Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jart synapse #709

Open
wants to merge 86 commits into
base: master
Choose a base branch
from
Open
Changes from 3 commits
Commits
Show all changes
86 commits
Select commit Hold shift + click to select a range
ba8a2a4
1st complie test
ZhenmingYu Jul 7, 2022
06343cf
partial fix
ZhenmingYu Jul 7, 2022
589da9a
compile success, need to hide write_noise_std
ZhenmingYu Jul 8, 2022
0bdca9d
fix pybind bugs
ZhenmingYu Jul 8, 2022
37ce140
fix weight granularity
ZhenmingYu Jul 11, 2022
a695df8
tried fixing
ZhenmingYu Jul 19, 2022
a47273c
fix nan
ZhenmingYu Jul 19, 2022
de4fe74
fix w_min
ZhenmingYu Jul 20, 2022
3aaceba
CPU version working!
ZhenmingYu Jul 26, 2022
d8057d6
merge with aihwkit 0.6.0
ZhenmingYu Jul 26, 2022
51839f0
cuda-stuff backup
ZhenmingYu Jul 29, 2022
3965a0d
Store Ndisc in w_persistent_
ZhenmingYu Aug 2, 2022
ec50faa
CUDA version, not able to change pytorch weights
ZhenmingYu Aug 13, 2022
1eb6369
remove redundent parameters
ZhenmingYu Aug 14, 2022
dd37c66
Merge branch 'JART_v1b_device' into old_dev_code_with_v0.5.1
ZhenmingYu Aug 15, 2022
bacc2c3
Merge pull request #1 from ZhenmingYu/old_dev_code_with_v0.5.1
ZhenmingYu Aug 15, 2022
6cf0580
fix merge issue
ZhenmingYu Aug 15, 2022
7923177
fix minor issue
ZhenmingYu Aug 15, 2022
6d3cb56
fix merge issues
ZhenmingYu Aug 15, 2022
0a02480
add JART v1b documentations
ZhenmingYu Aug 15, 2022
cae4f8a
write_noise_std does not fit our device, removed
ZhenmingYu Aug 16, 2022
4a34388
polishing documentation for JART v1b
ZhenmingYu Aug 16, 2022
6a07a5c
change basic test name
ZhenmingYu Aug 16, 2022
c13e089
update test for cuda
ZhenmingYu Aug 16, 2022
422bdfc
Simplify calculations, CUDA somehow works
ZhenmingYu Aug 28, 2022
be48182
fix CUDA version with a delay hack
ZhenmingYu Aug 31, 2022
610d906
Add support for noise
ZhenmingYu Sep 2, 2022
e8ea3cd
improve test useability
ZhenmingYu Sep 5, 2022
bca2d2c
debug test-releted problems
ZhenmingYu Sep 7, 2022
98e8f67
add dtod bounds/ctoc noise to uniform distribution
ZhenmingYu Sep 9, 2022
d5bfbf9
update test configs and ploting scripts
ZhenmingYu Sep 11, 2022
addf5e9
ignore figures
ZhenmingYu Sep 11, 2022
b673afc
bug fix
ZhenmingYu Sep 20, 2022
b9eaf2d
update test
ZhenmingYu Sep 20, 2022
397e4df
fix ctoc variation
ZhenmingYu Oct 11, 2022
0992e53
Update test configs
ZhenmingYu Oct 12, 2022
9fcb103
Merge pull request #1 from ZhenmingYu/JART_v1b_device
ZhenmingYu Oct 12, 2022
007c138
small plot fix
ZhenmingYu Oct 13, 2022
768f496
Merge pull request #2 from ZhenmingYu/JART_v1b_device
ZhenmingYu Oct 13, 2022
4b0a1de
Include JART v1b test instructions
ZhenmingYu Oct 18, 2022
f76383f
Merge pull request #3 from ZhenmingYu/JART_v1b_device
ZhenmingYu Oct 18, 2022
e443dbf
fix bugs tyops and documentations
ZhenmingYu Oct 24, 2022
25158a2
New release prep. Changes to version number and CHANGELOG
Jan 5, 2023
5410864
Merge branch 'release/0.7.0' of https://github.com/IBM-AI-Hardware-Ce…
Jan 30, 2023
d9ba361
Merge branch 'IBM-AI-Hardware-Center-release/0.7.0'
Jan 30, 2023
55f03f3
Merge branch 'master' of https://github.com/IBM/aihwkit
Mar 20, 2023
a9bec5a
upstream CHANGELOG
Mar 20, 2023
877b41f
Merge branch 'master' of https://github.com/IBM/aihwkit
Mar 22, 2023
73a98f0
Merge branch 'master' of https://github.com/IBM/aihwkit
Mar 23, 2023
147827c
Merge branch 'master' of https://github.com/IBM/aihwkit
Mar 24, 2023
b4d1ad8
Merge branch 'master' of https://github.com/IBM/aihwkit
Mar 24, 2023
121f041
plot script
ZhenmingYu May 2, 2023
a0896c6
update documentation to include citations
ZhenmingYu May 2, 2023
b97578f
Merge pull request #4 from ZhenmingYu/JART_v1b_device
ZhenmingYu May 2, 2023
5752105
yaml_loader that holds the configuration function
ZhenmingYu May 4, 2023
740d763
fix cuda __syncthreads
ZhenmingYu May 5, 2023
1cfc864
optimizing CUDA speed
ZhenmingYu May 5, 2023
6a0b28c
Merge branch 'master' of https://github.com/IBM/aihwkit
May 30, 2023
eb29d62
Merge branch 'master' of https://github.com/IBM/aihwkit
Jun 26, 2023
ea91677
Merge branch 'master' of https://github.com/IBM/aihwkit
Jun 27, 2023
843fc0e
Merge branch 'master' of https://github.com/IBM/aihwkit
Jul 10, 2023
ba88bc2
Merge branch 'master' of https://github.com/IBM/aihwkit
Jul 10, 2023
b442780
Merge branch 'master' of https://github.com/IBM/aihwkit
Jul 14, 2023
28b5dca
Merge branch 'master' of https://github.com/IBM/aihwkit
Aug 25, 2023
7ad577a
Merge branch 'master' of https://github.com/IBM/aihwkit
Sep 14, 2023
09f9ac3
Merge branch 'master' of https://github.com/IBM/aihwkit
Sep 15, 2023
e7dcc77
Merge branch 'master' of https://github.com/IBM/aihwkit
Sep 15, 2023
66eba8b
Merge branch 'master' of https://github.com/IBM/aihwkit
Sep 18, 2023
ae66db9
Merge branch 'master' of https://github.com/IBM/aihwkit
Dec 1, 2023
921e2c4
Merge branch 'master' of https://github.com/IBM/aihwkit
Dec 8, 2023
f8b1ef3
Merge branch 'master' of https://github.com/IBM/aihwkit
Dec 15, 2023
3e21469
Merge branch 'master' of https://github.com/IBM/aihwkit
Dec 19, 2023
60febbf
Merge branch 'master' of https://github.com/IBM/aihwkit
Dec 21, 2023
0d2dd78
Merge branch 'master' of https://github.com/IBM/aihwkit
Dec 27, 2023
d6d6f31
Merge branch 'master' of https://github.com/IBM/aihwkit
Dec 28, 2023
a263359
Merge branch 'master' into HEAD
Dec 28, 2023
67c6db9
merge
Dec 28, 2023
eb08472
still needs work
Jan 4, 2024
2376df4
Merge branch 'master' into jart-synapse
kaoutar55 May 8, 2024
a1f3134
Merge branch 'master' into jart-synapse
Borjagodoy Aug 6, 2024
d65d123
fixing src/rpucuda/rpu_JART_v1b_device.h
ZhenmingYu Aug 8, 2024
49971ae
Build success?
ZhenmingYu Aug 9, 2024
aa832bd
Add examples
ZhenmingYu Jan 20, 2025
8cd563d
update examples
ZhenmingYu Jan 21, 2025
032de01
problem solved by changing to public?
ZhenmingYu Jan 22, 2025
54fcb62
Merge branch 'IBM:master' into jart-synapse
ZhenmingYu Jan 30, 2025
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
29 changes: 17 additions & 12 deletions CHANGELOG.md
Original file line number Diff line number Diff line change
@@ -12,7 +12,7 @@ The format is based on [Keep a Changelog], and this project adheres to
* `Fixed` for any bug fixes.
* `Security` in case of vulnerabilities.

## Unreleased
## [0.7.0] - 2023/01/04

### Added

@@ -28,35 +28,40 @@ The format is based on [Keep a Changelog], and this project adheres to
* Added ``SoftBoundsReferenceDevice`` for easy subtraction of symmetry
point (\#456)
* Added stand-alone functions for applying inference drift to any model (\#419)
* Example 24: analog inference and hardware-aware training on BERT (\#440)
* Example 23: how to use ``AnalogTile`` directly to implement an
* Added Example 24: analog inference and hardware-aware training on BERT (\#440)
* Added Example 23: how to use ``AnalogTile`` directly to implement an
analog matrix-vector product without using pytorch modules. (\#393)
* Example 22: 2 layer LSTM network trained on War and Peace dataset. (\#391)
* Notebook for exploring analog sensitivities. (\#380)
* Added Example 22: 2 layer LSTM network trained on War and Peace dataset. (\#391)
* Added a new notebook for exploring analog sensitivities. (\#380)
* Remapping functionality for ``InferenceRPUConfig``. (\#388)
* Inference experiment and runners. (\#410)
* Added ``analog_modules`` generator in ``AnalogSequential``. (\#410)
* Added ``SKIP_CUDA_TESTS`` to manually switch off the CUDA tests
* Added ``SKIP_CUDA_TESTS`` to manually switch off the CUDA tests.
* Enabling comparisons of ``RPUConfig`` instances. (\#410)
* Specific use-defined function for layer-wise setting for RPUConfigs
in conversions. (\#412)
* Added stochastic rounding options for ``MixedPrecisionCompound``. (\#418)
* New `remap` parameter field and functionality in
``InferenceRPUConfig`` (\#423).
* Tile-level weight getter and setter have `apply_weight_scaling`
argument (\#423)
* Pre- and post update / backward / forward methods in `BaseTile` for
argument. (\#423)
* Pre and post update / backward / forward methods in `BaseTile` for
easier user-defined modification of pre and/or post-processings of a tile. (\#423)
* Type-checking for `RPUConfig` fields. (\#424)

### Fixed

<<<<<<< HEAD
* ``RPUCuda`` backend update with many fixes (\#456)
* Missing zero-grad call in example 02 (\#446)
* Indexing error in ``OneSidedDevice`` for CPU (\#447)
=======
* Missing zero-grad call in example 02. (\#446)
* Indexing error in ``OneSidedDevice`` for CPU. (\#447)
>>>>>>> 25158a2da88b714aaa60a7672fa32d01cce14443
* Analog summary error when model is on cuda device. (\#392)
* Index error when loading the state dict with a model use previously. (\#387)
* Weights that were not contiguous could have set wrongly. (\#388)
* Weights that were not contiguous could have been set wrongly. (\#388)
* Programming noise would not be applied if drift compensation was not
used. (\#389)
* Loading a new model state dict for inference does not overwrite the noise
@@ -80,14 +85,14 @@ The format is based on [Keep a Changelog], and this project adheres to
* All configs related classes can now be imported from
``aihwkit.simulator.config`` (\#456)
* Weight noise visualization now shows the programming noise and drift
noise difference. (\#389)
noise differences. (\#389)
* Concatenate the gradients before applying to the tile update
function (some speedup for CUDA expected). (\#390)
* Drift compensation uses eye instead of ones for readout. (\#412)
* `weight_scaling_omega_columnwise` parameter in `MappingParameter` is now called
`weight_scaling_columnwise`. (\#423)
* Tile-level weight getter and setter now use Tensors instead of numpy
arrays (\#423)
arrays. (\#423)
* Output scaling and mapping scales are now distiniguished, only the
former is learnable. (\#423)
* Renamed `learn_out_scaling_alpha` parameter in `MappingParameter` to
@@ -101,7 +106,7 @@ The format is based on [Keep a Changelog], and this project adheres to
### Removed

* The `_scaled` versions of the weight getter and setter methods are
removed (\#423)
removed. (\#423)


## [0.6.0] - 2022/05/16
2 changes: 1 addition & 1 deletion src/aihwkit/VERSION.txt
Original file line number Diff line number Diff line change
@@ -1 +1 @@
0.6.2
0.7.0