Instructions to use Tamazight/TrOCR-Tifinagh-Small with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Tamazight/TrOCR-Tifinagh-Small with Transformers:

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

pipe = pipeline("image-text-to-text", model="Tamazight/TrOCR-Tifinagh-Small")

# Load model directly
from transformers import AutoTokenizer, AutoModelForImageTextToText

tokenizer = AutoTokenizer.from_pretrained("Tamazight/TrOCR-Tifinagh-Small")
model = AutoModelForImageTextToText.from_pretrained("Tamazight/TrOCR-Tifinagh-Small")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Tamazight/TrOCR-Tifinagh-Small with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/Tamazight/TrOCR-Tifinagh-Small

SGLang

How to use Tamazight/TrOCR-Tifinagh-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 "Tamazight/TrOCR-Tifinagh-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": "Tamazight/TrOCR-Tifinagh-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 "Tamazight/TrOCR-Tifinagh-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": "Tamazight/TrOCR-Tifinagh-Small",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Tamazight/TrOCR-Tifinagh-Small with Docker Model Runner:
```
docker model run hf.co/Tamazight/TrOCR-Tifinagh-Small
```

TrOCR-Tifinagh-Small

File size: 935 Bytes

6b7196e

{
  "_from_model_config": false,
  "assistant_confidence_threshold": 0.4,
  "assistant_lookbehind": 10,
  "bos_token_id": 2,
  "decoder_start_token_id": 2,
  "diversity_penalty": 0.0,
  "do_sample": false,
  "early_stopping": true,
  "encoder_no_repeat_ngram_size": 0,
  "encoder_repetition_penalty": 1.0,
  "eos_token_id": [
    3,
    3
  ],
  "epsilon_cutoff": 0.0,
  "eta_cutoff": 0.0,
  "length_penalty": 2.0,
  "max_length": 64,
  "min_length": 0,
  "no_repeat_ngram_size": 3,
  "num_assistant_tokens": 20,
  "num_assistant_tokens_schedule": "constant",
  "num_beam_groups": 1,
  "num_beams": 4,
  "num_return_sequences": 1,
  "output_scores": false,
  "pad_token_id": 0,
  "remove_invalid_values": false,
  "repetition_penalty": 1.0,
  "return_dict_in_generate": false,
  "target_lookbehind": 10,
  "temperature": 1.0,
  "top_k": 50,
  "top_p": 1.0,
  "transformers_version": "5.0.0",
  "typical_p": 1.0,
  "use_cache": false
}