|
--- |
|
tags: |
|
- text-to-text |
|
- coding |
|
- finetuned r1 deepseek |
|
base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B |
|
license: mit |
|
--- |
|
|
|
# R1 DeepSeek Qwen 1.5B Finetuned |
|
|
|
This model has been finetuned on code extracted from over 200 of my personal GitHub repositories, building upon the `deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B` base model. It’s optimized for both text-to-text and coding applications. |
|
|
|
**Training Details:** |
|
- **Estimated Tokens Trained On:** ~20 million tokens |
|
|
|
Dataset generation was achieved using the [GIT_HISTORY](https://github.com/Himasnhu-AT/git-history) tool. |
|
|
|
|