Instructions to use rushai-dev/THAI-TrOCR with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use rushai-dev/THAI-TrOCR with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="rushai-dev/THAI-TrOCR")# Load model directly from transformers import AutoTokenizer, AutoModelForImageTextToText tokenizer = AutoTokenizer.from_pretrained("rushai-dev/THAI-TrOCR") model = AutoModelForImageTextToText.from_pretrained("rushai-dev/THAI-TrOCR") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use rushai-dev/THAI-TrOCR with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "rushai-dev/THAI-TrOCR" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "rushai-dev/THAI-TrOCR", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/rushai-dev/THAI-TrOCR
- SGLang
How to use rushai-dev/THAI-TrOCR 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 "rushai-dev/THAI-TrOCR" \ --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": "rushai-dev/THAI-TrOCR", "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 "rushai-dev/THAI-TrOCR" \ --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": "rushai-dev/THAI-TrOCR", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use rushai-dev/THAI-TrOCR with Docker Model Runner:
docker model run hf.co/rushai-dev/THAI-TrOCR
Commit ·
37caca0
1
Parent(s): 26b5dfd
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,25 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
| 4 |
+
|
| 5 |
+
```python
|
| 6 |
+
from transformers import TrOCRProcessor, AutoTokenizer, ViTFeatureExtractor, VisionEncoderDecoderModel
|
| 7 |
+
|
| 8 |
+
encode = 'rushai-dev/THAI-TrOCR'
|
| 9 |
+
decode = "xlm-roberta-base"
|
| 10 |
+
|
| 11 |
+
tokenizer = AutoTokenizer.from_pretrained(decode)
|
| 12 |
+
feature_extractor = ViTFeatureExtractor.from_pretrained(encode)
|
| 13 |
+
processor = TrOCRProcessor(feature_extractor=feature_extractor, tokenizer=tokenizer)
|
| 14 |
+
model = VisionEncoderDecoderModel.from_pretrained(encode)
|
| 15 |
+
```
|
| 16 |
+
|
| 17 |
+
```python
|
| 18 |
+
from PIL import Image
|
| 19 |
+
image = Image.open("xxxxxxx.png").convert("RGB")
|
| 20 |
+
image
|
| 21 |
+
pixel_values = processor(image, return_tensors="pt").pixel_values
|
| 22 |
+
generated_ids = model.generate(pixel_values)
|
| 23 |
+
generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
|
| 24 |
+
generated_text
|
| 25 |
+
```
|