loaded library: /usr/lib/x86_64-linux-gnu/libibverbs.so.1 loaded library: /usr/lib/x86_64-linux-gnu/libibverbs.so.1 ------------------------ arguments ------------------------ batch_size ...................................... 65536 batch_size_per_proc ............................. 65536 data_dir ........................................ /dataset/f9f659c5/wdl_ofrecord data_part_name_suffix_length .................... 5 data_part_num ................................... 256 dataset_format .................................. ofrecord ddp ............................................. True deep_dropout_rate ............................... 0.5 deep_embedding_vec_size ......................... 16 deep_vocab_size ................................. 2322444 eval_after_training ............................. False eval_batchs ..................................... 20 eval_interval ................................... 0 execution_mode .................................. eager hidden_size ..................................... 1024 hidden_units_num ................................ 2 learning_rate ................................... 0.001 loss_print_every_n_iter ......................... 100 max_iter ........................................ 1100 model_load_dir .................................. model_save_dir .................................. ./checkpoint num_deep_sparse_fields .......................... 26 num_dense_fields ................................ 13 num_wide_sparse_fields .......................... 2 save_initial_model .............................. False save_model_after_each_eval ...................... False test_name ....................................... noname_test wide_vocab_size ................................. 2322444 -------------------- end of arguments --------------------- [rank:0] iter: 100/1100, loss: 0.5047056674957275, latency(ms): 162.2926598042249680 | 2022-04-21 01:50:21.297 timestamp, name, driver_version, utilization.gpu [%], utilization.memory [%], memory.total [MiB], memory.free [MiB], memory.used [MiB] 2022/04/21 01:50:21.327, Tesla V100-SXM2-32GB, 470.57.02, 64 %, 26 %, 32510 MiB, 28966 MiB, 3544 MiB [rank:0] iter: 200/1100, loss: 0.4694474935531616, latency(ms): 114.4534153863787651 | 2022-04-21 01:50:32.742 [rank:0] iter: 300/1100, loss: 0.4632295966148376, latency(ms): 115.0134257972240448 | 2022-04-21 01:50:44.244 [rank:0] iter: 400/1100, loss: 0.4601241946220398, latency(ms): 113.7951760366559029 | 2022-04-21 01:50:55.623 [rank:0] iter: 500/1100, loss: 0.4572415649890900, latency(ms): 115.2302493900060654 | 2022-04-21 01:51:07.146 [rank:0] iter: 600/1100, loss: 0.4522928297519684, latency(ms): 118.2617050781846046 | 2022-04-21 01:51:18.972 [rank:0] iter: 700/1100, loss: 0.4466758072376251, latency(ms): 114.9625991284847260 | 2022-04-21 01:51:30.469 [rank:0] iter: 800/1100, loss: 0.4475266337394714, latency(ms): 118.1749506667256355 | 2022-04-21 01:51:42.286 [rank:0] iter: 900/1100, loss: 0.4468182027339935, latency(ms): 116.5101883932948112 | 2022-04-21 01:51:53.937 [rank:0] iter: 1000/1100, loss: 0.4463140070438385, latency(ms): 121.2108919024467468 | 2022-04-21 01:52:06.058 [rank:0] iter: 1100/1100, loss: 0.4464303851127625, latency(ms): 122.3694422468543053 | 2022-04-21 01:52:18.295