TensorFlow苗条的预训练模型及其权重保存在tf.float32中。为了提高内存效率，我想在tf.float16中加载一个预先训练的模型，然后在tf.float16中附加一些其他模块之后运行一些训练操作。
我知道由于张量数据类型不匹配，通常无法执行此操作。但是，是否有专门用于解决此问题的资源？ 有人曾经尝试过类似的东西吗？
对于记录，以下是我用来验证是否可以在tf.float16中恢复预训练模型并且不起作用的摘要。
import tensorflow as tf
from inception_v4 import inception_v4, inception_v4_arg_scope
import numpy as np
slim = tf.contrib.slim
dtype = tf.float16
input = tf.placeholder(dtype, [None, 299,299,3])
with slim.arg_scope(inception_v4_arg_scope()):
logits, endpoints = inception_v4(input)
saver = tf.train.Saver()
init_op = tf.global_variables_initializer()
img = np.random.randn(10, 299, 299, 3)
with tf.Session() as sess:
sess.run(init_op)
saver.restore(sess, './inception_v4.ckpt')
l = sess.run(logits, feed_dict={input : img})
输出为
2018-06-26 21:14:51.708190: I tensorflow/core/platform/cpu_feature_guard.cc:140] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
Traceback (most recent call last):
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1361, in _do_call
return fn(*args)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1340, in _run_fn
target_list, status, run_metadata)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/framework/errors_impl.py", line 516, in __exit__
c_api.TF_GetCode(self.status.status))
tensorflow.python.framework.errors_impl.InvalidArgumentError: Expected to restore a tensor of type half, got a tensor of type float instead: tensor_name = InceptionV4/AuxLogits/Aux_logits/biases
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_HALF, DT_HALF, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_HALF, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_HALF], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "test.py", line 18, in <module>
saver.restore(sess, './inception_v4.ckpt')
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/training/saver.py", line 1755, in restore
{self.saver_def.filename_tensor_name: save_path})
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 905, in run
run_metadata_ptr)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1137, in _run
feed_dict_tensor, options, run_metadata)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1355, in _do_run
options, run_metadata)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1374, in _do_call
raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.InvalidArgumentError: Expected to restore a tensor of type half, got a tensor of type float instead: tensor_name = InceptionV4/AuxLogits/Aux_logits/biases
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_HALF, DT_HALF, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_HALF, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_HALF], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
Caused by op 'save/RestoreV2', defined at:
File "test.py", line 10, in <module>
saver = tf.train.Saver()
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/training/saver.py", line 1293, in __init__
self.build()
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/training/saver.py", line 1302, in build
self._build(self._filename, build_save=True, build_restore=True)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/training/saver.py", line 1339, in _build
build_save=build_save, build_restore=build_restore)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/training/saver.py", line 796, in _build_internal
restore_sequentially, reshape)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/training/saver.py", line 449, in _AddRestoreOps
restore_sequentially)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/training/saver.py", line 847, in bulk_restore
return io_ops.restore_v2(filename_tensor, names, slices, dtypes)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/ops/gen_io_ops.py", line 1030, in restore_v2
shape_and_slices=shape_and_slices, dtypes=dtypes, name=name)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/framework/op_def_library.py", line 787, in _apply_op_helper
op_def=op_def)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/framework/ops.py", line 3271, in create_op
op_def=op_def)
File "/Users/meetukme/anaconda/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/framework/ops.py", line 1650, in __init__
self._traceback = self._graph._extract_stack() # pylint: disable=protected-access
InvalidArgumentError (see above for traceback): Expected to restore a tensor of type half, got a tensor of type float instead: tensor_name = InceptionV4/AuxLogits/Aux_logits/biases
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_HALF, DT_HALF, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_HALF, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_HALF], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]