---
language:
- zh
- en
license: apache-2.0
tags:
- text-generation-inference
- transformers
- unsloth
- trl
- sft
- yi
base_model: cognitivecomputations/dolphin-2.9.1-yi-1.5-9b
datasets:
- TouchNight/HumanlikeRP
---

# HumanlikeRP

It is an attempt to build a Humanlike chatbot.  
Designed to make it give short reply like a real human.

It is a failure, the dataset used to train it has weak context relevancy. So it often generates irrelevant answer. And it is also overfitting.

### Chat Format

This model has been trained to use ChatML format.

```
<|im_start|>system
{{system}}<|im_end|>
<|im_start|>{{char}}
{{message}}<|im_end|>
<|im_start|>{{user}}
{{message}}<|im_end|>
```

# Uploaded  model

- **Developed by:** TouchNight
- **License:** apache-2.0
- **Finetuned from model :** cognitivecomputations/dolphin-2.9.1-yi-1.5-9b

This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.

[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)