Hugging Face's logo

CharlesLi
/
OpenELM-1_1B-DPO-full-max-6-reward

Text Generation
Transformers
TensorBoard
Safetensors
openelm
trl
dpo
alignment-handbook
Generated from Trainer
conversational
custom_code
Model card Files Metrics Community
CharlesLi commited on
Commit
30bb95c
·
verified ·
1 Parent(s): 9ad5984

Model save

Browse files
Files changed (9) hide show
  1. README.md +38 -37
  2. all_results.json +17 -4
  3. eval_results.json +16 -0
  4. model.safetensors +1 -1
  5. runs/Oct05_10-19-21_xe8545-a100-14/events.out.tfevents.1728129481.xe8545-a100-14.960812.1 +3 -0
  6. runs/Oct06_18-38-13_xe8545-a100-22/events.out.tfevents.1728233315.xe8545-a100-22.2425886.0 +3 -0
  7. train_results.json +4 -4
  8. trainer_state.json +0 -0
  9. training_args.bin +1 -1
README.md CHANGED
@@ -3,6 +3,7 @@ library_name: transformers
3
  tags:
4
  - trl
5
  - dpo
 
6
  - generated_from_trainer
7
  model-index:
8
  - name: OpenELM-1_1B-DPO-full-max-6-reward
@@ -16,15 +17,15 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  This model was trained from scratch on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
- - Loss: 1.8393
20
- - Rewards/chosen: -14.6875
21
- - Rewards/rejected: -16.375
22
- - Rewards/accuracies: 0.5938
23
- - Rewards/margins: 1.7266
24
- - Logps/rejected: -1928.0
25
- - Logps/chosen: -1784.0
26
- - Logits/rejected: 4.75
27
- - Logits/chosen: 3.0
28
 
29
  ## Model description
30
 
@@ -61,34 +62,34 @@ The following hyperparameters were used during training:
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
- | 0.5941 | 0.1047 | 100 | 0.6786 | -1.2578 | -1.4453 | 0.5723 | 0.1846 | -434.0 | -444.0 | -12.0625 | -12.3125 |
65
- | 0.5458 | 0.2094 | 200 | 0.7028 | -1.6016 | -1.8203 | 0.5977 | 0.2129 | -470.0 | -478.0 | -11.625 | -12.125 |
66
- | 0.5451 | 0.3141 | 300 | 0.7219 | -3.5 | -3.8438 | 0.6016 | 0.3516 | -672.0 | -668.0 | -9.125 | -9.75 |
67
- | 0.5221 | 0.4188 | 400 | 0.8154 | -3.2188 | -3.7656 | 0.5859 | 0.5430 | -664.0 | -640.0 | -5.2812 | -6.375 |
68
- | 0.5128 | 0.5236 | 500 | 0.7688 | -3.3281 | -3.9219 | 0.6094 | 0.5859 | -680.0 | -652.0 | -5.5312 | -6.5938 |
69
- | 0.5178 | 0.6283 | 600 | 0.7540 | -3.8594 | -4.3438 | 0.5820 | 0.4902 | -724.0 | -704.0 | -5.5938 | -6.6875 |
70
- | 0.4831 | 0.7330 | 700 | 0.7995 | -4.6562 | -5.3438 | 0.6055 | 0.6836 | -824.0 | -784.0 | -4.0625 | -5.3125 |
71
- | 0.5004 | 0.8377 | 800 | 0.7362 | -4.2812 | -4.9062 | 0.6152 | 0.6094 | -776.0 | -748.0 | -6.125 | -7.4375 |
72
- | 0.4651 | 0.9424 | 900 | 0.8777 | -6.8438 | -7.5938 | 0.5879 | 0.7461 | -1048.0 | -1000.0 | -3.4375 | -5.0 |
73
- | 0.1407 | 1.0471 | 1000 | 1.0909 | -8.125 | -9.0 | 0.5801 | 0.9297 | -1192.0 | -1128.0 | -0.5938 | -2.2188 |
74
- | 0.1332 | 1.1518 | 1100 | 1.0452 | -8.5 | -9.375 | 0.5859 | 0.8945 | -1224.0 | -1168.0 | -0.7070 | -2.6562 |
75
- | 0.1397 | 1.2565 | 1200 | 1.0393 | -7.6875 | -8.5625 | 0.6016 | 0.8594 | -1144.0 | -1088.0 | -0.2539 | -2.0938 |
76
- | 0.1353 | 1.3613 | 1300 | 1.0452 | -7.875 | -8.8125 | 0.5938 | 0.9609 | -1168.0 | -1104.0 | -1.7656 | -3.6406 |
77
- | 0.122 | 1.4660 | 1400 | 1.1136 | -9.125 | -10.125 | 0.5957 | 0.9766 | -1304.0 | -1232.0 | -0.2578 | -2.1875 |
78
- | 0.1192 | 1.5707 | 1500 | 1.1467 | -9.75 | -10.9375 | 0.5938 | 1.1719 | -1376.0 | -1296.0 | 3.0312 | 1.1953 |
79
- | 0.12 | 1.6754 | 1600 | 1.2342 | -9.75 | -11.0625 | 0.6074 | 1.2656 | -1392.0 | -1296.0 | 1.2031 | -0.7930 |
80
- | 0.0994 | 1.7801 | 1700 | 1.1657 | -9.4375 | -10.625 | 0.6172 | 1.2266 | -1352.0 | -1256.0 | 1.9219 | 0.1504 |
81
- | 0.1305 | 1.8848 | 1800 | 1.0662 | -7.9375 | -8.875 | 0.5801 | 0.9648 | -1176.0 | -1112.0 | 1.3438 | -0.4492 |
82
- | 0.1205 | 1.9895 | 1900 | 1.1693 | -9.8125 | -10.9375 | 0.5840 | 1.1328 | -1384.0 | -1296.0 | 2.5625 | 0.8477 |
83
- | 0.0161 | 2.0942 | 2000 | 1.7204 | -14.1875 | -15.75 | 0.5918 | 1.5625 | -1864.0 | -1736.0 | 4.625 | 2.8438 |
84
- | 0.0152 | 2.1990 | 2100 | 1.5836 | -13.125 | -14.6875 | 0.5918 | 1.5938 | -1760.0 | -1624.0 | 4.0625 | 2.3281 |
85
- | 0.0138 | 2.3037 | 2200 | 1.7812 | -14.125 | -15.6875 | 0.5879 | 1.6016 | -1856.0 | -1728.0 | 4.5625 | 2.8125 |
86
- | 0.0157 | 2.4084 | 2300 | 1.7642 | -14.1875 | -15.75 | 0.5820 | 1.5547 | -1864.0 | -1736.0 | 4.5 | 2.8125 |
87
- | 0.0147 | 2.5131 | 2400 | 1.8053 | -14.5 | -16.25 | 0.5938 | 1.6953 | -1912.0 | -1768.0 | 4.8125 | 3.0938 |
88
- | 0.0161 | 2.6178 | 2500 | 1.9172 | -15.25 | -17.0 | 0.5898 | 1.75 | -1992.0 | -1848.0 | 4.875 | 3.1562 |
89
- | 0.014 | 2.7225 | 2600 | 1.8166 | -14.375 | -16.125 | 0.5957 | 1.6953 | -1896.0 | -1760.0 | 4.625 | 2.8906 |
90
- | 0.0166 | 2.8272 | 2700 | 1.8285 | -14.625 | -16.375 | 0.5938 | 1.7109 | -1920.0 | -1776.0 | 4.7188 | 3.0 |
91
- | 0.018 | 2.9319 | 2800 | 1.8393 | -14.6875 | -16.375 | 0.5938 | 1.7266 | -1928.0 | -1784.0 | 4.75 | 3.0 |
92
 
93
 
94
  ### Framework versions
 
3
  tags:
4
  - trl
5
  - dpo
6
+ - alignment-handbook
7
  - generated_from_trainer
8
  model-index:
9
  - name: OpenELM-1_1B-DPO-full-max-6-reward
 
17
 
18
  This model was trained from scratch on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
+ - Loss: 1.8412
21
+ - Rewards/chosen: -15.5625
22
+ - Rewards/rejected: -17.375
23
+ - Rewards/accuracies: 0.6035
24
+ - Rewards/margins: 1.7891
25
+ - Logps/rejected: -2024.0
26
+ - Logps/chosen: -1872.0
27
+ - Logits/rejected: 1.625
28
+ - Logits/chosen: -0.2451
29
 
30
  ## Model description
31
 
 
62
 
63
  | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
64
  |:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
65
+ | 0.5954 | 0.1047 | 100 | 0.6902 | -1.1172 | -1.2578 | 0.5742 | 0.1416 | -414.0 | -430.0 | -10.75 | -11.0625 |
66
+ | 0.5396 | 0.2094 | 200 | 0.6928 | -2.5156 | -2.8594 | 0.6387 | 0.3438 | -576.0 | -572.0 | -9.6875 | -10.25 |
67
+ | 0.5587 | 0.3141 | 300 | 0.7014 | -3.7188 | -4.1562 | 0.6094 | 0.4355 | -704.0 | -692.0 | -8.5625 | -9.4375 |
68
+ | 0.5067 | 0.4188 | 400 | 0.7869 | -3.875 | -4.375 | 0.5996 | 0.4961 | -724.0 | -708.0 | -14.0 | -14.9375 |
69
+ | 0.525 | 0.5236 | 500 | 0.8500 | -4.625 | -5.3438 | 0.6094 | 0.7109 | -824.0 | -780.0 | -6.8438 | -8.25 |
70
+ | 0.5173 | 0.6283 | 600 | 0.7292 | -6.25 | -6.7188 | 0.5723 | 0.4688 | -960.0 | -944.0 | -8.875 | -10.0 |
71
+ | 0.4944 | 0.7330 | 700 | 0.7881 | -5.1562 | -5.7812 | 0.6035 | 0.6445 | -868.0 | -832.0 | -6.4062 | -8.0 |
72
+ | 0.5113 | 0.8377 | 800 | 0.7106 | -4.7812 | -5.3438 | 0.6113 | 0.5586 | -824.0 | -796.0 | -9.4375 | -10.8125 |
73
+ | 0.4589 | 0.9424 | 900 | 0.8807 | -7.4375 | -8.1875 | 0.6094 | 0.7656 | -1112.0 | -1064.0 | -6.1875 | -8.125 |
74
+ | 0.1368 | 1.0471 | 1000 | 1.1006 | -8.375 | -9.4375 | 0.5879 | 1.0547 | -1232.0 | -1160.0 | -3.4531 | -5.375 |
75
+ | 0.138 | 1.1518 | 1100 | 1.0286 | -8.375 | -9.375 | 0.5977 | 0.9531 | -1224.0 | -1160.0 | -4.0938 | -5.9688 |
76
+ | 0.1376 | 1.2565 | 1200 | 1.0962 | -8.6875 | -9.75 | 0.6035 | 1.0312 | -1264.0 | -1192.0 | -1.2266 | -3.0781 |
77
+ | 0.1434 | 1.3613 | 1300 | 1.1220 | -9.375 | -10.5 | 0.5801 | 1.1172 | -1336.0 | -1256.0 | -3.7031 | -5.6875 |
78
+ | 0.1386 | 1.4660 | 1400 | 1.0638 | -9.4375 | -10.375 | 0.6230 | 0.9570 | -1328.0 | -1256.0 | -3.5 | -5.4688 |
79
+ | 0.1258 | 1.5707 | 1500 | 1.1923 | -10.5 | -11.75 | 0.6016 | 1.1953 | -1464.0 | -1368.0 | -2.4062 | -4.5625 |
80
+ | 0.1269 | 1.6754 | 1600 | 1.2009 | -9.4375 | -10.625 | 0.6074 | 1.1562 | -1352.0 | -1264.0 | -2.8438 | -5.2188 |
81
+ | 0.0967 | 1.7801 | 1700 | 1.1723 | -10.0 | -11.125 | 0.5996 | 1.0859 | -1400.0 | -1320.0 | -1.6328 | -3.6406 |
82
+ | 0.112 | 1.8848 | 1800 | 1.0807 | -9.75 | -10.75 | 0.5898 | 0.9805 | -1360.0 | -1296.0 | -2.5 | -4.5 |
83
+ | 0.1158 | 1.9895 | 1900 | 1.1470 | -10.875 | -12.0625 | 0.5938 | 1.2109 | -1496.0 | -1400.0 | -1.5391 | -3.5625 |
84
+ | 0.0172 | 2.0942 | 2000 | 1.6192 | -14.1875 | -15.6875 | 0.6055 | 1.5078 | -1864.0 | -1736.0 | 0.8438 | -1.1172 |
85
+ | 0.012 | 2.1990 | 2100 | 1.7070 | -14.6875 | -16.375 | 0.6016 | 1.6953 | -1928.0 | -1792.0 | 0.5117 | -1.4688 |
86
+ | 0.0145 | 2.3037 | 2200 | 1.6657 | -14.0625 | -15.625 | 0.5957 | 1.5547 | -1856.0 | -1728.0 | 0.6875 | -1.2891 |
87
+ | 0.0161 | 2.4084 | 2300 | 1.8217 | -15.5625 | -17.25 | 0.6035 | 1.7344 | -2016.0 | -1872.0 | 1.0 | -0.9141 |
88
+ | 0.0161 | 2.5131 | 2400 | 1.7852 | -15.0 | -16.625 | 0.6055 | 1.6641 | -1960.0 | -1824.0 | 1.6328 | -0.2471 |
89
+ | 0.0182 | 2.6178 | 2500 | 1.9600 | -16.25 | -18.125 | 0.5957 | 1.8125 | -2096.0 | -1952.0 | 1.7578 | -0.1089 |
90
+ | 0.0121 | 2.7225 | 2600 | 1.8076 | -15.125 | -16.875 | 0.6113 | 1.7656 | -1976.0 | -1832.0 | 1.4922 | -0.4238 |
91
+ | 0.016 | 2.8272 | 2700 | 1.8344 | -15.5 | -17.25 | 0.6055 | 1.7891 | -2016.0 | -1872.0 | 1.6016 | -0.2773 |
92
+ | 0.0144 | 2.9319 | 2800 | 1.8412 | -15.5625 | -17.375 | 0.6035 | 1.7891 | -2024.0 | -1872.0 | 1.625 | -0.2451 |
93
 
94
 
95
  ### Framework versions
all_results.json CHANGED
@@ -1,9 +1,22 @@
1
  {
2
  "epoch": 3.0,
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  "total_flos": 0.0,
4
- "train_loss": 0.22500331487643158,
5
- "train_runtime": 12392.8387,
6
  "train_samples": 61119,
7
- "train_samples_per_second": 14.795,
8
- "train_steps_per_second": 0.231
9
  }
 
1
  {
2
  "epoch": 3.0,
3
+ "eval_logits/chosen": 3.015625,
4
+ "eval_logits/rejected": 4.75,
5
+ "eval_logps/chosen": -1784.0,
6
+ "eval_logps/rejected": -1928.0,
7
+ "eval_loss": 1.839203119277954,
8
+ "eval_rewards/accuracies": 0.595703125,
9
+ "eval_rewards/chosen": -14.6875,
10
+ "eval_rewards/margins": 1.71875,
11
+ "eval_rewards/rejected": -16.375,
12
+ "eval_runtime": 46.9056,
13
+ "eval_samples": 2000,
14
+ "eval_samples_per_second": 42.639,
15
+ "eval_steps_per_second": 0.682,
16
  "total_flos": 0.0,
17
+ "train_loss": 0.22363413207193944,
18
+ "train_runtime": 12304.9,
19
  "train_samples": 61119,
20
+ "train_samples_per_second": 14.901,
21
+ "train_steps_per_second": 0.233
22
  }
eval_results.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_logits/chosen": 3.015625,
4
+ "eval_logits/rejected": 4.75,
5
+ "eval_logps/chosen": -1784.0,
6
+ "eval_logps/rejected": -1928.0,
7
+ "eval_loss": 1.839203119277954,
8
+ "eval_rewards/accuracies": 0.595703125,
9
+ "eval_rewards/chosen": -14.6875,
10
+ "eval_rewards/margins": 1.71875,
11
+ "eval_rewards/rejected": -16.375,
12
+ "eval_runtime": 46.9056,
13
+ "eval_samples": 2000,
14
+ "eval_samples_per_second": 42.639,
15
+ "eval_steps_per_second": 0.682
16
+ }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:53856c34129c4199094ac017a85281027298aa522e4f68b3762e6f94c22fcb20
3
  size 2159808696
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bde8085f5fc935e0166b933dad4ed4c22bfa1702ff312bd9b85cd0a8582aeee8
3
  size 2159808696
runs/Oct05_10-19-21_xe8545-a100-14/events.out.tfevents.1728129481.xe8545-a100-14.960812.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3485f65d26a8df8b67fc932b28d6c0819dfc1c10ec963a33eea1de0d2329e824
3
+ size 828
runs/Oct06_18-38-13_xe8545-a100-22/events.out.tfevents.1728233315.xe8545-a100-22.2425886.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:482f164e80096c75b7a397f83f1516669a8495bad7f69327209187761e0435a7
3
+ size 225707
train_results.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "epoch": 3.0,
3
  "total_flos": 0.0,
4
- "train_loss": 0.22500331487643158,
5
- "train_runtime": 12392.8387,
6
  "train_samples": 61119,
7
- "train_samples_per_second": 14.795,
8
- "train_steps_per_second": 0.231
9
  }
 
1
  {
2
  "epoch": 3.0,
3
  "total_flos": 0.0,
4
+ "train_loss": 0.22363413207193944,
5
+ "train_runtime": 12304.9,
6
  "train_samples": 61119,
7
+ "train_samples_per_second": 14.901,
8
+ "train_steps_per_second": 0.233
9
  }
trainer_state.json CHANGED
The diff for this file is too large to render. See raw diff
 
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:159476079e4e1992c50475f53405778e870792e4663fca96206177c6b9b0a67d
3
  size 7672
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:547a123c1713088e9e5b6c28a6e2a133b83277b263db3cf075ce86652aa78afc
3
  size 7672