這篇文章將為大家詳細(xì)講解有關(guān)如何解決Tensorflow sess.run導(dǎo)致的內(nèi)存溢出問題,小編覺得挺實(shí)用的,因此分享給大家做個(gè)參考,希望大家閱讀完這篇文章后可以有所收獲。
發(fā)展壯大離不開廣大客戶長(zhǎng)期以來的信賴與支持,我們將始終秉承“誠信為本、服務(wù)至上”的服務(wù)理念,堅(jiān)持“二合一”的優(yōu)良服務(wù)模式,真誠服務(wù)每家企業(yè),認(rèn)真做好每個(gè)細(xì)節(jié),不斷完善自我,成就企業(yè),實(shí)現(xiàn)共贏。行業(yè)涉及成都PE包裝袋等,在網(wǎng)站建設(shè)公司、營銷型網(wǎng)站、WAP手機(jī)網(wǎng)站、VI設(shè)計(jì)、軟件開發(fā)等項(xiàng)目上具有豐富的設(shè)計(jì)經(jīng)驗(yàn)。下面是調(diào)用模型進(jìn)行批量測(cè)試的代碼(出現(xiàn)溢出),開始以為導(dǎo)致溢出的原因是數(shù)據(jù)讀入方式問題引起的,用了tf , PIL和cv等方式讀入圖片數(shù)據(jù),發(fā)現(xiàn)越來越慢,內(nèi)存占用飆升,調(diào)試時(shí)發(fā)現(xiàn)是sess.run這里出了問題(隨著for循環(huán)進(jìn)行速度越來越慢)。
# Creates graph from saved GraphDef create_graph(pb_path) # Init tf Session config = tf.ConfigProto() config.gpu_options.allow_growth = True sess = tf.Session(config=config) init = tf.global_variables_initializer() sess.run(init) input_image_tensor = sess.graph.get_tensor_by_name("create_inputs/batch:0") output_tensor_name = sess.graph.get_tensor_by_name("conv6/out_1:0") for filename in os.listdir(image_dir): image_path = os.path.join(image_dir, filename) start = time.time() image_data = cv2.imread(image_path) image_data = cv2.resize(image_data, (w, h)) image_data_1 = image_data - IMG_MEAN input_image = np.expand_dims(image_data_1, 0) raw_output_up = tf.image.resize_bilinear(output_tensor_name, size=[h, w], align_corners=True) raw_output_up = tf.argmax(raw_output_up, axis=3) predict_img = sess.run(raw_output_up, feed_dict={input_image_tensor: input_image}) # 1,height,width predict_img = np.squeeze(predict_img) # height, width voc_palette = visual.make_palette(3) masked_im = visual.vis_seg(image_data, predict_img, voc_palette) cv2.imwrite("%s_pred.png" % (save_dir + filename.split(".")[0]), masked_im) print(time.time() - start) print(">>>>>>Done")
下面是解決溢出問題的代碼(將部分代碼放在for循環(huán)外)
# Creates graph from saved GraphDef create_graph(pb_path) # Init tf Session config = tf.ConfigProto() config.gpu_options.allow_growth = True sess = tf.Session(config=config) init = tf.global_variables_initializer() sess.run(init) input_image_tensor = sess.graph.get_tensor_by_name("create_inputs/batch:0") output_tensor_name = sess.graph.get_tensor_by_name("conv6/out_1:0") ############################################################################################################## raw_output_up = tf.image.resize_bilinear(output_tensor_name, size=[h, w], align_corners=True) raw_output_up = tf.argmax(raw_output_up, axis=3) ############################################################################################################## for filename in os.listdir(image_dir): image_path = os.path.join(image_dir, filename) start = time.time() image_data = cv2.imread(image_path) image_data = cv2.resize(image_data, (w, h)) image_data_1 = image_data - IMG_MEAN input_image = np.expand_dims(image_data_1, 0) predict_img = sess.run(raw_output_up, feed_dict={input_image_tensor: input_image}) # 1,height,width predict_img = np.squeeze(predict_img) # height, width voc_palette = visual.make_palette(3) masked_im = visual.vis_seg(image_data, predict_img, voc_palette) cv2.imwrite("%s_pred.png" % (save_dir + filename.split(".")[0]), masked_im) print(time.time() - start) print(">>>>>>Done")
關(guān)于“如何解決Tensorflow sess.run導(dǎo)致的內(nèi)存溢出問題”這篇文章就分享到這里了,希望以上內(nèi)容可以對(duì)大家有一定的幫助,使各位可以學(xué)到更多知識(shí),如果覺得文章不錯(cuò),請(qǐng)把它分享出去讓更多的人看到。