Search results

🐛 [Bug] Encountered "interrupted by signal 8:SIGFPE" using Torch-TensorRT

bugSomething isn't workingSomething isn't working

Status: Open.

#4165 In pytorch/TensorRT;

· shilovav3 opened on Apr 6, 2026

convert_exported_program_to_serialized_trt_engine fails with AssertionError on TorchExportableModuleWithStaticCache exports

Status: Open.

#4162 In pytorch/TensorRT;

· Mgluhovskoi opened on Apr 2, 2026

[Bug] 17 dynamo test failures across all GPUs on CUDA 13.x — SDP use_fp32_acc, BERT accuracy, AutomaticPlugin, refit

bugSomething isn't workingSomething isn't working

Status: Open.

#4154 In pytorch/TensorRT;

· apbose opened on Mar 30, 2026

[Bug] MutableTorchTRTModule refit flag stuck at NEEDS_REFIT, never transitions to LIVE on B100/H100 with CUDA 13.x

bugSomething isn't workingSomething isn't working

Status: Open.

#4153 In pytorch/TensorRT;

· apbose opened on Mar 30, 2026

[Bug] C++ API and core tests fail with "assert 3 == 0" across all GPUs and CUDA 13.x versions

bugSomething isn't workingSomething isn't working

Status: Open.

#4152 In pytorch/TensorRT;

· apbose opened on Mar 30, 2026

[Bug] TestTorchTensorRTModule::test_get_layer_info AssertionError on H100 with CUDA 13.x

bugSomething isn't workingSomething isn't working

Status: Open.

#4151 In pytorch/TensorRT;

· apbose opened on Mar 30, 2026

[Bug] TestWeightStrippedEngine OOM on RTX 3070 (8GB) with CUDA 12.x

bugSomething isn't workingSomething isn't working

Status: Open.

#4150 In pytorch/TensorRT;

· apbose opened on Mar 30, 2026

🐛 [Bug] add_attention fails with dynamic head dimension when exporting GQA models via torch.export

bugSomething isn't workingSomething isn't working

Status: Open.

#4145 In pytorch/TensorRT;

· chohk88 opened on Mar 24, 2026

[Bug] index_copy_ in transformers StaticCache causes TRT conversion failure (index_put broadcast error)

story: Operator Coverage & ConvertersMissing ATen op converters, converter accuracy bugs, type-casting issues, and new operator support rMissing ATen op converters, converter accuracy bugs, type-casting issues, and new operator support r

Status: Open.

#4142 In pytorch/TensorRT;

· Mgluhovskoi opened on Mar 19, 2026

[Bug] All-zero logits from convert_exported_program_to_serialized_trt_engine with use_explicit_typing=False

Story: Runtime & Memory & SerializationEngine execution, CUDA graphs, weight streaming, model save/load, refit, memory management, and runEngine execution, CUDA graphs, weight streaming, model save/load, refit, memory management, and run

Status: Open.

#4141 In pytorch/TensorRT;

· Mgluhovskoi opened on Mar 19, 2026

✨[Feature] Refit Perf Study and Remove Weight Name Map

feature requestNew feature or requestNew feature or request

Story: Runtime & Memory & SerializationEngine execution, CUDA graphs, weight streaming, model save/load, refit, memory management, and runEngine execution, CUDA graphs, weight streaming, model save/load, refit, memory management, and run

Status: Open.

#4140 In pytorch/TensorRT;

· cehongwang opened on Mar 19, 2026

🐛 [Bug] aten::index_put converter fails with "Dynamic shape in free dimensions is not supported" when a non-indexed dimension is dynamic

bugSomething isn't workingSomething isn't working

story: Operator Coverage & ConvertersMissing ATen op converters, converter accuracy bugs, type-casting issues, and new operator support rMissing ATen op converters, converter accuracy bugs, type-casting issues, and new operator support r

Status: Open.

#4139 In pytorch/TensorRT;

· narendasan opened on Mar 19, 2026