latte

README.md

@@ -58,17 +58,21 @@
 
 数据结构，这里为示例数据（仅展示UCF-101），完整数据请按如下结构准备。
 
-    UCF-101_tiny
-    ├── ApplyEyeMakeup
-    │   └── v_ApplyEyeMakeup_g01_c01.avi
-    ├── ApplyLipstick
-    │   └── v_ApplyLipstick_g01_c01.avi
-    ├── Archery
-    │   └── v_Archery_g01_c01.avi
-    ├── BabyCrawling
-    │   └── v_BabyCrawling_g01_c01.avi
-    ├── BalanceBeam
-    ├── .....
+    train_datasets/
+    └── UCF-101_tiny
+        ├── ApplyEyeMakeup
+        │   └── v_ApplyEyeMakeup_g01_c01.avi
+        ├── ApplyLipstick
+        │   └── v_ApplyLipstick_g01_c01.avi
+        ├── Archery
+        │   └── v_Archery_g01_c01.avi
+        ├── BabyCrawling
+        │   └── v_BabyCrawling_g01_c01.avi
+        ├── BalanceBeam
+        │   └── v_BalanceBeam_g01_c01.avi
+        ├── BandMarching
+        │   └── v_BandMarching_g01_c01.avi
+
 
 
 ## 训练

models/latte_t2v.py

@@ -541,7 +541,7 @@ class LatteT2V(ModelMixin, ConfigMixin):
                 f"Has to define `in_channels`: {in_channels}, `num_vector_embeds`: {num_vector_embeds}, or patch_size:"
                 f" {patch_size}. Make sure that `in_channels`, `num_vector_embeds` or `num_patches` is not None."
             )
-
+            
         # 2. Define input layers
         if self.is_input_continuous:
             self.in_channels = in_channels