julius/prototorch_models
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
303 Commits 7 Branches 22 Tags 1 MiB
7d3f59e54b
Commit Graph

181 Commits

Author SHA1 Message Date
Jensun Ravichandran
e62a8e6582 [BUGFIX] Log loss in NG and GNG 2021-06-11 18:50:14 +02:00
Jensun Ravichandran
57f8bec270 Update SOM 2021-06-09 18:21:12 +02:00
Jensun Ravichandran
022d791ea5 Route initialized prototypes 2021-06-07 21:18:08 +02:00
Alexander Engelsberger
43fc7d1678 [QA] Remove unused argument from CBC 2021-06-07 21:00:58 +02:00
Jensun Ravichandran
c7b5c88776 [WIP] Add SOM 2021-06-07 18:44:15 +02:00
Jensun Ravichandran
b031382072 Update NG 2021-06-07 18:35:08 +02:00
Jensun Ravichandran
d558fa6a4a [REFACTOR] Clean up GLVQ-types 2021-06-07 17:00:38 +02:00
Jensun Ravichandran
34ffeb95bc Update how the model is printed 2021-06-07 16:58:00 +02:00
Jensun Ravichandran
bed753a6e9 Minor aesthetic change 2021-06-05 01:23:58 +02:00
Jensun Ravichandran
016fcb4060 [REFACTOR] Major cleanup 2021-06-04 22:20:32 +02:00
Jensun Ravichandran
20471bfb1c [FEATURE] Update pruning callback to re-add pruned prototypes 2021-06-04 15:56:46 +02:00
Jensun Ravichandran
42d974e08c No implicit learning rate scheduling 2021-06-04 15:55:06 +02:00
Alexander Engelsberger
47db1965ee [BUGFIX] GNG Example 2021-06-03 15:42:54 +02:00
Alexander Engelsberger
0bc385fe7b [BUGFIX] Neural Gas gets prototype intiailizer from kwargs 2021-06-03 15:24:17 +02:00
Alexander Engelsberger
358f27257d [REFACTOR] Remove prototype_initializer function from GLVQ 
Fixes #9
 2021-06-03 15:15:22 +02:00
Alexander Engelsberger
bda88149d4 [BUGFIX] Growing neural gas 2021-06-03 15:13:38 +02:00
Alexander Engelsberger
7379c61966 [BUGFIX] Fix image visualization for some parameter combination 
image visualization was broken if add_embeding was False, but data visualization was on.
 2021-06-03 15:12:51 +02:00
Alexander Engelsberger
e209bf73d5 [BUGFIX] Pruning example works on GPU now 2021-06-03 14:35:24 +02:00
Alexander Engelsberger
1b09b1d57b [BUGFIX] Probabilistic Models work on GPU now 2021-06-03 14:05:44 +02:00
Alexander Engelsberger
459f7c24be [REFACTOR] Probabilistic loss signs changed 2021-06-03 14:00:47 +02:00
Alexander Engelsberger
3b02d99ebe [BUGFIX] Early stopping example works now 2021-06-03 13:38:16 +02:00
Jensun Ravichandran
ef6bcc1079 [BUG] Early stopping does not seem to work 
The early stopping callback does not work as expected, and crashes at the end of
max_epochs with:

```
~/miniconda3/envs/py38/lib/python3.8/site-packages/pytorch_lightning/trainer/callback_hook.py in on_train_end(self)
    155         """Called when the train ends."""
    156         for callback in self.callbacks:
--> 157             callback.on_train_end(self, self.lightning_module)
    158
    159     def on_pretrain_routine_start(self) -> None:

~/work/repos/prototorch_models/prototorch/models/callbacks.py in on_train_end(self, trainer, pl_module)
     18     def on_train_end(self, trainer, pl_module):
     19         # instead, do it at the end of training loop
---> 20         self._run_early_stopping_check(trainer, pl_module)
     21
     22

TypeError: _run_early_stopping_check() takes 2 positional arguments but 3 were given
```
 2021-06-02 12:44:34 +02:00
Jensun Ravichandran
8851d1bbc9 [FEATURE] Add PruneLoserPrototypes Callback 2021-06-02 03:52:41 +02:00
Jensun Ravichandran
8f7deb75dd [FEATURE] Log prototype win ratios over all training batches 2021-06-02 02:32:54 +02:00
Jensun Ravichandran
7743c50725 Tweak repr 2021-06-02 01:07:48 +02:00
Jensun Ravichandran
fcf3a4979c Remove unused import 2021-06-02 00:49:36 +02:00
Jensun Ravichandran
d46fe4a393 [WIP] Update CBC example 2021-06-02 00:45:33 +02:00
Jensun Ravichandran
88cfd5762e Remove unused imports in models/cbc.py 2021-06-02 00:44:35 +02:00
Jensun Ravichandran
aa42b9e331 [BUGFIX] Import missing module 
models/unsupervised.py uses `pt` in line 37, but `pt` is undefined in the file.
I wonder why Python doesn't complain about this. Perhaps because unsupervised.py
is never run in isolation and `pt` is otherwise available in the namespace of
the example scripts that use unsupervised.py.
 2021-06-02 00:31:57 +02:00
Jensun Ravichandran
91b57b01b1 [REFACTOR] neighbour -> neighbor 2021-06-02 00:29:45 +02:00
Jensun Ravichandran
98c198d463 [REFACTOR] Use LambdaLayer instead of EuclideanDistance 2021-06-02 00:21:11 +02:00
Jensun Ravichandran
757f4e980d Add Local-Matrix LVQ 
Also remove the use of `self.distance_fn` in favor of `self.distance_layer`.
 2021-06-01 23:44:16 +02:00
Jensun Ravichandran
5ec2dd47cd Remove unused import 2021-06-01 23:40:56 +02:00
Jensun Ravichandran
e8cd4d765c Remove unused variable 2021-06-01 23:39:39 +02:00
Jensun Ravichandran
8403b01081 Move CELVQ to probabilistic.py 2021-06-01 23:39:06 +02:00
Jensun Ravichandran
aff6aedd60 Use the add_components API for adding prototypes 2021-06-01 23:37:45 +02:00
Jensun Ravichandran
1b6843dbbb Remove unused imports 2021-06-01 19:31:03 +02:00
Jensun Ravichandran
21023a88d7 [BUGFIX] Fix RSLVQ 2021-06-01 17:44:10 +02:00
Alexander Engelsberger
9c1a41997b [FEATURE] Add Growing Neural Gas 2021-06-01 17:19:43 +02:00
Jensun Ravichandran
1636c84778 Rename rslvq example 2021-05-31 17:56:45 +02:00
Jensun Ravichandran
27eccf44d4 Use LambdaLayer from ProtoTorch 2021-05-31 16:53:04 +02:00
Alexander Engelsberger
2a218c0ede Add example for dynamic components in callbacks 2021-05-31 11:39:24 +02:00
Alexander Engelsberger
db064b5af1 Improvement of model __repr__ 2021-05-31 11:19:06 +02:00
Alexander Engelsberger
0ac4ced85d [refactor] Use functional variant of accuracy 
Prevents Accuracy in `__repr__` of the models.
 2021-05-31 11:12:27 +02:00
Jensun Ravichandran
7b7bc3693d Merge branch 'dev' of github.com:si-cim/prototorch_models into dev 2021-05-31 00:32:49 +02:00
Jensun Ravichandran
cd73f6c427 Add examples/dynamic_components.py 2021-05-31 00:32:27 +02:00
Alexander Engelsberger
a60337ff27 [refactor] Move probabilistic to Prototorch 2021-05-28 20:39:32 +02:00
Alexander Engelsberger
e3392ee952 [refactor] DRY Probabilistic models 2021-05-28 17:13:06 +02:00
Alexander Engelsberger
dade502686 Add MNIST datamodule and training mixin factory. 2021-05-28 16:33:31 +02:00
Jensun Ravichandran
b7edee02c3 [WIP] Add CELVQ 
TODO Ensure that the distances/probs corresponding to the plabels are sorted
like the target labels.
 2021-05-27 17:40:16 +02:00
Alexander Engelsberger
66e3e51a52 Add references to the documentation. 2021-05-26 21:20:17 +02:00
Jensun Ravichandran
d411e52be4 Refactor non-gradient-lvq models into lvq.py 2021-05-25 20:37:34 +02:00
Alexander Engelsberger
32d6f95db0 Add RSLVQ and LikelihoodLVQ 2021-05-25 20:26:15 +02:00
Jensun Ravichandran
139109804f [BUGFIX] Use _forward in LVQ1 and LVQ21 2021-05-25 17:43:37 +02:00
Alexander Engelsberger
72e064338c Use 'num_' in all variable names 2021-05-25 15:41:10 +02:00
Alexander Engelsberger
8ce18f83ce Add prototype_initializer function to GLVQ 
This allows overwriting it inside subclasses.
 2021-05-21 17:13:10 +02:00
Alexander Engelsberger
7b4f7d84e0 Update Documentation 
Clean up project
 2021-05-21 15:42:45 +02:00
Jensun Ravichandran
a5e086ce0d Refactor code 2021-05-21 13:33:57 +02:00
Jensun Ravichandran
0611f81aba Update models namespace 2021-05-21 13:11:59 +02:00
Jensun Ravichandran
a9382dcd9b Add get_prototype_grid method 2021-05-21 13:11:48 +02:00
Jensun Ravichandran
0933a88a1b Fix ImageCBC bug 2021-05-21 13:11:36 +02:00
Jensun Ravichandran
88a34a06ef [WIP] Update CBC implementation to use SiameseGLVQ 2021-05-20 17:36:00 +02:00
Jensun Ravichandran
de63eaf15a Fix numpy issue in vis.py 2021-05-20 17:33:19 +02:00
Jensun Ravichandran
16dc3cf4eb Update image visualization 2021-05-20 16:07:16 +02:00
Jensun Ravichandran
df061cc2ff Refactor code 2021-05-20 14:40:02 +02:00
Alexander Engelsberger
969fb34cc3 Accumulate test loss 2021-05-20 14:20:23 +02:00
Alexander Engelsberger
0204f5eab6 Log test accuracy. 2021-05-20 14:03:31 +02:00
Alexander Engelsberger
b7fc5df386 Log test loss. 2021-05-20 13:47:20 +02:00
Alexander Engelsberger
faf1a88f99 [Bugfix] Remove optimzer_idx from validation and test. 2021-05-20 13:17:27 +02:00
Jensun Ravichandran
5ffbd43a7c Refactor into shared_step 2021-05-19 16:57:51 +02:00
Jensun Ravichandran
fdf9443a2c Add validation and test logic 2021-05-19 16:30:19 +02:00
Jensun Ravichandran
eefec19c9b Custom non-gradient training 2021-05-18 19:49:16 +02:00
Jensun Ravichandran
4957e821f6 Close matplotlib figure on train end 2021-05-18 10:13:22 +02:00
Jensun Ravichandran
81346785bd Cleanup models 
Siamese architectures no longer accept a `backbone_module`. They have to be
initialized with an pre-initialized backbone object instead. This is so that the
visualization callbacks could use the very same object for visualization
purposes. Also, there's no longer a dependent copy of the backbone. It is
managed simply with `requires_grad` instead.
 2021-05-17 17:00:23 +02:00
Jensun Ravichandran
7a87636ad7 Update KNN 2021-05-17 16:59:35 +02:00
Jensun Ravichandran
77b7b59bad Clean visualization callbacks 2021-05-17 16:59:22 +02:00
Jensun Ravichandran
6e7d80be88 [BUGFIX] Fix siamese visualization callback 2021-05-15 12:52:44 +02:00
Jensun Ravichandran
b7684ae512 predict_latent no longer returns numpy 2021-05-15 12:52:16 +02:00
Alexander Engelsberger
0eac2ce326 Examples use GPUs if available. 2021-05-13 15:22:01 +02:00
Jensun Ravichandran
8f9c29bd2b [BUGFIX] Remove incorrect import statement 2021-05-12 16:45:01 +02:00
Jensun Ravichandran
ca39aa00d5 Stop passing component initializers as hparams 
Pass the component initializer as an hparam slows down the script very much. The
API has now been changed to pass it as a kwarg to the models instead.

The example scripts have also been updated to reflect the new changes.

Also, ImageGMLVQ and an example script `gmlvq_mnist.py` that uses it have also
been added.
 2021-05-12 16:36:22 +02:00
Alexander Engelsberger
1498c4bde5 Bump version: 0.1.6 → 0.1.7 2021-05-11 17:18:29 +02:00
Jensun Ravichandran
59b8ab6643 Add knn 2021-05-11 17:22:02 +02:00
Jensun Ravichandran
eab1ec72c2 Change optimizer using kwargs 2021-05-11 16:13:00 +02:00
Jensun Ravichandran
b38acd58a8 [BUGFIX] Fix visualization callbacks bug 2021-05-11 16:09:27 +02:00
Alexander Engelsberger
e87563e10d Bump version: 0.1.5 → 0.1.6 2021-05-11 13:41:26 +02:00
Alexander Engelsberger
3fa6378c4d Add LVQ1 and LVQ2.1 Models. 2021-05-11 13:26:13 +02:00
Alexander Engelsberger
30ee287ecc Bump version: 0.1.4 → 0.1.5 2021-05-10 17:13:00 +02:00
Alexander Engelsberger
f49db0bf2c Bump version: 0.1.3 → 0.1.4 2021-05-10 17:06:28 +02:00
Alexander Engelsberger
54a8494d86 Bump version: 0.1.2 → 0.1.3 2021-05-10 17:04:20 +02:00
Alexander Engelsberger
bf310be97c Bump version: 0.1.1 → 0.1.2 2021-05-10 16:47:33 +02:00
Alexander Engelsberger
1ae2b41edd Bump version: 0.1.0 → 0.1.1 2021-05-10 16:26:21 +02:00
Alexander Engelsberger
f6e3a37e2b Bump version: 0.0.0 → 0.1.0 2021-05-10 16:01:46 +02:00
Alexander Engelsberger
6873927349 Setup bumpversion 2021-05-10 15:34:43 +02:00
Jensun Ravichandran
49100f43f5 Example to save and reload a model 2021-05-10 14:30:02 +02:00
Jensun Ravichandran
ed03ab168e [BUGFIX] Fix lambda_matrix property in GMLVQ 2021-05-10 14:09:25 +02:00
Jensun Ravichandran
c6e06ceaa4 Properly initialize prototypes in LVQMLN 2021-05-09 20:55:28 +02:00
Jensun Ravichandran
ca4c9da10a Add the namespace hook for GMLVQ in the model class 2021-05-09 20:53:31 +02:00
Jensun Ravichandran
ff7a1e93d2 Refactor visualization callbacks 2021-05-09 20:53:03 +02:00
Jensun Ravichandran
11b3e53ecb Return prototypes as torch tensor 2021-05-07 15:45:37 +02:00