fix senta to senta_bilstm

demo/senta/cli_demo.sh

-python ../../paddlehub/commands/hub.py run senta --input_file test/test.txt
+python ../../paddlehub/commands/hub.py run senta_bilstm --input_file test/test.txt

demo/senta/predict.py

@@ -21,7 +21,7 @@ args = parser.parse_args()
 
 if __name__ == '__main__':
     # loading Paddlehub senta pretrained model
-    module = hub.Module(name="senta")
+    module = hub.Module(name="senta_bilstm")
     inputs, outputs, program = module.context(trainable=True)
 
     # Sentence classification  dataset reader
@@ -32,13 +32,11 @@ if __name__ == '__main__':
     place = fluid.CUDAPlace(0) if args.use_gpu else fluid.CPUPlace()
     exe = fluid.Executor(place)
     with fluid.program_guard(program):
-        # Use "sequence_output" for classification tasks on an entire sentence.
-        # Use "sequence_outputs" for token-level output.
-        sequence_output = outputs["sequence_output"]
+        sent_feature = outputs["sentence_feature"]
 
         # Define a classfication finetune task by PaddleHub's API
         cls_task = hub.create_text_cls_task(
-            feature=sequence_output, num_classes=dataset.num_labels)
+            feature=sent_feature, num_classes=dataset.num_labels)
 
         # Setup feed list for data feeder
         # Must feed all the tensor of senta's module need

demo/senta/run_finetune.sh

-export CUDA_VISIBLE_DEVICES=0
+export CUDA_VISIBLE_DEVICES=5
 
 DATASET="chnsenticorp"
 CKPT_DIR="./ckpt_${DATASET}"
 
 python -u senta_finetune.py \
                    --batch_size=24 \
-                   --use_gpu=False \
+                   --use_gpu=True \
                    --checkpoint_dir=${CKPT_DIR} \
                    --num_epoch=3

demo/senta/senta_finetune.py

@@ -15,7 +15,7 @@ args = parser.parse_args()
 
 if __name__ == '__main__':
     # Step1: load Paddlehub senta pretrained model
-    module = hub.Module(name="senta")
+    module = hub.Module(name="senta_bilstm")
     inputs, outputs, program = module.context(trainable=True)
 
     # Step2: Download dataset and use TextClassificationReader to read dataset
@@ -24,7 +24,7 @@ if __name__ == '__main__':
     reader = hub.reader.LACClassifyReader(
         dataset=dataset, vocab_path=module.get_vocab_path())
 
-    sent_feature = outputs["sequence_output"]
+    sent_feature = outputs["sentence_feature"]
 
     # Define a classfication finetune task by PaddleHub's API
     cls_task = hub.create_text_cls_task(
@@ -35,7 +35,7 @@ if __name__ == '__main__':
     feed_list = [inputs["words"].name, cls_task.variable('label').name]
 
     strategy = hub.finetune.strategy.AdamWeightDecayStrategy(
-        learning_rate=1e-3, weight_decay=0.01, warmup_proportion=0.01)
+        learning_rate=1e-4, weight_decay=0.01, warmup_proportion=0.05)
 
     config = hub.RunConfig(
         use_cuda=args.use_gpu,