rom1504

#!/bin/bash
#SBATCH --partition=gpu
#SBATCH --job-name=gputest
#SBATCH --nodes 1
#SBATCH --ntasks-per-node 8
#SBATCH --cpus-per-gpu=6
#SBATCH --gres=gpu:8
#SBATCH --nodelist gpu-st-p4d-24xlarge-42 
#SBATCH --output=%x_%j.out
#SBATCH --open-mode=append

rom1504

python3.8 -m venv .env
source .env/bin/activate
pip install -U pip
pip install "jax[cuda11_cudnn82]" -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html

python
import jax
jax.default_backend()
jax.devices()

rom1504

import json
import pandas as pd
import subprocess
import sys

def get_msg(
    backticks=True   # whether to add backticks for Discord formatting or not
    ):
    "gets a list of cluster usage from squeue and creates a text message from it"
    a = json.loads(subprocess.check_output(['squeue','--json']).decode("utf8"))

rom1504

Change paths then run keep_evaling.sh

Will keep evaluating openclip during training

to send to wandb, can also run while [ 1 ]; do python3 eval_to_wandb.py; sleep 300; done

rom1504

create an env:

python3.8 -m venv .env
source .env/bin/activate
pip install -U pip
pip3 install torch torchvision --extra-index-url https://download.pytorch.org/whl/cu113

fix the paths in simple.sh

rom1504

Install

git clone https://github.com/mlfoundations/open_clip.git
cd open_clip
python3.8 -m venv .env
source .env/bin/activate
pip install -U pip
pip install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu113
pip install -e .

rom1504

class WDSDataset(data.IterableDataset):
    def __init__(self, min_size, transform=None, target_transform=None):
        self.min_size = min_size
        self.transform = transform if transform is not None else nn.Identity()
        self.target_transform = target_transform if target_transform is not None else nn.Identity()
        self.kv = OnDiskKV(file='/home/ubuntu/laion5B-watermark-safety-ordered', key_format='q', value_format='ee')
        self.kv_aesthetic = OnDiskKV(file='/home/ubuntu/laion5B-aesthetic-tags-kv', key_format='q', value_format='e')
        self.pwatermark_threshold = 0.8
        self.punsafe_threshold = 0.5
        self.aesthetic_threshold = 5.

rom1504

https://wandb.ai/rom1504/dalle2_train_decoder/runs/mic5buox/files/decoder_config.json

get dalle2

get the config file

get these 2 .sh

run sbatch start_big.sh

rom1504

import pandas as pd

df = pd.read_parquet("aethetic_multi/0000.parquet")

buckets = [(i, i+1) for i in range(10)]

html= "<h1>Aesthetic subsets in Laion2B-multi</h1>"
for [a,b] in buckets:
    total_part = df[(df["prediction"] >= a) & (df["prediction"] <= b)]
    count_part = len(total_part) / len(df) * 100

rom1504

A red colored car.
A black colored car.
A pink colored car.
A black colored dog.
A red colored dog.
A blue colored dog.
A green colored banana.
A red colored banana.
A black colored banana.
A white colored sandwich.