Stepan Kargaltsev

utt2spk is copied twice in augment_data_dir.py

4

https://github.com/kaldi-asr/kaldi/blob/master/egs/wsj/s5/steps/data/augment_data_dir.py#L191 and https://github.com/kaldi-asr/kaldi/blob/master/egs/wsj/s5/steps/data/augment_data_dir.py#L187 Perhaps, spk2utt should be copied instead?

Fix armv8 linux build

2

Related to [issue 112](https://github.com/pytorch/cpuinfo/issues/112) To be honest, I am quite new to working with ARM, so here are some links to support my claims: 1. > In ARMv8, AArch32 state...

enc_dec: prompt_embedding_table not passed to encoder model

3

### System Info Tensorrt-LLM commit: 2a115dae84f13daaa54727534daa837c534eceb4 TensorRT-LLM version: 0.11.0.dev2024061800 ### Who can help? _No response_ ### Information - [X] The official example scripts - [X] My own modified scripts ###...

prompt_vocab_size is ignored in executor API

1

### System Info It seems like executor API ignores `prompt_vocab_size` argument and passes `max_prompt_embedding_table_size` to trt engine instead. I observe such behaviour using either 0.10.0 python api (well, ModelRunnerCpp to...

TLDR: When trying to forward FP8Linear layer twice, and error occurs ``` ... File "/usr/local/lib/python3.10/dist-packages/tensorrt_llm/module.py", line 40, in __call__ output = self.forward(*args, **kwargs) File "/usr/local/lib/python3.10/dist-packages/tensorrt_llm/quantization/layers.py", line 892, in forward alpha...

[BUG] failing assert in 57_hopper_grouped_gemm example

5

**Describe the bug** When running 57_hopper_grouped_gemm using large enough output [pipeline_check_is_consumer] (https://github.com/NVIDIA/cutlass/blob/main/include/cutlass/pipeline/sm90_pipeline.hpp#L84) assert is failing. It is masked by NDEBUG being defined by default, but if one were to, e.g,...