Bring back quantization_config parameter

nikita-savelyevv · nikita-savelyevv · commit d8017ab843f4 · 2024-05-31T13:05:45.000+02:00
diff --git a/optimum/intel/openvino/modeling_decoder.py b/optimum/intel/openvino/modeling_decoder.py
@@ -761,6 +761,7 @@ def _from_pretrained(
             config=config,
             model_save_dir=model_cache_path.parent,
             compile=enable_compilation,
+            quantization_config=quantization_config,
             **kwargs,
         )
 

Original file line number	Diff line number	Diff line change
`@@ -761,6 +761,7 @@ def _from_pretrained(`
`761`	`761`	`config=config,`
`762`	`762`	`model_save_dir=model_cache_path.parent,`
`763`	`763`	`compile=enable_compilation,`
	`764`	`+ quantization_config=quantization_config,`
`764`	`765`	`**kwargs,`
`765`	`766`	`)`
`766`	`767`