Reasoning Work
Collection
A collection models trained to think like DeepSeek R1 using online learning - Group Relative Policy Optimization (GRPO) introduced by DeepSeekMath
•
6 items
•
Updated
A Qwen2.5 3Billion parameter model trained to "think" like DeepSeek's R1 using GRPO to be able deduce a disease using patients' complaints in one-shot!
Tiny but really impressive model. Training to think and reason has also resulted significant boost in general ELO of the model.
Uploaded model
- Developed by: dumbequation
- License: apache-2.0
- Finetuned from model : unsloth/qwen2.5-3b-instruct-unsloth-bnb-4bit
This qwen2 model was trained 2x faster with Unsloth and Huggingface's TRL library.
- Downloads last month
- 52
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.