Search results

[Bug]: Memory leak: Qwen3 30B A3B pytorch backend

bugSomething isn't workingSomething isn't working

MemoryMemory utilization in TRTLLM: leak/OOM handling, footprint optimization, memory profiling.Memory utilization in TRTLLM: leak/OOM handling, footprint optimization, memory profiling.

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

Status: Open.

#13777 In NVIDIA/TensorRT-LLM;

· anurags25 opened on May 5, 2026

[chore] AutoDeploy: Remove Patches

AutoDeploy<NV> AutoDeploy Backend<NV> AutoDeploy Backend

Status: Open.

#13776 In NVIDIA/TensorRT-LLM;

· govind-ramnarayan opened on May 5, 2026

[Feature]: AutoDeploy: Merge Models with Substantial Infrastructure Changes

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.

#13775 In NVIDIA/TensorRT-LLM;

· govind-ramnarayan opened on May 5, 2026

[chore] AutoDeploy: Remove Unused Patches

AutoDeploy<NV> AutoDeploy Backend<NV> AutoDeploy Backend

Status: Open.

#13774 In NVIDIA/TensorRT-LLM;

· govind-ramnarayan opened on May 5, 2026

CleanUp and merge DeepSeek v4 support

deepseek-v4

Status: Open.

#13770 In NVIDIA/TensorRT-LLM;

· bmarimuthu-nv opened on May 5, 2026

[AutoDeploy][Feature]: sharding ir early testing

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.

#13769 In NVIDIA/TensorRT-LLM;

· greg-kwasniewski1 opened on May 5, 2026

[Feature]: [AutoDeploy] BTK: Migrate to BTK2.0

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.

#13764 In NVIDIA/TensorRT-LLM;

· galagam opened on May 5, 2026

[AutoDeploy]: BTK pipeline report skill/script

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.

#13763 In NVIDIA/TensorRT-LLM;

· galagam opened on May 5, 2026

ad-executor - analyze multimodal metadata induced synchronize

MultimodalLabel for issues & PRs regarding Multimodal related objectsLabel for issues & PRs regarding Multimodal related objects

Status: Open.

#13742 In NVIDIA/TensorRT-LLM;

· bmarimuthu-nv opened on May 4, 2026

Add automatic doc generation for AD transforms

Doc<NV>TRTLLM's textual/illustrative materials: API refs, guides, tutorials. Improvement & clarity.<NV>TRTLLM's textual/illustrative materials: API refs, guides, tutorials. Improvement & clarity.

Status: Open.

#13726 In NVIDIA/TensorRT-LLM;

· bmarimuthu-nv opened on May 4, 2026

[AutoDeploy][Feature]: Stateful MoE all-to-all (instance cache + runtime max-tokens)

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.

#13718 In NVIDIA/TensorRT-LLM;

· greg-kwasniewski1 opened on May 3, 2026

[Bug]: AutoDeploy: Qwen3Next crashes during build_and_run_ad

AutoDeploy<NV> AutoDeploy Backend<NV> AutoDeploy Backend

bugSomething isn't workingSomething isn't working

Status: Open.

#13699 In NVIDIA/TensorRT-LLM;

· govind-ramnarayan opened on May 2, 2026