tf 环境问题

model_training_nnn_tpu/rnn_model_tf.py

@@ -795,34 +795,31 @@ def create_tpu_strategy():
         print(f"   {key}={value}")
 
     try:
-        # Try different TPU resolver configurations
+        # Use official TPU initialization pattern (simplified and reliable)
-        resolver = None
+        print("🚀 Using official TensorFlow TPU initialization...")
 
-        if tpu_address:
+        # Use your tested official TPU initialization code
-            print(f"🚀 Attempting TPU connection with address: {tpu_address}")
-            resolver = tf.distribute.cluster_resolver.TPUClusterResolver(tpu=tpu_address)
-        elif tpu_name:
-            print(f"🚀 Attempting TPU connection with name: {tpu_name}")
-            resolver = tf.distribute.cluster_resolver.TPUClusterResolver(tpu=tpu_name)
-        else:
-            # Try auto-detection
-            print("🚀 Attempting TPU auto-detection...")
         resolver = tf.distribute.cluster_resolver.TPUClusterResolver()
-
-        # Initialize TPU
-        print("⚡ Connecting to TPU cluster...")
         tf.config.experimental_connect_to_cluster(resolver)
-
+        # This is the TPU initialization code that has to be at the beginning.
-        print("🔧 Initializing TPU system...")
         tf.tpu.experimental.initialize_tpu_system(resolver)
 
+        # Verify TPU devices (following official example)
+        tpu_devices = tf.config.list_logical_devices('TPU')
+        print("All devices: ", tpu_devices)
+
+        if not tpu_devices:
+            raise RuntimeError("No TPU devices found!")
+
+        print(f"✅ Found {len(tpu_devices)} TPU devices")
+
         # Create TPU strategy
         print("🎯 Creating TPU strategy...")
         strategy = tf.distribute.TPUStrategy(resolver)
 
         print(f"✅ TPU initialized successfully!")
         print(f"🎉 Number of TPU cores: {strategy.num_replicas_in_sync}")
-        print(f"🏃 TPU cluster: {resolver.cluster_spec()}")
+
         return strategy
 
     except Exception as e: