This is a replicate of DeepSeek-R1-Zero and DeepSeek-R1 training on small models with limited data
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).