fix requirements.txt

app.py

@@ -7,7 +7,6 @@ import glob
 import numpy as np
 from PIL import Image
 import time
-import tqdm
 import copy
 import sys
 
@@ -187,7 +186,7 @@ BASE_CONFIG = "configs/train_co3d_concept.yaml"
 BASE_CKPT = "pretrained-models/sd_xl_base_1.0.safetensors"
 
 start_time = time.time()
-# base_model = load_base_model(BASE_CONFIG, ckpt=BASE_CKPT, verbose=False)
+base_model = load_base_model(BASE_CONFIG, ckpt=BASE_CKPT, verbose=False)
 print(f"Time taken to load base model: {time.time() - start_time:.2f}s")
 
 global curr_camera_dict 
@@ -334,6 +333,9 @@ with gr.Blocks(head=head,
             <a href='https://customdiffusion360.github.io/index.html' style="padding: 10px;">
                <img src='https://img.shields.io/badge/Project%20Page-8A2BE2'>
             </a>
+            <a href='https://arxiv.org/abs/2404.12333'>
+                <img src="https://img.shields.io/badge/arXiv-2404.12333-red">
+            </a>
             <a class="link" href='https://github.com/customdiffusion360/custom-diffusion360' style="padding: 10px;">
                 <img src='https://img.shields.io/badge/Github-%23121011.svg'>
             </a>
@@ -343,9 +345,6 @@ with gr.Blocks(head=head,
     visible=True
     )
 
-    gr
-
-
     if SPACE_ID == ORIGINAL_SPACE_ID:
         gr.Markdown(SHARED_UI_WARNING)
       

requirements.txt

@@ -1,3 +1,5 @@
+gradio
+plotly
 omegaconf
 einops
 fire

sampling_for_demo.py

@@ -1,5 +1,3 @@
-import glob
-import os
 import sys
 import copy
 from typing import List
@@ -275,20 +273,7 @@ def process_camera_json(camera_json, example_cam):
     eye = torch.tensor([camera_dict["eye"]["x"], camera_dict["eye"]["y"], camera_dict["eye"]["z"]], dtype=torch.float32).unsqueeze(0)
     up = torch.tensor([camera_dict["up"]["x"], camera_dict["up"]["y"], camera_dict["up"]["z"]], dtype=torch.float32).unsqueeze(0)
     center = torch.tensor([camera_dict["center"]["x"], camera_dict["center"]["y"], camera_dict["center"]["z"]], dtype=torch.float32).unsqueeze(0)
-    new_R, new_T = look_at_view_transform(eye=eye, at=center, up=up)
-
-    ## temp
-    # new_R = torch.tensor([[[ 0.4988,  0.2666,  0.8247],
-    #      [-0.1917, -0.8940,  0.4049],
-    #      [ 0.8453, -0.3601, -0.3948]]], dtype=torch.float32)
-    # new_T = torch.tensor([[ 0.0739, -0.0013,  0.9973]], dtype=torch.float32)
-
-
-    # new_R = torch.tensor([[[ 0.2530,  0.2989,  0.9201],
-    #                         [-0.2652, -0.8932,  0.3631],
-    #                         [ 0.9304, -0.3359, -0.1467],]], dtype=torch.float32)
-    # new_T = torch.tensor([[ 0.0081,  0.0337,  1.0452]], dtype=torch.float32)
-    
+    new_R, new_T = look_at_view_transform(eye=eye, at=center, up=up)  
 
     print("focal length", example_cam.focal_length)
     print("principal point", example_cam.principal_point)
@@ -475,4 +460,3 @@ def sample(model, data,
 
     print("generation done")
     return image
-