---
datasets:
- gozfarb/bluemoon_roleplay_300k_vicuna
language:
- en
---
## General
Bluemoon roleplay finetune of LLaMA 13B (2 roleplayers only).

*Note.* This is an intermediate version which has not been trained for sufficiently long to reach a satisfactory final loss value. The repository will be updated later with a model trained over additional epochs.

## Models
Two models are provided, labeled (1) `4k-epoch6` and (2) `epoch3`. In case of the (1), the training is extended over more epochs to reduce the high training loss observed in (2). This release also tests a longer 4k context token size achieved with AliBi.

*GGML 4-bit for llama.cpp*<br/>

1. ggml-bluemoonrp-13b-4k-epoch6-q5_0.bin
2. ggml-bluemoonrp-13b-epoch3-q5_0.bin

*GPTQ 4-bit CUDA:*<br/>

1. bluemoonrp-13b-4k-epoch6-4bit-128g.safetensors<br/>
2. bluemoonrp-13b-epoch3-4bit-128g.safetensors<br/>

## Remarks
This model has been trained using the following prompt (Vicuna 1.1 format):
```
A transcript of a roleplay between two players, LEAD and ASSOCIATE. LEAD sets up a scenario and the characters, from which ASSOCIATE then assumes a character role and continues the story for that role in response to description given by LEAD. The story and characters are developed by exchange of detailed event descriptions and character dialogs, successively given by both LEAD and ASSOCIATE.
LEAD: [role1 message]
ASSOCIATE: [role2 message]</s>
```