[feat]: Add format auto fixer to main branch (#124)

.github/workflows/codespell.yml

@@ -0,0 +1,45 @@
+name: codespell
+
+on:
+  # Trigger the workflow on push or pull request,
+  # but only for the main branch
+  push:
+    branches:
+      - main
+    paths:
+      - "**/*.py"
+      - "**/*.md"
+      - "**/*.rst"
+      - pyproject.toml
+      - requirements-lint.txt
+      - .github/workflows/codespell.yml
+  pull_request:
+    branches:
+      - main
+    paths:
+      - "**/*.py"
+      - "**/*.md"
+      - "**/*.rst"
+      - pyproject.toml
+      - requirements-lint.txt
+      - .github/workflows/codespell.yml
+
+jobs:
+  codespell:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check out repository
+        uses: actions/checkout@v3
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.12'  # or any version you need
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install -r requirements-lint.txt
+      - name: Spelling check with codespell
+        run: |
+          # Refer to the above environment variable here
+          codespell --toml pyproject.toml $CODESPELL_EXCLUDES

.github/workflows/ruff.yml

@@ -0,0 +1,50 @@
+name: ruff
+
+on:
+  # Trigger the workflow on push or pull request,
+  # but only for the main branch
+  push:
+    branches:
+      - main
+    paths:
+      - "**/*.py"
+      - pyproject.toml
+      - requirements-lint.txt
+      - .github/workflows/matchers/ruff.json
+      - .github/workflows/ruff.yml
+  pull_request:
+    branches:
+      - main
+    # This workflow is only relevant when one of the following files changes.
+    # However, we have github configured to expect and require this workflow
+    # to run and pass before github with auto-merge a pull request. Until github
+    # allows more flexible auto-merge policy, we can just run this on every PR.
+    # It doesn't take that long to run, anyway.
+    #paths:
+    #  - "**/*.py"
+    #  - pyproject.toml
+    #  - requirements-lint.txt
+    #  - .github/workflows/matchers/ruff.json
+    #  - .github/workflows/ruff.yml
+
+jobs:
+  ruff:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check out repository
+        uses: actions/checkout@v3
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.12'  # or any version you need
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install -r requirements-lint.txt
+      - name: Analysing the code with ruff
+        run: |
+          ruff check .
+      - name: Run isort
+        run: |
+          isort . --check-only

.github/workflows/yapf.yml

@@ -0,0 +1,38 @@
+name: yapf
+
+on:
+  # Trigger the workflow on push or pull request,
+  # but only for the main branch
+  push:
+    branches:
+      - main
+    paths:
+      - "**/*.py"
+      - .github/workflows/yapf.yml
+  pull_request:
+    branches:
+      - main
+    paths:
+      - "**/*.py"
+      - .github/workflows/yapf.yml
+
+jobs:
+  yapf:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check out repository
+        uses: actions/checkout@v3
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.12'  # or any version you need
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install yapf==0.32.0
+          pip install toml==0.10.2
+      - name: Running yapf
+        run: |
+          yapf --diff --recursive .

README.md

@@ -116,7 +116,7 @@ Ensure your data is prepared and preprocessed in the format specified in [data_p
 ```bash
 python scripts/huggingface/download_hf.py --repo_id=FastVideo/Mochi-Black-Myth --local_dir=data/Mochi-Black-Myth --repo_type=dataset
 ```
-Download the original model weights as specificed in [Distill Section](#-distill):
+Download the original model weights as specified in [Distill Section](#-distill):
 
 Then you can run the finetune with:
 ```

demo/gradio_web_demo.py

@@ -1,13 +1,15 @@
+import argparse
+import os
+import tempfile
+
 import gradio as gr
 import torch
-from fastvideo.models.mochi_hf.pipeline_mochi import MochiPipeline
-from fastvideo.models.mochi_hf.modeling_mochi import MochiTransformer3DModel
 from diffusers import FlowMatchEulerDiscreteScheduler
 from diffusers.utils import export_to_video
+
 from fastvideo.distill.solver import PCMFMScheduler
-import tempfile
-import os
-import argparse
+from fastvideo.models.mochi_hf.modeling_mochi import MochiTransformer3DModel
+from fastvideo.models.mochi_hf.pipeline_mochi import MochiPipeline
 
 
 def init_args():
@@ -21,7 +23,9 @@ def init_args():
     parser.add_argument("--model_path", type=str, default="data/mochi")
     parser.add_argument("--seed", type=int, default=12345)
     parser.add_argument("--transformer_path", type=str, default=None)
-    parser.add_argument("--scheduler_type", type=str, default="pcm_linear_quadratic")
+    parser.add_argument("--scheduler_type",
+                        type=str,
+                        default="pcm_linear_quadratic")
     parser.add_argument("--lora_checkpoint_dir", type=str, default=None)
     parser.add_argument("--shift", type=float, default=8.0)
     parser.add_argument("--num_euler_timesteps", type=int, default=50)
@@ -32,7 +36,6 @@ def init_args():
 
 
 def load_model(args):
-    device = "cuda" if torch.cuda.is_available() else "cpu"
     if args.scheduler_type == "euler":
         scheduler = FlowMatchEulerDiscreteScheduler()
     else:
@@ -47,15 +50,15 @@ def load_model(args):
         )
 
     if args.transformer_path:
-        transformer = MochiTransformer3DModel.from_pretrained(args.transformer_path)
+        transformer = MochiTransformer3DModel.from_pretrained(
+            args.transformer_path)
     else:
         transformer = MochiTransformer3DModel.from_pretrained(
-            args.model_path, subfolder="transformer/"
-        )
+            args.model_path, subfolder="transformer/")
 
-    pipe = MochiPipeline.from_pretrained(
-        args.model_path, transformer=transformer, scheduler=scheduler
-    )
+    pipe = MochiPipeline.from_pretrained(args.model_path,
+                                         transformer=transformer,
+                                         scheduler=scheduler)
     pipe.enable_vae_tiling()
     # pipe.to(device)
     # if args.cpu_offload:
@@ -76,7 +79,7 @@ def generate_video(
     randomize_seed=False,
 ):
     if randomize_seed:
-        seed = torch.randint(0, 1000000, (1,)).item()
+        seed = torch.randint(0, 1000000, (1, )).item()
 
     generator = torch.Generator(device="cuda").manual_seed(seed)
 
@@ -134,9 +137,11 @@ def generate_video(
                     step=32,
                     value=args.height,
                 )
-                width = gr.Slider(
-                    label="Width", minimum=256, maximum=1024, step=32, value=args.width
-                )
+                width = gr.Slider(label="Width",
+                                  minimum=256,
+                                  maximum=1024,
+                                  step=32,
+                                  value=args.width)
 
             with gr.Row():
                 num_frames = gr.Slider(
@@ -159,19 +164,20 @@ def generate_video(
                 )
 
             with gr.Row():
-                use_negative_prompt = gr.Checkbox(
-                    label="Use negative prompt", value=False
-                )
+                use_negative_prompt = gr.Checkbox(label="Use negative prompt",
+                                                  value=False)
             negative_prompt = gr.Text(
                 label="Negative prompt",
                 max_lines=1,
                 placeholder="Enter a negative prompt",
                 visible=False,
             )
 
-            seed = gr.Slider(
-                label="Seed", minimum=0, maximum=1000000, step=1, value=args.seed
-            )
+            seed = gr.Slider(label="Seed",
+                             minimum=0,
+                             maximum=1000000,
+                             step=1,
+                             value=args.seed)
             randomize_seed = gr.Checkbox(label="Randomize seed", value=True)
             seed_output = gr.Number(label="Used Seed")
 
@@ -201,4 +207,4 @@ def generate_video(
     )
 
 if __name__ == "__main__":
-    demo.queue(max_size=20).launch(server_name="0.0.0.0", server_port=7860)
+    demo.queue(max_size=20).launch(server_name="0.0.0.0", server_port=7860)

env_setup.sh

@@ -6,5 +6,7 @@ pip install torch==2.5.0 torchvision --index-url https://download.pytorch.org/wh
 # install FA2 and diffusers
 pip install packaging ninja && pip install flash-attn==2.7.0.post2 --no-build-isolation 
 
+pip install -r requirements-lint.txt
+
 # install fastvideo
 pip install -e .