OpenImagingLab/FlashVSR_Plus

选用指南

AI 根据 README 整理 · 2026/10/22 14:11:54

使用方式

提供命令行工具和 Python API，可选择 Turbo、Base、V1 三种模式，支持自定义输出尺寸和帧处理参数。

查看 README 依据

- **Reward-Yield Distribution Matching Distillation (RY-DMD):** Enables preference alignment with black-box rewards without differentiating through the reward model, improving perceptual quality and generalizing to text-to-image generation. We release the inference code, pretrained models, and training dataset to facilitate further research and applications. ## Highlights - **Three presets:** run Turbo, Base, or V1 through the same CLI and Python API. V1 uses the same inference configuration as Base with its own checkpoints.

`prompt_tensor/posi_prompt.pth` is included in this code repository. Keep it at the path shown above when using the default presets. `Wan2.1_VAE.pth` is only required when using the VAE decoder through the unified Python API or `inference.cli`. ## Command-Line Inference Run from the repository root after installing dependencies and placing the checkpoints under `ckpts/`. ### Turbo ```bash python -m inference.run_stream \ --input examples/case_0.mp4 \ --mode turbo \ --output outputs/case_0_turbo.mp4 \ --height 2048 \ --width 3840 ``` ### Base ```bash

Run this example from the repository root with the model files in place. ```python from pathlib import Path from inference import InferenceRequest, build_inference_pipeline from inference.presets import model_paths_for_mode, stream_config_for_mode config = stream_config_for_mode("turbo", device="cuda:0") paths = model_paths_for_mode(Path.cwd(), "turbo") pipeline = build_inference_pipeline(config, paths) result = pipeline.infer( InferenceRequest( input_path="input.mp4", output_path="outputs/restored.mp4", height=384, width=768, ) ) ```

部署要求

需要 Linux、NVIDIA GPU 和 CUDA，Python 3.10，输入视频的高和宽必须能被 128 整除。推理仅支持 CUDA，不支持 CPU。

查看 README 依据

- **Large-motion support:** lower `--temporal-compression` from `4` (default) to `2` or `1` to better preserve motion details in videos with large motion. - **Long-video option:** a numeric `--decode-every` reads and writes video in segments; the default `auto` mode loads the full input. - **Attention backends:** use Block Sparse Attention, the included Triton operator, or PyTorch SDPA. ## Requirements - Linux - NVIDIA GPU with CUDA - Conda/Miniconda or Python `venv` - Python 3.10 - Input height and width must both be divisible by 128

The released inference pipeline is CUDA-only and does not support CPU inference. ## Installation ### 1. Create the environment #### Option A: Conda (recommended) ```bash conda create -n flashvsr_plus python=3.10 -y conda activate flashvsr_plus pip install -r requirements.txt pip install -e . --no-deps ``` The editable installation exposes the `flashvsr_plus` Python package and keeps the `python -m inference.run_stream` command available. `--no-deps` is intentional. It prevents the package installation from modifying an existing PyTorch/CUDA environment. #### Option B: Python `venv`

已知限制

推理仅支持 CUDA，不支持 CPU。输入尺寸必须被 128 整除。FlashAttention 2/3 不支持，必须使用特定后端。

查看 README 依据

The released inference pipeline is CUDA-only and does not support CPU inference. ## Installation ### 1. Create the environment #### Option A: Conda (recommended) ```bash conda create -n flashvsr_plus python=3.10 -y conda activate flashvsr_plus pip install -r requirements.txt pip install -e . --no-deps ``` The editable installation exposes the `flashvsr_plus` Python package and keeps the `python -m inference.run_stream` command available. `--no-deps` is intentional. It prevents the package installation from modifying an existing PyTorch/CUDA environment. #### Option B: Python `venv`

- **Large-motion support:** lower `--temporal-compression` from `4` (default) to `2` or `1` to better preserve motion details in videos with large motion. - **Long-video option:** a numeric `--decode-every` reads and writes video in segments; the default `auto` mode loads the full input. - **Attention backends:** use Block Sparse Attention, the included Triton operator, or PyTorch SDPA. ## Requirements - Linux - NVIDIA GPU with CUDA - Conda/Miniconda or Python `venv` - Python 3.10 - Input height and width must both be divisible by 128

| `--attn-kernel NAME` | `auto` | Stream attention backend: `auto`, `sdpa`, `block_sparse`, or `triton_block_sparse`. FlashAttention 2/3 cannot represent the required causal/sparse block masks. | | `--tiny-decoder-mode {auto,serial,parallel}` | `auto` | TinyDecoder execution strategy. Serial mode uses less workspace memory. | | `--serial-tiny-decoder` | off | Shorthand for serial TinyDecoder execution. | | `--tiny-decoder-channels-last/--no-tiny-decoder-channels-last` | on | Use channels-last BF16 convolution storage. |