把日记二级类别拆分

tensnsorflow/pipe.sh

@@ -33,15 +33,15 @@ rm ${DATA_PATH}/nearby/nearby_*
 
 
 echo "train..."
-${PYTHON_PATH} ${MODEL_PATH}/train.py --ctr_task_wgt=0.3 --learning_rate=0.0001 --deep_layers=256,128 --dropout=0.8,0.5 --optimizer=Adam --num_epochs=1 --embedding_size=16 --batch_size=1024 --field_size=10 --feature_size=2000 --l2_reg=0.005 --log_steps=100 --num_threads=36 --model_dir=${DATA_PATH}/model_ckpt/DeepCvrMTL/ --data_dir=${DATA_PATH} --task_type=train
+${PYTHON_PATH} ${MODEL_PATH}/train.py --ctr_task_wgt=0.3 --learning_rate=0.0001 --deep_layers=256,128 --dropout=0.8,0.5 --optimizer=Adam --num_epochs=1 --embedding_size=16 --batch_size=1024 --field_size=11 --feature_size=2000 --l2_reg=0.005 --log_steps=100 --num_threads=36 --model_dir=${DATA_PATH}/model_ckpt/DeepCvrMTL/ --data_dir=${DATA_PATH} --task_type=train
 
 
 echo "infer native..."
-${PYTHON_PATH} ${MODEL_PATH}/train.py --ctr_task_wgt=0.3 --learning_rate=0.0001 --deep_layers=256,128 --dropout=0.8,0.5 --optimizer=Adam --num_epochs=1 --embedding_size=16 --batch_size=1024 --field_size=10 --feature_size=2000 --l2_reg=0.005 --log_steps=100 --num_threads=36 --model_dir=${DATA_PATH}/model_ckpt/DeepCvrMTL/ --data_dir=${DATA_PATH}/native --task_type=infer > ${DATA_PATH}/infer.log
+${PYTHON_PATH} ${MODEL_PATH}/train.py --ctr_task_wgt=0.3 --learning_rate=0.0001 --deep_layers=256,128 --dropout=0.8,0.5 --optimizer=Adam --num_epochs=1 --embedding_size=16 --batch_size=1024 --field_size=11 --feature_size=2000 --l2_reg=0.005 --log_steps=100 --num_threads=36 --model_dir=${DATA_PATH}/model_ckpt/DeepCvrMTL/ --data_dir=${DATA_PATH}/native --task_type=infer > ${DATA_PATH}/infer.log
 
 
 echo "infer nearby..."
-${PYTHON_PATH} ${MODEL_PATH}/train.py --ctr_task_wgt=0.3 --learning_rate=0.0001 --deep_layers=256,128 --dropout=0.8,0.5 --optimizer=Adam --num_epochs=1 --embedding_size=16 --batch_size=1024 --field_size=10 --feature_size=2000 --l2_reg=0.005 --log_steps=100 --num_threads=36 --model_dir=${DATA_PATH}/model_ckpt/DeepCvrMTL/ --data_dir=${DATA_PATH}/nearby --task_type=infer > ${DATA_PATH}/infer.log
+${PYTHON_PATH} ${MODEL_PATH}/train.py --ctr_task_wgt=0.3 --learning_rate=0.0001 --deep_layers=256,128 --dropout=0.8,0.5 --optimizer=Adam --num_epochs=1 --embedding_size=16 --batch_size=1024 --field_size=11 --feature_size=2000 --l2_reg=0.005 --log_steps=100 --num_threads=36 --model_dir=${DATA_PATH}/model_ckpt/DeepCvrMTL/ --data_dir=${DATA_PATH}/nearby --task_type=infer > ${DATA_PATH}/infer.log
 
 echo "sort and 2sql"
 ${PYTHON_PATH} ${OLD_PATH}/Model_pipline/sort_and_2sql.py

tensnsorflow/test.py

@@ -28,19 +28,15 @@ def gen_tfrecords(in_file):
     df = pd.read_csv(in_file)
 
     for i in range(df.shape[0]):
+        feats = ["ucity_id", "clevel1_id", "ccity_name", "device_type", "manufacturer",
+                "channel", "top", "l1", "time", "stat_date","l2"]
+        id = np.array([])
+        for j in feats:
+            id = np.append(id,df[j][i])
         features = tf.train.Features(feature={
             "y": tf.train.Feature(float_list=tf.train.FloatList(value=[df["y"][i]])),
             "z": tf.train.Feature(float_list=tf.train.FloatList(value=[df["z"][i]])),
-            "top": tf.train.Feature(int64_list=tf.train.Int64List(value=[df["top"][i]])),
-            "channel": tf.train.Feature(int64_list=tf.train.Int64List(value=[df["channel"][i]])),
-            "ucity_id": tf.train.Feature(int64_list=tf.train.Int64List(value=[df["ucity_id"][i]])),
-            "clevel1_id": tf.train.Feature(int64_list=tf.train.Int64List(value=[df["clevel1_id"][i]])),
-            "ccity_name": tf.train.Feature(int64_list=tf.train.Int64List(value=[df["ccity_name"][i]])),
-            "device_type": tf.train.Feature(int64_list=tf.train.Int64List(value=[df["device_type"][i]])),
-            "manufacturer": tf.train.Feature(int64_list=tf.train.Int64List(value=[df["manufacturer"][i]])),
-            "level2_ids": tf.train.Feature(int64_list=tf.train.Int64List(value=[df["level2_ids"][i]])),
-            "time": tf.train.Feature(int64_list=tf.train.Int64List(value=[df["time"][i]])),
-            "stat_date": tf.train.Feature(int64_list=tf.train.Int64List(value=[df["stat_date"][i]]))
+            "ids": tf.train.Feature(int64_list=tf.train.Int64List(value=id.astype(np.int)))
         })
 
         example = tf.train.Example(features = features)

tensnsorflow/train.py

@@ -53,16 +53,7 @@ def input_fn(filenames, batch_size=32, num_epochs=1, perform_shuffle=False):
         features = {
             "y": tf.FixedLenFeature([], tf.float32),
             "z": tf.FixedLenFeature([], tf.float32),
-            "top": tf.FixedLenFeature([], tf.int64),
-            "channel": tf.FixedLenFeature([], tf.int64),
-            "ucity_id": tf.FixedLenFeature([], tf.int64),
-            "clevel1_id": tf.FixedLenFeature([], tf.int64),
-            "ccity_name": tf.FixedLenFeature([], tf.int64),
-            "device_type": tf.FixedLenFeature([], tf.int64),
-            "manufacturer": tf.FixedLenFeature([], tf.int64),
-            "level2_ids": tf.FixedLenFeature([], tf.int64),
-            "time": tf.FixedLenFeature([], tf.int64),
-            "stat_date": tf.FixedLenFeature([], tf.int64)
+            "ids": tf.FixedLenFeature([11], tf.int64)
         }
 
         parsed = tf.parse_single_example(record, features)