Instructions to use kavinh07/nid-ocr-vit-xlmroberta with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use kavinh07/nid-ocr-vit-xlmroberta with Transformers:

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

pipe = pipeline("image-text-to-text", model="kavinh07/nid-ocr-vit-xlmroberta")

# Load model directly
from transformers import AutoTokenizer, AutoModelForImageTextToText

tokenizer = AutoTokenizer.from_pretrained("kavinh07/nid-ocr-vit-xlmroberta")
model = AutoModelForImageTextToText.from_pretrained("kavinh07/nid-ocr-vit-xlmroberta")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use kavinh07/nid-ocr-vit-xlmroberta with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "kavinh07/nid-ocr-vit-xlmroberta"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kavinh07/nid-ocr-vit-xlmroberta",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/kavinh07/nid-ocr-vit-xlmroberta

SGLang

How to use kavinh07/nid-ocr-vit-xlmroberta 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 "kavinh07/nid-ocr-vit-xlmroberta" \
    --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": "kavinh07/nid-ocr-vit-xlmroberta",
		"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 "kavinh07/nid-ocr-vit-xlmroberta" \
        --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": "kavinh07/nid-ocr-vit-xlmroberta",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use kavinh07/nid-ocr-vit-xlmroberta with Docker Model Runner:
```
docker model run hf.co/kavinh07/nid-ocr-vit-xlmroberta
```

nid-ocr-vit-xlmroberta

This model is a fine-tuned version of microsoft/trocr-base-stage1 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 8.1543
Cer: 0.9993
Wer: 1.0

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: 9e-06
train_batch_size: 16
eval_batch_size: 32
seed: 42
distributed_type: multi-GPU
gradient_accumulation_steps: 4
total_train_batch_size: 64
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 1000
num_epochs: 1000

Training results

Training Loss	Epoch	Step	Validation Loss	Cer	Wer
No log	0	0	27.5340	3.6546	4.6053
7.6216	0.0801	250	7.6858	1.0	1.0
6.5641	0.1602	500	7.1390	1.0	1.0
6.2881	0.2402	750	6.9612	1.0	1.0
6.1757	0.3203	1000	6.8755	1.0	1.0
6.0908	0.4004	1250	6.8165	1.0	1.0
6.0241	0.4805	1500	6.7642	1.0	1.0
5.9962	0.5605	1750	6.7470	1.0	1.0
5.9477	0.6406	2000	6.7396	1.0	1.0
5.8893	0.7207	2250	6.8078	1.0	1.0
5.8618	0.8008	2500	6.9414	1.0	1.0
5.7779	0.8808	2750	7.1235	1.0	1.0
5.7539	0.9609	3000	7.2403	1.0	1.0
5.6984	1.0410	3250	7.5010	1.0	1.0
5.6579	1.1211	3500	7.9130	1.0	1.0
5.5958	1.2012	3750	8.1207	1.0	1.0
5.5785	1.2812	4000	8.3609	0.9998	1.0
5.5205	1.3613	4250	8.2284	0.9989	1.0
5.4865	1.4414	4500	8.1543	0.9993	1.0

Framework versions

Transformers 4.54.1
Pytorch 2.7.1+cu126
Datasets 4.5.0
Tokenizers 0.21.4

Downloads last month: 3

Safetensors

Model size

0.6B 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 kavinh07/nid-ocr-vit-xlmroberta

Base model

microsoft/trocr-base-stage1

Finetuned

(19)

this model