Instructions to use Reza2kn/surya-ocr-2-nvfp4a16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Reza2kn/surya-ocr-2-nvfp4a16 with Transformers:

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

pipe = pipeline("image-text-to-text", model="Reza2kn/surya-ocr-2-nvfp4a16")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("Reza2kn/surya-ocr-2-nvfp4a16")
model = AutoModelForMultimodalLM.from_pretrained("Reza2kn/surya-ocr-2-nvfp4a16", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Reza2kn/surya-ocr-2-nvfp4a16 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Reza2kn/surya-ocr-2-nvfp4a16"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Reza2kn/surya-ocr-2-nvfp4a16",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/Reza2kn/surya-ocr-2-nvfp4a16

SGLang

How to use Reza2kn/surya-ocr-2-nvfp4a16 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 "Reza2kn/surya-ocr-2-nvfp4a16" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Reza2kn/surya-ocr-2-nvfp4a16",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "Reza2kn/surya-ocr-2-nvfp4a16" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Reza2kn/surya-ocr-2-nvfp4a16",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use Reza2kn/surya-ocr-2-nvfp4a16 with Docker Model Runner:
```
docker model run hf.co/Reza2kn/surya-ocr-2-nvfp4a16
```

Reza2kn commited on Jun 20

Commit

cbb230d

verified ·

1 Parent(s): 0f7a886

Update model card with mini benchmark limitations

Browse files

Files changed (1) hide show

README.md +34 -18

README.md CHANGED Viewed

@@ -12,38 +12,54 @@ tags:
 - ocr
 - document-ai
 - surya
 ---
 # Surya OCR 2 NVFP4A16
-This repository contains a converted/quantized artifact derived from [datalab-to/surya-ocr-2](https://huggingface.co/datalab-to/surya-ocr-2).
-## What is included
-- Source model: `datalab-to/surya-ocr-2`
-- Runtime/format: llm-compressor / NVIDIA NVFP4-capable runtimes
-- Quantization: NVFP4A16 4-bit float weight quantization; selected unsupported/sensitive modules remain bf16
-- Vision weights included: yes, included; the current recipe preserves vision tower weights in bf16 rather than dropping them
-- Created for: local OCR/document-understanding experiments and parity testing
-## Validation status
-Mini olmOCR-bench score: **79.2% ± 6.2%**. This did not meet the current >98% parity target against the source mini baseline. Failures concentrated on long tiny text and old scans.
-## Known caveats
-Early artifact. The recipe intentionally avoids quantizing unsupported/sensitive visual and linear-attention modules. Use only in runtimes that understand the serialized NVFP4 format.
-## Files
-Quantized safetensors, config, tokenizer, processor, generation config, chat template, and quantization recipe metadata.
-## Usage
-Load with a compatible NVIDIA/NVFP4 runtime stack. Generic Transformers may not execute this format directly.
-## Provenance
-This artifact was generated non-destructively from the original Hugging Face checkpoint. It is not a new fine-tune.
-If you need production parity, compare against the original model on your own document distribution before deployment.

 - ocr
 - document-ai
 - surya
+- experimental
 ---
 # Surya OCR 2 NVFP4A16
+    This repository contains an **experimental quantized** artifact derived from [datalab-to/surya-ocr-2](https://huggingface.co/datalab-to/surya-ocr-2).
+    This NVFP4 artifact is useful for NVIDIA/NVFP4 runtime experimentation. On the mini benchmark it matches the 8-bit MLX split profile: strong on most clean/layout-heavy sections, weak on long tiny text, and not usable for old degraded scans yet.
+    ## What is included
+    - Source model: `datalab-to/surya-ocr-2`
+    - Runtime/format: llm-compressor / NVIDIA NVFP4-capable runtimes
+    - Quantization: NVFP4A16 4-bit float weight quantization; sensitive/unsupported modules remain bf16
+    - Vision weights included: Yes. Vision weights are included; the current recipe preserves the vision tower in bf16 rather than dropping it.
+    - Processor/tokenizer assets: included
+    ## Mini olmOCR-bench results
+    | Candidate | Overall | Arxiv math | Headers/footers | Long tiny text | Multi-column | Old scans | Old scans math | Tables | Baseline |
+|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|
+| Source mini baseline | 91.0% ± 6.3% | 100.0% | 100.0% | 100.0% | 100.0% | 33.3% | 100.0% | 100.0% | 94.7% |
+| Surya OCR 2 NVFP4A16 | 79.2% ± 6.2% | 100.0% | 100.0% | 33.3% | 100.0% | 0.0% | 100.0% | 100.0% | 100.0% |
+## How to read the benchmark table
+This is an early quant release with transparent limitations. The table uses our local 40-test mini slice of `allenai/olmOCR-bench`, with 3 samples from each named section plus the benchmark baseline checks. It is **not** the full public score and it is **not** a claim of >98% parity.
+The useful signal is the split behavior: this artifact is currently strong on clean academic/math, headers/footers, multi-column layouts, tables, old-scan math, and baseline OCR checks, but it should not be used for old degraded scans and is weak on long tiny text.
+## Recommended use
+Use this checkpoint for local experimentation and constrained OCR workloads whose documents resemble the passing sections above. Avoid using it as a production replacement for the original model on degraded historical scans, very small dense body text, or workloads requiring full benchmark parity.
+    ## Loading
+    Load with a runtime stack that understands NVFP4A16 serialized weights. Generic Transformers runtimes may not execute this checkpoint without NVFP4 support.
+    ## Limitations
+    - This is not a full-parity release yet.
+    - Do **not** use this artifact for degraded old scans; the current mini split score is 0.0% there.
+    - Do **not** use this artifact for long tiny text unless you independently validate your data; the current mini split score is 33.3%.
+    - Math-heavy and table/layout-heavy mini examples looked good in this slice, but full olmOCR-bench is still pending.
+    ## Provenance
+    Generated non-destructively from the original Hugging Face checkpoint. This is not a fine-tune. The goal of publishing this artifact now is transparency: the files are usable for the passing workload slices above, and the known failing slices are documented clearly.