File size: 3,826 Bytes
8ce4a12 f95f6d1 8ce4a12 8350f8e 0b64ff7 57b8136 5c0d530 137e1fd 8ce4a12 9cfe2f6 8ce4a12 9cfe2f6 8ce4a12 47620f1 8ce4a12 |
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 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 |
---
license: llama3.1
datasets:
- Borcherding/OARC_Commander_v001
language:
- en
base_model:
- unsloth/Llama-3.2-3B-bnb-4bit
---
# 🎮 OARC_Commander_v0.1-Llama-3.2-3B 🎮
Part of the [🎮OARC_Commander🎮 Collection](https://huggingface.co/collections/Borcherding/oarc-commander-66e7e6cfa4bbc652ff6c0699)
Curated and trained by [Leo Borcherding](https://huggingface.co/Borcherding)
[](https://discord.gg/mNeQZzBHuW)
OARC Discord: https://discord.gg/mNeQZzBHuW
[](https://github.com/Leoleojames1/ollama_agent_roll_cage)
OARC Github: https://github.com/Leoleojames1/ollama_agent_roll_cage
<img src="https://cdn-uploads.huggingface.co/production/uploads/66004e81b1e509e1e47ef31d/UpVFp2JTXJvgx38en2G7o.webp" alt="Image description" width="500" height="200">
## What is OARC_Commander_v0.1?
OARC_Commander_v0.1 is the first generation of the OARC_Commander series of models.
OARC_Commander aims to be a general purpose Llama-3.2-3B model, while also having been finetuned on set of user commands in Ollama Agent Roll Cage (OARC).
## Chat Template
We use ChatML for the chat template.
```
<|im_start|>system
You are OARC_Commander, a helpful AI assistant that has been trained on the command list from ollama agent roll cage. You will guide the OARC user, and help them understand the meaning of any oarc information that they will request from you.<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## System Prompt
In OARC_Commander, the system prompt is what you can use to set the tone and alignment of the responses for OARC_Commander if you wish.
Example use of system prompt:
```
<|im_start|>system
You are OARC_Commander, a helpful AI assistant that has been trained on the command list from ollama agent roll cage. You will guide the OARC user, and help them understand the meaning of any oarc information that they will request from you.<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Sample Outputs
Here are some sample outputs from OARC_Commander_v001, showcasing the commanding data from the Borcherding/OARC_Commander_v001 training data, I am aware that there are like greek and korean tokens being randomly output, I am working on figuring out why this is, and will update the repo with the fixed version once I have ironed this out.
## How to use
There are many ways to use a huggingface model including:
- ollama
- Huggingface Transformers library
- vllm
- LM Studio
### ollama usage
- [Install ollama](https://ollama.com/download)
- ```ollama run hf.co/Borcherding/OARC_Commander_v001_beta:Q8_0```
- ```instead of Q8_0, you can try F16, or Q4_K_M as well, these commands can be found in the top right under "Use this model"```
- ```once you have ollama loaded try asking the mode something like:```
- ```"what is the purpose of the /swap command?"```
- ```you can also try:```
- ```/set system <your system prompt>```
## Evals
TBD
## Appreciation
Respect and thanks to the creators of the open source datasets that were used:
Special thanks to
- Meta, and @mlabonne who wrote papers and published models that were instrumental in creating OARC_Commander_v001.
|