Skip to content

GitLab

Unverified Commit 7cee07a0 authored by Ke Bao's avatar Ke Bao Committed by GitHub
Browse files

Fix skip layer in get_quant_method (#12632)

parent bb517fe3
Hide whitespace changes
Inline Side-by-side
Showing with 4 additions and 1 deletion
python/sglang/srt/layers/quantization/compressed_tensors/compressed_tensors.py
View file @ 7cee07a0
...@@ -123,7 +123,10 @@ class CompressedTensorsConfig(QuantizationConfig): ...@@ -123,7 +123,10 @@ class CompressedTensorsConfig(QuantizationConfig):
if should_ignore_layer( if should_ignore_layer(
prefix, ignore=self.ignore, fused_mapping=self.packed_modules_mapping prefix, ignore=self.ignore, fused_mapping=self.packed_modules_mapping
): ):
return UnquantizedLinearMethod() if isinstance(layer, LinearBase):
return UnquantizedLinearMethod()
return None
if isinstance(layer, LinearBase): if isinstance(layer, LinearBase):
if CompressedTensorsConfig.DeepSeekFP8Config is not None: if CompressedTensorsConfig.DeepSeekFP8Config is not None:
return Fp8LinearMethod(CompressedTensorsConfig.DeepSeekFP8Config) return Fp8LinearMethod(CompressedTensorsConfig.DeepSeekFP8Config)
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment