Hugging Face's logo

CharlesLi
/
OpenELM-1_1B-DPO-full-max-4-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
7d13ca1
•
1 Parent(s): 1643ab1

Model save

Browse files
Files changed (9) hide show
  1. README.md +45 -44
  2. all_results.json +16 -3
  3. eval_results.json +16 -0
  4. model.safetensors +1 -1
  5. runs/Oct05_06-10-41_xe8545-a100-14/events.out.tfevents.1728114897.xe8545-a100-14.957316.1 +3 -0
  6. runs/Oct06_14-30-18_xe8545-a100-22/events.out.tfevents.1728218423.xe8545-a100-22.2422281.0 +3 -0
  7. train_results.json +3 -3
  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-4-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: 1.5952
20
- - Rewards/chosen: -13.125
21
- - Rewards/rejected: -14.4375
22
- - Rewards/accuracies: 0.6035
23
- - Rewards/margins: 1.3047
24
- - Logps/rejected: -1728.0
25
- - Logps/chosen: -1632.0
26
- - Logits/rejected: 2.4062
27
- - Logits/chosen: 0.5391
28
 
29
  ## Model description
30
 
@@ -61,41 +62,41 @@ 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
- | 0.6374 | 0.0838 | 80 | 0.6876 | -0.6875 | -0.7773 | 0.5664 | 0.0908 | -366.0 | -386.0 | -9.8125 | -10.125 |
65
- | 0.6205 | 0.1675 | 160 | 0.6953 | -1.2266 | -1.375 | 0.5840 | 0.1475 | -426.0 | -440.0 | -11.4375 | -11.75 |
66
- | 0.6209 | 0.2513 | 240 | 0.7235 | -1.6016 | -1.7578 | 0.5762 | 0.1553 | -464.0 | -478.0 | -12.125 | -12.375 |
67
- | 0.6605 | 0.3351 | 320 | 0.7567 | -2.7969 | -3.0469 | 0.5996 | 0.2461 | -592.0 | -600.0 | -12.375 | -12.6875 |
68
- | 0.6471 | 0.4188 | 400 | 0.7148 | -2.7031 | -2.875 | 0.5801 | 0.1816 | -576.0 | -588.0 | -12.0625 | -12.5 |
69
- | 0.6121 | 0.5026 | 480 | 0.7704 | -4.1562 | -4.5625 | 0.5879 | 0.3945 | -744.0 | -736.0 | -7.4062 | -8.1875 |
70
- | 0.6059 | 0.5864 | 560 | 0.7471 | -4.125 | -4.4688 | 0.5957 | 0.3320 | -736.0 | -732.0 | -11.25 | -11.9375 |
71
- | 0.5698 | 0.6702 | 640 | 0.7281 | -3.3125 | -3.7344 | 0.6367 | 0.4297 | -660.0 | -648.0 | -14.1875 | -15.1875 |
72
- | 0.6151 | 0.7539 | 720 | 0.7413 | -2.8438 | -3.1562 | 0.5840 | 0.3105 | -604.0 | -604.0 | -12.625 | -13.0625 |
73
- | 0.5859 | 0.8377 | 800 | 0.7781 | -4.7812 | -5.2812 | 0.6074 | 0.4844 | -816.0 | -796.0 | -8.5 | -9.8125 |
74
- | 0.6049 | 0.9215 | 880 | 0.7388 | -2.8281 | -3.125 | 0.5977 | 0.2930 | -600.0 | -600.0 | -12.3125 | -12.6875 |
75
- | 0.4217 | 1.0052 | 960 | 0.7678 | -5.1875 | -5.8438 | 0.6465 | 0.6562 | -872.0 | -836.0 | -6.3125 | -7.8125 |
76
- | 0.1894 | 1.0890 | 1040 | 1.0973 | -6.9688 | -7.625 | 0.6074 | 0.6719 | -1056.0 | -1016.0 | -4.3438 | -6.0 |
77
- | 0.1959 | 1.1728 | 1120 | 0.9770 | -6.75 | -7.5 | 0.6133 | 0.7422 | -1040.0 | -992.0 | -3.9062 | -5.625 |
78
- | 0.17 | 1.2565 | 1200 | 1.0293 | -7.2188 | -7.9062 | 0.6094 | 0.6719 | -1080.0 | -1040.0 | -4.5625 | -6.0938 |
79
- | 0.1857 | 1.3403 | 1280 | 0.9556 | -7.0625 | -7.8125 | 0.5996 | 0.7578 | -1072.0 | -1024.0 | -5.75 | -7.3125 |
80
- | 0.1872 | 1.4241 | 1360 | 0.9190 | -7.25 | -8.0625 | 0.5938 | 0.8359 | -1096.0 | -1040.0 | -4.7812 | -6.4375 |
81
- | 0.1445 | 1.5079 | 1440 | 1.0569 | -9.0 | -9.9375 | 0.5996 | 0.9258 | -1280.0 | -1216.0 | -3.7656 | -5.3438 |
82
- | 0.136 | 1.5916 | 1520 | 1.0663 | -9.5625 | -10.4375 | 0.6191 | 0.9219 | -1336.0 | -1272.0 | -2.3125 | -4.0312 |
83
- | 0.1765 | 1.6754 | 1600 | 1.0288 | -8.0625 | -8.9375 | 0.6133 | 0.875 | -1184.0 | -1128.0 | -2.6562 | -4.5312 |
84
- | 0.1661 | 1.7592 | 1680 | 1.0917 | -8.0625 | -8.9375 | 0.6035 | 0.8633 | -1184.0 | -1128.0 | -2.7656 | -4.6562 |
85
- | 0.1451 | 1.8429 | 1760 | 1.0870 | -8.375 | -9.25 | 0.5957 | 0.8867 | -1216.0 | -1152.0 | -2.9688 | -4.6875 |
86
- | 0.1712 | 1.9267 | 1840 | 1.0650 | -8.6875 | -9.6875 | 0.6172 | 0.9922 | -1256.0 | -1184.0 | -2.4375 | -4.3125 |
87
- | 0.0278 | 2.0105 | 1920 | 1.0530 | -8.875 | -9.875 | 0.6152 | 0.9805 | -1272.0 | -1208.0 | -1.8906 | -3.8125 |
88
- | 0.0225 | 2.0942 | 2000 | 1.4602 | -11.5 | -12.5625 | 0.6035 | 1.0312 | -1544.0 | -1472.0 | 0.4297 | -1.5547 |
89
- | 0.0182 | 2.1780 | 2080 | 1.5544 | -12.6875 | -13.8125 | 0.5977 | 1.1172 | -1672.0 | -1592.0 | 1.7266 | -0.1621 |
90
- | 0.0385 | 2.2618 | 2160 | 1.5476 | -13.0 | -14.1875 | 0.6016 | 1.1953 | -1712.0 | -1616.0 | 1.5859 | -0.2598 |
91
- | 0.0162 | 2.3455 | 2240 | 1.5637 | -12.8125 | -13.9375 | 0.6016 | 1.1641 | -1688.0 | -1600.0 | 1.8984 | 0.0913 |
92
- | 0.0239 | 2.4293 | 2320 | 1.4822 | -11.9375 | -13.0625 | 0.5938 | 1.1797 | -1600.0 | -1512.0 | 1.2422 | -0.6172 |
93
- | 0.0264 | 2.5131 | 2400 | 1.6307 | -13.375 | -14.6875 | 0.6035 | 1.3047 | -1760.0 | -1656.0 | 2.4688 | 0.6406 |
94
- | 0.024 | 2.5969 | 2480 | 1.5421 | -12.3125 | -13.5625 | 0.5996 | 1.2188 | -1640.0 | -1552.0 | 1.7422 | -0.1631 |
95
- | 0.0217 | 2.6806 | 2560 | 1.6061 | -13.0 | -14.25 | 0.6035 | 1.2656 | -1720.0 | -1616.0 | 2.2656 | 0.3770 |
96
- | 0.0208 | 2.7644 | 2640 | 1.5995 | -13.1875 | -14.4375 | 0.6055 | 1.2812 | -1736.0 | -1632.0 | 2.5156 | 0.6602 |
97
- | 0.0206 | 2.8482 | 2720 | 1.5964 | -13.1875 | -14.4375 | 0.6055 | 1.2969 | -1736.0 | -1632.0 | 2.4688 | 0.6133 |
98
- | 0.0163 | 2.9319 | 2800 | 1.5952 | -13.125 | -14.4375 | 0.6035 | 1.3047 | -1728.0 | -1632.0 | 2.4062 | 0.5391 |
99
 
100
 
101
  ### 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-4-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: 1.5703
21
+ - Rewards/chosen: -13.1875
22
+ - Rewards/rejected: -14.6875
23
+ - Rewards/accuracies: 0.5898
24
+ - Rewards/margins: 1.5
25
+ - Logps/rejected: -1760.0
26
+ - Logps/chosen: -1640.0
27
+ - Logits/rejected: 8.0
28
+ - Logits/chosen: 6.0312
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
+ | 0.6384 | 0.0838 | 80 | 0.6899 | -0.7109 | -0.7969 | 0.5566 | 0.0854 | -368.0 | -390.0 | -9.8125 | -10.125 |
66
+ | 0.6225 | 0.1675 | 160 | 0.7090 | -1.4844 | -1.6094 | 0.5508 | 0.1279 | -450.0 | -466.0 | -9.25 | -9.6875 |
67
+ | 0.6067 | 0.2513 | 240 | 0.7372 | -2.1875 | -2.375 | 0.5625 | 0.1865 | -528.0 | -540.0 | -6.75 | -7.5312 |
68
+ | 0.63 | 0.3351 | 320 | 0.7781 | -2.9688 | -3.1719 | 0.5723 | 0.2051 | -604.0 | -616.0 | -4.3125 | -5.2188 |
69
+ | 0.6294 | 0.4188 | 400 | 0.7330 | -2.5938 | -2.8594 | 0.5977 | 0.2773 | -576.0 | -576.0 | -4.6562 | -5.75 |
70
+ | 0.6199 | 0.5026 | 480 | 0.7532 | -3.3438 | -3.6719 | 0.5820 | 0.3301 | -656.0 | -652.0 | -7.9688 | -8.9375 |
71
+ | 0.5935 | 0.5864 | 560 | 0.7359 | -4.0625 | -4.4062 | 0.5859 | 0.3262 | -728.0 | -724.0 | -7.9062 | -8.875 |
72
+ | 0.5859 | 0.6702 | 640 | 0.7149 | -3.6719 | -4.125 | 0.6191 | 0.4648 | -704.0 | -684.0 | -5.125 | -6.4062 |
73
+ | 0.6127 | 0.7539 | 720 | 0.7057 | -3.7031 | -4.0312 | 0.6016 | 0.3320 | -692.0 | -688.0 | -6.875 | -8.0 |
74
+ | 0.6012 | 0.8377 | 800 | 0.7512 | -3.6406 | -3.9375 | 0.5938 | 0.2988 | -684.0 | -684.0 | -10.75 | -11.125 |
75
+ | 0.5948 | 0.9215 | 880 | 0.7496 | -3.0938 | -3.4531 | 0.5918 | 0.3457 | -632.0 | -628.0 | -4.9688 | -6.875 |
76
+ | 0.434 | 1.0052 | 960 | 0.7349 | -4.4062 | -4.9688 | 0.6289 | 0.5703 | -784.0 | -760.0 | -2.8906 | -4.75 |
77
+ | 0.1948 | 1.0890 | 1040 | 1.0377 | -7.0 | -7.75 | 0.6211 | 0.7539 | -1064.0 | -1016.0 | 1.125 | -0.8477 |
78
+ | 0.2139 | 1.1728 | 1120 | 0.9297 | -7.7188 | -8.375 | 0.6133 | 0.6562 | -1128.0 | -1088.0 | 3.2969 | 1.5469 |
79
+ | 0.1666 | 1.2565 | 1200 | 1.0332 | -8.375 | -9.0625 | 0.5840 | 0.7109 | -1200.0 | -1152.0 | 2.9219 | 1.0859 |
80
+ | 0.1994 | 1.3403 | 1280 | 0.9937 | -7.4688 | -8.25 | 0.5918 | 0.7969 | -1112.0 | -1064.0 | 0.4863 | -1.6797 |
81
+ | 0.1836 | 1.4241 | 1360 | 0.9652 | -7.4688 | -8.3125 | 0.6074 | 0.8359 | -1120.0 | -1064.0 | 1.3984 | -0.7383 |
82
+ | 0.153 | 1.5079 | 1440 | 1.1797 | -9.6875 | -10.625 | 0.5938 | 0.9102 | -1352.0 | -1288.0 | 4.0 | 1.9219 |
83
+ | 0.1398 | 1.5916 | 1520 | 1.0459 | -9.0 | -9.9375 | 0.5996 | 0.9609 | -1280.0 | -1216.0 | 4.1875 | 2.1562 |
84
+ | 0.1634 | 1.6754 | 1600 | 1.0131 | -9.0625 | -10.0 | 0.5977 | 0.9375 | -1288.0 | -1224.0 | 2.4062 | 0.1006 |
85
+ | 0.1777 | 1.7592 | 1680 | 1.0902 | -8.6875 | -9.625 | 0.625 | 0.9219 | -1248.0 | -1192.0 | 3.5312 | 1.1875 |
86
+ | 0.1536 | 1.8429 | 1760 | 1.0046 | -9.125 | -10.0625 | 0.6074 | 0.9180 | -1296.0 | -1232.0 | 3.8281 | 1.6797 |
87
+ | 0.1713 | 1.9267 | 1840 | 1.0627 | -9.3125 | -10.375 | 0.6230 | 1.0781 | -1328.0 | -1248.0 | 5.2812 | 3.0781 |
88
+ | 0.0282 | 2.0105 | 1920 | 1.0638 | -9.0625 | -10.125 | 0.6133 | 1.0625 | -1304.0 | -1224.0 | 4.125 | 1.8281 |
89
+ | 0.0234 | 2.0942 | 2000 | 1.3751 | -11.625 | -12.8125 | 0.5957 | 1.2266 | -1568.0 | -1480.0 | 6.5938 | 4.5 |
90
+ | 0.0167 | 2.1780 | 2080 | 1.4855 | -12.625 | -14.0 | 0.5977 | 1.375 | -1688.0 | -1584.0 | 7.5 | 5.4375 |
91
+ | 0.0366 | 2.2618 | 2160 | 1.4665 | -12.1875 | -13.5625 | 0.5957 | 1.3594 | -1648.0 | -1536.0 | 7.3125 | 5.25 |
92
+ | 0.0268 | 2.3455 | 2240 | 1.5465 | -12.8125 | -14.1875 | 0.5996 | 1.3594 | -1712.0 | -1600.0 | 7.6875 | 5.7188 |
93
+ | 0.0241 | 2.4293 | 2320 | 1.5025 | -12.25 | -13.6875 | 0.5996 | 1.4219 | -1656.0 | -1544.0 | 7.25 | 5.1875 |
94
+ | 0.0261 | 2.5131 | 2400 | 1.5371 | -12.75 | -14.1875 | 0.5938 | 1.4375 | -1712.0 | -1592.0 | 7.5312 | 5.5 |
95
+ | 0.0245 | 2.5969 | 2480 | 1.5284 | -12.625 | -14.0625 | 0.5977 | 1.4453 | -1696.0 | -1584.0 | 7.625 | 5.5625 |
96
+ | 0.0249 | 2.6806 | 2560 | 1.5392 | -12.8125 | -14.25 | 0.5938 | 1.4766 | -1712.0 | -1600.0 | 7.7812 | 5.7188 |
97
+ | 0.02 | 2.7644 | 2640 | 1.5571 | -13.0625 | -14.5625 | 0.5898 | 1.4922 | -1744.0 | -1624.0 | 7.9375 | 5.9375 |
98
+ | 0.0194 | 2.8482 | 2720 | 1.5688 | -13.1875 | -14.6875 | 0.5898 | 1.5 | -1760.0 | -1640.0 | 8.0625 | 6.0312 |
99
+ | 0.0163 | 2.9319 | 2800 | 1.5703 | -13.1875 | -14.6875 | 0.5898 | 1.5 | -1760.0 | -1640.0 | 8.0 | 6.0312 |
100
 
101
 
102
  ### Framework versions
all_results.json CHANGED
@@ -1,9 +1,22 @@
1
  {
2
  "epoch": 3.0,
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  "total_flos": 0.0,
4
- "train_loss": 0.27085763049583367,
5
- "train_runtime": 12609.4045,
6
  "train_samples": 61119,
7
- "train_samples_per_second": 14.541,
8
  "train_steps_per_second": 0.227
9
  }
 
1
  {
2
  "epoch": 3.0,
3
+ "eval_logits/chosen": 0.53125,
4
+ "eval_logits/rejected": 2.40625,
5
+ "eval_logps/chosen": -1632.0,
6
+ "eval_logps/rejected": -1728.0,
7
+ "eval_loss": 1.5954219102859497,
8
+ "eval_rewards/accuracies": 0.60546875,
9
+ "eval_rewards/chosen": -13.125,
10
+ "eval_rewards/margins": 1.3046875,
11
+ "eval_rewards/rejected": -14.4375,
12
+ "eval_runtime": 46.8279,
13
+ "eval_samples": 2000,
14
+ "eval_samples_per_second": 42.71,
15
+ "eval_steps_per_second": 0.683,
16
  "total_flos": 0.0,
17
+ "train_loss": 0.27230105622692674,
18
+ "train_runtime": 12616.2292,
19
  "train_samples": 61119,
20
+ "train_samples_per_second": 14.533,
21
  "train_steps_per_second": 0.227
22
  }
eval_results.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_logits/chosen": 0.53125,
4
+ "eval_logits/rejected": 2.40625,
5
+ "eval_logps/chosen": -1632.0,
6
+ "eval_logps/rejected": -1728.0,
7
+ "eval_loss": 1.5954219102859497,
8
+ "eval_rewards/accuracies": 0.60546875,
9
+ "eval_rewards/chosen": -13.125,
10
+ "eval_rewards/margins": 1.3046875,
11
+ "eval_rewards/rejected": -14.4375,
12
+ "eval_runtime": 46.8279,
13
+ "eval_samples": 2000,
14
+ "eval_samples_per_second": 42.71,
15
+ "eval_steps_per_second": 0.683
16
+ }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b5279a4114db3c0cf02d4d005fdbae5f5cba8e81cccad39af698472c5339bd01
3
  size 2159808696
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21670fe06ce3aa8d2e6bf2fd073f536f60f474a108bdf2afaada591e3297982b
3
  size 2159808696
runs/Oct05_06-10-41_xe8545-a100-14/events.out.tfevents.1728114897.xe8545-a100-14.957316.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9744ee6e45130ec88ac281a55659c8406be289a443c5126b4690febb197445c9
3
+ size 828
runs/Oct06_14-30-18_xe8545-a100-22/events.out.tfevents.1728218423.xe8545-a100-22.2422281.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9529b1e99fb367ab84445ab28666e76311f93d4cd35c4342e6fbaace9271134f
3
+ size 230886
train_results.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "epoch": 3.0,
3
  "total_flos": 0.0,
4
- "train_loss": 0.27085763049583367,
5
- "train_runtime": 12609.4045,
6
  "train_samples": 61119,
7
- "train_samples_per_second": 14.541,
8
  "train_steps_per_second": 0.227
9
  }
 
1
  {
2
  "epoch": 3.0,
3
  "total_flos": 0.0,
4
+ "train_loss": 0.27230105622692674,
5
+ "train_runtime": 12616.2292,
6
  "train_samples": 61119,
7
+ "train_samples_per_second": 14.533,
8
  "train_steps_per_second": 0.227
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 sha256:9a0e09b2bdd18fec963608b7d04e2cfaee9c4a87d680eff665e03af3b9e815a8
3
  size 7672
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbd3173082e109bde939d1add2fe75f094a12ea0232ad4b3ed9313a57e7e5a80
3
  size 7672