Running
Phi 2 Fine Tuned With GRPO
๐
Using DeepSeek's GRPO to fine tune Microsoft Phi-2!
Using DeepSeek's GRPO to fine tune Microsoft Phi-2!
Microsoft Phi-2 model fine tuned on Open Assistant dataset.
A replica of smollm2-135M trained on smollm corpus.
A decoder trained starting with GPT2 weights.
A basic image classifier app built on ResNet 50 model.