protonicage

### System Info For this case not necessary, but I use the 25.09 ngc tensorrt llm container for triton inference server. ### Who can help? @juney-nvidia @kaiyux ### Information -...

feat: increase performance - only calculate embeddings on tensor channels which are not zero

1

From my understanding currently embeddings are calculated on the whole batch, even all zero channels, this code addresses that to increase performance of the embedding extraction.

Python backend: Starting multiple GPU instances is currently not feasible/possible with a global config

1

**Description** Lets say you want to start multiple GPU instances for a python triton model. How do you do it? Short answer: I think it is currently not possible. Example:...

**Description** I want to use batching or dynamic batching with a decoupled python model. However the usual approach of iterating over requests and appending tensor to a global list does...

Onnx backend nout found

1

**Description** I get the the error: `"failed to load 'ambernet' version 1: Invalid argument: unable to find backend library for backend 'onnxruntime', try specifying runtime on the model configuration."` **Triton...

**Is your feature request related to a problem? Please describe.** So I used your script to build my own onnx backend and integrated it successfully into the triton server I...

Delta releases are gone?

1

So Mozilla launched this new approach to download the common voice dataset, which is fine. However I would like to know, how I can get a delta release now? Why?...