Instructions to use ConwayResearch/dsv4-flash-0731-vision with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ConwayResearch/dsv4-flash-0731-vision with Transformers:

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

pipe = pipeline("image-text-to-text", model="ConwayResearch/dsv4-flash-0731-vision", trust_remote_code=True)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("ConwayResearch/dsv4-flash-0731-vision", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ConwayResearch/dsv4-flash-0731-vision with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ConwayResearch/dsv4-flash-0731-vision"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ConwayResearch/dsv4-flash-0731-vision",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/ConwayResearch/dsv4-flash-0731-vision

SGLang

How to use ConwayResearch/dsv4-flash-0731-vision 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 "ConwayResearch/dsv4-flash-0731-vision" \
    --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": "ConwayResearch/dsv4-flash-0731-vision",
		"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 "ConwayResearch/dsv4-flash-0731-vision" \
        --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": "ConwayResearch/dsv4-flash-0731-vision",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use ConwayResearch/dsv4-flash-0731-vision with Docker Model Runner:
```
docker model run hf.co/ConwayResearch/dsv4-flash-0731-vision
```

DSV4-Flash-0731 Vision

Repository for the DSV4-Flash-0731 vision graft.

The language base is the exact public deepseek-ai/DeepSeek-V4-Flash-0731 checkpoint. It is intentionally not duplicated here. The vision stack uses the frozen public google/siglip2-so400m-patch14-384 tower, a trained two-layer projector, and an attention-only Stage-2 LoRA.

Recovery contents

stage1/projector.safetensors: completed Stage-1 projector.
stage1/metadata.json: exact projector iteration and sample counters.
provenance/: evaluation, source, and pipeline receipts.
recovery_manifest.json: SHA-256 and byte size for every uploaded artifact.
stage2/: final attention LoRA, projector, and release receipts after Stage-2 training completes.

Current state

Stage 1 is complete. Its full 11,703-row held-out baseline scored:

MMBench EN dev: 0.385077
TextVQA exact match: 0.04762
MME: 847.5577

Stage 2 is still in progress. Until its final artifacts and evaluation receipts appear under stage2/, this repository contains the basic Stage-1 vision graft, not the final OCR/chart/GUI-enhanced release.

The reproducible pipeline is checkpointed in the project repository at Git commit 7007b6b (Implement verified DSV4 vision training pipeline).

Reconstruction

Materialize the exact public DSV4-Flash-0731 and SigLIP2 checkpoints.
Use the DSV4 vision-graft pipeline to build the zero-copy multimodal view.
Load the matching Stage-1 or Stage-2 projector from this repository.
For Stage 2, load the attention LoRA together with its same-iteration projector. The projector and adapter are a required pair.
Verify all files against recovery_manifest.json before loading.

This repository is private and intended for recovery and controlled evaluation.

Downloads last month: 66

Safetensors

Model size

285B params

Tensor type

F32

I32

BF16

Model tree for ConwayResearch/dsv4-flash-0731-vision

Base model

deepseek-ai/DeepSeek-V4-Flash-0731

Adapter

(7)

this model