Spaces:
Running
on
Zero
Running
on
Zero
File size: 3,555 Bytes
1b80e0f |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 |
import argparse
from PIL import Image
from moondream.moondream import Moondream
from huggingface_hub import snapshot_download
import torch
from huggingface_hub import hf_hub_download
import os
from threading import Thread
from transformers import CodeGenTokenizerFast as Tokenizer
import numpy as np
def detect_device():
"""
Detects the appropriate device to run on, and return the device and dtype.
"""
if torch.cuda.is_available():
return torch.device("cuda"), torch.float16
elif torch.backends.mps.is_available():
return torch.device("mps"), torch.float16
else:
return torch.device("cpu"), torch.float32
# Tensor to PIL
def tensor2pil(image):
return Image.fromarray(np.clip(255. * image.cpu().numpy().squeeze(), 0, 255).astype(np.uint8))
# Convert PIL to Tensor
def pil2tensor(image):
return torch.from_numpy(np.array(image).astype(np.float32) / 255.0).unsqueeze(0)
# get_torch_device()
moondream = None
tokenizer = None
models_base_path = r"K:\VARIE\ComfyUI_windows_portable\ComfyUI\models\GPTcheckpoints\moondream"
def main2(image_path, prompt):
# get_torch_device()
moondream = None
tokenizer = None
image = Image.open(image_path)
#image_embeds = vision_encoder(image)
dtype = torch.float32
selected_device='cuda'
if selected_device=='cuda':
device = torch.device("cuda")
elif selected_device!=device.type:
device=torch.device("cpu")
if moondream ==None:
config_json=os.path.join(models_base_path,'config.json')
if os.path.exists(config_json)==False:
hf_hub_download("vikhyatk/moondream1",
local_dir=models_base_path,
filename="config.json",
endpoint='https://hf-mirror.com')
model_safetensors=os.path.join(models_base_path,'model.safetensors')
if os.path.exists(model_safetensors)==False:
hf_hub_download("vikhyatk/moondream1",
local_dir=models_base_path,
filename="model.safetensors",
endpoint='https://hf-mirror.com')
tokenizer_json=os.path.join(models_base_path,'tokenizer.json')
if os.path.exists(tokenizer_json)==False:
hf_hub_download("vikhyatk/moondream1",
local_dir=models_base_path,
filename="tokenizer.json",
endpoint='https://hf-mirror.com')
tokenizer = Tokenizer.from_pretrained(models_base_path)
moondream = Moondream.from_pretrained(models_base_path).to(device=device, dtype=dtype)
moondream.eval()
im=image
im=pil2tensor(im)
im=tensor2pil(im)
image_embeds = moondream.encode_image(im)
# streamer = TextIteratorStreamer(tokenizer, skip_special_tokens=True)
print(tokenizer)
print(moondream)
res=moondream.answer_question(image_embeds, prompt,tokenizer)
print(res)
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--image", type=str, required=True)
parser.add_argument("--prompt", type=str, required=False)
args = parser.parse_args()
main2(args.image, args.prompt) |