Repository navigation
Expand file tree
/
Copy pathpyproject.toml
More file actions
51 lines (44 loc) · 1.51 KB
/
Copy pathpyproject.toml
File metadata and controls
51 lines (44 loc) · 1.51 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
[build-system]
requires = ["setuptools>=68"]
build-backend = "setuptools.build_meta"
[project]
name = "fastframes"
version = "0.1.0"
description = "FastFrames: interactive-latency autoregressive video diffusion"
readme = "README.md"
license = { text = "Apple Sample Code License" }
requires-python = ">=3.10"
dependencies = [
# The paper's numbers were produced with exactly these PyTorch builds (CUDA 12.8).
"torch==2.9.1+cu128",
"torchvision==0.24.1+cu128",
"transformers>=5.10", # umT5 tokenizer; DINOv3 for decoder GAN training
"safetensors>=0.4",
"einops",
"ftfy",
"regex",
"pyyaml",
"numpy",
]
[project.optional-dependencies]
wandb = ["wandb"]
lingbot = ["diffusers==0.31.0", "easydict", "imageio", "scipy", "tqdm"] # modules of the upstream clone
demo = ["websockets>=15,<17", "Pillow>=12.3"] # install together with the lingbot extra
# modules of the baseline clones; setuptools for the Inductor build of their CPU-side code (latency)
baselines = ["accelerate", "diffusers==0.31.0", "easydict", "omegaconf", "setuptools", "tqdm"]
dev = ["pytest", "ruff"]
[tool.setuptools.packages.find]
include = ["fastframes*"]
# uv sync installs the pinned CUDA 12.8 PyTorch wheels automatically.
[[tool.uv.index]]
name = "pytorch-cu128"
url = "https://download.pytorch.org/whl/cu128"
explicit = true
[tool.uv.sources]
torch = { index = "pytorch-cu128" }
torchvision = { index = "pytorch-cu128" }
[tool.ruff]
line-length = 120
extend-exclude = ["evaluation/vbench/VBench"]
[tool.ruff.lint]
select = ["E", "F", "W", "I"]