executorch
executorch copied to clipboard
On-device AI across mobile, embedded and edge for PyTorch
Summary: Identify builds with kv cache Differential Revision: D54300827
Summary: Add warning if a seemingly fairseq2 checkpoint is passed to model constructor but --fairseq2 was not specified. (Heuristic.) Differential Revision: D54209318
Summary: Resolve recurring errors where query is c10::Half and key and value float. This should ideally work from first principles, but somehow it does not. We need to fix the...
Summary: Add options for embedding quantization: bitwidth, group_size on CLI Differential Revision: D54159472
Summary: improvements: asymmetric group size less militant checking out_dtype added Differential Revision: D53990647
Summary: Add debug info to op_embedding Differential Revision: D54047417
Summary: Add embedding op to llama build Reviewed By: shoumikhin Differential Revision: D53985792
Summary: improvements: asymmetric group size less militant checking out_dtype added legacy compatibility Differential Revision: D54074895
Summary: Compute size and bleu score for different quantization combinations Reviewed By: mavlyutovr Differential Revision: D53687784