gpu memory usage
ArashHosseini opened this issue · comments
@nimish1512 know a solution to minimize the GPU memory usage?
thank you
Assuming you have 12GB of GPU memory and want to allocate 6GB:
gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=0.5)
with tf.Session(config=tf.ConfigProto(gpu_options=gpu_options)) as sess:
This will force tensorflow to use only 50% of your GPU memory. Similarly, you can change the fraction value as per need
@nimish1512 thx,
this will not fix it sadly, i changed in trainer/restore(from 448) already the following lines
config = tf.ConfigProto()
tflearn_conf = tf.get_collection(tf.GraphKeys.GRAPH_CONFIG)
if tflearn_conf:
config = tflearn_conf[0]
#config.gpu_options.allow_growth=True
config.gpu_options.per_process_gpu_memory_fraction = 0.5
self.session = tf.Session(config=config)
and still using 9GB, i solved it with changes on trainer/init(from 114)
if not session:
config = tf.ConfigProto()
config.gpu_options.allow_growth=True
self.session = tf.Session(config=config)