Fix users of NumPy APIs that are removed in NumPy 2.0 (#2640)

BUG=cl/655772545

tensorflow/lite/micro/tools/layer_by_layer_debugger.py

diff --git a/tensorflow/lite/micro/tools/layer_by_layer_debugger.py b/tensorflow/lite/micro/tools/layer_by_layer_debugger.py
index e41e551..8aa263b 100644
--- a/tensorflow/lite/micro/tools/layer_by_layer_debugger.py
+++ b/tensorflow/lite/micro/tools/layer_by_layer_debugger.py
@@ -101,7 +101,7 @@
       layer_schema_fb.TensorTypes.INT64:
       np.int64,
       layer_schema_fb.TensorTypes.STRING:
-      np.string_,
+      np.bytes_,
       layer_schema_fb.TensorTypes.BOOL:
       np.bool_,
       layer_schema_fb.TensorTypes.INT16:

tensorflow/lite/micro/tools/model_transforms_utils.py

diff --git a/tensorflow/lite/micro/tools/model_transforms_utils.py b/tensorflow/lite/micro/tools/model_transforms_utils.py
index c713f01..4a59d88 100644
--- a/tensorflow/lite/micro/tools/model_transforms_utils.py
+++ b/tensorflow/lite/micro/tools/model_transforms_utils.py
@@ -264,7 +264,7 @@
       schema_fb.TensorType.INT32: np.int32,
       schema_fb.TensorType.UINT8: np.uint8,
       schema_fb.TensorType.INT64: np.int64,
-      schema_fb.TensorType.STRING: np.string_,
+      schema_fb.TensorType.STRING: np.bytes_,
       schema_fb.TensorType.BOOL: np.bool_,
       schema_fb.TensorType.INT16: np.int16,
       schema_fb.TensorType.COMPLEX64: np.complex64,

tensorflow/lite/micro/tools/requantize_flatbuffer_utils.py

diff --git a/tensorflow/lite/micro/tools/requantize_flatbuffer_utils.py b/tensorflow/lite/micro/tools/requantize_flatbuffer_utils.py
index 5709ff2..c4a3b10 100644
--- a/tensorflow/lite/micro/tools/requantize_flatbuffer_utils.py
+++ b/tensorflow/lite/micro/tools/requantize_flatbuffer_utils.py
@@ -24,7 +24,7 @@
     TensorType.INT32: np.int32,
     TensorType.UINT8: np.uint8,
     TensorType.INT64: np.int64,
-    TensorType.STRING: np.string_,
+    TensorType.STRING: np.bytes_,
     TensorType.BOOL: np.bool_,
     TensorType.INT16: np.int16,
     TensorType.COMPLEX64: np.complex64,