2026-10-19 · America/Los_Angeles · 社区动态 · #16

Taco: cross-platform LLM compression that doesn't break when you change your GPU or CPU. [P]

The problem: an LLM compressor needs the decoder to compute exactly the same probabilities as the encoder. Floats differ across hardware; float GPT-2 encoded on a Mac decoded 0 of 4 windows on x86. The approach: integer-only inference (int8 weights, fixed rounding, lookup tables); an exact float32 → integer converter (mantissa × 2^exp); conformance vectors. The proof: one hash across 7 backends (Mac CPU/Metal, Linux CPU/CUDA, Windows CPU/CUDA/Vulkan https://github.com/tac0mania/taco

热度 44.3 / 100；排名与评分保留该期记录。