a fork of shap-e for gc
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 

108 lines
2.7 KiB

{
"cells": [
{
"cell_type": "code",
"execution_count": null,
"id": "964ccced",
"metadata": {},
"outputs": [],
"source": [
"import torch\n",
"\n",
"from shap_e.diffusion.sample import sample_latents\n",
"from shap_e.diffusion.gaussian_diffusion import diffusion_from_config\n",
"from shap_e.models.download import load_model, load_config\n",
"from shap_e.util.notebooks import create_pan_cameras, decode_latent_images, gif_widget\n",
"from shap_e.util.image_util import load_image"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "8eed3a76",
"metadata": {},
"outputs": [],
"source": [
"device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "2d922637",
"metadata": {},
"outputs": [],
"source": [
"xm = load_model('transmitter', device=device)\n",
"model = load_model('image300M', device=device)\n",
"diffusion = diffusion_from_config(load_config('diffusion'))"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "53d329d0",
"metadata": {},
"outputs": [],
"source": [
"batch_size = 4\n",
"guidance_scale = 3.0\n",
"\n",
"image = load_image(\"example_data/corgi.png\")\n",
"\n",
"latents = sample_latents(\n",
" batch_size=batch_size,\n",
" model=model,\n",
" diffusion=diffusion,\n",
" guidance_scale=guidance_scale,\n",
" model_kwargs=dict(images=[image] * batch_size),\n",
" progress=True,\n",
" clip_denoised=True,\n",
" use_fp16=True,\n",
" use_karras=True,\n",
" karras_steps=64,\n",
" sigma_min=1e-3,\n",
" sigma_max=160,\n",
" s_churn=0,\n",
")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "633da2ec",
"metadata": {},
"outputs": [],
"source": [
"render_mode = 'nerf' # you can change this to 'stf' for mesh rendering\n",
"size = 64 # this is the size of the renders; higher values take longer to render.\n",
"\n",
"cameras = create_pan_cameras(size, device)\n",
"for i, latent in enumerate(latents):\n",
" images = decode_latent_images(xm, latent, cameras, rendering_mode=render_mode)\n",
" display(gif_widget(images))"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.9.9"
}
},
"nbformat": 4,
"nbformat_minor": 5
}