calling mtq.quantize causes AttributeError: 'tuple' object has no attribute 'dim'

Make sure you already checked the examples and documentation before submitting an issue.

How would you like to use ModelOpt

running python Model-Optimizer/examples/diffusers/quantization/quantize.py --model sdxl-turbo --format fp8 --n-steps 4 --calib-size 128

results in

File "c:\python.venv\lib\site-packages\torch\nn\functional.py", line 3040, in group_norm
if input.dim() < 2:
AttributeError: 'tuple' object has no attribute 'dim'

the full output:

c:\python.venv\lib\site-packages\torch\jit_script.py:1491: FutureWarning: torch.jit.script is deprecated. Please switch to torch.compile or torch.export.
warnings.warn(
c:\python.venv\lib\site-packages\modelopt\torch_init_.py:55: UserWarning: transformers 4.56.0 is not tested with current version of modelopt and may cause issues. Please install recommended version with pip install -U nvidia-modelopt[hf] if working with HF models.
_warnings.warn(
2026-09-13 18:24:43 | INFO | main | Starting Enhanced Diffusion Model Quantization
2026-09-13 18:24:43 | INFO | main | Validating configurations...
2026-09-13 18:24:43 | INFO | main | Creating pipeline for sdxl-turbo
2026-09-13 18:24:43 | INFO | main | Model path: stabilityai/sdxl-turbo
2026-09-13 18:24:43 | INFO | main | Data type: {'default': torch.float16}
Loading pipeline components...: 57%|█████████████████████████████▋ | 4/7 [00:02<00:01, 1.70it/s]torch_dtype is deprecated! Use dtype instead!
Loading pipeline components...: 100%|████████████████████████████████████████████████████| 7/7 [00:05<00:00, 1.35it/s]
2026-09-13 18:24:56 | INFO | main | Pipeline created successfully
2026-09-13 18:24:56 | INFO | main | Moving pipeline to CUDA
2026-09-13 18:24:57 | INFO | main | Initializing calibration...
2026-09-13 18:24:57 | INFO | main | Loading calibration prompts from {'name': 'Gustavosta/Stable-Diffusion-Prompts', 'split': 'train', 'column': 'Prompt'}
2026-09-13 18:25:00 | INFO | main | Quantizing backbone: unet
2026-09-13 18:25:00 | INFO | main | Building quantization config for fp8
2026-09-13 18:25:00 | INFO | main | Quant config {'quant_cfg': [{'quantizer_name': '*', 'enable': False}, {'quantizer_name': '*weight_quantizer', 'cfg': {'num_bits': (4, 3), 'axis': None, 'trt_high_precision_dtype': 'Half'}}, {'quantizer_name': '*input_quantizer', 'cfg': {'num_bits': (4, 3), 'axis': None, 'trt_high_precision_dtype': 'Half'}}, {'quantizer_name': '*output_quantizer', 'enable': False}, {'quantizer_name': '*softmax_quantizer', 'cfg': {'num_bits': (4, 3), 'axis': None, 'trt_high_precision_dtype': 'Half'}}], 'algorithm': {'method': 'max'}}
2026-09-13 18:25:00 | INFO | main | Checking for LoRA layers...
2026-09-13 18:25:00 | INFO | main | Starting model quantization for unet...
Inserted 3502 quantizers
2026-09-13 18:25:01 | INFO | main | Starting calibration with 64 batches
Calibration: 0%| | 0/64 [00:00<?, ?batch/s]Token indices sequence length is longer than the specified maximum sequence length for this model (101 > 77). Running this sequence through the model will result in indexing errors
The following part of your input was truncated because CLIP can only handle sequences up to 77 tokens: ['incev, in style of lee souder, in plastic, dark atmosphere, tilt shift, depth of field,', 'trending on art station <|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|>']
Token indices sequence length is longer than the specified maximum sequence length for this model (101 > 77). Running this sequence through the model will result in indexing errors
The following part of your input was truncated because CLIP can only handle sequences up to 77 tokens: ['incev, in style of lee souder, in plastic, dark atmosphere, tilt shift, depth of field,', 'trending on art station <|endoftext|>!!!!!!!!!!!!!!!!!!!']
Calibration: 0%| | 0/64 [00:03<?, ?batch/s]
2026-09-13 18:25:05 | ERROR | main | Quantization failed: 'tuple' object has no attribute 'dim'
Traceback (most recent call last):
File "C:\python\Model-Optimizer\examples\diffusers\quantization\quantize.py", line 706, in main
quantizer.quantize_model(
File "C:\python\Model-Optimizer\examples\diffusers\quantization\quantize.py", line 239, in quantize_model
mtq.quantize(backbone, quant_config, forward_loop)
File "c:\python.venv\lib\site-packages\modelopt\torch\quantization\model_quant.py", line 247, in quantize
return calibrate(model, config.get("algorithm"), forward_loop=forward_loop)
File "c:\python.venv\lib\site-packages\modelopt\torch\quantization\model_quant.py", line 110, in calibrate
apply_mode(
File "c:\python.venv\lib\site-packages\modelopt\torch\opt\conversion.py", line 419, in apply_mode
model, metadata = get_mode(m).convert(model, config, **kwargs) # type: ignore [call-arg]
File "c:\python.venv\lib\site-packages\modelopt\torch\quantization\mode.py", line 351, in wrapped_func
return wrapped_calib_func(
File "c:\python.venv\lib\site-packages\modelopt\torch\quantization\mode.py", line 273, in wrapped_calib_func
func(model, forward_loop=forward_loop, **kwargs)
File "c:\python.venv\lib\site-packages\torch\utils_contextlib.py", line 124, in decorate_context
return func(*args, **kwargs)
File "c:\python.venv\lib\site-packages\modelopt\torch\quantization\model_calib.py", line 354, in max_calibrate
forward_loop(model)
File "C:\python\Model-Optimizer\examples\diffusers\quantization\quantize.py", line 704, in forward_loop
calibrator.run_calibration(batched_prompts)
File "C:\python\Model-Optimizer\examples\diffusers\quantization\calibration.py", line 103, in run_calibration
self.pipe(**common_args, **extra_args).images
File "c:\python.venv\lib\site-packages\torch\utils_contextlib.py", line 124, in decorate_context
return func(*args, **kwargs)
File "c:\python.venv\lib\site-packages\diffusers\pipelines\stable_diffusion_xl\pipeline_stable_diffusion_xl.py", line 1292, in call
image = self.vae.decode(latents, return_dict=False)[0]
File "c:\python.venv\lib\site-packages\diffusers\utils\accelerate_utils.py", line 46, in wrapper
return method(self, *args, **kwargs)
File "c:\python.venv\lib\site-packages\diffusers\models\autoencoders\autoencoder_kl.py", line 294, in decode
decoded = self._decode(z).sample
File "c:\python.venv\lib\site-packages\diffusers\models\autoencoders\autoencoder_kl.py", line 265, in _decode
dec = self.decoder(z)
File "c:\python.venv\lib\site-packages\torch\nn\modules\module.py", line 1783, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "c:\python.venv\lib\site-packages\torch\nn\modules\module.py", line 1794, in _call_impl
return forward_call(*args, **kwargs)
File "c:\python.venv\lib\site-packages\diffusers\models\autoencoders\vae.py", line 298, in forward
sample = self.mid_block(sample, latent_embeds)
File "c:\python.venv\lib\site-packages\torch\nn\modules\module.py", line 1783, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "c:\python.venv\lib\site-packages\torch\nn\modules\module.py", line 1794, in _call_impl
return forward_call(*args, **kwargs)
File "c:\python.venv\lib\site-packages\diffusers\models\unets\unet_2d_blocks.py", line 746, in forward
hidden_states = resnet(hidden_states, temb)
File "c:\python.venv\lib\site-packages\torch\nn\modules\module.py", line 1783, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "c:\python.venv\lib\site-packages\torch\nn\modules\module.py", line 1794, in _call_impl
return forward_call(*args, **kwargs)
File "c:\python.venv\lib\site-packages\diffusers\models\resnet.py", line 327, in forward
hidden_states = self.norm1(hidden_states)
File "c:\python.venv\lib\site-packages\torch\nn\modules\module.py", line 1783, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "c:\python.venv\lib\site-packages\torch\nn\modules\module.py", line 1794, in _call_impl
return forward_call(*args, **kwargs)
File "c:\python.venv\lib\site-packages\torch\nn\modules\normalization.py", line 334, in forward
return F.group_norm(input, self.num_groups, self.weight, self.bias, self.eps)
File "c:\python.venv\lib\site-packages\torch\nn\functional.py", line 3040, in group_norm
if input.dim() < 2:
AttributeError: 'tuple' object has no attribute 'dim'

Who can help?

• ?

System information

• Container used (if applicable): ?
• OS (e.g., Ubuntu 22.04, CentOS 7, Windows 10): ? Windows 11
• CPU architecture (x86_64, aarch64): Core i7 13700K
• GPU name (e.g. H100, A100, L40S): RTX5080
• GPU memory size: 32gb
• Number of GPUs: 1
• Library versions (if applicable):
  • Python: 3.10.14
  • ModelOpt version or commit hash: 0.46.1
  • CUDA: 13.2
  • PyTorch: 2.14.0+cu132
  • Transformers: 4.56.0
  • TensorRT-LLM: ?
  • ONNXRuntime: 1.23.2
  • TensorRT: 10.16.1.11
• Any other details that may help: ?