Developer chat

fastai fastai dev
442

No, the current factory methods will stay (as they are useful for beginners) and DataBunch.create is what we use all the time behind the scene whenever we build a databunch, so that one will stay too.

2 Likes
443

Spacy is by far the biggest lib depencency in fastai… around 1Gb. For comparisson, torch is about 250Mb.
It seems that we use it basically for training, is it possible to somehow prevent loading it when we only want/need to predict?

In our study group we wanted to deploy our language model in AWS Lambda but there is a limit on code size and we had to not use fastai, used torch directly.

copied from: https://forums.fast.ai/t/lesson-4-advanced-discussion/30319/19?u=fredguth

444

There is two misspelling on the doc https://docs.fast.ai/data_block.html#Invisible-step:-preprocessing : “vlaidation” and “isntance”

445

Feel free to open a PR to fix them :wink:

446

Regression is here. Whatever your application, you can now easily get your data ready for regression by

  • doing nothing if your target is just an array of floats of dimension 1 since the API should detect it automatically
  • by forcing it with label_cls = FloatList when your call your label_from_*** method
447

I implemented the jupyter notebook experiments module to maximize memory utilization, as discussed here.

Please have a look: https://github.com/stas00/ipyexperiments

Your feedback is sought after and if you do have some, please send it to this thread.

Thank you.

2 Likes
448

Okey, I only have to change the file in docs_src and you take care of the conversion right?

449

This is great!

450

Yes indeed.

451

Small breaking changes:

  • removed TextFilesList to replace it by TextList like in everywhere else.
  • put col everywhere there was a col or cols argument in the data block API.
452

While you’re changing the API, perhaps these could be normalized?

def language_model_learner(data:DataBunch, bptt:int=70, emb_sz:int=400, nh:int=1150, nl:int=3, pad_token:int=1,
def text_classifier_learner(data:DataBunch, bptt:int=70, max_len:int=70*20, emb_sz:int=400, nh:int=1150, nl:int=3,
def get_tabular_learner(data:DataBunch, layers:Collection[int], emb_szs:Dict[str,int]=None, metrics=None,
def get_collab_learner(ratings:DataFrame, n_factors:int, pct_val:float=0.2, user_name:Optional[str]=None,

have get_ everywhere, or nowhere?

Also the first two could have their argument positions synced. text_classifier_learner injects max_len before other arguments - could probably go after, to stay similar.

and then we have:

def create_cnn(data:DataBunch, arch:Callable, cut:Union[int,Callable]=None, pretrained:bool=True,

it also returns a learner object, but the name is completely different. get_cnn_learner?

And this one has no action - get/create in the name:

def simple_cnn(actns:Collection[int], kernel_szs:Collection[int]=None,

and we use ‘get_’ in:

def get_embedding(ni:int,nf:int) -> nn.Module:
2 Likes
453

Nice, that was my first PR ever :smile:

4 Likes
454

Here is another questionable API:

def series2cat(df:DataFrame, *col_names):

it does in place edit, returns nothing - should it be series2cat_ instead?

455

That one will very likely disappear once I’ve refactored collab as it’s only used there.

456

awesome. and when you do also can you replace *col_names with normal list argument, so one could pass a list and not needing to expand it with *cols. Thank you.

457

After discussing with Jeremy, I changed again the names of the arguments from col to cols in the data block API when you can pass one or more columns. If you can pass only one, the name is col, if you can pass several, the name is cols. Example:

data_clas = (TextList.from_csv(imdb, 'texts.csv', cols='text')
                     .split_from_df(col='is_valid')
                     .label_from_df(cols='label'))

In the first and last function, you can pass multiple columns (if you have multiple text fields or multiple labels), but one the second one, you can only pass one column.

1 Like
459

a few datablock api questions:
i really like it in general but am not sure how to do a couple things…

  1. it’s not clear to me through the datablock api how one would change the sampler or set shuffle=False
  2. how do you add an unlabeled test set? in general, inference on outside data is confusing to me

apologies if there are examples or docs i missed.

1 Like
460
  1. All the arguments for dataloader have to be passed to the call databunch(). It delegates to DataBunch.create which doesn’t take shuffle or sampler for now. My advice would be to write a customized subclass of DataBunch for your needs (like the TextLMDataBunch or TextClasDataBunch).
  2. You add the test set with the add_test method, after the labelling. It can either take an array of items or an ItemList.
1 Like
461

Thanks

462

So to continue on the renamings:

  • get_tabular_learner is now tabular_learner
  • get_collab_learner is now collab_learner
  • get_embedding is now embedding

Args of text_classifier_learner and language_model_learner are aligned.

2 Likes