Search results

Block‑AP Training Performance on H100

Status: Open.

#34 In OpenGVLab/EfficientQAT;

· bimalmagar10 opened on Jan 19, 2026

About training setting of Llama3-8B

Status: Open.

#33 In OpenGVLab/EfficientQAT;

· HXuan-Wang opened on Dec 24, 2025

about the training data format of e2e training

Status: Open.

#32 In OpenGVLab/EfficientQAT;

· brisker opened on Dec 2, 2025

In main_e2e_qp.py, the default value of max_steps must be set to -1!

Status: Open.

#31 In OpenGVLab/EfficientQAT;

· efsotr opened on Nov 12, 2025

About Dataset

Status: Open.

#30 In OpenGVLab/EfficientQAT;

· nostalgiaa-a opened on Aug 14, 2025

How are the quantization parameters of block_ap passed to the e2e-qp stage?

Status: Open.

#29 In OpenGVLab/EfficientQAT;

· tayton42 opened on May 26, 2025

Release artifacts (models, dataset) on Hugging Face

Status: Open.

#28 In OpenGVLab/EfficientQAT;

· NielsRogge opened on May 20, 2025

colab t4

Status: Open.

#27 In OpenGVLab/EfficientQAT;

· werruww opened on Feb 24, 2025

其它模型是否支持？

Status: Open.

#26 In OpenGVLab/EfficientQAT;

· Code-KingTT opened on Dec 3, 2024

Llava 1.6 Instruction Finetuning

Status: Open.

#25 In OpenGVLab/EfficientQAT;

· gdsaikrishna opened on Nov 28, 2024

Is 7B llama speed expected to be slow?

Status: Open.

#19 In OpenGVLab/EfficientQAT;

· w32zhong opened on Aug 30, 2024

why model output is too much slow?

Status: Open.

#18 In OpenGVLab/EfficientQAT;

· dahwin opened on Aug 27, 2024