torchchat
torchchat copied to clipboard
pytorch
Run PyTorch LLMs locally on servers, desktop and mobile
This PR is in progress but to checkpoint progress: 1 - Now uses torchtune model builders to create the model architecture (i.e. llama 3 8b). Previously it used pure HF...
### 🐛 Describe the bug I am using the streamlit version of streamlit 1.37.1 and while I run the browser version of torchat for llama2. Getting the following error `2024-08-15...
I want to use torchchat to load the trained model directly from the local. How to change the torchchat/config/data/models.json? Need to change download _ and _ convert in download.py?And, what...
### 🚀 The feature, motivation and pitch Select a method for enforcing types in Python dataclasses (i.e. pydantic) ### Alternatives Don't enforce types - may cause type errors when decoding...
### 🚀 The feature, motivation and pitch torchchat provides quantization functionality but the interaction isn't ideal. ### Currently Currently you can run generate with a --quantization flag and it quantize...
### 🚀 The feature, motivation and pitch torchchat supports adding models to the "known_model" list and has CLI support for local models not hosted in torchchat's, but this can be...
Introduced in https://github.com/pytorch/torchchat/pull/1046, the `torchchat/edge` path header reflects the structure in the repo, but is unnecessary when abd'ing the android local path. This PR just undo's that extra indirection
This PR is a major restructuring of the files in the torchchat repo. In order to make this safe and easy to review, each of the commits are squashed into...