Merge pull request #4 from nateraw/revert-3-package

Revert "add pypi package"

README.md

@@ -26,43 +26,41 @@ The app is built with [Gradio](https://gradio.app/), which allows you to interac
     - Set the `num_walk_steps` - for testing you can use a small number like 3 or 5, but to get great results you'll want to use something larger (60-200 steps). 
     - You can (and should) use the `name` input to separate out where the images/videos are saved. (Note that currently ffmpeg will not overwrite if you already made a video with the same name. You'll have to use ffmpeg to create the video yourself if the app fails to do so.)
 
-### Python Package
+### The Script
 
 #### Setup
 
-Install the package
-
-```
-pip install stable_diffusion_videos
-```
-
-Authenticate with Hugging Face
-
 ```
-huggingface-cli login
+git clone https://github.com/nateraw/stable-diffusion-videos
+cd stable-diffusion-videos
+pip install -r requirements.txt
 ```
 
 #### Usage
 
-```python
-from stable_diffusion_videos import walk
-
-walk(
-    prompts=['a cat', 'a dog'],
-    seeds=[42, 1337],
-    output_dir='dreams',     # Where images/videos will be saved
-    name='animals_test',     # Subdirectory of output_dir where images/videos will be saved
-    guidance_scale=8.5,      # Higher adheres to prompt more, lower lets model take the wheel
-    num_steps=5,             # Change to 60-200 for better results...3-5 for testing
-    num_inference_steps=50, 
-    scheduler='klms',        # One of: "klms", "default", "ddim"
-    disable_tqdm=False,      # Set to True to disable tqdm progress bar
-    make_video=True,         # If false, just save images
-    use_lerp_for_text=True,  # Use lerp for text embeddings instead of slerp
-    do_loop=False,           # Change to True if you want last prompt to loop back to first prompt
-)
+If you would prefer to use the `stable_diffusion_walk.py` script directly, you can do so by running:
+
+Run with `num_steps` set to 3 or 5 for testing, then up it to something like 60-200 for better results.
+
+```bash
+python stable_diffusion_walk.py \
+    --prompts "['a cat', 'a dog', 'a horse']" \
+    --seeds 903,123,42 \
+    --output_dir dreams \
+    --name animals_test \
+    --guidance_scale 8.5 \
+    --num_steps 5 \
+    --height 512 \
+    --width 512 \
+    --num_inference_steps 50 \
+    --scheduler klms \
+    --disable_tqdm \
+    --make_video \
+    --use_lerp_for_text \
+    --do_loop
 ```
 
+
 ## Credits
 
 This work built off of [a script](https://gist.github.com/karpathy/00103b0037c5aaea32fe1da1af553355

stable_diffusion_videos/app.py → app.py

@@ -1,7 +1,7 @@
 import gradio as gr
 import torch
 
-from .stable_diffusion_walk import SCHEDULERS, pipeline, walk
+from stable_diffusion_walk import SCHEDULERS, pipeline, walk
 
 
 def fn_images(

requirements.txt

@@ -1,5 +1,5 @@
 transformers
-diffusers @ git+https://github.com/huggingface/diffusers@f085d2f5c6569a1c0d90327c51328622036ef76e
+git+https://github.com/huggingface/diffusers@f085d2f5c6569a1c0d90327c51328622036ef76e
 scipy
 fire
 gradio