Improve model card: add metadata, paper link, and base model reference
Browse filesHi! I'm Niels, part of the community science team at Hugging Face.
This pull request aims to improve the model card for **Qwen2.5-Coder-0.5B-Instruct-diffusion-mdlm-v0.1**. I have added YAML metadata for `pipeline_tag`, `library_name`, and `base_model` to make the model easier to find and use. I've also added an explicit link to the **dLLM** paper and maintained the existing codebase references.
These changes help users identify the model's lineage and provide quick access to the associated research and framework.
README.md
CHANGED
|
@@ -1,5 +1,12 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
| 4 |
|
| 5 |
<center> <div style="text-align: center;"> <img src="https://raw.githubusercontent.com/ZHZisZZ/dllm/main/assets/logo.gif" width="400" />
|
|
@@ -7,7 +14,7 @@ license: apache-2.0
|
|
| 7 |
|
| 8 |
# Qwen2.5-Coder-0.5B-Instruct-diffusion-mdlm-v0.1
|
| 9 |
|
| 10 |
-
Qwen2.5-Coder-0.5B-Instruct-diffusion-mdlm-v0.1 is a diffusion-based language model adapted from [Qwen2.5-Coder-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-0.5B-Instruct) using [MDLM](https://arxiv.org/abs/2406.07524) (masked diffusion)
|
| 11 |
|
| 12 |
|
| 13 |
## Model Overview
|
|
@@ -165,7 +172,8 @@ new_tokens = [
|
|
| 165 |
text[i, prompt_lens[i] : prompt_lens[i] + max_new_tokens].tolist() for i in range(text.size(0))
|
| 166 |
]
|
| 167 |
for idx, decoded in enumerate(tokenizer.batch_decode(new_tokens, skip_special_tokens=False)):
|
| 168 |
-
print(f"
|
|
|
|
| 169 |
print(decoded)
|
| 170 |
|
| 171 |
```
|
|
@@ -251,4 +259,4 @@ If you use Qwen2.5-Coder-0.5B-Instruct-diffusion-mdlm-v0.1 or dLLM, please cite:
|
|
| 251 |
primaryClass={cs.CL},
|
| 252 |
url={https://arxiv.org/abs/2602.22661},
|
| 253 |
}
|
| 254 |
-
```
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
+
library_name: transformers
|
| 4 |
+
base_model: Qwen/Qwen2.5-Coder-0.5B-Instruct
|
| 5 |
+
pipeline_tag: text-generation
|
| 6 |
+
tags:
|
| 7 |
+
- code
|
| 8 |
+
- diffusion-lm
|
| 9 |
+
- mdlm
|
| 10 |
---
|
| 11 |
|
| 12 |
<center> <div style="text-align: center;"> <img src="https://raw.githubusercontent.com/ZHZisZZ/dllm/main/assets/logo.gif" width="400" />
|
|
|
|
| 14 |
|
| 15 |
# Qwen2.5-Coder-0.5B-Instruct-diffusion-mdlm-v0.1
|
| 16 |
|
| 17 |
+
Qwen2.5-Coder-0.5B-Instruct-diffusion-mdlm-v0.1 is a diffusion-based language model adapted from [Qwen2.5-Coder-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-0.5B-Instruct) using [MDLM](https://arxiv.org/abs/2406.07524) (masked diffusion). It was trained using the [dLLM](https://github.com/ZHZisZZ/dllm) framework, as presented in the paper [dLLM: Simple Diffusion Language Modeling](https://huggingface.co/papers/2602.22661).
|
| 18 |
|
| 19 |
|
| 20 |
## Model Overview
|
|
|
|
| 172 |
text[i, prompt_lens[i] : prompt_lens[i] + max_new_tokens].tolist() for i in range(text.size(0))
|
| 173 |
]
|
| 174 |
for idx, decoded in enumerate(tokenizer.batch_decode(new_tokens, skip_special_tokens=False)):
|
| 175 |
+
print(f"
|
| 176 |
+
[Sample {idx}]")
|
| 177 |
print(decoded)
|
| 178 |
|
| 179 |
```
|
|
|
|
| 259 |
primaryClass={cs.CL},
|
| 260 |
url={https://arxiv.org/abs/2602.22661},
|
| 261 |
}
|
| 262 |
+
```
|