OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-07 00:21:07 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity
16,718 Commits 8,304 Branches 150 Tags 1.5 GiB
54b33503ec
Commit Graph

10 Commits

Author SHA1 Message Date
Orion Reblitz-Richardson
1d5780d42c Remove Apache headers from source. 
* LICENSE file contains details, so removing from individual source files.
 2018-03-27 13:10:18 -07:00
Kutta Srinivasan
ed05ca9fec Clean up naming of FP16-related code, add comments 2018-03-09 13:51:42 -08:00
Yangqing Jia
8286ce1e3a Re-license to Apache 
Summary: Closes https://github.com/caffe2/caffe2/pull/1260

Differential Revision: D5906739

Pulled By: Yangqing

fbshipit-source-id: e482ba9ba60b5337d9165f28f7ec68d4518a0902
 2017-09-28 16:22:00 -07:00
Luke Yeager
2579be1227 Skip fp16 initializer test for CPU-only builds 
Summary:
Working towards https://github.com/caffe2/caffe2/pull/817.
```
E           AttributeError: Method FloatToHalf is not a registered operator. Did you mean: []
```
https://travis-ci.org/caffe2/caffe2/jobs/243867951

/cc slayton58
Closes https://github.com/caffe2/caffe2/pull/829

Differential Revision: D5276796

Pulled By: akyrola

fbshipit-source-id: 34edca6090a9ce7ab39ae1fdc0e83b5c3b7e4f49
 2017-06-19 12:21:25 -07:00
Andrey Malevich
e05173a476 Create ExternalInitializer to simplify logic around init_params = False 
Summary:
This diff is creating new type of Initializer - ExternalInitializer. This
initializer is supposed to be used in cases when the parameter blob is already
expected to exist in the workspace.

Reviewed By: dzhulgakov

Differential Revision: D5171322

fbshipit-source-id: d27861f0f80afdea93c235d49f63da19adccc92c
 2017-06-02 18:22:50 -07:00
Simon Layton
58874ad5bf Fp16 training initializers 
Summary:
Re-open for re-importing :)
Closes https://github.com/caffe2/caffe2/pull/721

Differential Revision: D5164345

Pulled By: akyrola

fbshipit-source-id: e80b32556cd25610602df91a4225b93edc0ca40b
 2017-06-01 08:34:46 -07:00
Aapo Kyrola
0f8c8f37a8 Revert D5159712: [caffe2][PR] Fp16 training initializers 
Summary: This reverts commit 60a889494d2e2f4df1d720331e19f638c5eb95cc

Differential Revision: D5159712

fbshipit-source-id: 16040c911b260648857f656f92b165f92c2daae0
 2017-06-01 00:17:14 -07:00
Simon Layton
2bfacff426 Fp16 training initializers 
Summary:
Adds support for generating and training pfp16 models. Added SGD optimizer for multi-precision trainers and a new callback to data_parallel_model in order to help multi-precision models keep their different copies of parameters in sync during training.
Closes https://github.com/caffe2/caffe2/pull/697

Differential Revision: D5159712

Pulled By: salexspb

fbshipit-source-id: 60a889494d2e2f4df1d720331e19f638c5eb95cc
 2017-05-31 17:46:58 -07:00
Simon Layton
2c3071fc4e Rework initializers to pass a class not object 
Summary:
Changed tests
Moved to WeightInitializer, BiasInitializer keywords
Closes https://github.com/caffe2/caffe2/pull/682

Reviewed By: Yangqing

Differential Revision: D5138769

Pulled By: salexspb

fbshipit-source-id: 81d266100b2a95c64c0196c16670dfd34ea03e02
 2017-05-30 09:06:56 -07:00
Alexander Sidorov
016f72537a ModelHelper.create_param, Initializer abstraction and ParameterInfo for optimizers 
Summary:
This is going to unblock Nvidia in their work on adding fp16
support to Caffe2. I discussed this with kennyhorror before to make
sure this fits into his work on parameter sharing.

Reviewed By: kennyhorror

Differential Revision: D5127797

fbshipit-source-id: 4db155d320b1862570c23b77c4252bdacbf2296f
 2017-05-25 22:03:15 -07:00