Instructions to use zhiqiulin/clip-flant5-xl with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use zhiqiulin/clip-flant5-xl with Transformers:

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

pipe = pipeline("text-generation", model="zhiqiulin/clip-flant5-xl")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("zhiqiulin/clip-flant5-xl")
model = AutoModelForSeq2SeqLM.from_pretrained("zhiqiulin/clip-flant5-xl", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use zhiqiulin/clip-flant5-xl with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "zhiqiulin/clip-flant5-xl"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zhiqiulin/clip-flant5-xl",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/zhiqiulin/clip-flant5-xl

SGLang

How to use zhiqiulin/clip-flant5-xl 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 "zhiqiulin/clip-flant5-xl" \
    --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": "zhiqiulin/clip-flant5-xl",
		"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 "zhiqiulin/clip-flant5-xl" \
        --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": "zhiqiulin/clip-flant5-xl",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use zhiqiulin/clip-flant5-xl with Docker Model Runner:
```
docker model run hf.co/zhiqiulin/clip-flant5-xl
```

CLIP-FlanT5-XL (VQAScore)

This model is a fine-tuned version of google/flan-t5-xl designed for image-text retrieval tasks, as presented in the VQAScore paper.

Model Description

Developed by: Zhiqiu Lin and collaborators
Model type: Vision-Language Generative Model
License: Apache-2.0
Finetuned from model: google/flan-t5-xxl

Model Sources [optional]

Repository: https://github.com/linzhiqiu/CLIP-FlanT5
Paper: https://arxiv.org/pdf/2404.01291
Demo: https://huggingface.co/spaces/zhiqiulin/VQAScore

Downloads last month: 90,251

Model tree for zhiqiulin/clip-flant5-xl

Base model

google/flan-t5-xl

Finetuned

(28)

this model

Space using zhiqiulin/clip-flant5-xl 1

Paper for zhiqiulin/clip-flant5-xl

Evaluating Text-to-Visual Generation with Image-to-Text Generation

Paper • 2404.01291 • Published Apr 1, 2024 • 6