Files
llm-arch-research/notebooks
Sergey Penkovsky 9898e8ee83 feat: add RoPE positional embeddings implementation in llama.ipynb
- Implement Rotary Positional Embeddings (RoPE) with separate cosine/sine components
- Add vectorized computation of inverse frequencies for RoPE
- Include tensor slicing utilities for even/odd column separation
- Update dependencies in pyproject.toml and uv.lock
2025-10-06 12:52:59 +03:00
..
2025-10-05 12:48:32 +03:00
2025-10-05 12:46:17 +03:00