--- a
+++ b/bilstm_crf_ner/README.md
@@ -0,0 +1,29 @@
+# BiLSTM + CRF for NER
+Before running anything, make sure you have generated the pre-processed dataset using the generate_data.py file with the command mentioned in the parent directory. 
+
+## Additional Requirements
+1.	Packages: Anaconda, Pytorch, AllenNLP  
+2.	Glove 300B embeddings
+
+## Training
+1. Change settings in model/config.py  
+2. Main settings to change: File directories, model hyperparameters etc.  
+3. Run build_data.py  
+    - Builds embedding dictionary, text file of words, chars tags, as well as idx to word and idx to char mapping for the model to read  
+        
+4. Run train.py  
+
+## Results
+|              |precision|    recall|  f1-score|
+|:---:|:---:|:---:|:---:|
+|         ADE |      0.1807 |     0.7168 |     0.2887  |
+|         Dosage |      0.9272 |     0.9123 |     0.9197  |
+|        Drug |      0.8898 |     0.9287|     0.9088  |
+|         Duration |      0.8882 |     0.7778 |     0.8293  |
+|         Form |      0.9840 |     0.9172 |     0.9494  |
+|         Frequency |      0.9412 |     0.9494 |     0.9453  |
+|         Reason |      0.7883 |     0.5238 |     0.6294  |
+|         Route |      0.9583 |     0.9226 |     0.9401  |
+|         Strength |      0.9769 |     0.9683 |     0.9726  |
+|   micro avg |      0.8708 |     0.8957 |     0.8831  |
+|   macro avg |      0.8624 |     0.8792 |     0.8684  |