[ROCm] Temporarily remove GPTQ ROCm support (#2138)

2023-12-15 17:13:58 -08:00
parent b81a6a6bb3
commit 2acd76f346
2 changed files with 2 additions and 2 deletions
--- a/vllm/config.py
+++ b/vllm/config.py
@@ -143,7 +143,7 @@ class ModelConfig:

    def _verify_quantization(self) -> None:
        supported_quantization = ["awq", "gptq", "squeezellm"]
-        rocm_not_supported_quantization = ["awq"]
+        rocm_not_supported_quantization = ["awq", "gptq"]
        if self.quantization is not None:
            self.quantization = self.quantization.lower()