All issues

Search results

Decoding cannot be constrained: no per-step hook in the sampling path, and logit_bias cannot substitute for one

Status: Open.

#218 In NVIDIA/TensorRT-Edge-LLM;

· filipemartinsubrobotics opened on Sep 17, 2026

Direct export/build fails for Qwen3 VL FP8 fine-grained 2D block-FP8 checkpoints

bugSomething isn't workingSomething isn't working

Status: Open.

#217 In NVIDIA/TensorRT-Edge-LLM;

· Sunk011 opened on Sep 17, 2026

tensorrt-edgellm-build --dense nvfp4-qdq SIGSEGVs in a libnvinfer compiler worker thread on Jetson Thor (CC 11.0) for Cosmos-Reason2-2B; --dense fp16 builds fine from the same NVFP4 checkpoint

bugSomething isn't workingSomething isn't working

Status: Open.

#216 In NVIDIA/TensorRT-Edge-LLM;

· TeramotoRYEC opened on Sep 17, 2026

llm_inference ignores temperature/top_p/top_k from the input JSON its own help tells you to set them in

Status: Open.

#211 In NVIDIA/TensorRT-Edge-LLM;

· filipemartinsubrobotics opened on Sep 12, 2026

Int4GroupwiseGemmPluginV2 fails silently when ENABLE_CUTE_DSL omits int4_fp16_gemm: "Failed to enqueue status -1" with no diagnostic

Status: Open.

#210 In NVIDIA/TensorRT-Edge-LLM;

· filipemartinsubrobotics opened on Sep 11, 2026

ONNX-less direct builder cannot resolve decoder tensors for multimodal cosmos3_edge, and renaming them yields a numerically wrong engine

Status: Open.

#208 In NVIDIA/TensorRT-Edge-LLM;

· filipemartinsubrobotics opened on Sep 11, 2026

SM103 (B300 / Blackwell Ultra) is missing from five SM allowlists across the build and the attention path

Status: Open.

#206 In NVIDIA/TensorRT-Edge-LLM;

· filipemartinsubrobotics opened on Sep 11, 2026

XQA NVRTC JIT passes no CUDA include path: cuda_fp16.h cannot find vector_types.h on x86 CUDA 12.x

Status: Open.

#204 In NVIDIA/TensorRT-Edge-LLM;

· filipemartinsubrobotics opened on Sep 11, 2026

x86 CuteDSL cubins missing in the repo

bugSomething isn't workingSomething isn't working

Status: Open.

#203 In NVIDIA/TensorRT-Edge-LLM;

· victoroliv2 opened on Sep 10, 2026

feat: Add base64 image support to llm_inference

Status: Open.

#202 In NVIDIA/TensorRT-Edge-LLM;

· FABallemand opened on Sep 10, 2026

Experimental OpenAI-compatible server fails to start with: ModuleNotFoundError: No module named 'tensorrt'

Status: Open.

#201 In NVIDIA/TensorRT-Edge-LLM;

· FABallemand opened on Sep 9, 2026

Qwen3-TTS CodePredictor (MTP) crashes with "No available kernel for the GQA" on Jetson AGX Orin

Status: Open.

#198 In NVIDIA/TensorRT-Edge-LLM;

· WonderW-hub opened on Sep 4, 2026