trtllm

Modules

modelopt.torch.export.trtllm.distribute	torch.distribute utils.
modelopt.torch.export.trtllm.hf_config_map	Define the config mapping between HF and modelopt.
modelopt.torch.export.trtllm.layer_utils	Utils for TensorRT-LLM model_config export.
modelopt.torch.export.trtllm.mcore_config_map	Define the config mapping between Mcore and modelopt.
modelopt.torch.export.trtllm.model_config	This module defines the TensorRT-LLM checkpoint model_config format.
modelopt.torch.export.trtllm.model_config_export	Code that export optimized models to the TensorRT-LLM checkpoint.
modelopt.torch.export.trtllm.model_config_utils	Common utils for the ModelConfig.
modelopt.torch.export.trtllm.postprocess	Utils to load and process model_config.
modelopt.torch.export.trtllm.quant_utils	Quantization helpers used only by the TensorRT-LLM checkpoint export path.
modelopt.torch.export.trtllm.tensorrt_llm_type	Code from TRT-LLM that export optimized models to the TensorRT-LLM checkpoint.
modelopt.torch.export.trtllm.tensorrt_llm_utils	Utils for TensorRT-LLM checkpoint export.

Checkpoint export logic for the TensorRT-LLM specific format.

Deprecation Notice: The export_tensorrt_llm_checkpoint API is deprecated as of 0.48.0 and will be removed in 0.49.0. Users are encouraged to transition to the unified HF export API (export_hf_checkpoint), which provides enhanced functionality and flexibility for exporting models to multiple inference frameworks including TensorRT-LLM, vLLM, and SGLang.