Lightning-AI/lightning-thunder
Vedi su GitHubquantization: process tensors on meta device directly, maybe implement CPU quantization (if it is easy)
Open
#1111 aperta il 6 set 2024
good first issuetransforms
Metriche repository
- Star
- (1460 star)
- Metriche merge PR
- (Metriche PR in attesa)
Descrizione
Currently the BitsAndBytesLinearQuant4bit for submodule always calls bitsandbytes.functional.quantize_4bit. This is somewhat touchy for CPU tensors because quantize_4bit only works on GPU tensors but it is outright not so nice for meta tensors, where we only would need to get the right shapes.