---
base_model:
- Qwen/Qwen3-14B-Base
datasets:
- MegaScience/MegaScience
language:
- en
license: apache-2.0
metrics:
- accuracy
pipeline_tag: text-generation
library_name: transformers
---

# [MegaScience: Pushing the Frontiers of Post-Training Datasets for Science Reasoning](https://arxiv.org/abs/2507.16812)

This repository contains the **Qwen3-14B-MegaScience** model, a large language model fine-tuned on the MegaScience dataset for enhanced scientific reasoning.

**Project Link**: [https://huggingface.co/MegaScience](https://huggingface.co/MegaScience) (Hugging Face Organization for MegaScience project)

**Code Repository**: [https://github.com/GAIR-NLP/lm-open-science-evaluation](https://github.com/GAIR-NLP/lm-open-science-evaluation)

## Usage

You can use this model with the `transformers` library for text generation:

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_id = "MegaScience/Qwen3-14B-MegaScience"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.bfloat16, # or torch.float16 if bfloat16 is not supported
    device_map="auto"
)

messages = [
    {"role": "system", "content": "You are a helpful and knowledgeable assistant."},
    {"role": "user", "content": "Explain the concept of quantum entanglement in simple terms."}
]

text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)

model_inputs = tokenizer(text, return_tensors="pt").to(model.device)

generated_ids = model.generate(
    model_inputs.input_ids,
    max_new_tokens=512,
    do_sample=True,
    temperature=0.7,
    top_p=0.9,
    eos_token_id=tokenizer.eos_token_id,
)

response = tokenizer.decode(generated_ids[0][model_inputs.input_ids.shape[1]:], skip_special_tokens=True)
print(response)
```

## Qwen3-14B-MegaScience

### Training Recipe

- **LR**: 5e-6
- **LR Schedule**: Cosine
- **Batch Size**: 512
- **Max Length**: 4,096
- **Warm Up Ratio**: 0.05
- **Epochs**: 3

### Evaluation Results

<div style="display: flex; justify-content: left; gap: 20px;">
<img src="https://cdn-uploads.huggingface.co/production/uploads/616bfc2b40e2f69baa1c7add/abIVZ2XB9D-o-TCyvOkDE.png" alt="Data Pipeline" style="width:80%;">
</div>

<div style="display: flex; justify-content: left; gap: 20px;">
<img src="https://cdn-uploads.huggingface.co/production/uploads/616bfc2b40e2f69baa1c7add/xFTJ7nevc3S4UYJxUS7ue.png" alt="Data Pipeline" style="width:80%;">
</div>

### More about MegaScience

<div style="display: flex; justify-content: left; gap: 20px;">
<img src="https://cdn-uploads.huggingface.co/production/uploads/616bfc2b40e2f69baa1c7add/VogIpBbjfNxXFP9DfVMms.png" alt="Data Pipeline" style="width:100%;">
</div>

## Citation

If you use our dataset or find our work useful, please cite

```
@article{fan2025megascience,
  title={MegaScience: Pushing the Frontiers of Post-Training Datasets for Science Reasoning},
  author={Fan, Run-Ze and Wang, Zengzhi and Liu, Pengfei},
  year={2025},
  journal={arXiv preprint arXiv:2507.16812},
  url={https://arxiv.org/abs/2507.16812}
}
```