Instructions to use TianHuiLab/Falcon-Single-Instruction-Large with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TianHuiLab/Falcon-Single-Instruction-Large with Transformers:

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

pipe = pipeline("image-text-to-text", model="TianHuiLab/Falcon-Single-Instruction-Large")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("TianHuiLab/Falcon-Single-Instruction-Large", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use TianHuiLab/Falcon-Single-Instruction-Large with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/TianHuiLab/Falcon-Single-Instruction-Large

SGLang

How to use TianHuiLab/Falcon-Single-Instruction-Large 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 "TianHuiLab/Falcon-Single-Instruction-Large" \
    --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": "TianHuiLab/Falcon-Single-Instruction-Large",
		"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 "TianHuiLab/Falcon-Single-Instruction-Large" \
        --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": "TianHuiLab/Falcon-Single-Instruction-Large",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use TianHuiLab/Falcon-Single-Instruction-Large with Docker Model Runner:
```
docker model run hf.co/TianHuiLab/Falcon-Single-Instruction-Large
```

Follow-Up on Falcon Model Access Request

by onurulu17 - opened Jun 15, 2025

Discussion

onurulu17

Jun 15, 2025

Dear Team,

I have successfully submitted my access request for the Falcon model, providing all the required information in accordance with the Meta Privacy Policy and the Falcon Community License Agreement.

I would like to kindly follow up on the status of my request. I look forward to your response and hope to receive approval soon.

Thank you for your time and consideration.

Best regards,
Muhammed Onur ULU
Master Student, Computer Engineering
Yıldız Technical University
mronurulu@gmail.com

dukeyu

Jun 30, 2025

the same request , looking forward to the approval of the model

onurulu17

Jun 30, 2025

Hello,

Thank you for making this vision-language model available on Hugging Face. I’m currently working on a project involving zero-shot object detection, and I would like to use this model for that purpose. However, I couldn’t find a usage guide or any example code for inference through Hugging Face (either in the model card or repository). I also tried following the GitHub README, but it does not seem to be compatible with the Hugging Face setup or lacks details on how to run inference directly.

Would it be possible to provide a minimal example (e.g., Python code snippet) showing how to perform zero-shot object detection using this model via the Hugging Face Transformers or diffusers ecosystem? Even just loading the model and passing a sample image and prompt would be extremely helpful.

Thanks in advance for your time and support!

Best regards

laxm12

Jul 22, 2025

The same request , looking forward to the approval of the model

JacoCronje

Mar 26

Hi, has anyone been able to get hold of the model and gain access to the model repository?

onurulu17

Mar 27

No. nevermind it. u can use other open-source vlms

JacoCronje

Mar 27

No. nevermind it. u can use other open-source vlms

Can you recommend a good open-source vlm for satellite processing that is smallish?
The falcon one looked promising because it is only 0.7B parameters.

onurulu17

Mar 27

there are so many image processing task. if u need a model for real-time object detection, u can use transfer learning YOLOv11 or YOLO26 but u need to some labelled data. u can train model on colab GPU. i dont recommend VLMs for object detection task bcz transformer based models are not succesfull for it but if u use VLMs for VQA task u can prefer Qwen has new models. maybe u research the latest articles on google scholar about your image processing purpose.

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