Rename variables #124
Rename variables #124
Conversation
Borda
left a comment
Borda
left a comment
There was a problem hiding this comment.
I think that replacing some abbreviations to full names is better since for example tng is not a standard shortcut and may be misunderstood with tunning... :)
Anyway tests are not passing and some fixes are required
| """ | ||
| Lightning calls this inside the training loop | ||
| :param data_batch: | ||
| :param batch: |
There was a problem hiding this comment.
add variable type and description like
:param tuple(ndarray,ndarray) batch: pair of data and desired labels...
There was a problem hiding this comment.
May as well use the typing module then since this project is 3.6+ anyway, and it's natively supported, plus adding that improves things like autocomplete.
There was a problem hiding this comment.
Also, I think that's a separate PR, since this is about renaming, not adding type info.
| return nll | ||
|
|
||
| def training_step(self, data_batch, batch_i): | ||
| def training_step(self, batch, batch_i): |
There was a problem hiding this comment.
maybe also bacth_i -> batch_idx of just idx
There was a problem hiding this comment.
i -> idx is a good improvement
|
|
||
| ``` {.python} | ||
| def training_step(self, data_batch, batch_nb) | ||
| def training_step(self, batch, batch_nb) |
There was a problem hiding this comment.
in this case, I would prefer to stay with data_batch of batch_data
There was a problem hiding this comment.
let's keep these consistent. i think we agreed on batch for x_step
williamFalcon
force-pushed
the
master
branch
2 times, most recently
from
7bb2fb9 to
c4ce347
Compare
|
@alok mind fixing these conflicts so we can merge? we’ll enable this new syntax for version 0.5.0 so there’s a clean version change for potentially annoying refactoring |
- data_batch → batch - batch_i → batch_idx - dataloader_i → dataloader_idx - tng → training - training_dataloader → train_dataloader - add_log_row_interval → row_log_interval - gradient_clip → gradient_clip_val - prog → progress - tqdm_dic → tqdm_dict
3 participants
No description provided.