New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fine Tuning Example for summarization #1
Comments
Hi, I think you need to use an Encoder-Decoder model for summarization whereas this model is just an Encoder like BERT. One possible option is to join two instances of this model and train them end-to-end. for more information take a look at huggingface encoder-decoder. What makes you want to train this with MLM? Is it because you want to increase accuracy on your downstream task through intermediate-finetuning? |
1- Yes, it is so. I have finetuned models: I know of a solution to strengthen the pretrain Bigbird model that the training should continue for a few more epochs, but I did not find the example Pytorch code to continue training. 2-One of the state-of-the-art methods for summarizing long texts is the BigBirdPegasusForConditionalGeneration architecture. 3- Is there any script for convert Bigbird pytorch to tf? Thank you for your guidance |
|
1- Yes, I use Huggingface.
This is my code for summarization that is not well trained and does not get good results. 2- Yes I am going to train this model, how can I get guidance from you? thanks. |
It would be better to discuss this somewhere else. this is my telegram ID |
Hi,
Thanks for publishing this model.
1-Is there an example of fine-tuning it for summarization?
2-How to fine-tune it on my dataset using Masked LM in pytorch?
Thanks
The text was updated successfully, but these errors were encountered: