Image-Text-to-Text
Transformers
Safetensors
Korean
unlimited-ocr
feature-extraction
ocr
document-parsing
vision-language
Mixture of Experts
korean
custom_code
Instructions to use schift-io/schift-ocr-1-beta with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use schift-io/schift-ocr-1-beta with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="schift-io/schift-ocr-1-beta", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("schift-io/schift-ocr-1-beta", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use schift-io/schift-ocr-1-beta with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "schift-io/schift-ocr-1-beta" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "schift-io/schift-ocr-1-beta", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/schift-io/schift-ocr-1-beta
- SGLang
How to use schift-io/schift-ocr-1-beta 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 "schift-io/schift-ocr-1-beta" \ --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": "schift-io/schift-ocr-1-beta", "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 "schift-io/schift-ocr-1-beta" \ --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": "schift-io/schift-ocr-1-beta", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use schift-io/schift-ocr-1-beta with Docker Model Runner:
docker model run hf.co/schift-io/schift-ocr-1-beta
Card: drop the slide lead claim (markup-sensitive scoring)
Browse files
README.md
CHANGED
|
@@ -42,7 +42,7 @@ We report the character error rate (CER). It is the edit distance divided by the
|
|
| 42 |
|
| 43 |
| Model | Printed forms | Slides | General |
|
| 44 |
|---|---|---|---|
|
| 45 |
-
| **schift-ocr-1-beta** | **0.067** |
|
| 46 |
| baidu/Unlimited-OCR (base) | 0.135 | 0.228 | 0.219 |
|
| 47 |
| MinerU2.5-Pro-2605-1.2B | 0.091 | 0.222 | 0.164 |
|
| 48 |
| PaddleOCR-VL-1.6 | 0.155 | 0.258 | 0.161 |
|
|
@@ -51,6 +51,8 @@ We report the character error rate (CER). It is the edit distance divided by the
|
|
| 51 |
|
| 52 |
Compared with the base model, errors drop by 50% on printed forms, 5% on slides and 13% on general documents.
|
| 53 |
|
|
|
|
|
|
|
| 54 |
## Usage
|
| 55 |
|
| 56 |
### vLLM
|
|
|
|
| 42 |
|
| 43 |
| Model | Printed forms | Slides | General |
|
| 44 |
|---|---|---|---|
|
| 45 |
+
| **schift-ocr-1-beta** | **0.067** | 0.217 | 0.191 |
|
| 46 |
| baidu/Unlimited-OCR (base) | 0.135 | 0.228 | 0.219 |
|
| 47 |
| MinerU2.5-Pro-2605-1.2B | 0.091 | 0.222 | 0.164 |
|
| 48 |
| PaddleOCR-VL-1.6 | 0.155 | 0.258 | 0.161 |
|
|
|
|
| 51 |
|
| 52 |
Compared with the base model, errors drop by 50% on printed forms, 5% on slides and 13% on general documents.
|
| 53 |
|
| 54 |
+
The Slides and General scores depend on how Markdown markup in each model's output is normalized, and several slide references are incomplete, so we do not claim a lead on those two sets.
|
| 55 |
+
|
| 56 |
## Usage
|
| 57 |
|
| 58 |
### vLLM
|