Update README.md

README.md

@@ -129,7 +129,7 @@ With sufficient pre-training data, it can surpass larger models. The results hig
 
 ## Training Details
 
-RoBERTaLexPT-base is pretrained from both data:
+RoBERTaLexPT-base is pretrained on:
 - [LegalPT](https://huggingface.co/datasets/eduagarcia/LegalPT_dedup) is a Portuguese legal corpus by aggregating diverse sources of up to 125GiB data.
 - [CrawlPT](https://huggingface.co/datasets/eduagarcia/CrawlPT_dedup) is a composition of three Portuguese general corpora: [brWaC](https://huggingface.co/datasets/brwac), [CC100 PT subset](https://huggingface.co/datasets/eduagarcia/cc100-pt), [OSCAR-2301 PT subset](https://huggingface.co/datasets/eduagarcia/OSCAR-2301-pt_dedup).
 
@@ -149,10 +149,10 @@ To ensure that domain models are not constrained by a generic vocabulary, we uti
 
 #### Training Hyperparameters
 
-The pretraining process involved training the model for 62,500 steps, with a batch size of 2048 and a learning rate of 4e-4, each sequence containing a maximum of 512 tokens.
-The weight initialization is random.
-We employed the masked language modeling objective, where 15\% of the input tokens were randomly masked.
-The optimization was performed using the AdamW optimizer with a linear warmup and a linear decay learning rate schedule.
+The pretraining process involved training the model for 62,500 steps, with a batch size of 2048 and a learning rate of 4e-4, each sequence containing a maximum of 512 tokens.  
+The weight initialization is random.  
+We employed the masked language modeling objective, where 15\% of the input tokens were randomly masked.  
+The optimization was performed using the AdamW optimizer with a linear warmup and a linear decay learning rate schedule.  
 
 For other parameters we adopted the standard [RoBERTa-base hyperparameters](https://huggingface.co/FacebookAI/roberta-base):