llama-anon's picture
Update README.md
f1d550d
|
raw
history blame
277 Bytes
---
license: agpl-3.0
---
LLaMA-13B merged with Instruct-13B weights, unlike the bare weights it does not output gibberish.
Quantized to 4bit (jarted)
Feel free to donate:
XMR: 86Z8nLSVPx3SZ5z7iWugeK5JruAeGPUJyExD9e3wdTSxUvFMhGXNG9ucPqCm8M29y1AxP6ta56GBQ4GiEUMzeew9MfX1yct