Michael Feil

Integrate multiple workflows as tabs or subpaths

3

 Idea: - have multiple chains / workflows running side-by-side one tab could then be - Q&A - one could be Google Query - Image Generation with DallE This would...

Fine-tuning of OpenKoala

3

Thanks for opening up LLama! Writing this as issue is easier said than the training itself, so propably this issue to you as authors of Koala.. As the diff weights...

Adding support for CTranslate2 acceleration

6

## Feature request The feature would be to support accelerated inference with the CTranslate2 framework https://github.com/OpenNMT/CTranslate2 ## Motivation Reasons to CTranslate2 #### faster float16 generation In my case, outperforms VLLM...

CodeGen Converter

1

This PR aims to integrate CodeGen. Work in progress, not ready.

Creating a Docker image or Dockerfile from this repo.

8

It would be awesome to build a docker image for this repo.

Split converted model.bin into multiple .bin

2

Something that might be coming up more often with Models such as LLama-2-70b / Bloom etc. Is there any way to split the model.bin into multiple subfiles Example: e.g in...

[VLLM ](https://github.com/vllm-project/vllm) implemented a mechanism called "PagedAttention", which helps in fast generation of long sequences. This is might be quite a large feature request. Blog: https://vllm.ai/ and maybe this https://github.com/vllm-project/vllm/blob/665c48963be11b2e5cb7209cd25f884129e5c284/vllm/model_executor/layers/attention.py#L16...

### 🚀 The feature I propose the integration of the [Infinity framework](https://github.com/michaelfeil/infinity) into embedchain to significantly speed up embedding inference. Infinity is a pure Python framework designed to enhance the...

I am building https://github.com/michaelfeil/infinity and would love to contribute to ollama. It is compatible with cuda, cpu and mps, with the option to run onnx models. Beta also for torch...

Warning : `load_model` does not return WordVectorModel or SupervisedModel any more, but a `FastText` object which is very similar. https://github.com/facebookresearch/fastText/issues/1056