Update README.md

README.md

@@ -161,9 +161,9 @@ Some instruction datasets are added for curiosity sake although model is not tra
 - The model does not perform well enough to tell educational value in instruction datasets.
 
 # 📈Analysis
-## 🤖Model training with and without classifier
+## 🤖Model Training With And Without Classifier
 The expectation is that the model trained with filter will outperform model trained without the filter.  
-Fineweb is filtered on the fly with Educational Value >= 1.   
+Fineweb is filtered on the fly with Educational Value >= 1.0.   
 
 Test 1:  
 Model params: 192M  
@@ -178,7 +178,7 @@ Training token: 3.1B training token, 6000 global steps
 |TruthfulQA| 45.88 | 45.20| 45.97|
 |Winogrande| 49.49 | 50.59 | 50.67 |
 
-The reasoning and commensense reasoning seems to be better when class, aligning with expectation. It is also close to Cosmopedia.  
+The reasoning and commensense reasoning seems to be better when filter is on, aligning with expectation. It is also close to Cosmopedia.  
 MMLU is better also; however it is close to random due to limitation in compute (both training time and model size).  
 Model of larger size will be trained to further validate this claim.  
 
@@ -192,3 +192,6 @@ The first 10M records have been analysed.  Full file in [here](https://drive.goo
 Below is the top 100 domain names, with no of record >= 100.
 ![image/png](https://cdn-uploads.huggingface.co/production/uploads/60e50ce5350d181892d5a636/3QNYYVbFIqaAUh-574lED.png)
 
+## 🧪Classifier Ranking Ordering
+Spearman rank-order correlation coefficient between Educational Value and that of test data is 0.7055, indicating a strong monotonic relationship. The Educational Value can be used for ranking.
+![image/png](https://cdn-uploads.huggingface.co/production/uploads/60e50ce5350d181892d5a636/dKV2oXRv3WpEsfDXy0bl7.png)