add proper detection and handling for nemotron VL model in ptq examples

Signed-off-by: Zhiyu Cheng <[email protected]>

Author: Edwardf0t1

examples/llm_ptq/example_utils.py

@@ -185,6 +185,17 @@ def get_model(
     if device == "cpu":
         device_map = "cpu"
 
+    # Special handling for vision-language models that may have device mapping issues
+    # Check if this is a VL model by looking at the model path
+    is_vl_model = any(
+        vl_keyword in ckpt_path.lower() for vl_keyword in ["vl", "vision", "nemotron-nano-vl"]
+    )
+    if is_vl_model:
+        print(
+            "Detected vision-language model. Disabling automatic device mapping to avoid device_map errors."
+        )
+        device_map = None
+
     config_kwargs = {"trust_remote_code": trust_remote_code} if trust_remote_code else {}
     if attn_implementation is not None:
         config_kwargs["attn_implementation"] = attn_implementation
@@ -282,6 +293,12 @@ def get_model(
                 **model_kwargs,
             )
     model.eval()
+
+    # If device_map was disabled (None), manually move model to target device
+    if device_map is None and device != "cpu":
+        print(f"Moving model to {device} device...")
+        model = model.to(device)
+
     if device == "cuda" and not is_model_on_gpu(model):
         print("Warning: Some parameters are not on a GPU. Calibration can be slow or hit OOM")
 

examples/llm_ptq/hf_ptq.py

@@ -275,6 +275,16 @@ def main(args):
 
     model_type = get_model_type(model)
 
+    # Special handling for Nemotron VL models that aren't detected by standard model type detection
+    # For HF export, we want to keep vision unquantized, so we treat it as a regular language model
+    # and only quantize the language components
+    if model_type != "mllama" and is_multimodal_model(model):
+        print(
+            f"Detected multimodal model: {type(model).__name__}. "
+            f"For HF export, will quantize language components only, keeping vision unquantized."
+        )
+        # Keep as regular model type to use text-only calibration
+
     device = model.device
     if hasattr(model, "model"):
         device = model.model.device
@@ -458,20 +468,41 @@ def main(args):
             KV_QUANT_CFG_CHOICES,
         )
 
+            # For Nemotron VL models, disable quantization of vision components
+            is_nemotron_vl = (
+                "nemotron" in args.pyt_ckpt_path.lower() and "vl" in args.pyt_ckpt_path.lower()
+            )
+            if is_nemotron_vl:
+                print("Disabling quantization for vision components in Nemotron VL model")
+                quant_cfg["quant_cfg"]["*vision*"] = {"enable": False}
+                quant_cfg["quant_cfg"]["*image*"] = {"enable": False}
+                # Also disable radio model components specifically
+                quant_cfg["quant_cfg"]["*radio*"] = {"enable": False}
+                quant_cfg["quant_cfg"]["*visual*"] = {"enable": False}
+
         if not model_is_already_quantized or calibration_only:
             # Only run single sample for preview
             input_ids = next(iter(calib_dataloader))[
                 "input_features" if model_type == "whisper" else "input_ids"
             ][0:1]
-            try:
-                generated_ids_before_ptq = full_model.generate(input_ids, max_new_tokens=100)
-            except Exception as e:
-                print(
-                    "Error during model generation. Please check if your transformers version is "
-                    "compatible with the model."
-                )
-                print(f"Error details: {e}")
-                raise
+
+            # Skip preview generation for Nemotron VL models that require special handling
+            is_nemotron_vl = (
+                "nemotron" in args.pyt_ckpt_path.lower() and "vl" in args.pyt_ckpt_path.lower()
+            )
+            if is_nemotron_vl:
+                print("Skipping preview generation for Nemotron VL model (requires image input)")
+                generated_ids_before_ptq = None
+            else:
+                try:
+                    generated_ids_before_ptq = full_model.generate(input_ids, max_new_tokens=100)
+                except Exception as e:
+                    print(
+                        "Error during model generation. Please check if your transformers version is "
+                        "compatible with the model."
+                    )
+                    print(f"Error details: {e}")
+                    raise
             if model_type == "gptoss" and args.qformat == "nvfp4_mlp_only":
                 print("Applying nvfp4 quantization (MoE only) for gpt-oss")
 
@@ -483,9 +514,13 @@ def main(args):
             # Run some samples
             torch.cuda.empty_cache()
             generated_ids_after_ptq = None
-            if model_type != "llama4":
+            if model_type != "llama4" and not is_nemotron_vl:
                 # Our fake quantizer may not be fully compatible with torch.compile.
                 generated_ids_after_ptq = full_model.generate(input_ids, max_new_tokens=100)
+            elif is_nemotron_vl:
+                print(
+                    "Skipping post-quantization generation for Nemotron VL model (requires image input)"
+                )
             else:
                 warnings.warn(
                     "Llama4 Maverick generation after quantization has a bug. Skipping generation sample."
@@ -548,9 +583,12 @@ def output_decode(generated_ids, input_shape):
             # Save original model config and the processor config to the export path for VLMs.
             print(f"Saving original model config to {export_path}")
 
-            AutoConfig.from_pretrained(
-                args.pyt_ckpt_path, trust_remote_code=args.trust_remote_code
-            ).save_pretrained(export_path)
+            config_kwargs = {"trust_remote_code": args.trust_remote_code}
+            if args.attn_implementation is not None:
+                config_kwargs["attn_implementation"] = args.attn_implementation
+            AutoConfig.from_pretrained(args.pyt_ckpt_path, **config_kwargs).save_pretrained(
+                export_path
+            )
 
             # Try to save processor config if available
             try: