Search results

[Feature]: Merge Remaining Models from Sprint

feature requestNew feature or request. This includes new model, dtype, functionality supportNew feature or request. This includes new model, dtype, functionality support

new modelRequest to add a new modelRequest to add a new model

Status: Open.

#13446 In NVIDIA/TensorRT-LLM;

· govind-ramnarayan opened on Apr 24, 2026

[Feat] Support DeepSeek-V4 in PyTorch backend

Pytorch<NV>Pytorch backend related issues<NV>Pytorch backend related issues

Status: Open.

#13435 In NVIDIA/TensorRT-LLM;

· weikuo0506 opened on Apr 24, 2026

[New Model]: DeepSeek V4 support https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro

new modelRequest to add a new modelRequest to add a new model

Status: Open.

#13431 In NVIDIA/TensorRT-LLM;

· inference666 opened on Apr 24, 2026

[AutoDeploy][Feature]: Switch nemotron-H, qwen3, qwen3.5-moe, deepseek to sharding IR (step 2)

AutoDeploy<NV> AutoDeploy Backend<NV> AutoDeploy Backend

feature requestNew feature or request. This includes new model, dtype, functionality supportNew feature or request. This includes new model, dtype, functionality support

Status: Open.

#13429 In NVIDIA/TensorRT-LLM;

· greg-kwasniewski1 opened on Apr 24, 2026

[Feature]: Fix Qwen3.5, Gemma4 CI

feature requestNew feature or request. This includes new model, dtype, functionality supportNew feature or request. This includes new model, dtype, functionality support

Testing<NV>Continuous integration, build system, and testing infrastructure issues<NV>Continuous integration, build system, and testing infrastructure issues

Status: Open.

#13390 In NVIDIA/TensorRT-LLM;

· bmarimuthu-nv opened on Apr 23, 2026

[Feature]: Minimax 2.7 Perf Improvements

feature requestNew feature or request. This includes new model, dtype, functionality supportNew feature or request. This includes new model, dtype, functionality support

Status: Open.

#13389 In NVIDIA/TensorRT-LLM;

· bmarimuthu-nv opened on Apr 23, 2026

[Feature]: Support Nemotron Nano Omni

feature requestNew feature or request. This includes new model, dtype, functionality supportNew feature or request. This includes new model, dtype, functionality support

new modelRequest to add a new modelRequest to add a new model

Status: Open.

#13388 In NVIDIA/TensorRT-LLM;

· bmarimuthu-nv opened on Apr 23, 2026

[Usage]: Is Qwen3.5 MTP currently supported in TensorRT-LLM?

questionFurther information is requestedFurther information is requested

Speculative Decoding<NV>MTP/Eagle/Medusa/Lookahead/Prompt-Lookup-Decoding/Draft-Target-Model/ReDrafter<NV>MTP/Eagle/Medusa/Lookahead/Prompt-Lookup-Decoding/Draft-Target-Model/ReDrafter

Status: Open.

#13361 In NVIDIA/TensorRT-LLM;

· minedec opened on Apr 23, 2026

[Feature]: Enhance AllToAll performance

AutoDeploy<NV> AutoDeploy Backend<NV> AutoDeploy Backend

feature requestNew feature or request. This includes new model, dtype, functionality supportNew feature or request. This includes new model, dtype, functionality support

Status: Open.

Task

#13360 In NVIDIA/TensorRT-LLM;

· taylor-yb-lee opened on Apr 23, 2026

[Feature]: AutoDeploy: deprecate export_to_onnx backend

AutoDeploy<NV> AutoDeploy Backend<NV> AutoDeploy Backend

feature requestNew feature or request. This includes new model, dtype, functionality supportNew feature or request. This includes new model, dtype, functionality support

Status: Open.

Task

#13353 In NVIDIA/TensorRT-LLM;

· lucaslie opened on Apr 22, 2026

[AutoDeploy][spec-dec] resize_kv_cache sample batch classified as decode-only when max_num_tokens // max_batch_size == 1

Speculative Decoding<NV>MTP/Eagle/Medusa/Lookahead/Prompt-Lookup-Decoding/Draft-Target-Model/ReDrafter<NV>MTP/Eagle/Medusa/Lookahead/Prompt-Lookup-Decoding/Draft-Target-Model/ReDrafter

Status: Open.

#13348 In NVIDIA/TensorRT-LLM;

· govind-ramnarayan opened on Apr 22, 2026

[Bug]: Re-enable piecewise cudagraph for Qwen3.5

bugSomething isn't workingSomething isn't working

CUDA Graph

Status: Open.

#13321 In NVIDIA/TensorRT-LLM;

· bmarimuthu-nv opened on Apr 22, 2026