Add "feature extraction" as our pipeline

README.md

@@ -1,12 +1,13 @@
 ---
-language: 
-  - multilingual
+language:
+- multilingual
 tags:
 - code
 - sentence embedding
-license: "mit"
+license: mit
 datasets:
 - CodeSearchNet
+pipeline_tag: feature-extraction
 ---
 
 # Model Card for CodeCSE
@@ -36,4 +37,4 @@ batch = prepare_inputs(nl_json, tokenizer, args)
 nl_inputs = batch[3]
 with torch.no_grad():        
     nl_vec = model(input_ids=nl_inputs, sent_emb="nl")[1] 
-```
+```