chosen": -93.71913146972656,
395
+ "logps/rejected": -612.4299926757812,
396
+ "loss": 0.0106,
397
+ "rewards/accuracies": 1.0,
398
+ "rewards/chosen": 0.6972833871841431,
399
+ "rewards/margins": 4.548653602600098,
400
+ "rewards/rejected": -3.851370334625244,
401
+ "step": 230
402
+ },
403
+ {
404
+ "epoch": 0.84,
405
+ "grad_norm": 0.1328125,
406
+ "learning_rate": 3.71612017236837e-07,
407
+ "logits/chosen": -2.7227025032043457,
408
+ "logits/rejected": -3.198974132537842,
409
+ "logps/chosen": -93.90065002441406,
410
+ "logps/rejected": -616.3263549804688,
411
+ "loss": 0.0101,
412
+ "rewards/accuracies": 1.0,
413
+ "rewards/chosen": 0.7073711156845093,
414
+ "rewards/margins": 4.595206260681152,
415
+ "rewards/rejected": -3.8878350257873535,
416
+ "step": 240
417
+ },
418
+ {
419
+ "epoch": 0.88,
420
+ "grad_norm": 0.1298828125,
421
+ "learning_rate": 2.2708004227369407e-07,
422
+ "logits/chosen": -2.7253494262695312,
423
+ "logits/rejected": -3.200887680053711,
424
+ "logps/chosen": -92.90316009521484,
425
+ "logps/rejected": -616.0049438476562,
426
+ "loss": 0.0102,
427
+ "rewards/accuracies": 1.0,
428
+ "rewards/chosen": 0.6983229517936707,
429
+ "rewards/margins": 4.588388919830322,
430
+ "rewards/rejected": -3.890066146850586,
431
+ "step": 250
432
+ },
433
+ {
434
+ "epoch": 0.91,
435
+ "grad_norm": 0.12890625,
436
+ "learning_rate": 1.1673489911451536e-07,
437
+ "logits/chosen": -2.6972110271453857,
438
+ "logits/rejected": -3.2123961448669434,
439
+ "logps/chosen": -92.834716796875,
440
+ "logps/rejected": -614.0274658203125,
441
+ "loss": 0.0103,
442
+ "rewards/accuracies": 1.0,
443
+ "rewards/chosen": 0.7053926587104797,
444
+ "rewards/margins": 4.57442569732666,
445
+ "rewards/rejected": -3.869033098220825,
446
+ "step": 260
447
+ },
448
+ {
449
+ "epoch": 0.95,
450
+ "grad_norm": 0.134765625,
451
+ "learning_rate": 4.223628142195929e-08,
452
+ "logits/chosen": -2.706958055496216,
453
+ "logits/rejected": -3.205293655395508,
454
+ "logps/chosen": -93.16792297363281,
455
+ "logps/rejected": -615.3536376953125,
456
+ "loss": 0.0101,
457
+ "rewards/accuracies": 1.0,
458
+ "rewards/chosen": 0.7087845802307129,
459
+ "rewards/margins": 4.595980167388916,
460
+ "rewards/rejected": -3.887195110321045,
461
+ "step": 270
462
+ },
463
+ {
464
+ "epoch": 0.98,
465
+ "grad_norm": 0.1318359375,
466
+ "learning_rate": 4.704717749627052e-09,
467
+ "logits/chosen": -2.6964120864868164,
468
+ "logits/rejected": -3.2064449787139893,
469
+ "logps/chosen": -93.33256530761719,
470
+ "logps/rejected": -614.2545166015625,
471
+ "loss": 0.0104,
472
+ "rewards/accuracies": 1.0,
473
+ "rewards/chosen": 0.6982468962669373,
474
+ "rewards/margins": 4.565683841705322,
475
+ "rewards/rejected": -3.8674368858337402,
476
+ "step": 280
477
+ },
478
+ {
479
+ "epoch": 1.0,
480
+ "step": 285,
481
+ "total_flos": 0.0,
482
+ "train_loss": 0.12933478820742222,
483
+ "train_runtime": 748.758,
484
+ "train_samples_per_second": 1.521,
485
+ "train_steps_per_second": 0.381
486
+ }
487
+ ],
488
+ "logging_steps": 10,
489
+ "max_steps": 285,
490
+ "num_input_tokens_seen": 0,
491
+ "num_train_epochs": 1,
492
+ "save_steps": 200,
493
+ "total_flos": 0.0,
494
+ "train_batch_size": 2,
495
+ "trial_name": null,
496
+ "trial_params": null
497
+ }