lxe

I was looking for a push-to-talk tool that allows me to just paste my speech transcription automatically into whatever application I'm using. I wasn't able to find anything simple enough that works, so I built my own.…

llm-companion is a rather fast and high-quality (I think) TTS push-to-talk chat interface with OpenAI-compatible LLM APIs, such as OpenAI, Oobabooga, vllm, llama.cpp, etc. It uses Whisper for transcription (local),…

A simple mobile web app inspired by Fuzzy-Search/realtime-bakllava that uses llama.cpp server backend with multimodal mode to describe and narrate what the phone camera sees. I built this thing in a few hours using a…

Using threaded emscripten to speed up the generation and offload the main loop. No SIMD or other optimizations. Might work faster with #enable-experimental-webassembly-features enabled. Tested in x86 Chrome and Firefox,…

Yes I know, this is yet another language model. I used PEFT to finetune the Cerebras-GPT-2.7B on Alpaca, which makes for a very very fast, coherent, albeit hallucinating model. It took 5 hours on a vast.ai machine at 29…

I've been playing around with https://github.com/zphang/minimal-llama/ and https://github.com/tloen/alpaca-lora/blob/main/finetune.py, and wanted to create a simple UI where you can just paste text, tweak the…