Enable native ModelOpt quantization support (1/3) (#7149)

Signed-off-by: Zhiyu Cheng <zhiyuc@nvidia.com>
2025-10-06 13:24:15 -07:00
parent eb30b888db
commit 155cbb51f0
11 changed files with 464 additions and 42 deletions
@@ -226,6 +226,9 @@ def get_quant_config(
                    return ModelOptFp4Config.from_config(config)
                else:
                    return quant_cls.from_config(config)
+        elif model_config.quantization == "modelopt_fp8":
+            if config["producer"]["name"] == "modelopt_fp8":
+                return quant_cls.from_config(config)
            else:
                raise ValueError(
                    f"Unsupported quantization config"