Add Todo + Pipeline example

README.md

@@ -14,8 +14,8 @@ Title generator based on Neo-GPT 125M fine-tuned on a dataset of 39k url's title
 
 # Pipeline example
 
-import pandas as pd
 ```python
+import pandas as pd
 from transformers import AutoModelForMaskedLM
 from transformers import GPT2Tokenizer, TrainingArguments, AutoModelForCausalLM, AutoConfig 
 
@@ -50,10 +50,16 @@ for i in list_title_gen:
     except:
         continue        
 
-list_title_gen = [sub.replace('�', ' ').replace('\r',' ').replace('\n',' ').replace('\t', ' ').replace('\xa0', '') for sub in list_title_gen]    
+list_title_gen = [sub.replace('�', ' ').replace('\\r',' ').replace('\
+',' ').replace('\\t', ' ').replace('\\xa0', '') for sub in list_title_gen]    
 list_title_gen = [sub if sub != '<|startoftext|> Electric car' else '' for sub in list_title_gen]  
 
 for i in list_title_gen:
     print(i)
 
 ```
+
+# Todo
+- Improve the quality of the training sample
+- Add more data
+