File size: 1,176 Bytes
abf9af7 35ea32b abf9af7 67ebe44 abf9af7 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 |
---
datasets:
- raincandy-u/Dextromethorphan-50k-v0.1
library_name: transformers
pipeline_tag: text-generation
license: mit
---
# phillama-3.8b-v1
Phillama is a model based on Phi-3-mini and trained on Llama-generated datasets to make it more "llama-like".
Also, this model is converted into Llama format, so it will work with any Llama-2/3 workflow.
# Dataset
| Source | Task | Number of examples（k） |
| :-----------: | :------: | :--------------: |
| lmsys-1m | Chat | 50 |
| dolphin-coder | Code | 10 |
| slimorca | Reasoning | 10 |
<h1>For more information include training details, see <a href="https://angelkawaii.xyz/2024/04/26/llama-3-finetune-1/">this blog post</a></h1>
# System prompt
`You are a humanoid AI assistant. You think step by step and give detailed long response.`
# Prompt template
```
<|system|>
You are a humanoid AI assistant. You think step by step and give detailed long response.<|end|>
<|user|>
Why people like llama?<|end|>
<|assistant|>
``` |