Search results

Deuqantize block-wise FP8 weight on CPU (G2-Specific)

good first issueGood for newcomersGood for newcomers

Status: Open.

#1196 In intel/auto-round;

· yiliu30 opened on Dec 25, 2025

Support FP8 MLA calibration (for Deepseek)

good first issueGood for newcomersGood for newcomers

Status: Open.

#1195 In intel/auto-round;

· yiliu30 opened on Dec 25, 2025

Align the first input scale of MoE experts for FP8 dispatch

good first issueGood for newcomersGood for newcomers

Status: Open.

#1194 In intel/auto-round;

· yiliu30 opened on Dec 25, 2025

[Feat] AR for vLLM-Gaudi

Status: Open.

#1193 In intel/auto-round;

· yiliu30 opened on Dec 25, 2025

dump additional keys into the model file to support different MXFP variants

Status: Open.

#1187 In intel/auto-round;

· wenhuach21 opened on Dec 24, 2025

Strengthen support for transformers diffusers, vllm-omni, and sglang-diffusion.

Status: Open.

#1186 In intel/auto-round;

· xin3he opened on Dec 24, 2025

·  

1.1.0

Re-distributing auto-round

Status: Open.

#1178 In intel/auto-round;

· isuruf opened on Dec 22, 2025

add kv quant example

Status: Open.

#1174 In intel/auto-round;

· mengniwang95 opened on Dec 19, 2025

Support w8a8 scheme in auto-round and add example

Status: Open.

#1173 In intel/auto-round;

· mengniwang95 opened on Dec 19, 2025

Add mixtral3 nvfp4 support

Status: Open.

#1172 In intel/auto-round;

· xin3he opened on Dec 19, 2025

·  

0.9.3

[LLMC] Enable auto-device

llmc

Status: Open.

#1169 In intel/auto-round;

· yiliu30 opened on Dec 19, 2025

LLMC Upstream Tasks

llmc

Status: Open.

#1168 In intel/auto-round;

· yiliu30 opened on Dec 19, 2025