Change Readme, as per the easier loading of .npz file

README.md

@@ -28,7 +28,14 @@ import cv2
 
 
 def main():
-	dataset = wds.WebDataset("https://huggingface.co/datasets/bridgeconn/sign-dictionary-isl/resolve/main/chunk_{00001..00005}.tar").decode()
+	buffer_size = 1024
+	dataset = (
+	    wds.WebDataset(
+            "https://huggingface.co/datasets/bridgeconn/sign-dictionary-isl/resolve/main/chunk_{00001..00002}.tar",
+            shardshuffle=False)
+	    .shuffle(buffer_size)
+	    .decode()
+	)
     for sample in dataset:
 		''' Each sample contains:
 			 'mp4', 'pose-animation.mp4', 
@@ -38,7 +45,7 @@ def main():
 		# print(sample.keys())
 
 		# JSON metadata
-		json_str = sample['json'].decode('utf-8')
+		json_str = sample['json']
         json_data = json.loads(json_str)
 		print(json_data['filename']) 
 		print(json_data['transcripts'])
@@ -53,8 +60,7 @@ def main():
 		process_video(pose_data)
 
 		# dwpose results
-        npz_bytes = sample["pose-dwpose.npz"]
-        dwpose_coords = = np.load(BytesIO(npz_bytes))
+        dwpose_coords = sample["pose-dwpose.npz"] 
 
 		frame_poses = dwpose_coords['frames'].tolist()
 		print(f"Frames in dwpose coords: {len(frame_poses)} poses")