---
pipeline_tag: image-text-to-text
language:
- multilingual
tags:
- deepseek
- vision-language
- ocr
- custom_code
license: mit
---
π Github |
π₯ Model Download |
π Paper Link |
π Arxiv Paper Link |
DeepSeek-OCR: Contexts Optical Compression
Explore the boundaries of visual-text compression.
## Usage
Inference using Huggingface transformers on NVIDIA GPUs. Requirements tested on python 3.12.9 + CUDA11.8οΌ
```
torch==2.6.0
transformers==4.46.3
tokenizers==0.20.3
einops
addict
easydict
pip install flash-attn==2.7.3 --no-build-isolation
```
```python
from transformers import AutoModel, AutoTokenizer
import torch
import os
os.environ["CUDA_VISIBLE_DEVICES"] = '0'
model_name = 'deepseek-ai/DeepSeek-OCR'
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
model = AutoModel.from_pretrained(model_name, _attn_implementation='flash_attention_2', trust_remote_code=True, use_safetensors=True)
model = model.eval().cuda().to(torch.bfloat16)
# prompt = "\nFree OCR. "
prompt = "\n<|grounding|>Convert the document to markdown. "
image_file = 'your_image.jpg'
output_path = 'your/output/dir'
# infer(self, tokenizer, prompt='', image_file='', output_path = ' ', base_size = 1024, image_size = 640, crop_mode = True, test_compress = False, save_results = False):
# Tiny: base_size = 512, image_size = 512, crop_mode = False
# Small: base_size = 640, image_size = 640, crop_mode = False
# Base: base_size = 1024, image_size = 1024, crop_mode = False
# Large: base_size = 1280, image_size = 1280, crop_mode = False
# Gundam: base_size = 1024, image_size = 640, crop_mode = True
res = model.infer(tokenizer, prompt=prompt, image_file=image_file, output_path = output_path, base_size = 1024, image_size = 640, crop_mode=True, save_results = True, test_compress = True)
```
## vLLM
Refer to [πGitHub](https://github.com/deepseek-ai/DeepSeek-OCR/) for guidance on model inference acceleration and PDF processing, etc.
[2025/10/23] πππ DeepSeek-OCR is now officially supported in upstream [vLLM](https://docs.vllm.ai/projects/recipes/en/latest/DeepSeek/DeepSeek-OCR.html#installing-vllm).
```shell
uv venv
source .venv/bin/activate
# Until v0.11.1 release, you need to install vLLM from nightly build
uv pip install -U vllm --pre --extra-index-url https://wheels.vllm.ai/nightly
```
```python
from vllm import LLM, SamplingParams
from vllm.model_executor.models.deepseek_ocr import NGramPerReqLogitsProcessor
from PIL import Image
# Create model instance
llm = LLM(
model="deepseek-ai/DeepSeek-OCR",
enable_prefix_caching=False,
mm_processor_cache_gb=0,
logits_processors=[NGramPerReqLogitsProcessor]
)
# Prepare batched input with your image file
image_1 = Image.open("path/to/your/image_1.png").convert("RGB")
image_2 = Image.open("path/to/your/image_2.png").convert("RGB")
prompt = "\nFree OCR."
model_input = [
{
"prompt": prompt,
"multi_modal_data": {"image": image_1}
},
{
"prompt": prompt,
"multi_modal_data": {"image": image_2}
}
]
sampling_param = SamplingParams(
temperature=0.0,
max_tokens=8192,
# ngram logit processor args
extra_args=dict(
ngram_size=30,
window_size=90,
whitelist_token_ids={128821, 128822}, # whitelist: , |
),
skip_special_tokens=False,
)
# Generate output
model_outputs = llm.generate(model_input, sampling_param)
# Print output
for output in model_outputs:
print(output.outputs[0].text)
```
## Visualizations
## Acknowledgement
We would like to thank [Vary](https://github.com/Ucas-HaoranWei/Vary/), [GOT-OCR2.0](https://github.com/Ucas-HaoranWei/GOT-OCR2.0/), [MinerU](https://github.com/opendatalab/MinerU), [PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR), [OneChart](https://github.com/LingyvKong/OneChart), [Slow Perception](https://github.com/Ucas-HaoranWei/Slow-Perception) for their valuable models and ideas.
We also appreciate the benchmarks: [Fox](https://github.com/ucaslcl/Fox), [OminiDocBench](https://github.com/opendatalab/OmniDocBench).
---
## π Summary of Changes
### 1. **Formatting & Clean-Up**
* Removed extra spaces, blank lines, and inconsistent indentation.
* Fixed small style issues (like missing spaces in comments).
* Added missing newline at the end of the file.
---
### 2. **Device and Dtype Handling**
* Added automatic device detection:
```python
model_device = next(self.parameters()).device
```
* Added adaptive dtype logic:
```python
image_dtype = torch.bfloat16 if model_device.type == "cuda" else torch.float32
```
* Replaced all hardcoded `.cuda()` and `.to(torch.bfloat16)` with:
```python
.to(model_device)
.to(image_dtype)
```
β
**Now works automatically on both GPU and CPU**, without device mismatch errors.
---
### 3. **Autocast and Inference Improvements**
* Wrapped generation in a conditional autocast block:
```python
use_autocast = model_device.type == "cuda"
if use_autocast:
with torch.autocast("cuda", dtype=torch.bfloat16):
with torch.no_grad():
...
else:
with torch.no_grad():
...
```
* Reduces memory usage and speeds up inference on GPU.
* Added `torch.no_grad()` for safer evaluation (no gradient tracking).
---
### 4. **Image Preprocessing**
* All image tensors now use:
```python
.to(image_dtype)
```
instead of hardcoded `torch.bfloat16`.
* Improves flexibility and prevents dtype errors when running on CPU.
---
### 5. **Generation Parameter Updates**
* Adjusted text generation settings:
```python
do_sample = False
num_beams = 1
max_new_tokens = 4096 # was 8192
min_new_tokens = 1
repetition_penalty = 1.2
pad_token_id = tokenizer.pad_token_id or tokenizer.eos_token_id
```
π§ Results: Faster, more controlled generation; avoids repetitive or runaway outputs.
---
### 6. **Safer Decoding**
* Cleaned up decoding logic:
```python
input_length = input_ids.unsqueeze(0).to(model_device).shape[1]
outputs = tokenizer.decode(output_ids[0, input_length:])
```
β
Avoids CUDA-specific assumptions, consistent across devices.
---
### 7. **Miscellaneous**
* Added helpful comments for clarity.
* Improved readability around image transformation and saving results.
* Added extra blank lines for cleaner structure.
---
## βοΈ Overall Impact
| Category | Before | After |
| --------------- | --------------------- | ---------------------------------------- |
| Device handling | Hardcoded `.cuda()` | Auto-detected and flexible |
| Dtype | Always bfloat16 | Adaptive: bfloat16 (GPU) / float32 (CPU) |
| Inference | Could crash on CPU | Runs safely everywhere |
| Generation | Unbounded, repetitive | Tuned and stable |
| Readability | Mixed formatting | Clean and consistent |
---
## Citation
```bibtex
@article{wei2025deepseek,
title={DeepSeek-OCR: Contexts Optical Compression},
author={Wei, Haoran and Sun, Yaofeng and Li, Yukun},
journal={arXiv preprint arXiv:2510.18234},
year={2025}
} 
