Commit
•
083ec94
1
Parent(s):
4cf971b
End of training
Browse files- README.md +41 -0
- generation_config.json +14 -0
README.md
ADDED
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
base_model: Qwen/Qwen2-0.5B-Instruct
|
3 |
+
datasets: trl-lib/ultrafeedback-prompt
|
4 |
+
library_name: transformers
|
5 |
+
model_name: xpo-qwen2
|
6 |
+
tags:
|
7 |
+
- trl
|
8 |
+
- generated_from_trainer
|
9 |
+
- xpo
|
10 |
+
licence: license
|
11 |
+
---
|
12 |
+
|
13 |
+
# Model Card for xpo-qwen2
|
14 |
+
|
15 |
+
This model is a fine-tuned version of [Qwen/Qwen2-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2-0.5B-Instruct) on the [trl-lib/ultrafeedback-prompt](https://huggingface.co/datasets/trl-lib/ultrafeedback-prompt) dataset.
|
16 |
+
It has been trained using [TRL](https://github.com/huggingface/trl).
|
17 |
+
|
18 |
+
## Quick start
|
19 |
+
|
20 |
+
```python
|
21 |
+
from transformers import pipeline
|
22 |
+
|
23 |
+
question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
|
24 |
+
generator = pipeline("text-generation", model="qgallouedec/xpo-qwen2", device="cuda")
|
25 |
+
output = generator([{"role": "user", "content": question}], max_new_tokens=500)[0]
|
26 |
+
print(output["generated_text"][1]["content"])
|
27 |
+
```
|
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/huggingface/huggingface/runs/9fvfu3z2)
|
32 |
+
|
33 |
+
This model was trained with XPO, a method introduced in [Exploratory Preference Optimization: Harnessing Implicit Q*-Approximation for Sample-Efficient RLHF](https://huggingface.co/papers/2405.21046).
|
34 |
+
|
35 |
+
### Framework versions
|
36 |
+
|
37 |
+
- TRL: 0.12.0.dev0
|
38 |
+
- Transformers: 4.45.0.dev0
|
39 |
+
- Pytorch: 2.4.1
|
40 |
+
- Datasets: 3.0.0
|
41 |
+
- Tokenizers: 0.19.1
|
generation_config.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token_id": 151643,
|
3 |
+
"do_sample": true,
|
4 |
+
"eos_token_id": [
|
5 |
+
151645,
|
6 |
+
151643
|
7 |
+
],
|
8 |
+
"pad_token_id": 151643,
|
9 |
+
"repetition_penalty": 1.1,
|
10 |
+
"temperature": 0.7,
|
11 |
+
"top_k": 20,
|
12 |
+
"top_p": 0.8,
|
13 |
+
"transformers_version": "4.45.0.dev0"
|
14 |
+
}
|