Create test

test

@@ -0,0 +1,22 @@
+import streamlit as st
+
+from transformers import T5ForConditionalGeneration, T5Tokenizer
+
+model = T5ForConditionalGeneration.from_pretrained("Unbabel/gec-t5_small")
+tokenizer = T5Tokenizer.from_pretrained('t5-small')
+
+sentence = st.text_area("enter some text")
+
+tokenized_sentence = tokenizer('gec: ' + sentence, max_length=128, truncation=True, padding='max_length', return_tensors='pt')
+corrected_sentence = tokenizer.decode(
+    model.generate(
+        input_ids = tokenized_sentence.input_ids,
+        attention_mask = tokenized_sentence.attention_mask, 
+        max_length=128,
+        num_beams=5,
+        early_stopping=True,
+    )[0],
+    skip_special_tokens=True, 
+    clean_up_tokenization_spaces=True
+)
+print(corrected_sentence)