--- a
+++ b/src/Parser/biomedner_server.py
@@ -0,0 +1,91 @@
+import os
+import json
+import socket
+import struct
+import argparse
+
+from datetime import datetime
+from biomedner_init import BioMedNER
+from ops import filter_entities, pubtator2dict_list
+
+def count_entities(data):
+    num_entities = 0
+    for d in data:
+        if 'entities' not in d:
+            continue
+        for ent_type, entities in d['entities'].items():
+            num_entities += len(entities)
+
+    return num_entities
+
+def biomedner_recognize(model, dict_path, base_name, args):
+    input_mt_ner = os.path.join(args.biomedner_home, 'input',
+                                f'{dict_path[2:]}.PubTator')
+    output_mt_ner = os.path.join(args.biomedner_home, 'output',
+                                f'{dict_path[2:]}.json')
+    
+    dict_list = pubtator2dict_list(input_mt_ner)
+
+    res = model.recognize(
+        input_dl=dict_list,
+        base_name=base_name
+    )
+
+    if res is None:
+        return None, 0
+
+    num_filtered_species_per_doc = filter_entities(res)
+    for n_f_spcs in num_filtered_species_per_doc:
+        if n_f_spcs[1] > 0:
+            print(datetime.now().strftime(args.time_format),
+                  '[{}] Filtered {} species'
+                  .format(base_name, n_f_spcs[1]))
+    num_entities = count_entities(res)
+
+    res[0]['num_entities'] = num_entities
+    # Write output str to a .PubTator format file
+    with open(output_mt_ner, 'w', encoding='utf-8') as f:
+        json.dump(res[0], f)
+
+def run_server(model, args):
+    host = args.biomedner_host
+    port = args.biomedner_port
+    with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
+        s.bind((host, port))
+        s.listen(1)
+        while True:
+            conn, addr = s.accept()
+            dict_path = conn.recv(512).decode('utf-8')
+            base_name = dict_path.split('.')[0]
+            # hotfix
+            base_name = base_name.replace("\x00A","")
+            
+            biomedner_recognize(model, dict_path, base_name, args)
+            
+            output_stream = struct.pack('>H', len(dict_path)) + dict_path.encode(
+                'utf-8')
+
+            conn.send(output_stream)
+            conn.close()
+
+
+if __name__ == "__main__":
+    argparser = argparse.ArgumentParser()
+    argparser.add_argument('--seed', type=int, help='random seed for initialization',
+                            default=1)
+    argparser.add_argument('--model_name_or_path', default='dmis-lab/bern2-ner')
+    argparser.add_argument('--max_seq_length', type=int, help='The maximum total input sequence length after tokenization. Sequences longer than this will be truncated, sequences shorter will be padded.',
+                            default=128)
+    argparser.add_argument('--biomedner_home',
+                           help='biomedical language model home')         
+    argparser.add_argument('--biomedner_host',
+                           help='biomedical language model host', default='localhost')
+    argparser.add_argument('--biomedner_port', type=int, 
+                           help='biomedical language model port', default=18894)
+    argparser.add_argument('--time_format',
+                            help='time format', default='[%d/%b/%Y %H:%M:%S.%f]')    
+    argparser.add_argument('--no_cuda', action="store_true", help="Avoid using CUDA when available")
+    args = argparser.parse_args()
+    mt_ner = BioMedNER(args)
+    
+    run_server(mt_ner, args)
\ No newline at end of file