onnxruntime
Route fp16 HQNBIT_CompInt8 (4-bit and 8-bit) through fp32 MLAS path in MatMulNBits
#27820

Merged

Commits

Route fp16 HQNBIT_CompInt8 through fp32 MLAS path for 4-bit and 8-bit

jambayk committed 91 days ago
Remove dead HQ4BitGemm_CompInt8 and HQ8BitGemm_CompInt8 MLAS code

jambayk committed 90 days ago
lint

jambayk committed 90 days ago
Fix HQNBIT_CompInt8 PrePack bugs for 4-bit and 8-bit

jambayk committed 90 days ago
Address review: ORT_ENFORCE for scales, move SQNBIT check to GetComputeType

jambayk committed 90 days ago