Skip to content

Commit

Permalink
Merge remote-tracking branch 'origin/master'
Browse files Browse the repository at this point in the history
  • Loading branch information
@YSaxon
YSaxon committed Sep 16, 2020
2 parents 3b9c7fe + d2f9402 commit c629414
Show file tree
Hide file tree
Showing 3 changed files with 131 additions and 117 deletions.
236 changes: 123 additions & 113 deletions docs_src/sidebar.json
View file
Original file line number Diff line number Diff line change
@@ -1,118 +1,128 @@
{
"Getting started": {
"Overview": "/",
"Quick start": "/quick_start.html"
},
"Tutorials": {
"Overview": "/tutorial",
"": {
"Beginners": {
"Vision tutorial": "/tutorial.vision",
"Text tutorial": "/tutorial.text",
"Tabular tutorial": "/tutorial.tabular",
"Collab tutorial": "/tutorial.collab"
}
"Getting started": {
"Overview": "/",
"Quick Start": "/quick_start.html"
},
"empty1": {
"Intermediate": {
"Data block tutorial": "/tutorial.datablock",
"Imagenette tutorial": "/tutorial.imagenette",
"Pets tutorial": "/tutorial.pets",
"Transformers tutorial": "/tutorial.transformers",
"Wikitext tutorial": "/tutorial.wikitext"
}
"Tutorials": {
"Overview": "/tutorial",
"": {
"Beginner": {
"Vision Tutorial": "/tutorial.vision",
"Text Tutorial": "/tutorial.text",
"Tabular Tutorial": "/tutorial.tabular",
"Collab Tutorial": "/tutorial.collab"
}
},
"empty1": {
"Intermediate": {
"DataBlock Tutorial": "/tutorial.datablock",
"Imagenette Tutorial": "/tutorial.imagenette",
"Medical Imaging Tutorial": "/tutorial.medical_imaging",
"Pets Tutorial": "/tutorial.pets",
"Transformers Tutorial": "/tutorial.transformers",
"Wikitext Tutorial": "/tutorial.wikitext"
}
},
"empty2": {
"Advanced": {
"Albumentions Tutorial": "/tutorial.albumentions",
"Siamese Tutorial": "/tutorial.siamese"
}
},
"empty3": {
"Migrating from Other Libraries": {
"Plain PyTorch": "/migrating_pytorch",
"Ignite": "/migrating_ignite",
"Lightning": "/migrating_lightning",
"Catalyst": "/migrating_catalyst"
}
}
},
"empty2": {
"Advanced": {
"Siamese tutorial": "/tutorial.siamese"
}
"Training": {
"Training Loop": "/learner",
"Optimizer": "/optimizer",
"Metrics": "/metrics",
"Interpretation": "/interpret",
"Distributed": "/distributed",
"": {
"Callbacks": {
"Core": "/callback.core",
"Hooks": "/callback.hook",
"Progress": "/callback.hook",
"Schedulers": "/callback.schedule",
"CutMix": "/callback.cutmix",
"Data": "/callback.data",
"Mixed Precision": "/callback.fp16",
"MixUp": "/callback.mixup",
"Predictions": "/callback.preds",
"RNN": "/callback.rnn",
"Tracker": "/callback.tracker",
"Training": "/callback.training"
}
}
},
"empty3": {
"Migrating from other libraries": {
"Plain PyTorch": "/migrating_pytorch",
"Ignite": "/migrating_ignite",
"Lightning": "/migrating_lightning",
"Catalyst": "/migrating_catalyst"
}
}
},
"Training": {
"Training loop": "/learner",
"Optimizer": "/optimizer",
"Metrics": "/metrics",
"Interpretation": "/interpret",
"Distributed": "/distributed",
"": {
"Callbacks": {
"Core": "/callback.core",
"Schedulers": "/callback.schedule",
"Hooks and callbacks": "/callback.hook",
"Mixed precision": "/callback.fp16",
"Mixup": "/callback.mixup",
"Cutmix": "/callback.cutmix",
"Tracker": "/callback.tracker",
"Training": "/callback.training",
"RNN": "/callback.rnn",
"Data": "/callback.data",
"Progress": "/callback.progress"
}
}
},
"Data": {
"Data Blocks": "/data.block",
"Data Transforms": "/data.transforms",
"Data External": "/data.external",
"Data Core": "/data.core",
"DataLoader": "/data.load"
},
"Core": {
"PyTorch Core": "/torch_core",
"Layers": "/layers"
},
"Vision": {
"Vision Learner": "/vision.learner",
"Vision Data Augmentation": "/vision.augment",
"Vision Data": "/vision.data",
"Vision Core": "/vision.core",
"Vision Utils": "/vision.utils",
"Vision Widgets": "/vision.widgets",
"GAN": "/vision.gan",
"": {
"Models": {
"XResnet": "/vision.models.xresnet",
"Unet": "/vision.models.unet"
}
}
},
"Text": {
"Text Core": "/text.core",
"Text Data": "/text.data",
"Text Learner": "/text.learner",
"": {
"Models": {
"Core": "/text.models.core",
"AWD LSTM": "/text.models.awdlstm",
"QRNN": "/text.models.qrnn"
}
"Data": {
"Blocks": "/data.block",
"Core": "/data.core",
"DataLoader": "/data.load",
"External": "/data.external",
"Transforms": "/data.transforms"
},
"Core": {
"PyTorch Core": "/torch_core",
"Layers": "/layers"
},

"Vision": {
"Core": "/vision.core",
"Data": "/vision.data",
"Data Augmentation": "/vision.augment",
"Learner": "/vision.learner",
"": {
"Models": {
"XResnet": "/vision.models.xresnet",
"Unet": "/vision.models.unet"
}
},
"GAN": "/vision.gan",
"Utilities": "/vision.utils",
"Widgets": "/vision.widgets"
},
"Text": {
"Core": "/text.core",
"Data": "/text.data",
"Learner": "/text.learner",
"": {
"Models": {
"Core": "/text.models.core",
"AWD LSTM": "/text.models.awdlstm",
"QRNN": "/text.models.qrnn"
}
}
},
"Tabular": {
"Core": "/tabular.core",
"Data": "/tabular.data",
"Learner": "/tabular.learner",
"Model": "/tabular.model",
"Collab": "/collab"
},
"Medical": {
"Imagery": "/medical.imaging",
"Text": "/medical.text"
},

"Integrations": {
"Neptune": "/callback.neptune",
"TensorBoard": "/callback.tensorboard",
"Weights & Biases": "/callback.wandb"
},
"fastai Development": {
"Developers Guide": "/dev-setup",
"Abbreviation Guide": "/dev/abbr",
"Style Guide": "/dev/style",
"GPU Notes": "/dev/gpu",
"Git Notes": "/dev/git",
"Notes": "/dev/develop"
}
},
"Tabular": {
"Tabular Core": "/tabular.core",
"Tabular Data": "/tabular.data",
"Tabular Learner": "/tabular.learner",
"Tabular Model": "/tabular.model",
"Collab": "/collab"
},
"Medical": {
"Medical imagery": "/medical.imaging",
"Medical text": "/medical.text"
},
"fastai development": {
"Developers guide": "/dev-setup",
"Abbreviation guide": "/dev/abbr",
"Style guide": "/dev/style",
"GPU Notes": "/dev/gpu",
"git notes": "/dev/git",
"Notes": "/dev/develop"
}
}
}
6 changes: 4 additions & 2 deletions fastai/text/learner.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -207,7 +207,8 @@ def language_model_learner(dls, arch, config=None, drop_mult=1., backwards=False
warn("There are no pretrained weights for that architecture yet!")
return learn
model_path = untar_data(meta[url] , c_key='model')
fnames = [list(model_path.glob(f'*.{ext}'))[0] for ext in ['pth', 'pkl']]
try: fnames = [list(model_path.glob(f'*.{ext}'))[0] for ext in ['pth', 'pkl']]
except IndexError: print(f'The model in {model_path} is incomplete, download again'); raise
learn = learn.load_pretrained(*fnames)
return learn

Expand All @@ -230,7 +231,8 @@ def text_classifier_learner(dls, arch, seq_len=72, config=None, backwards=False,
warn("There are no pretrained weights for that architecture yet!")
return learn
model_path = untar_data(meta[url], c_key='model')
fnames = [list(model_path.glob(f'*.{ext}'))[0] for ext in ['pth', 'pkl']]
try: fnames = [list(model_path.glob(f'*.{ext}'))[0] for ext in ['pth', 'pkl']]
except IndexError: print(f'The model in {model_path} is incomplete, download again'); raise
learn = learn.load_pretrained(*fnames, model=learn.model[0])
learn.freeze()
return learn
Expand Down
6 changes: 4 additions & 2 deletions nbs/37_text.learner.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -580,7 +580,8 @@
" warn(\"There are no pretrained weights for that architecture yet!\")\n",
" return learn\n",
" model_path = untar_data(meta[url] , c_key='model')\n",
" fnames = [list(model_path.glob(f'*.{ext}'))[0] for ext in ['pth', 'pkl']]\n",
" try: fnames = [list(model_path.glob(f'*.{ext}'))[0] for ext in ['pth', 'pkl']]\n",
" except IndexError: print(f'The model in {model_path} is incomplete, download again'); raise\n",
" learn = learn.load_pretrained(*fnames)\n",
" return learn"
]
Expand Down Expand Up @@ -714,7 +715,8 @@
" warn(\"There are no pretrained weights for that architecture yet!\")\n",
" return learn\n",
" model_path = untar_data(meta[url], c_key='model')\n",
" fnames = [list(model_path.glob(f'*.{ext}'))[0] for ext in ['pth', 'pkl']]\n",
" try: fnames = [list(model_path.glob(f'*.{ext}'))[0] for ext in ['pth', 'pkl']]\n",
" except IndexError: print(f'The model in {model_path} is incomplete, download again'); raise\n",
" learn = learn.load_pretrained(*fnames, model=learn.model[0])\n",
" learn.freeze()\n",
" return learn"
Expand Down

0 comments on commit c629414

Please sign in to comment.