PTQ-Level Efficiency, QAT-Level Accuracy | 训练后量化的效率,量化感知训练的精度
Commits on main
Update BackendInfos for AutoGPTQ based on transformer version and save g_idx for gptqmodel backend (#1429)
a187610 Xin He committed 9 months ago
remove require_intel_extension_for_pytorch and fix TypeError: unhashable type: 'set' (#1425)
319da04 Xin He committed 9 months ago
Refactor evaluation in tests to use evaluate_accuracy function (#1402)
70a6bcc Xin He committed 9 months ago
fix: update cleanup logic for quantized model path in tests (#1419)
ebb534a Xin He committed 9 months ago
update auto-round-kernel package name to auto-round-lib (#1410)
bab0993 chen, suyue committed 9 months ago
[fix]AttributeError: 'Autotuner' object has no attribute '_cache_lock' (#1414)
6093abc Xin He committed 9 months ago
fix Qwen3-VL model auto_awq export, add auto_awq vllm ut (#1384)
40bd641 Weiwei committed 9 months ago
better warning for not support mllm while quant_non_text_module (#1411)
96cfbc9 Heng Guo committed 9 months ago
fix: update model imports from AutoModelForVision2Seq to AutoModelFor… (#1412)
b2bf11d Xin He committed 9 months ago
fix cuda ut fail: gptqmodel and gguf format with fp8 input model (#1399)
4dc04f4 Heng Guo committed 9 months ago
Support quant for `meituan-longcat/LongCat-Flash-Lite` (#1388)
a12e3f5 Tang Kaihui committed 9 months ago