Update README.md

README.md

@@ -12,9 +12,15 @@ pipeline_tag: text-classification
 ### Model Description
 Finetuned xlm-roberta-base for Sentiment Analysis
 ### Training results
-I forgot to save the metric but I remember it's quite good (accuracy~96%, loss ~0.1)
+Trained on TPU VM v4-8 for ~3 hours
+|epoch|step|train_accuracy|train_loss|val_accuracy|val_loss|
+| --- | --- | --- | --- | --- | --- |
+|0|5391|0.955597997|0.118527733|0.963498533|0.098501749|
+|1|10783|0.965486944|0.092906699|0.964689374|0.094814248|
+|2|16175|0.968293846|0.085916176|0.965770006|0.093040377|
 ### Training procedure
 For replication, go to [GitHub page](https://github.com/carant-ai/tpu-sentiment-trainer/tree/main)
 ### Special Thanks
-1. [carlesoctav](https://huggingface.co/carlesoctav) for making the training script workable in TPU VM
-2. Google’s TPU Research Cloud (TRC) for providing Cloud TPU VM.
+1. Google’s TPU Research Cloud (TRC) for providing Cloud TPU VM.
+2. [carlesoctav](https://huggingface.co/carlesoctav) for making the training script on TPU VM
+3. [thonyyy](https://huggingface.co/thonyyy) for gathering the sentiment dataset