Update README.md

README.md

@@ -58,12 +58,12 @@ print(decoded)
 
 ## Custom pipeline
 
-#### The easiest way to generate text with more control and without length restrictions is by using a custom pipeline. To reduce the number of similar questions, the similarity_threshold used to increase and decrease the similarity of questions returned. Additionally, the stride argument is used to overlap data in the context parts.
+The easiest way to generate text with more control and without length restrictions is by using a custom pipeline. To reduce the number of similar questions, the similarity_threshold used to increase and decrease the similarity of questions returned. Additionally, the stride argument is used to overlap data in the context parts.
 ```shell
 pip install git+https://github.com/ESgarbi/t5-qa-builder
 ```
 
-#### Run the custom pipeline. Weights will be downloaded from the HF hub automatically. All special tokens will be handled internally, no need to supply the <qa_builder_context> token to initialise the task.
+Run the custom pipeline. Weights will be downloaded from the HF hub automatically. All special tokens will be handled internally, no need to supply the <qa_builder_context> token to initialise the task.
 
 ```python
 
@@ -97,7 +97,7 @@ print(result)
 ```
 
 
-#### Output:
+Output:
 ```json
 Generating QA pairs: 100%|██████████| 1/1 [00:13<00:00, 13.66s/it]
 {