|
--- |
|
license: other |
|
tags: |
|
- chatbot |
|
- gptq |
|
- storywriting |
|
--- |
|
|
|
# chronos-13b-8K-4bit |
|
|
|
The original Chronos-13B model was merged with a LoRA trained on a majority of 1500 samples in the 8000 token range in the same style, with a cutoff of 8k tokens in full 8bit. It is meant to be used standalone, but if you would like the LoRA to merge/combine on your own, you can find it here https://huggingface.co/ZeusLabs/chronos-13b-8k-lora |
|
|
|
The `config.json` includes modifications allowing extended context so you will need to use it with `trust_remote_code` if not using Exllama. |
|
|
|
4bit (int4) quantized version using `true-sequential` and `groupsize 128` of https://huggingface.co/elinas/chronos-13b plus https://huggingface.co/ZeusLabs/chronos-13b-8k-lora |
|
|
|
This model is primarily focused on chat, roleplay, and storywriting, but can accomplish other tasks such as simple reasoning and coding. |
|
|
|
Chronos generates very long outputs with coherent text, largely due to the human inputs it was trained on. |
|
|
|
This model uses Alpaca formatting, so for optimal model performance, use: |
|
``` |
|
### Instruction: |
|
Your instruction or question here. |
|
### Response: |
|
``` |
|
|
|
[Zeus Labs Discord](https://discord.gg/76e2HBzRKD) |