dnrops
/
burn
mirror of https://github.com/tracel-ai/burn.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
873 Commits 39 Branches 19 Tags 101 MiB
Commit Graph

330 Commits

Author SHA1 Message Date
polina guseva 64090a582b
Add static full method for tensor intialization with custom values (#486) 
---------

Co-authored-by: Dilshod Tadjibaev <939125+antimora@users.noreply.github.com>
 2023-07-24 13:03:27 -04:00
Louis Fortier-Dubois 9aca1837c2
Example/wgpu/mnist (#514) 
* add wgpu for mnist

* auto graphics api

* fix display tests

* clipy
 2023-07-20 17:12:13 -04:00
Nathaniel Simard d7ce52f0da
Feat/wgpu/conv (#512) 2023-07-20 15:14:42 -04:00
Louis Fortier-Dubois 4b60c0e7a0
continuous to contiguous (#511) 2023-07-20 11:28:35 -04:00
Nathaniel Simard c4afff182f
Feat/wgpu/max pool2d (#500) 2023-07-14 13:58:08 -04:00
Dilshod Tadjibaev ef421f0ae9
Add arange with steps op for int tensor (#490) 2023-07-13 17:09:57 -04:00
Nathaniel Simard a2ac2057d8
Fix: wgpu scatter with different shapes (#489) 2023-07-12 13:21:19 -04:00
Jacob cb1ac26bb3
Add SiLU activation function (#482) 2023-07-09 11:28:14 -04:00
Nathaniel Simard 04ad14a32a
refactor: wgpu reductions (#471) 2023-07-06 11:40:37 -04:00
Dilshod Tadjibaev e62ee1269b
Fix burn-tch's random implementation for standard dist (#469) 2023-07-06 08:50:50 -04:00
Nathaniel Simard 65bf6c1cbb
Refactor index => slice (#466) 2023-07-05 16:30:11 -04:00
Nathaniel Simard 98d4abd9c0
Perf/wgpu/comparison (#460) 2023-07-04 09:33:05 -04:00
Nathaniel Simard a1c9970373
Refactor/wgpu/mask (#453) 2023-07-02 12:59:22 -04:00
Nathaniel Simard 77219f7262
Perf/unary binary (#446) 2023-06-29 08:56:35 -04:00
Louis Fortier-Dubois f99fe0fadd
Matmul 2D Tiling (#442) 2023-06-28 16:48:15 -04:00
Nathaniel Simard 13836ac599
Fix: wgpu powf (#443) 2023-06-28 12:50:16 -04:00
Nathaniel Simard dc216f574b
fix ndarray index select assign (#441) 2023-06-27 12:48:01 -04:00
Nathaniel Simard d26dead299
fix arange (#440) 2023-06-27 12:47:49 -04:00
Dilshod Tadjibaev 61291cc30c
Revert temp half-rs fix for no_std (#430) 
Fixes #268
 2023-06-26 15:59:06 -04:00
Nathaniel Simard 6c834d3b22
Feat/wgpu/index select (#425) 2023-06-23 09:31:49 -04:00
Dilshod Tadjibaev 825aaa9977
Add missing documents (#424) 2023-06-23 09:28:34 -04:00
Nathaniel Simard c4e4c25fef
Feat/wgpu/mask (#422) 2023-06-21 13:01:18 -04:00
Yu Sun 9c5afa3469
fix(loss): Declare all contents of reduction as public (#423) 2023-06-21 13:00:57 -04:00
Dilshod Tadjibaev fce45f51be
Doc fixes (#418) 2023-06-21 12:32:50 -04:00
Nathaniel Simard 3b9c997513
feat: provide embedding default impl (#417) 2023-06-20 16:16:31 -04:00
Nathaniel Simard b6b684b9f2
Feat/wgpu/cat (#416) 2023-06-20 16:16:10 -04:00
Nathaniel Simard 4d40bde7b9
feat: argmax + argmin (#412) 2023-06-20 10:03:00 -04:00
Nathaniel Simard 323261b594
Feat/wgpu/comparisons (#411) 2023-06-19 19:11:49 -04:00
Louis Fortier-Dubois e5cc50a837
refactor/initializer (#398) 
* add tests to linear forward

* swap linear weight dims

* rename uniformdefault

* wip bias in init

* refactor initializer

* formatting

* fix get(0) to first

* wip refactor

* builder pattern + kaiming init

* kaiming initializer and major initializer refactor

* fix fan out

* easier fixes

* remove initializer options

* clippy fix

* revert to input*output linear

* gru uses xavier init

* should be normal actually

---------

Co-authored-by: louisfd <louisfortier-dubois@Louiss-MacBook-Pro.local>
 2023-06-19 16:55:11 -04:00
Nathaniel Simard 40743c477a
Feat/onnx/more ops (#410) 2023-06-19 16:10:50 -04:00
Nathaniel Simard a8624590af
feat: mask_where (#409) 2023-06-18 15:04:28 -04:00
Dilshod Tadjibaev 834c7ecc1f
Clean up cargo descriptions and formatting (#403) 2023-06-15 09:20:53 -04:00
Louis Fortier-Dubois fd88398ce4
feat/Xavier glorot (#394) 2023-06-11 09:54:44 -04:00
Nathaniel Simard 8c9802c363
Feat/wgpu/reduction (#392) 2023-06-08 16:54:36 -04:00
Nathaniel Simard cb4f049eae
fix: improve assert approx (#390) 2023-06-06 14:34:20 -04:00
Mathias Insley 8a88a868ee
Feat/lstm (#370) 2023-06-06 14:33:22 -04:00
Nathaniel Simard bff752b1a8
Feat/wgpu/index (#386) 2023-06-06 12:29:14 -04:00
Nathaniel Simard c1e1e38a79
Add more ops (#387) 2023-06-06 12:21:20 -04:00
Nathaniel Simard ecc67c58f9
Feat/wgpu/swap dims (#381) 2023-06-04 19:34:35 -04:00
Nathaniel Simard 0a205a3603
feat: add basic matmul for wgpu backend (#379) 2023-06-03 10:32:57 -04:00
Nathaniel Simard 974fdfaba1
Feat/wgpu backend setup (#376) 2023-06-02 11:52:47 -04:00
Nathaniel Simard 2a4ba5a6ab
Feat/gather scatter (#367) 2023-05-27 11:40:04 -04:00
Nathaniel Simard ba0ac112e1
Refactor: Burn Import use BurnGraph (#355) 2023-05-21 13:09:27 -04:00
Nathaniel Simard 510d53f37e
Improve softmax doc (#352) 2023-05-16 14:39:42 -04:00
Nathaniel Simard 976102fec0
Feat/avg pool1d (#349) 2023-05-15 08:29:45 -04:00
Nathaniel Simard 8bbfbd8a8f
perf: softmax (#346) 2023-05-12 09:37:42 -04:00
Dilshod Tadjibaev 05763e1878
Bump version to the next minor to indicate dev (#344) 2023-05-10 18:02:08 -04:00
Nathaniel Simard 73f99ef79f
Feat/maxmin numeric (#340) 2023-05-09 16:35:55 -04:00
Nathaniel Simard a88357ce1d
feat: add max & min ops (#339) 2023-05-09 15:49:33 -04:00
Nathaniel Simard 69001b0d69
Feat/activation ops (#338) 
* perf: GELU

* Refactor relu
 2023-05-09 08:32:35 -04:00
Nathaniel Simard 29eecd6383
Prepare next release (#335) 2023-05-06 10:32:23 -04:00
Nathaniel Simard b54f9302c7
Feat/avg pool2d (#318) 2023-04-30 12:25:14 -04:00
Sunny Gonnabathula 02abc373d3
add burn-tch support for bf16 (#303) 2023-04-24 11:24:36 -04:00
Nathaniel Simard c5e31b272f
Feat/group conv (#306) 2023-04-22 15:00:41 -04:00
Nathaniel Simard 78ac09fb7a
Support dilation in convolution operations (#301) 2023-04-18 10:01:11 -04:00
Nathaniel Simard bd58922784
Feat/conv stride (#300) 2023-04-16 22:03:16 -04:00
Yu Sun 26cf555612
Feat: add new tensor ops mask_scatter (#258) 2023-04-13 08:58:31 -04:00
Sunny Gonnabathula 69954c14ec
add bf16 element (#295) 2023-04-12 12:36:03 -04:00
Nathaniel Simard 04bcf9550a
Fix/text gen example (#292) 2023-04-11 17:18:45 -04:00
Nathaniel Simard 2220965b5c
Feat/add tensor checks (#283) 2023-04-10 19:16:15 -04:00
Nathaniel Simard f04fe101d8
Feat/module no grad (#274) 2023-04-07 09:01:27 -04:00
Mathias Insley d8f64ce1dd
Pretty Print Tensors (#257) 2023-04-06 20:06:38 -04:00
Dilshod Tadjibaev 4e9e6d2706
Move unsqueeze op to the tensor's base (#261) 2023-04-01 14:20:48 -04:00
Dilshod Tadjibaev 7364d09d32
Add flatten op to the tensor base (#260) 2023-03-31 16:44:48 -04:00
Nathaniel Simard 7d7504686a
Refactor/init modules (#250) 2023-03-24 15:38:02 -04:00
Nathaniel Simard a74e4cd0bc
fix: conv bias backward (#248) 2023-03-23 16:10:26 -04:00
Nathaniel Simard 6f43d983f7
State serialization/deserialization overhaul (#247) 2023-03-23 11:02:46 -04:00
nathaniel 00625d1527 fix: add version to path dependencies 2023-03-21 10:13:44 -04:00
Nathaniel Simard 4e28e2a776
chore: prepare release v0.6.0 (#246) 2023-03-21 09:47:37 -04:00
Dilshod Tadjibaev bf9d33e6fc
Add MNIST inference on the web demo crate (#228) 
* Add MNIST inference on the web demo crate

* Fix problems identified during a PR review
 2023-03-13 19:51:32 -04:00
Nathaniel Simard 403500f018
feat(burn-core): refactor cross entropy (#229) 2023-03-13 13:47:19 -04:00
Nathaniel Simard d09ab44979
Feat/index_select (#227) 2023-03-12 17:44:22 -04:00
Nathaniel Simard 9655b74b22
Feat/index_select_dim ops (#225) 2023-03-11 16:14:57 -05:00
Nathaniel Simard 860051ca5c
Perf/ndarray maxpool (#223) 2023-03-10 19:14:18 -05:00
Hariganesh Srinivasan 740b554047
Implemented sigmoid and log_sigmoid, Resolves #171 (#221) 
* implemented sigmoid and log_sigmoid with tests

* added test for overflow and computation.
 2023-03-10 19:03:01 -05:00
Nathaniel Simard 0544a915eb
Refactor/migrate more numeric func (#220) 2023-03-10 10:47:15 -05:00
Nathaniel Simard a2ec774c37
draft: Perf/ndarray matmul (#214) 2023-03-09 14:00:35 -05:00
Nathaniel Simard be96160065
refactor: elems (#206) 2023-03-06 18:39:49 -05:00
Nathaniel Simard 019c5f9c44
Refactor/int backend (#197) 
* Update burn-tensor API

* Migrate burn-autodiff

* Update burn-tch

* Update burn-ndarray

* Add some doc
 2023-03-06 14:45:58 -05:00
Nathaniel Simard 15ec42dd6f
Refactor/backend bool tensor (#192) 2023-03-05 11:23:46 -05:00
Nathaniel Simard ffd3d35176
Refactor/tensor api (#191) 2023-03-05 09:23:42 -05:00
Nathaniel Simard bddc1461ef
fix: running state (#190) 2023-03-01 17:33:32 -05:00
Nathaniel Simard d4c298c221
Refactor/burn core (#188) 2023-03-01 16:10:58 -05:00
Nathaniel Simard e6e7f4de42
feat: inplace tensor api. (#187) 2023-03-01 10:55:51 -05:00
Nathaniel Simard 25deb5a13b
Refactor/autodiff (#186) 2023-02-28 20:01:26 -05:00
Dilshod Tadjibaev fb925acc73
Make burn and burn-core packages no_std compatible (#168) (#173) 
* Make burn-ndarray and burn-tensor no_std compatible (#168)
 2023-02-25 09:38:01 -05:00
Dilshod Tadjibaev 9091363ada
Make burn-ndarray and burn-tensor no_std compatible (#168) (#169) 2023-02-21 08:35:24 -05:00
Nathaniel Simard 7d2f43dfca
Refactor Tensor API (#163) 2023-02-17 17:31:20 -05:00
Nathaniel Simard 2401d8ad96
Prepare next release (#161) 2023-02-12 15:32:29 -05:00
Yu Sun 0b85cb0eed
feat(trait-TensorOps): add log1p (#160) 2023-02-11 13:30:50 -05:00
Nathaniel Simard c7963d8485
refactor: device functions (#157) 2023-01-27 18:37:21 -05:00
Nathaniel Simard 2d4e514b41
Refactor/shape function (#156) 2023-01-27 15:18:55 -05:00
Makro f6f0d0e4f3
Add cos, sin and tanh operations (#155) 
* Add cos, sin and tanh operations

* Add tests

* Fix formatting
 2023-01-24 19:40:30 -05:00
Nathaniel Simard 34d233cd3e
Feat/max pooling backend (#152) 2023-01-21 15:39:21 -05:00
Nathaniel Simard 745c88f0a0
Feat/conv (#147) 2023-01-11 18:33:09 -05:00
Nathaniel Simard 2f179f12c9
Bump versions (#141) 2022-12-30 15:15:51 -05:00
Nathaniel Simard eea5a263bf
Feat/adam optimizer (#140) 2022-12-30 15:02:43 -05:00
Nathaniel Simard 248039da0a
Refactor/metric adaptor (#139) 2022-12-26 16:30:25 -05:00
Visual 567adfb93e
refactor: fix all clippy warnings (#137) 2022-12-25 11:22:25 -05:00
Visual 85f98b9d54
refactor, feat: clean Cargo.toml files, upgrade tch to 0.10 (#131) 
* Clean Cargo.toml files, upgrade tch to 0.10

* Add pull_request hook to test.yml workflow
 2022-12-25 10:36:23 -05:00
Nathaniel Simard 1ec35a9e1b
feat: from floats (#87) 2022-12-25 10:11:20 -05:00
Nathaniel Simard 3a9dfe6097
feat: cross entropy loss (#130) 2022-12-25 10:10:22 -05:00
Nathaniel Simard 1a1d86dc3e
refactor: save and load state (#129) 2022-12-24 13:02:37 -05:00
Nathaniel Simard 3a91c2c48e
Feat/multi device (#128) 2022-12-20 18:01:58 -05:00
Nathaniel Simard a599eaed88
Feat/module visitor (#127) 2022-12-17 14:25:36 -05:00
Nathaniel Simard d9592411c2
Feat/text generation example (#126) 2022-12-16 19:23:51 -05:00
Nathaniel Simard 63d8d39517
Feat/autoregressive transformer (#125) 2022-12-10 15:47:57 -05:00
Nathaniel Simard b99b23e1a7
refactor: backends (#124) 2022-12-02 19:28:34 -05:00
Nathaniel Simard 7c38a980c1
feat: improve bool tensor (#122) 2022-12-01 19:43:36 -05:00
Nathaniel Simard 8bd0b17296
Feat/mha (#118) 2022-11-26 15:48:26 -05:00
Nathaniel Simard 46d06f0c90
feat: module init (#117) 2022-11-25 22:02:26 -05:00
Nathaniel Simard acb14adc29
Fix/named tensor (#115) 2022-11-25 19:30:35 -05:00
Nathaniel Simard e0e787f87d
Experimental/named tensor (#113) 2022-11-23 19:05:46 -05:00
nathaniel a8e75f6164 fix/publish-crates-io 2022-11-20 13:08:54 -05:00
Nathaniel Simard 9ecd1be992
chore: get ready for next release (#111) 2022-11-20 12:59:10 -05:00
Nathaniel Simard ca94a9f105
refactor: autodiff gradients types (#107) 2022-11-19 19:43:49 -05:00
Nathaniel Simard dda067e79b
Refactor/backend autodiff (#106) 2022-11-19 12:37:06 -05:00
Nathaniel Simard d45d674a04
Refactor/backend ndarray (#105) 2022-11-18 20:37:38 -05:00
Nathaniel Simard 713f078602
refactor: burn tensor testgen (#104) 2022-11-16 21:02:32 -05:00
Nathaniel Simard ab51c22a55
Refactor/extract tch backend (#103) 2022-11-15 21:06:40 -05:00
Nathaniel Simard 23677b8e89
Refactor/zeros ones elems (#102) 2022-11-14 18:41:26 -05:00
Nathaniel Simard 1a45368878
refactor: relu ops (#101) 2022-11-12 13:28:45 -05:00
Nathaniel Simard da7a8e3f6a
refactor: cat ops (#100) 2022-11-12 13:02:10 -05:00
Nathaniel Simard ab39b8779b
refactor: erf ops (#99) 2022-11-12 12:27:31 -05:00
Nathaniel Simard ef01a4ed3f
refactor: pow ops (#98) 2022-11-12 12:06:53 -05:00
Nathaniel Simard 8c050c2904
refactor: exp + log ops (#97) 2022-11-12 11:50:47 -05:00
Nathaniel Simard 7684857282
refactor: args ops (#96) 2022-11-12 11:29:42 -05:00
Nathaniel Simard 0b77ef5dbc
refactor: precision ops (#95) 2022-11-12 11:13:47 -05:00
Nathaniel Simard 9d832a802a
refactor: aggregation ops (#94) 2022-11-12 10:23:00 -05:00
Nathaniel Simard e7094b92ac
refactor: detach-ops (#93) 2022-11-12 09:44:59 -05:00
Nathaniel Simard cba0db14db
refactor: comparison ops (#88) 2022-11-08 19:57:58 -05:00
Nathaniel Simard c5213b6c32
Fix: mnist example (#85) 2022-11-07 17:58:50 -05:00
Nathaniel Simard eee8bf4599
Refactor/mask fill (#74) 2022-11-05 22:09:40 -04:00
Nathaniel Simard 4111c46d6d
Refactor/index (#73) 2022-11-05 21:26:05 -04:00
Nathaniel Simard d369388036
Refactor/reshape (#72) 2022-11-05 20:49:44 -04:00
Nathaniel Simard e6541298b9
refactor: transpose (#71) 2022-11-05 20:18:31 -04:00
Nathaniel Simard ad23898d23
refactor: neg-ops (#70) 2022-11-05 17:19:04 -04:00
Nathaniel Simard 94b0283bac
refactor: matmul-ops (#69) 2022-11-05 16:29:52 -04:00
Nathaniel Simard 10d1c13c88
refactor/div-ops (#68) 2022-11-05 16:13:55 -04:00
Nathaniel Simard ee61e843a5
refactor/mul-ops (#67) 2022-11-05 14:21:52 -04:00
Nathaniel Simard 2bdad6fa00
refactor: sub ops (#66) 2022-11-05 10:00:52 -04:00
Nathaniel Simard 0f4c1e4c3e
refactor: add ops (#65) 2022-11-01 20:59:31 -04:00
Nathaniel Simard 0c4c657854
feat: repeat (#63) 2022-10-24 18:25:53 -04:00
Nathaniel Simard a78886d51e
Feat/arange (#62) 2022-10-23 10:32:44 -04:00
Nathaniel Simard 847243ddae
Feat/embedding (#61) 2022-10-23 10:18:35 -04:00
Nathaniel Simard b1df39e7fc
Refactor/device ops (#60) 2022-10-21 11:36:51 -04:00
Nathaniel Simard 0e1b0accd6
Refactor/tensor ops (#58) 2022-10-16 12:52:38 -04:00
Nathaniel Simard 72e44336b5
Refactor/ad backend decorator (#57) 2022-10-13 21:33:40 -04:00
nathaniel 31d512ed8f fix: dataset + optim 2022-10-10 11:09:09 -04:00
Nathaniel Simard f2f4fa8a92
Add swap dims (#56) 2022-10-08 11:54:53 -04:00