🧠 Train a 64M-parameter LLM from scratch in just 2h!
Repository Index
82,088 visible repositories in the index
1–2 of 2 · Most stars
Implementing DeepSeek R1's GRPO algorithm from scratch
82,088 visible repositories in the index
1–2 of 2 · Most stars
🧠 Train a 64M-parameter LLM from scratch in just 2h!
Implementing DeepSeek R1's GRPO algorithm from scratch