Update README.md

README.md

@@ -151,7 +151,7 @@ model-index:
 
 # deberta-v3-base for Extractive QA
 
-This is the [deberta-v3-base](https://huggingface.co/microsoft/deberta-v3-base) model, fine-tuned using the [SQuAD2.0](https://huggingface.co/datasets/squad_v2) dataset. It's been trained on question-answer pairs, including unanswerable questions, for the task of Extractive Question Answering.
+This is the [deberta-v3-base](https://huggingface.co/microsoft/deberta-v3-base) model, fine-tuned using the SQuAD 2.0, MRQA, AdversarialQA, and SynQA datasets. It's been trained on question-answer pairs, including unanswerable questions, for the task of Extractive Question Answering.
 
 ## Overview
 **Language model:** deberta-v3-base  
@@ -199,41 +199,17 @@ answer = tokenizer.decode(tokenizer.convert_tokens_to_ids(answer_tokens))
 # 'London'
 ```
 
-## Metrics
+## Dataset Preparation
 
-```bash
-# Squad v2
-{
-    "eval_HasAns_exact": 84.36234817813765,
-    "eval_HasAns_f1": 90.09079905537246,
-    "eval_HasAns_total": 5928,
-    "eval_NoAns_exact": 74.61732548359966,
-    "eval_NoAns_f1": 74.61732548359966,
-    "eval_NoAns_total": 5945,
-    "eval_best_exact": 79.45759285774446,
-    "eval_best_exact_thresh": 0.0,
-    "eval_best_f1": 82.31771724081922,
-    "eval_best_f1_thresh": 0.0,
-    "eval_exact": 79.48286027120358,
-    "eval_f1": 82.34298465427844,
-    "eval_runtime": 109.7262,
-    "eval_samples": 11951,
-    "eval_samples_per_second": 108.917,
-    "eval_steps_per_second": 4.539,
-    "eval_total": 11873
-}
+The MRQA dataset was updated to fix some errors and formatting to work with the `run_qa.py` example script provided in the Hugging Face Transformers library.
+The changes included
+- Updating incorrect answer starts locations (usually off by a few characters)
+- Updating the answer text to match the text found in the context
+The script used to process the MRQA dataset is provided in this repo at XXX.
+
+### MRQA
+- The answer
 
-# Squad
-{
-    "eval_exact": 85.89403973509934,
-    "eval_f1": 91.2982923196374,
-    "eval_runtime": 96.6499,
-    "eval_samples": 10618,
-    "eval_samples_per_second": 109.86,
-    "eval_steps_per_second": 4.584,
-    "eval_total": 10570
-}
-```
 
 ## Training procedure