update links from sgnlp to sgnlp-models

README.md

@@ -74,7 +74,7 @@ from sgnlp.models.lif_3way_ap.modules.allennlp.dataset_reader import (
 )
 
 model = Lif3WayApModel.from_pretrained(
-    "https://storage.googleapis.com/sgnlp/models/lif_3way_ap/model.tar.gz",
+    "https://storage.googleapis.com/sgnlp-models/models/lif_3way_ap/model.tar.gz",
     predictor_name="lif_3way_ap_predictor",
 )
 
@@ -112,10 +112,10 @@ It can be downloaded from the [author's github repository](https://github.com/nu
 
 #### Training Results
 - **Training Time:** ~12 hours for 13 epochs on a single V100 GPU.
-- **Training Config:** [link](https://storage.googleapis.com/sgnlp/models/lif_3way_ap/config.json)
+- **Training Config:** [link](https://storage.googleapis.com/sgnlp-models/models/lif_3way_ap/config.json)
 
 # Model Parameters
-- **Model Weights:** [link](https://storage.googleapis.com/sgnlp/models/lif_3way_ap/model.tar.gz)
+- **Model Weights:** [link](https://storage.googleapis.com/sgnlp-models/models/lif_3way_ap/model.tar.gz)
 - **Model Inputs:** Context, historical questions and answers pairs, and candidate question.
 - **Model Outputs:** Probability score of whether candidate question is a valid follow-up question.
 - **Model Size:** ~130MB