reset num_hidden_layers into encoder_layers and decoder_layers

set tf.random_seed to initial parameters

.idea/deployment.xml

+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="PublishConfigData" autoUpload="Always" serverName="39.104.93.174">
+    <serverData>
+      <paths name="39.104.93.174">
+        <serverdata>
+          <mappings>
+            <mapping deploy="/WMT19" local="$PROJECT_DIR$" web="/" />
+          </mappings>
+        </serverdata>
+      </paths>
+    </serverData>
+    <option name="myAutoUpload" value="ALWAYS" />
+  </component>
+</project>
\ No newline at end of file

.idea/vcs.xml

+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="VcsDirectoryMappings">
+    <mapping directory="$PROJECT_DIR$" vcs="Git" />
+  </component>
+</project>
\ No newline at end of file

.idea/webServers.xml

+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="WebServers">
+    <option name="servers">
+      <webServer id="660dec94-3859-4796-9d39-b7fec4547030" name="39.104.93.174" url="http://39.104.93.174">
+        <fileTransfer host="39.104.93.174" port="22" rootFolder="/media/libei" accessType="SFTP">
+          <advancedOptions>
+            <advancedOptions dataProtectionLevel="Private" />
+          </advancedOptions>
+          <option name="port" value="22" />
+        </fileTransfer>
+      </webServer>
+    </option>
+  </component>
+</project>
\ No newline at end of file

tensor2tensor/models/common_hparams.py

@@ -41,6 +41,8 @@ def basic_params1():
       # [8, 10, 12, 14, 16, 20, 24 ... (max_length or batch_size)]
       # [8, 10, 12, 14, 16, 20, 24 ... (max_length or batch_size)]
       batching_mantissa_bits=1,
       batching_mantissa_bits=1,
       num_hidden_layers=4,
       num_hidden_layers=4,
+      encoder_layers=4,
+      decoder_layers=4,
       kernel_height=3,
       kernel_height=3,
       kernel_width=1,
       kernel_width=1,
       hidden_size=64,
       hidden_size=64,

tensor2tensor/models/transformer.py

@@ -231,7 +231,7 @@ def transformer_decoder(decoder_input,
   # Summaries don't work in multi-problem setting yet.
   # Summaries don't work in multi-problem setting yet.
   summaries = "problems" not in hparams.values() or len(hparams.problems) == 1
   summaries = "problems" not in hparams.values() or len(hparams.problems) == 1
   with tf.variable_scope(name):
   with tf.variable_scope(name):
-    for layer in xrange(hparams.num_hidden_layers):
+    for layer in xrange(hparams.decoder_layers):
       with tf.variable_scope("layer_%d" % layer):
       with tf.variable_scope("layer_%d" % layer):
         # self-attention network
         # self-attention network
         residual = x
         residual = x
@@ -361,7 +361,8 @@ def transformer_base():
   hparams.learning_rate = 0.1
   hparams.learning_rate = 0.1
   hparams.learning_rate_warmup_steps = 4000
   hparams.learning_rate_warmup_steps = 4000
   hparams.initializer_gain = 1.0
   hparams.initializer_gain = 1.0
-  hparams.num_hidden_layers = 6
+  hparams.encoder_layers = 6
+  hparams.decoder_layers = 6
   hparams.initializer = "uniform_unit_scaling"
   hparams.initializer = "uniform_unit_scaling"
   hparams.weight_decay = 0.0
   hparams.weight_decay = 0.0
   hparams.optimizer_adam_beta1 = 0.9
   hparams.optimizer_adam_beta1 = 0.9

tensor2tensor/utils/trainer_utils.py

@@ -196,7 +196,7 @@ def create_experiment_components(hparams, output_dir, data_dir, model_name):
           model_dir=output_dir,
           model_dir=output_dir,
           gpu_memory_fraction=FLAGS.worker_gpu_memory_fraction,
           gpu_memory_fraction=FLAGS.worker_gpu_memory_fraction,
           session_config=session_config(),
           session_config=session_config(),
-          #tf_random_seed=FLAGS.random_seed,
+          tf_random_seed=FLAGS.random_seed,
           keep_checkpoint_max=FLAGS.keep_checkpoint_max,
           keep_checkpoint_max=FLAGS.keep_checkpoint_max,
           save_checkpoints_secs=FLAGS.save_checkpoint_secs))
           save_checkpoints_secs=FLAGS.save_checkpoint_secs))
   # Store the hparams in the estimator as well
   # Store the hparams in the estimator as well