make sure readme runs

README.md

@@ -20,7 +20,11 @@ $ pip install meshgpt-pytorch
 
 ```python
 import torch
-from meshgpt_pytorch import MeshAutoencoder
+
+from meshgpt_pytorch import (
+    MeshAutoencoder,
+    MeshTransformer
+)
 
 # autoencoder
 
@@ -37,12 +41,17 @@ vertices = torch.randn((2, 121, 3))
 faces = torch.randint(0, 121, (2, 64, 3))
 face_edges = torch.randint(0, 64, (2, 2, 96))
 
+face_len = torch.randint(1, 64, (2,))
+face_edges_len = torch.randint(1, 96, (2,))
+
 # forward in the faces
 
 loss = autoencoder(
     vertices = vertices,
     faces = faces,
-    face_edges = face_edges
+    face_edges = face_edges,
+    face_len = face_len,
+    face_edges_len = face_edges_len
 )
 
 loss.backward()
@@ -52,14 +61,17 @@ loss.backward()
 face_vertex_codes = autoencoder.tokenize(
     vertices = vertices,
     faces = faces,
-    face_edges = face_edges
+    face_edges = face_edges,
+    face_len = face_len,
+    face_edges_len = face_edges_len
 )
 
 # now train your transformer to generate this sequence of codes
 
 transformer = MeshTransformer(
     autoencoder,
-    dim = 512
+    dim = 512,
+    max_seq_len = 768
 )
 
 loss = transformer(face_vertex_codes)
@@ -71,7 +83,6 @@ faces_coordinates = transformer.generate()
 
 # (batch, num faces, vertices (3), coordinates (3))
 # now post process for the generated 3d asset
-
 ```
 
 ## Todo

meshgpt_pytorch/meshgpt_pytorch.py

@@ -31,6 +31,8 @@
 
 from torch_geometric.nn.conv import SAGEConv
 
+from tqdm import tqdm
+
 # helper functions
 
 def exists(v):
@@ -308,7 +310,7 @@ def quantize(
 
         vertices = torch.zeros((batch, num_vertices, vertex_dim), device = device)
 
-        # create pad vertex, due to variable lenghted faces
+        # create pad vertex, due to variable lengthed faces
 
         pad_vertex_id = num_vertices
         vertices = F.pad(vertices, (0, 0, 0, 1), value = 0.)
@@ -579,7 +581,7 @@ def generate(
 
         curr_length = codes.shape[-1]
 
-        for i in range(curr_length, self.max_seq_len):
+        for i in tqdm(range(curr_length, self.max_seq_len)):
             can_eos = divisible_by(i + 1, self.num_quantizers * 3)  # only allow for eos to be decoded at the end of each face, defined as 3 vertices with D residusl VQ codes
 
             logits = self.forward(codes, return_loss = False, append_eos = False)

setup.py

@@ -3,7 +3,7 @@
 setup(
   name = 'meshgpt-pytorch',
   packages = find_packages(exclude=[]),
-  version = '0.0.12',
+  version = '0.0.14',
   license='MIT',
   description = 'MeshGPT Pytorch',
   author = 'Phil Wang',
@@ -27,7 +27,8 @@
     'torch_geometric',
     'torchtyping',
     'vector-quantize-pytorch>=1.11.8',
-    'x-transformers>=1.26.0'
+    'x-transformers>=1.26.0',
+    'tqdm'
   ],
   classifiers=[
     'Development Status :: 4 - Beta',