Instructions to use Serialtechlab/dhivehi-trocr-small with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Serialtechlab/dhivehi-trocr-small with Transformers:

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

pipe = pipeline("image-text-to-text", model="Serialtechlab/dhivehi-trocr-small")

# Load model directly
from transformers import AutoTokenizer, AutoModelForImageTextToText

tokenizer = AutoTokenizer.from_pretrained("Serialtechlab/dhivehi-trocr-small")
model = AutoModelForImageTextToText.from_pretrained("Serialtechlab/dhivehi-trocr-small")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Serialtechlab/dhivehi-trocr-small with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/Serialtechlab/dhivehi-trocr-small

SGLang

How to use Serialtechlab/dhivehi-trocr-small 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 "Serialtechlab/dhivehi-trocr-small" \
    --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": "Serialtechlab/dhivehi-trocr-small",
		"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 "Serialtechlab/dhivehi-trocr-small" \
        --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": "Serialtechlab/dhivehi-trocr-small",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Serialtechlab/dhivehi-trocr-small with Docker Model Runner:
```
docker model run hf.co/Serialtechlab/dhivehi-trocr-small
```

dhivehi-trocr-small / README.md

Serialtechlab

End of training

b4b85ec verified about 2 months ago

preview code

raw

history blame contribute delete

3.55 kB

metadata

library_name: transformers
base_model: microsoft/trocr-small-printed
tags:
  - generated_from_trainer
model-index:
  - name: dhivehi-trocr-small
    results: []

dhivehi-trocr-small

This model is a fine-tuned version of microsoft/trocr-small-printed on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.0606
Cer: 0.1147

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: 4e-05
train_batch_size: 32
eval_batch_size: 32
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 6
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Cer	Validation Loss
2.0301	0.1337	1000	0.4438	1.9695
1.8511	0.2673	2000	0.4465	1.7758
1.7348	0.4010	3000	0.4479	1.6484
1.6226	0.5346	4000	0.4414	1.5418
1.5449	0.6683	5000	0.4396	1.4657
1.4864	0.8019	6000	0.4335	1.3844
1.3952	0.9356	7000	0.4282	1.3176
1.3033	1.0692	8000	0.4102	1.1908
1.1261	1.2029	9000	0.3844	1.0153
0.9671	1.3365	10000	0.3496	0.9206
0.8140	1.4702	11000	0.3190	0.9407
0.6986	1.6038	12000	0.2873	0.6398
0.6410	1.7375	13000	0.2570	0.5260
0.5472	1.8712	14000	0.2511	0.7460
0.4880	2.0048	15000	0.2288	0.4803
0.4308	2.1385	16000	0.2219	0.5494
0.4045	2.2721	17000	0.2115	0.4733
0.3729	2.4058	18000	0.2059	0.4428
0.3587	2.5394	19000	0.1910	0.3564
0.3318	2.6731	20000	0.1906	0.4022
0.3069	2.8067	21000	0.1927	0.5067
0.3014	2.9404	22000	0.1782	0.3107
0.3963	1.2283	23000	0.2111	0.6155
0.3335	1.2817	24000	0.1934	0.3839
0.3227	1.3351	25000	0.1777	0.2538
0.2834	1.3885	26000	0.1752	0.2734
0.2660	1.4419	27000	0.1617	0.2125
0.2424	1.4953	28000	0.1682	0.2456
0.1846	3.2041	30000	0.3984	0.1719
0.1262	3.7381	35000	0.1237	0.1373
0.0982	4.2721	40000	0.0775	0.1236
0.0712	4.8062	45000	0.1080	0.1265
0.0616	5.3402	50000	0.0868	0.1191
0.0527	5.8742	55000	0.0606	0.1147

Framework versions

Transformers 5.0.0
Pytorch 2.10.0+cu128
Datasets 4.0.0
Tokenizers 0.22.2