update cuda version in replicate docker image

This commit is contained in:
Brett Kuprel 2022-07-05 15:07:32 -04:00
parent 3aa67b4f5d
commit cc0450337c
4 changed files with 5 additions and 6 deletions

2
README.md vendored
View File

@ -6,7 +6,7 @@
 
[![Join us on Discord](https://img.shields.io/discord/823813159592001537?color=5865F2&logo=discord&logoColor=white)](https://discord.gg/xBPBXfcFHd)
This is a fast, minimal implementation of Boris Dayma's [DALL·E Mega](https://github.com/borisdayma/dalle-mini). It has been stripped down for inference and converted to PyTorch. The only third party dependencies are numpy, requests, pillow and torch.
This is a fast, minimal port of Boris Dayma's [DALL·E Mega](https://github.com/borisdayma/dalle-mini). It has been stripped down for inference and converted to PyTorch. The only third party dependencies are numpy, requests, pillow and torch.
To generate a 4x4 grid of DALL·E Mega images it takes:
- 89 sec with a T4 in Colab

6
cog.yaml vendored
View File

@ -1,13 +1,13 @@
build:
cuda: "11.0"
cuda: "11.5.1"
gpu: true
python_version: "3.8"
python_version: "3.10"
system_packages:
- "libgl1-mesa-glx"
- "libglib2.0-0"
python_packages:
- "min-dalle==0.2.27"
run:
- pip install torch==1.10.0+cu113 -f https://download.pytorch.org/whl/torch_stable.html
- pip install torch==1.12.0+cu116 -f https://download.pytorch.org/whl/torch_stable.html
predict: "replicate_predictor.py:ReplicatePredictor"

View File

@ -1,7 +1,6 @@
import argparse
import os
from PIL import Image
from matplotlib.pyplot import grid
from min_dalle import MinDalle

2
min_dalle.ipynb vendored
View File

@ -183,7 +183,7 @@
"text = \"Dali painting of WALL·E\" #@param {type:\"string\"}\n",
"intermediate_outputs = True #@param {type:\"boolean\"}\n",
"grid_size = 3 #@param {type:\"integer\"}\n",
"supercondition_factor = 8 #@param [\"2\", \"4\", \"8\", \"16\", \"32\", \"64\"] {type:\"raw\"}\n",
"supercondition_factor = 16 #@param [\"2\", \"4\", \"8\", \"16\", \"32\", \"64\"] {type:\"raw\"}\n",
"log2_mid_count = 3 if intermediate_outputs else 0\n",
"\n",
"image_stream = model.generate_image_stream(\n",