-
# Replicate the issue
1. git clone fastai2 and nbdev and fastcore
2. editable install the latest fastai2 and nbdev and fastcore
3. `nbdev_install_git_hooks` in new cloned fastai2
4. `nbdev_build_l…
-
I've been trying to use Optuna with fastai. When I run the study, I get the following values
```
[I 2020-03-03 19:37:32,469] Finished trial#44 resulted in value: 0.5143198370933533. Current best v…
-
I have a Django app running locally when I tried deploying it doesn't work on apache unless I comment the import of the library ptporch/fastai in my views.
the app works locally on the same environme…
-
@jjhelmus it seems you were able to build pytorch GPU without needing to have variants
https://anaconda.org/anaconda/pytorch/files?version=1.0.1
Is that true?
If so, what challenges do you se…
-
hello, i was following the tutorial. and in this code give me an error
```
trn_dl = LanguageModelLoader(pd.DataFrame(trn_lm,columns=['x']), bs, bptt)
val_dl = LanguageModelLoader(pd.DataFrame(trn…
-
**Describe the bug**
Expected input batch_size (17664) to match target batch_size (32)
**Provide your installation details**
```text
=== Software ===
python : 3.7.3
f…
-
I understand the desire to use abstraction to make the library easier to use out-of-the-box, especially for newer users who just want to call run_model.py with some hyparameters.
However, as someon…
-
**Describe the bug**
`get_preds(Test)` outputs zero category for all the items in the dataset. Individual `predict()` is ok.
Probabilities returned by both do match.
**Provide your …
-
## Describe the proposal
It'd be nice to add an mlflow.pytorch.autolog() API for automatically logging metrics, params & models generated via PyTorch. In particular, I'd like this to be incorporate…
-
Hi,
First of all, thanks for the great work!
I adapted the image by installing conda on top of it. I need conda to install the fast.ai library and all its dependencies. I do this on the base conda e…