Search results

GGUF q2ks accuracy regression

high priority

Status: Open.

Bug

#1902 In intel/auto-round;

· XuehaoSun opened on Jun 10, 2026

·  

0.13.1

[Bug]: test_cuda/models/test_fp8_model.py::TestAutoRound::test_fp8_model_gguf_q4 failed in CI

bugSomething isn't workingSomething isn't working

CI-known-issue

high priority

Status: Open.

#1898 In intel/auto-round;

· chensuyue opened on Jun 9, 2026

[Feature]: AutoScheme is currently underperforming compared to Unsloth's model for gguf

enhancementNew feature or requestNew feature or request

Status: Open.

#1896 In intel/auto-round;

· wenhuach21 opened on Jun 8, 2026

[Bug]: llmc's ut test failed with pr 1879 but passed at local test

bugSomething isn't workingSomething isn't working

Status: Open.

#1894 In intel/auto-round;

· wenhuach21 opened on Jun 8, 2026

·  

0.14.0

[Bug]: Subprocess dataset preprocessing always failed on mac

bugSomething isn't workingSomething isn't working

Status: Open.

#1890 In intel/auto-round;

· wenhuach opened on Jun 4, 2026

[Regreesion]: multi-cards tuning+eval causes exceptions at eval phase

bugSomething isn't workingSomething isn't working

Status: Open.

#1889 In intel/auto-round;

· wenhuach21 opened on Jun 4, 2026

[Feature]: support teq algorithm

enhancementNew feature or requestNew feature or request

Status: Open.

#1888 In intel/auto-round;

· wenhuach21 opened on Jun 4, 2026

stepfun-ai/Step-3.7-Flash model size issue

Model Support

Status: Open.

#1881 In intel/auto-round;

· XuehaoSun opened on Jun 2, 2026

[Bug] Fusing MoE for GGUF before tuning breaks quantization logic because MoE is no longer a QuantLinear layer withQDQ simulation

high priority

Status: Open.

#1878 In intel/auto-round;

· wenhuach21 opened on Jun 1, 2026

·  

0.14.0

Quality Assurance for v0.14.0

Status: Open.

#1874 In intel/auto-round;

· chensuyue opened on May 31, 2026

·  

0.14.0

[Bug]: Cuda CI failed with test_sglang

bugSomething isn't workingSomething isn't working

CI-known-issue

Status: Open.

#1873 In intel/auto-round;

· chensuyue opened on May 31, 2026

·  

0.14.0

[Feature gap][LLMC]: missing args

bugSomething isn't workingSomething isn't working

high priority

llmc-only

Status: Open.

#1855 In intel/auto-round;

· chensuyue opened on May 26, 2026

·  

0.14.0