sglang/docs/platforms/ascend_npu_quantization.md

Quantization on Ascend.

To load already quantized models, simply load the model weights and config. Again, if the model has been quantized offline, there's no need to add --quantization argument when starting the engine. The quantization method will be automatically parsed from the downloaded quant_model_description.json or config.json config.

ModelSlim on Ascend support:

• W4A4 dynamic linear
• W8A8 static linear
• W8A8 dynamic linear
• W4A4 dynamic MOE
• W4A8 dynamic MOE
• W8A8 dynamic MOE

AWQ on Ascend support:

• W4A16 linear
• W8A16 linear # Need to test
• W4A16 MOE # Need to test

Compressed-tensors (LLM Compressor) on Ascend support:

• W4A8 dynamic MOE with/without activation clip # Need to test
• W4A16 MOE
• W8A8 dynamic linear
• W8A8 dynamic MOE

Diffusion model modelslim quantization on Ascend support:

• W4A4 dynamic linear
• W8A8 static linear
• W8A8 dynamic linear