Update README.md

README.md

@@ -7,11 +7,11 @@ tags:
 
 ---
 
-# {MODEL_NAME}
+# Quora Sentence Similarity
 
 This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
 
-<!--- Describe your model here -->
+For more information on how it was created, check out the following link: https://github.com/DunnBC22/NLP_Projects/blob/main/Semantic_Similarity/Semantic%20Similarity-large.ipynb
 
 ## Usage (Sentence-Transformers)
 
@@ -32,11 +32,31 @@ embeddings = model.encode(sentences)
 print(embeddings)
 ```
 
-
-
 ## Evaluation Results
 
-<!--- Describe how your model was evaluated -->
+| Metric | Measure | Value | Notes |
+| :--------: | :--------: | :--------: | :--------: |
+| Accuracy | **Cosine-Similarity** | 88.72	| Threshold: 0.8397 |
+| F1 | Cosine-Similarity | 85.22 | Threshold: 0.8223 |
+| Precision | Cosine-Similarity | 80.72 | - |
+| Recall | Cosine-Similarity | 90.25 | - |
+| Average Precision | Cosine-Similarity | 89.75 | - |
+| Accuracy | **Manhattan-Distance** | 88.71	| Threshold: 12.4351 |
+| F1 | Manhattan-Distance | 85.22 | Threshold: 13.2209 |
+| Precision | Manhattan-Distance | 80.58 | - |
+| Recall | Manhattan-Distance | 90.42 | - |
+| Average Precision | Manhattan-Distance | 89.74 | - |
+| Accuracy | **Euclidean-Distance** | 88.72	| Threshold: 0.5662 |
+| F1 | Euclidean-Distance | 85.22 | Threshold: 0.5962 |
+| Precision | Euclidean-Distance | 80.72 | - |
+| Recall | Euclidean-Distance | 90.25 | - |
+| Average Precision | Euclidean-Distance | 89.75 | - |
+| Accuracy | **Dot-Product** | 88.72 | Threshold: 0.8397 |
+| F1 | Dot-Product | 85.22 | Threshold: 0.8223 |
+| Precision | Dot-Product | 80.72 | - |
+| Recall | Dot-Product | 90.25 | - |
+| Average Precision | Dot-Product | 89.75 | - |
+
 
 For an automated evaluation of this model, see the *Sentence Embeddings Benchmark*: [https://seb.sbert.net](https://seb.sbert.net?model_name={MODEL_NAME})
 
@@ -73,6 +93,19 @@ Parameters of the fit()-Method:
 }
 ```
 
+**Potential Improvements**
+
+One way to improve the results of this model is to use a larger checkpoint of T5. This was trained with the T5-large checkpoint.
+
+The larger checkpoints are:
+
+| Checkpoint | # of Train Params |
+| :--------: | :--------: |
+| T5-Base | 220 Million |
+| T5-Large | 770 Million* |
+| T5-3B | 3 Billion |
+| T5-11B | 11 Billion |
+
 
 ## Full Model Architecture
 ```
@@ -86,4 +119,4 @@ SentenceTransformer(
 
 ## Citing & Authors
 
-<!--- Describe where people can find more information -->
+Dataset Source: https://www.kaggle.com/datasets/quora/question-pairs-dataset