Instructions to use QuixiAI/WizardLM-Uncensored-Falcon-40b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use QuixiAI/WizardLM-Uncensored-Falcon-40b with Transformers:

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

pipe = pipeline("text-generation", model="QuixiAI/WizardLM-Uncensored-Falcon-40b", trust_remote_code=True)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("QuixiAI/WizardLM-Uncensored-Falcon-40b", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use QuixiAI/WizardLM-Uncensored-Falcon-40b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "QuixiAI/WizardLM-Uncensored-Falcon-40b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "QuixiAI/WizardLM-Uncensored-Falcon-40b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/QuixiAI/WizardLM-Uncensored-Falcon-40b

SGLang

How to use QuixiAI/WizardLM-Uncensored-Falcon-40b 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 "QuixiAI/WizardLM-Uncensored-Falcon-40b" \
    --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": "QuixiAI/WizardLM-Uncensored-Falcon-40b",
		"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 "QuixiAI/WizardLM-Uncensored-Falcon-40b" \
        --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": "QuixiAI/WizardLM-Uncensored-Falcon-40b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use QuixiAI/WizardLM-Uncensored-Falcon-40b with Docker Model Runner:
```
docker model run hf.co/QuixiAI/WizardLM-Uncensored-Falcon-40b
```

Thank you for the amazing model!

by max-fry - opened Jun 3, 2023

Discussion

max-fry

Jun 3, 2023

I was looking for uncensored falcon model for a while and now it's here! Thanks again!

ehartford

Quixi AI org Jun 5, 2023

More precisely, this is Uncensored WizardLM trained on Falcon base model. Not "uncensored falcon"

max-fry

Jun 5, 2023

@ehartford As far as I understand, tiiuae/falcon-40b is uncensored. It's tiiuae/falcon-40b-instruct that is censored. Am I wrong?

ehartford

Quixi AI org Jun 5, 2023

That is correct.
The foundational model has bias, but it is not censored, by definition, it is the source of knowledge.
An instructional model such as tiiuae/falcon-40b-instruct is usually censored - meaning it has refusals, avoidance, and bias.
So something like wizardlm-uncensored in trained with an instruction dataset that doesn't have (or at least has much less of) refusals, avoidance, and bias.

max-fry changed discussion status to closed Jun 5, 2023

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment