Instructions to use Image-Captioning-ML/Vit-GPT2-UCA-UCF-07 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Image-Captioning-ML/Vit-GPT2-UCA-UCF-07 with Transformers:

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

pipe = pipeline("image-text-to-text", model="Image-Captioning-ML/Vit-GPT2-UCA-UCF-07")

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("Image-Captioning-ML/Vit-GPT2-UCA-UCF-07")
model = AutoModelForMultimodalLM.from_pretrained("Image-Captioning-ML/Vit-GPT2-UCA-UCF-07")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Image-Captioning-ML/Vit-GPT2-UCA-UCF-07 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Image-Captioning-ML/Vit-GPT2-UCA-UCF-07"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Image-Captioning-ML/Vit-GPT2-UCA-UCF-07",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Image-Captioning-ML/Vit-GPT2-UCA-UCF-07

SGLang

How to use Image-Captioning-ML/Vit-GPT2-UCA-UCF-07 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 "Image-Captioning-ML/Vit-GPT2-UCA-UCF-07" \
    --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": "Image-Captioning-ML/Vit-GPT2-UCA-UCF-07",
		"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 "Image-Captioning-ML/Vit-GPT2-UCA-UCF-07" \
        --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": "Image-Captioning-ML/Vit-GPT2-UCA-UCF-07",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Image-Captioning-ML/Vit-GPT2-UCA-UCF-07 with Docker Model Runner:
```
docker model run hf.co/Image-Captioning-ML/Vit-GPT2-UCA-UCF-07
```

Vit-GPT2-UCA-UCF-07

This model is a fine-tuned version of NourFakih/Vit-GPT2-UCA-UCF-06 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.1968
Rouge1: 34.6433
Rouge2: 13.5351
Rougel: 29.5099
Rougelsum: 30.0007
Gen Len: 16.002

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-05
train_batch_size: 4
eval_batch_size: 4
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 16
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.4617	0.5469	500	0.1655	34.1712	12.9219	29.0744	29.6374	16.407
0.4256	1.0930	1000	0.1755	34.2664	13.121	29.2664	29.8242	15.724
0.3498	1.6399	1500	0.1807	34.9169	13.5342	29.5801	30.157	16.269
0.3158	2.1859	2000	0.1921	33.9586	12.8412	28.6693	29.1732	16.157
0.2768	2.7328	2500	0.1968	34.6433	13.5351	29.5099	30.0007	16.002

Framework versions

Transformers 4.47.0
Pytorch 2.5.1+cu121
Datasets 3.3.1
Tokenizers 0.21.0

Downloads last month: 2

Safetensors

Model size

0.2B params

Tensor type

F32

Inference Providers NEW

Image-Text-to-Text

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Image-Captioning-ML/Vit-GPT2-UCA-UCF-07

Base model

nlpconnect/vit-gpt2-image-captioning

Finetuned

Image-Captioning-ML/Vit-GPT2-UCA-UCF-06

Finetuned

(1)

this model