[Nov 16th Event] Lewis Tunstall: Simple Training with the 🤗 Transformers Trainer

Course
1

Use this topic to ask your questions to Lewis Tunstall during his talk: Simple Training with the :hugs: Transformers Trainer

You can watch it on YouTube or on Twitch at 8am PST

Link to Notebook

2

Is it possible to share the link to the notebook Lewis is working on here?

1 Like
3

You suggest changing bert-base for MiniLM as baseline pre-trained model. Is there an analogous recommendation for multilingual models? Which checkpoint should be the go-to?

4

I added it on the first post :slight_smile:

1 Like
5

During the technical glitch Lewis explained the weighted loss
Can he show it again?

1 Like
6

Is there a page in the docs / chapter in the course / an external resource with an overview of different loss functions and their advantages/disadvantages?

1 Like
7

How can you use the dataset/trainer/pipeline approach with your own data?

8

Does the way the initialized model was made (as in made using tensorflow or pytorch) affect its performance or cause any issue?

9

This is answered at 25:00 in the main stream.

10

Like Sylvain and Lewis mentioned, this would be a great resource to have. I’d love to collaborate on this

11

This is shown at 27:05 in the main stream.

12

This is answered at 28:37 in the main stream.

13

This is answered at 29:37 on the main stream.