hathibelagal-dev

{import torch}
{import diffusers@[StableDiffusionPipeline]}

# segmind/tiny-sd is a very small distilled model (~600MB)
model_id = "segmind/tiny-sd"

{print "Loading Tiny-SD pipeline..."}
pipe = {StableDiffusionPipeline.from_pretrained 
            model_id 

hathibelagal-dev

import sys, os

comfy_path = "/content/ComfyUI"
sys.path.insert(0, comfy_path)

import importlib.util

utils_path = os.path.join(comfy_path, "utils")
spec = importlib.util.spec_from_file_location("utils", os.path.join(utils_path, "__init__.py") if os.path.exists(os.path.join(utils_path, "__init__.py")) else os.devnull)
utils = importlib.util.module_from_spec(spec)

hathibelagal-dev

ffmpeg -i input.mp4 -vf "
crop=iw-24:ih-24:
12+12*(random(1)-0.5):
12+12*(random(2)-0.5),
scale=iw:ih" output.mp4

hathibelagal-dev

import json
import sys

if len(sys.argv) != 2:
    print("ERROR: Provide input file")

try:
    with open(sys.argv[1]) as f:
        data = f.read()
except:

hathibelagal-dev

import random
import math
from typing import List, Optional

def total_conflicts(state: List[int]) -> int:
    """Count total number of attacking pairs (each pair counted once)."""
    n = len(state)
    conflicts = 0
    for i in range(n):
        for j in range(i + 1, n):

hathibelagal-dev

import jax
import jax.numpy as jnp
import random
import sys
import numpy as np
from thrml.block_management import Block
from thrml.block_sampling import BlockGibbsSpec, sample_states, SamplingSchedule
from thrml.pgm import CategoricalNode
from thrml.factor import AbstractFactor, FactorSamplingProgram
from thrml.interaction import InteractionGroup

hathibelagal-dev

hathibelagal-dev

hathibelagal-dev

hathibelagal-dev