Deprecations

• Deprecated modelopt.torch._deploy.utils.get_onnx_bytes API. Please use modelopt.torch._deploy.utils.get_onnx_bytes_and_metadata instead to access the ONNX model bytes with external data. See examples/onnx_ptq/download_example_onnx.py for example usage.

New Features

• Added flag op_types_to_exclude_fp16 in ONNX quantization to exclude ops from being converted to FP16/BF16. Alternatively, for custom TensorRT ops, this can also be done by indicating 'fp32' precision in trt_plugins_precision.
• Added LoRA mode support for MCore in a new peft submodule: modelopt.torch.peft.update_model(model, LORA_CFG).
• Supported PTQ and fakequant in vLLM for fast evaluation of arbitrary quantization formats. See examples/vllm_serve for more details.
• Added support for nemotron-post-training-dataset-v2 and nemotron-post-training-dataset-v1 in examples/llm_ptq. Defaults to a mix of cnn_dailymail and nemotron-post-training-dataset-v2 (gated dataset accessed using the HF_TOKEN environment variable) if no dataset is specified.
• Allows specifying calib_seq in examples/llm_ptq to set the maximum sequence length for calibration.
• Added support for MCore MoE PTQ/QAT/QAD.
• Added support for multi-node PTQ and export with FSDP2 in examples/llm_ptq/multinode_ptq.py. See examples/llm_ptq/README.md for more details.
• Added support for Nemotron Nano VL v1 & v2 models in FP8/NVFP4 PTQ workflow.
• Added flags nodes_to_include and op_types_to_include in AutoCast to force-include nodes in low precision, even if they would otherwise be excluded by other rules.
• Added support for torch.compile and benchmarking in examples/diffusers/quantization/diffusion_trt.py.
• Enabled native ModelOpt quantization support for FP8 and NVFP4 formats in SGLang. See SGLang quantization documentation for more details.
• Added ModelOpt quantized checkpoints in vLLM/SGLang CI/CD pipelines (PRs are under review).
• Added support for exporting QLoRA checkpoints finetuned using ModelOpt.

Documentation

• Added general guidelines for Minitron pruning and distillation. See examples/pruning/README.md for more details.
• Added example for exporting QLoRA checkpoints for vLLM deployment. Refer to examples/llm_qat/README.md for more details.

Additional Announcements

• ModelOpt will change its versioning from odd minor versions to all consecutive versions from next release. This means next release will be named 0.40.0 instead of 0.41.0

Deprecations

• Deprecated ModelOpt's custom docker images. Please use the PyTorch, TensorRT-LLM, or TensorRT docker image directly or refer to the installation guide for more details.
• Deprecated quantize_mode argument in examples/onnx_ptq/evaluate.py to support strong typing. Use engine_precision instead.
• Deprecated TRT-LLM's TRT backend in examples/llm_ptq and examples/vlm_ptq. Tasks build and benchmark support are removed and replaced with quant. engine_dir is replaced with checkpoint_dir in examples/llm_ptq and examples/vlm_ptq. For performance evaluation, please use trtllm-bench directly.
• The --export_fmt flag in examples/llm_ptq is removed. By default, we export to the unified Hugging Face checkpoint format.
• Deprecated examples/vlm_eval as it depends on the deprecated TRT-LLM's TRT backend.

New Features

• high_precision_dtype defaults to fp16 in ONNX quantization, i.e., quantized output model weights are now FP16 by default.
• Upgraded TensorRT-LLM dependency to 1.1.0rc2.
• Support for Phi-4-multimodal and Qwen2.5-VL quantized HF checkpoint export in examples/vlm_ptq.
• Support storing and restoring Minitron pruning activations and scores for re-pruning without running the forward loop again.
• Added Minitron pruning example for the Megatron-LM framework. See examples/megatron-lm for more details.

• Import fixes

Deprecations

• Deprecate torch<2.6 support.
• Deprecate NeMo 1.0 model support.

Bug Fixes

• Fix attention head ranking logic for pruning Megatron Core GPT models.

New Features

• ModelOpt now supports PTQ and QAT for GPT-OSS models. See examples/gpt_oss for end-to-end PTQ/QAT example.
• Add support for QAT with HuggingFace + DeepSpeed. See examples/gpt_oss for an example.
• Add support for QAT with LoRA. The LoRA adapters can be folded into the base model after QAT and deployed just like a regular PTQ model. See examples/gpt_oss for an example.
• ModelOpt provides convenient trainers such as :class:QATTrainer, :class:QADTrainer, :class:KDTrainer, :class:QATSFTTrainer which inherits from Huggingface trainers.
  ModelOpt trainers can be used as drop in replacement of the corresponding Huggingface trainer. See usage examples in examples/gpt_oss, examples/llm_qat or examples/llm_distill.
• (Experimental) Add quantization support for custom TensorRT op in ONNX models.
• Add support for Minifinetuning (MFT; https://arxiv.org/abs/2506.15702) self-corrective distillation, which enables training on small datasets with severely mitigated catastrophic forgetting.
• Add tree decoding support for Megatron Eagle models.
• For most VLMs, we now explicitly disable quant on the vision part so we add them to the excluded_modules during HF export.
• Add support for mamba_num_heads, mamba_head_dim, hidden_size and num_layers pruning for Megatron Core Mamba or Hybrid Transformer Mamba models in mcore_minitron (previously mcore_gpt_minitron) mode.
• Add example for QAT/QAD training with LLaMA Factory <https://github.com/hiyouga/LLaMA-Factory/tree/main>_. See examples/llm_qat/llama_factory for more details.
• Upgrade TensorRT-LLM dependency to 1.0.0rc6.
• Add unified HuggingFace model export support for quantized NVFP4 GPT-OSS models.

Bug Fixes

• Fix a Qwen3 MOE model export issue.

Backward Breaking Changes

• PyTorch dependencies for modelopt.torch features are no longer optional and pip install nvidia-modelopt is now same as pip install nvidia-modelopt[torch].

New Features

• Upgrade TensorRT-LLM dependency to 0.20.
• Add new CNN QAT example to demonstrate how to use ModelOpt for QAT.
• Add support for ONNX models with custom TensorRT ops in Autocast.
• Add quantization aware distillation (QAD) support in llm_qat example.
• Add support for BF16 in ONNX quantization.
• Add per node calibration support in ONNX quantization.
• ModelOpt now supports quantization of tensor-parallel sharded Huggingface transformer models. This requires transformers>=4.52.0.
• Support quantization of FSDP2 wrapped models and add FSDP2 support in the llm_qat example.
• Add NeMo 2 Simplified Flow examples for quantization aware training/distillation (QAT/QAD), speculative decoding, pruning & distillation.

Backward Breaking Changes

• NeMo and Megatron-LM distributed checkpoint (torch-dist) stored with legacy version can no longer be loaded. The remedy is to load the legacy distributed checkpoint with 0.29 and store a torch checkpoint and resume with 0.31 to convert to a new format. The following changes only apply to storing and resuming distributed checkpoint.
  • quantizer_state of :class:TensorQuantizer <modelopt.torch.quantization.nn.modules.TensorQuantizer> is now stored in extra_state of :class:QuantModule <modelopt.torch.quantization.nn.module.QuantModule> where it used to be stored in the sharded modelopt_state.
  • The dtype and shape of amax and pre_quant_scale stored in the distributed checkpoint are now retored. Some dtype and shape are previously changed to make all decoder layers to have homogeneous structure in the checkpoint.
  • Togather with megatron.core-0.13, quantized model will store and resume distributed checkpoint in a heterogenous format.
• auto_quantize API now accepts a list of quantization config dicts as the list of quantization choices.
  • This API previously accepts a list of strings of quantization format names. It was therefore limited to only pre-defined quantization formats unless through some hacks.
  • With this change, now user can easily use their own custom quantization formats for auto_quantize.
  • In addition, the quantization_formats now exclude None (indicating "do not quantize") as a valid format because the auto_quantize internally always add "do not quantize" as an option anyway.
• Model export config is refactored. The quant config in hf_quant_config.json is converted and saved to config.json. hf_quant_config.json will be deprecated soon.

Deprecations

• Deprecate Python 3.9 support.

New Features

• Upgrade LLM examples to use TensorRT-LLM 0.19.
• Add new model support in the llm_ptq example: Qwen3 MoE.
• ModelOpt now supports advanced quantization algorithms such as AWQ, SVDQuant and SmoothQuant for cpu-offloaded Huggingface models.
• Add AutoCast tool to convert ONNX models to FP16 or BF16.
• Add --low_memory_mode flag in the llm_ptq example support to initialize HF models with compressed weights and reduce peak memory of PTQ and quantized checkpoint export.

Backward Breaking Changes

• Refactor SequentialQuantizer to improve its implementation and maintainability while preserving its functionality.

Deprecations

• Deprecate torch<2.4 support.

New Features

• Upgrade LLM examples to use TensorRT-LLM 0.18.
• Add new model support in the llm_ptq example: Gemma-3, Llama-Nemotron.
• Add INT8 real quantization support.
• Add an FP8 GEMM per-tensor quantization kernel for real quantization. After PTQ, you can leverage the mtq.compress <modelopt.torch.quantization.compress> API to accelerate evaluation of quantized models.
• Use the shape of Pytorch parameters and buffers of TensorQuantizer <modelopt.torch.quantization.nn.modules.TensorQuantizer> to initialize them during restore. This makes quantized model restoring more robust.
• Support adding new custom quantization calibration algorithms. Please refer to mtq.calibrate <modelopt.torch.quantization.model_quant.calibrate> or custom calibration algorithm doc for more details.
• Add EAGLE3 (LlamaForCausalLMEagle3) training and unified ModelOpt checkpoint export support for Megatron-LM.
• Add support for --override_shapes flag to ONNX quantization.
  • --calibration_shapes is reserved for the input shapes used for calibration process.
  • --override_shapes is used to override the input shapes of the model with static shapes.
• Add support for UNet ONNX quantization.
• Enable concat_elimination pass by default to improve the performance of quantized ONNX models.
• Enable Redundant Cast elimination pass by default in moq.quantize <modelopt.onnx.quantization.quantize>.
• Add new attribute parallel_state to DynamicModule <modelopt.torch.opt.dynamic.DynamicModule> to support distributed parallelism such as data parallel and tensor parallel.
• Add MXFP8, NVFP4 quantized ONNX export support.
• Add new example for torch quantization to ONNX for MXFP8, NVFP4 precision.

Add experimental quantization support for Llama4, QwQ and Qwen MOE models.

Deprecations

• Deprecate real quantization configs, please use mtq.compress <modelopt.torch.quantization.compress> API for model compression after quantization.

New Features

• New model support in the llm_ptq example: OpenAI Whisper.
• Blockwise FP8 quantization support in unified model export.
• Add quantization support to the Transformer Engine Linear module.
• Add support for SVDQuant. Currently, only simulation is available; real deployment (for example, TensorRT deployment) support is coming soon.
• To support distributed checkpoint resume expert-parallel (EP), modelopt_state in Megatron Core distributed checkpoint (used in NeMo and Megatron-LM) is stored differently. The legacy modelopt_state in the distributed checkpoint generated by previous modelopt version can still be loaded in 0.27 and 0.29 but will need to be stored in the new format.
• Add triton-based NVFP4 quantization kernel that delivers approximately 40% performance improvement over the previous implementation.
• Add a new API mtq.compress <modelopt.torch.quantization.compress> for model compression for weights after quantization.
• Add option to simplify ONNX model before quantization is performed.
• (Experimental) Improve support for ONNX models with custom TensorRT op:
  • Add support for --calibration_shapes flag.
  • Add automatic type and shape tensor propagation for full ORT support with TensorRT EP.

Known Issues

• Quantization of T5 models is broken. Please use nvidia-modelopt==0.25.0 with transformers<4.50 meanwhile.