Skip to content

Open-Sora: Democratizing Efficient Video Production for All

Python

[feat] update requirements

zhengzangwb93a254abb5f
diff --git a/README.md b/README.md
index 60a900a..3ca35ef 100644
--- a/README.md
+++ b/README.md
@@ -146,7 +146,7 @@ Other useful documents and links are listed below.
### Install from Source
-For CUDA 12.1, you can install the dependencies with the following commands. Otherwise, please refer to [Installation](docs/installation.md) for more instructions.
+For CUDA 12.1, you can install the dependencies with the following commands. Otherwise, please refer to [Installation](docs/installation.md) for more instructions on different cuda version, and additional dependency for data preprocessing.
```bash
# create a virtual env and activate (conda as an example)
@@ -159,9 +159,24 @@ pip install -r requirements/requirements_cu121.txt
# install this project
git clone https://github.com/hpcaitech/Open-Sora
cd Open-Sora
+
+# the default installation is for inference only
pip install -v .
```
+(Optional, recommended for fast speed, especially for training) To enable `layernorm_kernel` and `flash_attn`, you need to install `apex` and `flash-attn` with the following commands.
+
+```bash
+# install flash attention
+# set enable_flash_attn=False in config to disable flash attention
+pip install packaging ninja
+pip install flash-attn --no-build-isolation
+
+# install apex
+# set enable_layernorm_kernel=False in config to disable apex
+pip install -v --disable-pip-version-check --no-cache-dir --no-build-isolation --config-settings "--build-option=--cpp_ext" --config-settings "--build-option=--cuda_ext" git+https://github.com/NVIDIA/apex.git
+```
+
### Use Docker
Run the following command to build a docker image from Dockerfile provided.
diff --git a/docs/installation.md b/docs/installation.md
index d74b923..9c41d0d 100644
--- a/docs/installation.md
+++ b/docs/installation.md
@@ -18,23 +18,11 @@ pip install torch==2.2.2 torchvision==0.17.2 --index-url https://download.pytorc
pip install xformers --index-url https://download.pytorch.org/whl/cu121
```
+## Different Dependencies
+
+The default installation is for inference only. Other optional dependencies are listed below.
+
```bash
-# install flash attention (optional)
-# set enable_flash_attn=False in config to avoid using flash attention
-pip install packaging ninja
-pip install flash-attn --no-build-isolation
-
-# install apex (optional)
-# set enable_layernorm_kernel=False in config to avoid using apex
-pip install -v --disable-pip-version-check --no-cache-dir --no-build-isolation --config-settings "--build-option=--cpp_ext" --config-settings "--build-option=--cuda_ext" git+https://github.com/NVIDIA/apex.git
+pip install -v .[data] # for data preprocessing
+pip install -v .[eval] # for evaluation
```
-
-gdown
-pre-commit
-pyarrow
-tensorboard
-transformers
-wandb
-pandarallel
-gradio
-spaces
diff --git a/notebooks/inference.ipynb b/notebooks/inference.ipynb
index 7ef8489..170d55b 100644
--- a/notebooks/inference.ipynb
+++ b/notebooks/inference.ipynb
@@ -16,7 +16,7 @@
},
{
"cell_type": "code",
- "execution_count": 10,
+ "execution_count": null,
"metadata": {},
"outputs": [],
"source": [
@@ -38,26 +38,9 @@
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": null,
"metadata": {},
- "outputs": [
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/home/zhengzangwei/.conda/envs/opensora/lib/python3.10/site-packages/transformers/utils/generic.py:441: UserWarning: torch.utils._pytree._register_pytree_node is deprecated. Please use torch.utils._pytree.register_pytree_node instead.\n",
- " _torch_pytree._register_pytree_node(\n",
- "/home/zhengzangwei/.conda/envs/opensora/lib/python3.10/site-packages/colossalai/pipeline/schedule/_utils.py:19: UserWarning: torch.utils._pytree._register_pytree_node is deprecated. Please use torch.utils._pytree.register_pytree_node instead.\n",
- " _register_pytree_node(OrderedDict, _odict_flatten, _odict_unflatten)\n",
- "/home/zhengzangwei/.conda/envs/opensora/lib/python3.10/site-packages/torch/utils/_pytree.py:254: UserWarning: <class 'collections.OrderedDict'> is already registered as pytree node. Overwriting the previous registration.\n",
- " warnings.warn(\n",
- "/home/zhengzangwei/.conda/envs/opensora/lib/python3.10/site-packages/transformers/utils/generic.py:309: UserWarning: torch.utils._pytree._register_pytree_node is deprecated. Please use torch.utils._pytree.register_pytree_node instead.\n",
- " _torch_pytree._register_pytree_node(\n",
- "/home/zhengzangwei/.conda/envs/opensora/lib/python3.10/site-packages/transformers/utils/generic.py:309: UserWarning: torch.utils._pytree._register_pytree_node is deprecated. Please use torch.utils._pytree.register_pytree_node instead.\n",
- " _torch_pytree._register_pytree_node(\n"
- ]
- }
- ],
+ "outputs": [],
"source": [
"import os\n",
"from pprint import pformat\n",
@@ -90,20 +73,9 @@
},
{
"cell_type": "code",
- "execution_count": 20,
+ "execution_count": null,
"metadata": {},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "42"
- ]
- },
- "execution_count": 20,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"torch.set_grad_enabled(False)\n",
"\n",
@@ -124,24 +96,9 @@
},
{
"cell_type": "code",
- "execution_count": 5,
+ "execution_count": null,
"metadata": {},
- "outputs": [
- {
- "data": {
- "application/vnd.jupyter.widget-view+json": {
- "model_id": "dbb26c0ba0824c69b1e1d293fbfe3d9a",
- "version_major": 2,
- "version_minor": 0
- },
- "text/plain": [
- "Loading checkpoint shards: 0%| | 0/2 [00:00<?, ?it/s]"
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
+ "outputs": [],
"source": [
"# == build text-encoder and vae ==\n",
"text_encoder = build_module(cfg.text_encoder, MODELS, device=device)\n",
@@ -174,7 +131,7 @@
},
{
"cell_type": "code",
- "execution_count": 26,
+ "execution_count": null,
"metadata": {},
"outputs": [],
"source": [
@@ -314,7 +271,7 @@
},
{
"cell_type": "code",
- "execution_count": 16,
+ "execution_count": null,
"metadata": {},
"outputs": [],
"source": [
@@ -513,7 +470,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
- "version": "3.10.9"
+ "version": "3.10.14"
}
},
"nbformat": 4,
diff --git a/requirements/requirements-data.txt b/requirements/requirements-data.txt
new file mode 100644
index 0000000..d427bea
--- /dev/null
+++ b/requirements/requirements-data.txt
@@ -0,0 +1,4 @@
+gdown
+
+# [aesthetic]
+clip @ git+https://github.com/openai/CLIP.git
diff --git a/requirements/requirements-eval.txt b/requirements/requirements-eval.txt
new file mode 100644
index 0000000..e69de29
diff --git a/requirements/requirements-fast.txt b/requirements/requirements-fast.txt
deleted file mode 100644
index 0c02389..0000000
--- a/requirements/requirements-fast.txt
+++ /dev/null
@@ -1,3 +0,0 @@
-packaging
-ninja
-flash-attn --no-build-isolation
diff --git a/requirements/requirements.txt b/requirements/requirements.txt
index e124aca..8a7b440 100644
--- a/requirements/requirements.txt
+++ b/requirements/requirements.txt
@@ -7,3 +7,20 @@ ftfy>=6.2.0 # for t5
diffusers==0.27.2 # for vae
accelerate==0.29.2 # for t5
av>=12.0.0
+
+# [gradio]
+gradio
+spaces
+
+# [notebook]
+ipykernel
+ipywidgets
+
+# [training]
+wandb
+tensorboard
+pandarallel
+pyarrow # for parquet
+
+# [dev]
+pre-commit
diff --git a/setup.py b/setup.py
index fea582c..9743539 100644
--- a/setup.py
+++ b/setup.py
@@ -77,6 +77,7 @@ setup(
"Topic :: System :: Distributed Computing",
],
extras_require={
- "fast": ["flash-attn --no-build-isolation"],
+ "data": fetch_requirements("requirements/requirements-data.txt"),
+ "eval": fetch_requirements("requirements/requirements-eval.txt"),
},
)