File size: 3,748 Bytes
bc1ecc4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 |
import random
import io
import zipfile
import numpy as np
from PIL.PngImagePlugin import PngInfo
from PIL import Image
from curl_cffi import requests
from tqdm import tqdm
jwt_token = ""
random_seed = random.randint(0, 2**32 - 1)
# Define the API URL
url = "https://image.novelai.net/ai/generate-image"
# Set the headers
headers = {
"Authorization": f"Bearer {jwt_token}",
"Accept": "application/json, text/plain, */*",
"Content-Type": "application/json",
"Origin": "https://novelai.net",
"Referer": "https://novelai.net/"
}
QUALITY_TAGS = "best quality, amazing quality, very aesthetic, absurdres"
# Define the payload
def generate(prompt="1girl, best quality, amazing quality, very aesthetic, absurdres"):
# neg_prompt = "nsfw, lowres, {bad}, error, fewer, extra, missing, worst quality, jpeg artifacts, bad quality, watermark, unfinished, displeasing, chromatic aberration, signature, extra digits, artistic error, username, scan, [abstract]"
payload = {
"action": "generate",
"input": f'{prompt}, best quality, amazing quality, very aesthetic, absurdres',
"model": "nai-diffusion-3",
"parameters": {
"width": 832,
"height": 1216,
"scale": 5,
"sampler": "k_euler_ancestral",
"steps": 28,
"n_samples": 1,
"ucPreset": 0,
"qualityToggle": True,
"add_original_image": False,
"cfg_rescale": 0,
"controlnet_strength": 1,
"dynamic_thresholding": False,
"legacy": False,
"noise_schedule": "karras",
"seed": 8888,
"sm": False,
"sm_dyn": False,
"uncond_scale": 1,
"negative_prompt":"nsfw, lowres, bad, error, fewer, extra, missing, worst quality, jpeg artifacts, bad quality, watermark, unfinished, displeasing, chromatic aberration, signature, extra digits, artistic error, username, scan, [abstract], lowres, bad, error, fewer, extra, missing, worst quality, jpeg artifacts, bad quality, unfinished, displeasing, chromatic aberration, signature, extra digits, artistic error, username, scan, [abstract], chibi,doll, +_+",
"legacy_v3_extend": False,
}
}
# Send the POST request
response = requests.post(url, impersonate="safari15_5", json=payload, headers=headers, timeout=120)
# Save the response content (assuming it's a zip file)
# with open('images.zip', 'wb') as f:
# f.write(response.content)
zipfile_in_memory = io.BytesIO(response.content)
with zipfile.ZipFile(zipfile_in_memory, 'r') as zip_ref:
# Extract the list of file names
file_names = zip_ref.namelist()
# Check if there are files in the zip
if file_names:
# Open the first file as an image
with zip_ref.open(file_names[0]) as file:
# Display the image
return Image.open(io.BytesIO(file.read())), payload
def process_image_and_save(image, path):
metadata = PngInfo()
image = image.convert('RGBA')
image = Image.fromarray(np.array(image)[:,:,:3])
image.save(path, pnginfo=metadata, quality=95, format="WEBP")
print(path)
# read prompts for testing
with open("prompts.csv") as f:
prompts = f.readlines()
# warmup
generate("abcd")
# generate images
for i, prompt in tqdm(enumerate(prompts), total=len(prompts)):
try:
image, payload = generate(prompt.strip())
image = image.convert('RGBA')
image = Image.fromarray(np.array(image)[:,:,:3])
fn = f"naiv3/{i+1}.webp"
image.save(fn, quality=95, format="WEBP")
except Exception as e:
print(e)
continue |