model.py

import gc
import tempfile

import numpy as np
import torch
import trimesh
from shap_e.diffusion.gaussian_diffusion import diffusion_from_config
from shap_e.diffusion.sample import sample_latents
from shap_e.models.download import load_config, load_model
from shap_e.models.nn.camera import (DifferentiableCameraBatch,
                                     DifferentiableProjectiveCamera)
from shap_e.models.transmitter.base import Transmitter, VectorDecoder
from shap_e.rendering.torch_mesh import TorchMesh
from shap_e.util.collections import AttrDict
from shap_e.util.image_util import load_image


# Copied from https://github.com/openai/shap-e/blob/d99cedaea18e0989e340163dbaeb4b109fa9e8ec/shap_e/util/notebooks.py#L15-L42
def create_pan_cameras(size: int,
                       device: torch.device) -> DifferentiableCameraBatch:
    origins = []
    xs = []
    ys = []
    zs = []
    for theta in np.linspace(0, 2 * np.pi, num=20):
        z = np.array([np.sin(theta), np.cos(theta), -0.5])
        z /= np.sqrt(np.sum(z**2))
        origin = -z * 4
        x = np.array([np.cos(theta), -np.sin(theta), 0.0])
        y = np.cross(z, x)
        origins.append(origin)
        xs.append(x)
        ys.append(y)
        zs.append(z)
    return DifferentiableCameraBatch(
        shape=(1, len(xs)),
        flat_camera=DifferentiableProjectiveCamera(
            origin=torch.from_numpy(np.stack(origins,
                                             axis=0)).float().to(device),
            x=torch.from_numpy(np.stack(xs, axis=0)).float().to(device),
            y=torch.from_numpy(np.stack(ys, axis=0)).float().to(device),
            z=torch.from_numpy(np.stack(zs, axis=0)).float().to(device),
            width=size,
            height=size,
            x_fov=0.7,
            y_fov=0.7,
        ),
    )


# Copied from https://github.com/openai/shap-e/blob/8625e7c15526d8510a2292f92165979268d0e945/shap_e/util/notebooks.py#LL64C1-L76C33
@torch.no_grad()
def decode_latent_mesh(
    xm: Transmitter | VectorDecoder,
    latent: torch.Tensor,
) -> TorchMesh:
    decoded = xm.renderer.render_views(
        AttrDict(cameras=create_pan_cameras(
            2, latent.device)),  # lowest resolution possible
        params=(xm.encoder if isinstance(xm, Transmitter) else
                xm).bottleneck_to_params(latent[None]),
        options=AttrDict(rendering_mode='stf', render_with_direction=False),
    )
    return decoded.raw_meshes[0]


class Model:
    def __init__(self):
        self.device = torch.device(
            'cuda' if torch.cuda.is_available() else 'cpu')
        self.xm = load_model('transmitter', device=self.device)
        self.diffusion = diffusion_from_config(load_config('diffusion'))
        self.model_name = ''
        self.model = None

    def load_model(self, model_name: str) -> None:
        assert model_name in ['text300M', 'image300M']
        if model_name == self.model_name:
            return
        self.model = load_model(model_name, device=self.device)
        self.model_name = model_name
        gc.collect()
        torch.cuda.empty_cache()

    def to_glb(self, latent: torch.Tensor) -> str:
        ply_path = tempfile.NamedTemporaryFile(suffix='.ply',
                                               delete=False,
                                               mode='w+b')
        decode_latent_mesh(self.xm, latent).tri_mesh().write_ply(ply_path)

        mesh = trimesh.load(ply_path.name)
        rot = trimesh.transformations.rotation_matrix(-np.pi / 2, [1, 0, 0])
        mesh = mesh.apply_transform(rot)
        rot = trimesh.transformations.rotation_matrix(np.pi, [0, 1, 0])
        mesh = mesh.apply_transform(rot)

        mesh_path = tempfile.NamedTemporaryFile(suffix='.glb', delete=False)
        mesh.export(mesh_path.name, file_type='glb')

        return mesh_path.name

    def run_text(self,
                 prompt: str,
                 seed: int = 0,
                 guidance_scale: float = 15.0,
                 num_steps: int = 64) -> str:
        self.load_model('text300M')
        torch.manual_seed(seed)

        latents = sample_latents(
            batch_size=1,
            model=self.model,
            diffusion=self.diffusion,
            guidance_scale=guidance_scale,
            model_kwargs=dict(texts=[prompt]),
            progress=True,
            clip_denoised=True,
            use_fp16=True,
            use_karras=True,
            karras_steps=num_steps,
            sigma_min=1e-3,
            sigma_max=160,
            s_churn=0,
        )
        return self.to_glb(latents[0])

    def run_image(self,
                  image_path: str,
                  seed: int = 0,
                  guidance_scale: float = 3.0,
                  num_steps: int = 64) -> str:
        self.load_model('image300M')
        torch.manual_seed(seed)

        image = load_image(image_path)
        latents = sample_latents(
            batch_size=1,
            model=self.model,
            diffusion=self.diffusion,
            guidance_scale=guidance_scale,
            model_kwargs=dict(images=[image]),
            progress=True,
            clip_denoised=True,
            use_fp16=True,
            use_karras=True,
            karras_steps=num_steps,
            sigma_min=1e-3,
            sigma_max=160,
            s_churn=0,
        )
        return self.to_glb(latents[0])