23 lines
366 B
Markdown
23 lines
366 B
Markdown
|
---
|
||
|
|
license: apache-2.0
|
||
|
|
datasets:
|
||
|
|
- bespokelabs/Bespoke-Stratos-17k
|
||
|
|
language:
|
||
|
|
- en
|
||
|
|
base_model:
|
||
|
|
- HuggingFaceTB/SmolLM2-1.7B-Instruct
|
||
|
|
pipeline_tag: text-generation
|
||
|
|
library_name: transformers
|
||
|
|
tags:
|
||
|
|
- reasoning
|
||
|
|
---
|
||
|
|
# SmolR1-SFT
|
||
|
|
|
||
|
|
Potential limitations:
|
||
|
|
|
||
|
|
* Endless repetition
|
||
|
|
* Mistakes in reasoning
|
||
|
|
|
||
|
|
Prompt format: ChatML
|
||
|
|
|
||
|
|
Trained using Hugging Face's Open-R1 framework.
|