Hugging Face's logo

nnheui
/
pythia-1.4b-dpo-full

Text Generation
Transformers
TensorBoard
Safetensors
gpt_neox
alignment-handbook
trl
dpo
Generated from Trainer
conversational
text-generation-inference
Inference Endpoints
Model card Files Metrics Community
nnheui commited on
Commit
c78836d
•
1 Parent(s): f6046e2

Model save

Browse files
Files changed (7) hide show
  1. README.md +28 -39
  2. all_results.json +7 -19
  3. generation_config.json +1 -1
  4. model.safetensors +1 -1
  5. runs/Jul08_06-23-29_42dbe5cf9ed4/events.out.tfevents.1720420363.42dbe5cf9ed4.848884.0 +2 -2
  6. train_results.json +7 -6
  7. trainer_state.json +0 -0
README.md CHANGED
@@ -1,4 +1,6 @@
1
  ---
 
 
2
  tags:
3
  - trl
4
  - dpo
@@ -13,17 +15,18 @@ should probably proofread and complete it, then remove this comment. -->
13
 
14
  # pythia-1.4b-dpo-full
15
 
16
- This model was trained from scratch on the None dataset.
17
  It achieves the following results on the evaluation set:
18
- - Loss: 0.6403
19
- - Rewards/chosen: 0.6094
20
- - Rewards/rejected: 0.4102
21
- - Rewards/accuracies: 0.5893
22
- - Rewards/margins: 0.2002
23
- - Logps/rejected: -2024.0
24
- - Logps/chosen: -2320.0
25
- - Logits/rejected: -0.6719
26
- - Logits/chosen: -0.6172
 
27
 
28
  ## Model description
29
 
@@ -44,12 +47,13 @@ More information needed
44
  The following hyperparameters were used during training:
45
  - learning_rate: 5e-07
46
  - train_batch_size: 5
47
- - eval_batch_size: 8
48
  - seed: 42
49
  - distributed_type: multi-GPU
50
  - num_devices: 6
51
- - total_train_batch_size: 30
52
- - total_eval_batch_size: 48
 
53
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
54
  - lr_scheduler_type: cosine
55
  - lr_scheduler_warmup_ratio: 0.1
@@ -57,33 +61,18 @@ The following hyperparameters were used during training:
57
 
58
  ### Training results
59
 
60
- | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
61
- |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
62
- | 0.684 | 0.05 | 100 | 0.6768 | 0.2314 | 0.1904 | 0.4494 | 0.0405 | -2048.0 | -2352.0 | -0.7227 | -0.6641 |
63
- | 0.663 | 0.1 | 200 | 0.6566 | 0.5977 | 0.4883 | 0.4940 | 0.1108 | -2016.0 | -2320.0 | -0.7266 | -0.6680 |
64
- | 0.6529 | 0.15 | 300 | 0.6513 | 0.625 | 0.4941 | 0.5149 | 0.1279 | -2016.0 | -2320.0 | -0.7188 | -0.6562 |
65
- | 0.6371 | 0.2 | 400 | 0.6491 | 0.6562 | 0.5 | 0.5595 | 0.1523 | -2016.0 | -2304.0 | -0.7266 | -0.6680 |
66
- | 0.6206 | 0.25 | 500 | 0.6466 | 0.5391 | 0.3945 | 0.5952 | 0.1445 | -2024.0 | -2320.0 | -0.7148 | -0.6562 |
67
- | 0.686 | 0.29 | 600 | 0.6446 | 0.5781 | 0.4180 | 0.5714 | 0.1592 | -2024.0 | -2320.0 | -0.7188 | -0.6602 |
68
- | 0.6459 | 0.34 | 700 | 0.6449 | 0.5508 | 0.3633 | 0.6012 | 0.1885 | -2032.0 | -2320.0 | -0.6875 | -0.6289 |
69
- | 0.6458 | 0.39 | 800 | 0.6421 | 0.5586 | 0.3867 | 0.5774 | 0.1709 | -2024.0 | -2320.0 | -0.6953 | -0.6406 |
70
- | 0.6451 | 0.44 | 900 | 0.6398 | 0.7109 | 0.5039 | 0.5685 | 0.2070 | -2016.0 | -2304.0 | -0.6719 | -0.6133 |
71
- | 0.6213 | 0.49 | 1000 | 0.6407 | 0.7734 | 0.5742 | 0.5714 | 0.2012 | -2008.0 | -2304.0 | -0.6602 | -0.6016 |
72
- | 0.6313 | 0.54 | 1100 | 0.6387 | 0.5391 | 0.3555 | 0.5893 | 0.1807 | -2032.0 | -2320.0 | -0.6680 | -0.6094 |
73
- | 0.6298 | 0.59 | 1200 | 0.6380 | 0.6953 | 0.4922 | 0.6042 | 0.2031 | -2016.0 | -2304.0 | -0.6523 | -0.5977 |
74
- | 0.6461 | 0.64 | 1300 | 0.6396 | 0.5586 | 0.3613 | 0.5863 | 0.1963 | -2032.0 | -2320.0 | -0.6914 | -0.6367 |
75
- | 0.6258 | 0.69 | 1400 | 0.6360 | 0.6914 | 0.4727 | 0.5923 | 0.2207 | -2016.0 | -2304.0 | -0.6758 | -0.6172 |
76
- | 0.6347 | 0.74 | 1500 | 0.6375 | 0.625 | 0.4141 | 0.5893 | 0.2100 | -2024.0 | -2320.0 | -0.6641 | -0.6094 |
77
- | 0.6185 | 0.79 | 1600 | 0.6382 | 0.5977 | 0.3926 | 0.6042 | 0.2051 | -2032.0 | -2320.0 | -0.6797 | -0.625 |
78
- | 0.6408 | 0.83 | 1700 | 0.6374 | 0.5977 | 0.3926 | 0.5952 | 0.2041 | -2024.0 | -2320.0 | -0.6719 | -0.6172 |
79
- | 0.662 | 0.88 | 1800 | 0.6355 | 0.6094 | 0.3984 | 0.6012 | 0.2119 | -2024.0 | -2320.0 | -0.6836 | -0.6289 |
80
- | 0.6385 | 0.93 | 1900 | 0.6379 | 0.6055 | 0.3926 | 0.625 | 0.2129 | -2024.0 | -2320.0 | -0.6758 | -0.6211 |
81
- | 0.6154 | 0.98 | 2000 | 0.6381 | 0.6094 | 0.4043 | 0.6012 | 0.2041 | -2024.0 | -2320.0 | -0.6758 | -0.6211 |
82
 
83
 
84
  ### Framework versions
85
 
86
- - Transformers 4.38.2
87
- - Pytorch 2.2.1
88
- - Datasets 2.14.6
89
- - Tokenizers 0.15.2
 
1
  ---
2
+ license: apache-2.0
3
+ base_model: nnheui/pythia-1.4b-sft-full
4
  tags:
5
  - trl
6
  - dpo
 
15
 
16
  # pythia-1.4b-dpo-full
17
 
18
+ This model is a fine-tuned version of [nnheui/pythia-1.4b-sft-full](https://huggingface.co/nnheui/pythia-1.4b-sft-full) on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
+ - Loss: 0.6259
21
+ - Rewards/chosen: -0.5234
22
+ - Rewards/rejected: -0.7852
23
+ - Rewards/accuracies: 0.6567
24
+ - Rewards/margins: 0.2617
25
+ - Logps/rejected: -416.0
26
+ - Logps/chosen: -446.0
27
+ - Logits/rejected: -1.2422
28
+ - Logits/chosen: -1.1953
29
+ - Logps/bottom Tokens: -0.0007
30
 
31
  ## Model description
32
 
 
47
  The following hyperparameters were used during training:
48
  - learning_rate: 5e-07
49
  - train_batch_size: 5
50
+ - eval_batch_size: 5
51
  - seed: 42
52
  - distributed_type: multi-GPU
53
  - num_devices: 6
54
+ - gradient_accumulation_steps: 4
55
+ - total_train_batch_size: 120
56
+ - total_eval_batch_size: 30
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
 
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 | Logps/bottom Tokens |
65
+ |:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|:-------------------:|
66
+ | 0.678 | 0.1963 | 100 | 0.6789 | -0.0275 | -0.0608 | 0.5881 | 0.0332 | -344.0 | -396.0 | -1.1562 | -1.0938 | -0.0009 |
67
+ | 0.645 | 0.3925 | 200 | 0.6489 | -0.2871 | -0.4238 | 0.6448 | 0.1367 | -380.0 | -422.0 | -1.2031 | -1.1562 | -0.0009 |
68
+ | 0.6396 | 0.5888 | 300 | 0.6304 | -0.4512 | -0.6797 | 0.6627 | 0.2275 | -406.0 | -438.0 | -1.2344 | -1.1875 | -0.0008 |
69
+ | 0.6102 | 0.7851 | 400 | 0.6268 | -0.5039 | -0.7617 | 0.6567 | 0.2578 | -414.0 | -444.0 | -1.2344 | -1.1875 | -0.0007 |
70
+ | 0.6084 | 0.9814 | 500 | 0.6259 | -0.5234 | -0.7852 | 0.6567 | 0.2617 | -416.0 | -446.0 | -1.2422 | -1.1953 | -0.0007 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
71
 
72
 
73
  ### Framework versions
74
 
75
+ - Transformers 4.40.0
76
+ - Pytorch 2.2.2+cu121
77
+ - Datasets 2.19.0
78
+ - Tokenizers 0.19.1
all_results.json CHANGED
@@ -1,21 +1,9 @@
1
  {
2
- "epoch": 1.0,
3
- "eval_logits/chosen": -0.6171875,
4
- "eval_logits/rejected": -0.671875,
5
- "eval_logps/chosen": -2320.0,
6
- "eval_logps/rejected": -2024.0,
7
- "eval_loss": 0.6403203010559082,
8
- "eval_rewards/accuracies": 0.5892857313156128,
9
- "eval_rewards/chosen": 0.609375,
10
- "eval_rewards/margins": 0.2001953125,
11
- "eval_rewards/rejected": 0.41015625,
12
- "eval_runtime": 85.4736,
13
- "eval_samples": 2000,
14
- "eval_samples_per_second": 23.399,
15
- "eval_steps_per_second": 0.491,
16
- "train_loss": 0.6502101503246783,
17
- "train_runtime": 8979.6364,
18
- "train_samples": 61135,
19
- "train_samples_per_second": 6.808,
20
- "train_steps_per_second": 0.227
21
  }
 
1
  {
2
+ "epoch": 0.9990186457311089,
3
+ "total_flos": 0.0,
4
+ "train_loss": 0.6464882252961105,
5
+ "train_runtime": 7224.3746,
6
+ "train_samples": 61134,
7
+ "train_samples_per_second": 8.462,
8
+ "train_steps_per_second": 0.07
 
 
 
 
 
 
 
 
 
 
 
 
9
  }
generation_config.json CHANGED
@@ -2,6 +2,6 @@
2
  "_from_model_config": true,
3
  "bos_token_id": 0,
4
  "eos_token_id": 0,
5
- "transformers_version": "4.38.2",
6
  "use_cache": false
7
  }
 
2
  "_from_model_config": true,
3
  "bos_token_id": 0,
4
  "eos_token_id": 0,
5
+ "transformers_version": "4.40.0",
6
  "use_cache": false
7
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:811b22c69714101dba0a13cf905313113be9cd2fb0da880dd2a6f3f027fb2922
3
  size 2829330208
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dee645992f24ee02b486f2e81b344b7a98df284d1f79aa4f2f1679fdd185f99d
3
  size 2829330208
runs/Jul08_06-23-29_42dbe5cf9ed4/events.out.tfevents.1720420363.42dbe5cf9ed4.848884.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:39f21f40447c4a76a762c3adaa79e5bdf2e91c9d12f7b57ad3c51ba1751abf1d
3
- size 47174
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e1e67859856cf3694e01f54ca2dccfaa3c3cd4bdb09f64aeaccc2bae9c86f7ec
3
+ size 47528
train_results.json CHANGED
@@ -1,8 +1,9 @@
1
  {
2
- "epoch": 1.0,
3
- "train_loss": 0.6502101503246783,
4
- "train_runtime": 8979.6364,
5
- "train_samples": 61135,
6
- "train_samples_per_second": 6.808,
7
- "train_steps_per_second": 0.227
 
8
  }
 
1
  {
2
+ "epoch": 0.9990186457311089,
3
+ "total_flos": 0.0,
4
+ "train_loss": 0.6464882252961105,
5
+ "train_runtime": 7224.3746,
6
+ "train_samples": 61134,
7
+ "train_samples_per_second": 8.462,
8
+ "train_steps_per_second": 0.07
9
  }
trainer_state.json CHANGED
The diff for this file is too large to render. See raw diff