Hugging Face's logo

Kyleyee
/
Qwen2-0.5B-DPO-imdb_kl

Text Generation
Transformers
Safetensors
qwen2
Generated from Trainer
trl
dpo
conversational
text-generation-inference
Model card Files
xet
Community
Kyleyee commited on
Commit
7b0d923
·
verified ·
1 Parent(s): cd8d395

Model save

Browse files
Files changed (5) hide show
  1. README.md +2 -3
  2. all_results.json +5 -5
  3. eval_results.json +5 -5
  4. model.safetensors +1 -1
  5. training_args.bin +1 -1
README.md CHANGED
@@ -1,6 +1,5 @@
1
  ---
2
  base_model: Kyleyee/Qwen2-0.5B-stf-imdb
3
- datasets: Kyleyee/train_data_imdb_for_drdpo_preference
4
  library_name: transformers
5
  model_name: Qwen2-0.5B-DPO-imdb_kl
6
  tags:
@@ -12,7 +11,7 @@ licence: license
12
 
13
  # Model Card for Qwen2-0.5B-DPO-imdb_kl
14
 
15
- This model is a fine-tuned version of [Kyleyee/Qwen2-0.5B-stf-imdb](https://huggingface.co/Kyleyee/Qwen2-0.5B-stf-imdb) on the [Kyleyee/train_data_imdb_for_drdpo_preference](https://huggingface.co/datasets/Kyleyee/train_data_imdb_for_drdpo_preference) dataset.
16
  It has been trained using [TRL](https://github.com/huggingface/trl).
17
 
18
  ## Quick start
@@ -28,7 +27,7 @@ print(output["generated_text"])
28
 
29
  ## Training procedure
30
 
31
- [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/Tripleleg/huggingface/runs/2bj6yof2)
32
 
33
 
34
  This model was trained with DPO, a method introduced in [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://huggingface.co/papers/2305.18290).
 
1
  ---
2
  base_model: Kyleyee/Qwen2-0.5B-stf-imdb
 
3
  library_name: transformers
4
  model_name: Qwen2-0.5B-DPO-imdb_kl
5
  tags:
 
11
 
12
  # Model Card for Qwen2-0.5B-DPO-imdb_kl
13
 
14
+ This model is a fine-tuned version of [Kyleyee/Qwen2-0.5B-stf-imdb](https://huggingface.co/Kyleyee/Qwen2-0.5B-stf-imdb).
15
  It has been trained using [TRL](https://github.com/huggingface/trl).
16
 
17
  ## Quick start
 
27
 
28
  ## Training procedure
29
 
30
+ [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/Tripleleg/huggingface/runs/b6a6n5g8)
31
 
32
 
33
  This model was trained with DPO, a method introduced in [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://huggingface.co/papers/2305.18290).
all_results.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "epoch": 3.0,
3
- "eval_/kl_divergence": 12.871448516845703,
4
- "eval_/mean_score": 0.850456953048706,
5
  "eval_loss": 0.0,
6
- "eval_runtime": 12.0741,
7
- "eval_samples_per_second": 8.282,
8
- "eval_steps_per_second": 0.331
9
  }
 
1
  {
2
  "epoch": 3.0,
3
+ "eval_/kl_divergence": 59.429229736328125,
4
+ "eval_/mean_score": 0.9832416772842407,
5
  "eval_loss": 0.0,
6
+ "eval_runtime": 12.1242,
7
+ "eval_samples_per_second": 8.248,
8
+ "eval_steps_per_second": 0.33
9
  }
eval_results.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "epoch": 3.0,
3
- "eval_/kl_divergence": 12.871448516845703,
4
- "eval_/mean_score": 0.850456953048706,
5
  "eval_loss": 0.0,
6
- "eval_runtime": 12.0741,
7
- "eval_samples_per_second": 8.282,
8
- "eval_steps_per_second": 0.331
9
  }
 
1
  {
2
  "epoch": 3.0,
3
+ "eval_/kl_divergence": 59.429229736328125,
4
+ "eval_/mean_score": 0.9832416772842407,
5
  "eval_loss": 0.0,
6
+ "eval_runtime": 12.1242,
7
+ "eval_samples_per_second": 8.248,
8
+ "eval_steps_per_second": 0.33
9
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6ea4ade42a8af8721658c44dbb6e1971c187b08cc8e0a6ec91cf6dc09702a29a
3
  size 1976163472
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d181637fdf3734c1f4b4a9f2f553e7828a73d1bf62fc079f0d11a8ca994b96e
3
  size 1976163472
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c3c7f62eccf53cb9acebead9ae718c836d81614205e977acbe6c0fba51d43376
3
  size 6264
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:542aa4d7836c53b34f92fae917aacc76bd7379b664c5cc864bd047a8ba85085b
3
  size 6264