Created an evaluation script which can be explicitly run on the cpu.

Eval script constantly loops and waits to eval on new checkpoints.

Added accuracy tracking to training script, although it causes a slowdown.

eval_squeezenet.py

@@ -0,0 +1,64 @@
+import tensorflow as tf
+import math
+from models.slim.datasets import dataset_factory
+from models.slim.preprocessing import preprocessing_factory
+
+import squeezenet
+
+slim = tf.contrib.slim
+
+BATCH_SIZE = 256
+CHECKPOINT_DIR = '/mnt/data1/squeezenet_test/wee/train'
+EVAL_DIR = CHECKPOINT_DIR[:-5] + 'test'
+DATA_DIR = '/mnt/data1/cifar'
+EVAL_DEVICE = '/cpu:0'
+
+tf.logging.set_verbosity(tf.logging.INFO)
+
+with tf.Graph().as_default() as g:
+    with g.device(EVAL_DEVICE):
+        dataset = dataset_factory.get_dataset('cifar10', 'test', DATA_DIR)
+
+        tf_global_step = slim.get_or_create_global_step()
+
+        provider = slim.dataset_data_provider.DatasetDataProvider(
+            dataset,
+            shuffle=False,
+            common_queue_capacity=2 * BATCH_SIZE,
+            common_queue_min=BATCH_SIZE)
+
+        [image, label] = provider.get(['image', 'label'])
+
+        image_preprocessing_fn = preprocessing_factory.get_preprocessing(
+            'cifarnet',
+            is_training=False)
+
+        image = image_preprocessing_fn(image, 32, 32)
+
+        images, labels = tf.train.batch(
+            [image, label],
+            batch_size=BATCH_SIZE,
+            num_threads=2,
+            capacity=5 * BATCH_SIZE)
+
+        logits, end_points = squeezenet.inference(images)
+
+        predictions = tf.argmax(logits, 1)
+
+        accuracy, update_op = slim.metrics.streaming_accuracy(predictions,
+                                                              labels)
+        tf.scalar_summary('eval/accuracy', accuracy)
+        summary_op = tf.merge_all_summaries()
+
+        num_batches = math.ceil(dataset.num_samples / float(BATCH_SIZE))
+
+        sess_config = tf.ConfigProto(allow_soft_placement=True)
+        slim.evaluation.evaluation_loop(
+            master='',
+            checkpoint_dir=CHECKPOINT_DIR,
+            logdir=EVAL_DIR,
+            num_evals=num_batches,
+            eval_op=update_op,
+            eval_interval_secs=45,
+            session_config=sess_config,
+            variables_to_restore=slim.get_variables_to_restore())