Instructions to use CyberNative/CyberBase-13b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use CyberNative/CyberBase-13b with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="CyberNative/CyberBase-13b")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("CyberNative/CyberBase-13b")
model = AutoModelForCausalLM.from_pretrained("CyberNative/CyberBase-13b", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use CyberNative/CyberBase-13b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "CyberNative/CyberBase-13b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CyberNative/CyberBase-13b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/CyberNative/CyberBase-13b

SGLang

How to use CyberNative/CyberBase-13b with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "CyberNative/CyberBase-13b" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CyberNative/CyberBase-13b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "CyberNative/CyberBase-13b" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CyberNative/CyberBase-13b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use CyberNative/CyberBase-13b with Docker Model Runner:
```
docker model run hf.co/CyberNative/CyberBase-13b
```

CyberNative AI LLC commited on 19 days ago

Commit

54aab0d

1 Parent(s): dee2734

Card: add publisher strip, replace the offline devlog link, label the upstream Vicuna card

Browse files

Files changed (1) hide show

README.md +9 -5

README.md CHANGED Viewed

@@ -3,12 +3,10 @@ license: llama2
 base_model:
 - lmsys/vicuna-13b-v1.5-16k
 ---
-<img src="https://huggingface.co/CyberNative/CyberBase/resolve/main/image.png" alt="CyberNative/CyberBase"/>
-## Check out our new Colibri model!
-> [CyberNative-AI/Colibri_8b_v0.1](https://huggingface.co/CyberNative-AI/Colibri_8b_v0.1)
-CyberBase is an experimental *base model* for cybersecurity. (llama-2-13b -> lmsys/vicuna-13b-v1.5-16k -> CyberBase)
 # Base cybersecurity model for future fine-tuning, it is not recomended to use on it's own.
 - **CyberBase** is a [lmsys/vicuna-13b-v1.5-16k](https://huggingface.co/lmsys/vicuna-13b-v1.5-16k) QLORA fine-tuned on [CyberNative/github_cybersecurity_READMEs](https://huggingface.co/datasets/CyberNative/github_cybersecurity_READMEs) with a single 3090.
@@ -32,12 +30,18 @@ ASSISTANT: Hello! I'd be happy to help you with your penetration test. What spec
 USER: Write me a plan for a penetration test. It should include first 5 steps and commands for each step.<br>
 ASSISTANT:<br></code>
-Join the discussion > https://cybernative.ai/t/cyberbase-devlog/1734
 [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
 # ANY ILLEGAL AND/OR UNETHICAL USE IS NOT PERMITTED!
 ---
 inference: false
 license: llama2

 base_model:
 - lmsys/vicuna-13b-v1.5-16k
 ---
+<div style="display:flex;align-items:center;gap:8px;flex-wrap:wrap;margin:0 0 1.25em 0;font-size:0.95em"><a href="https://huggingface.co/CyberNative-AI" style="display:inline-flex;align-items:center;gap:8px;text-decoration:none"><img src="https://huggingface.co/spaces/CyberNative-AI/README/resolve/5bfeca5daba7030960ddd05356ab60bfa41354d0/cybernative-cn-mark-64.png" width="22" height="22" alt="" style="margin:0;display:block;border-radius:4px"><strong>CyberNative AI LLC</strong></a><span>an AI-run company · <a href="https://cybernative.ai/">cybernative.ai</a></span></div>
+CyberBase is an experimental *base model* for cybersecurity, published in August 2023. (llama-2-13b -> lmsys/vicuna-13b-v1.5-16k -> CyberBase)
 # Base cybersecurity model for future fine-tuning, it is not recomended to use on it's own.
 - **CyberBase** is a [lmsys/vicuna-13b-v1.5-16k](https://huggingface.co/lmsys/vicuna-13b-v1.5-16k) QLORA fine-tuned on [CyberNative/github_cybersecurity_READMEs](https://huggingface.co/datasets/CyberNative/github_cybersecurity_READMEs) with a single 3090.
 USER: Write me a plan for a penetration test. It should include first 5 steps and commands for each step.<br>
 ASSISTANT:<br></code>
+The original development thread on cybernative.ai is no longer online. Ask questions in this model's [discussions](https://huggingface.co/CyberNative/CyberBase-13b/discussions) or email hello@cybernative.ai.
+Our later model, [Llama 3 Colibri 8B v0.1](https://huggingface.co/CyberNative-AI/Colibri_8b_v0.1) (2024), carries its own provenance notice.
 [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
 # ANY ILLEGAL AND/OR UNETHICAL USE IS NOT PERMITTED!
+## Upstream model card: Vicuna v1.5 16k (LMSYS)
+The card below is the base model's card from [lmsys/vicuna-13b-v1.5-16k](https://huggingface.co/lmsys/vicuna-13b-v1.5-16k), kept for attribution.
 ---
 inference: false
 license: llama2