8789850eea
Init Conext supports lazy allocate model memory ( #842 )
2022-04-22 18:03:35 +08:00
943982d29a
[unittest] refactored unit tests for change in dependency ( #838 )
2022-04-22 15:39:07 +08:00
01e9f834f5
[dependency] removed torchvision ( #833 )
...
* [dependency] removed torchvision
* fixed transforms
2022-04-22 15:24:35 +08:00
cb5a4778e1
Revert "[WIP] Applying ColoTensor on TP-1D-row Linear. ( #831 )" ( #835 )
...
This reverts commit ac88de6dfc
2022-04-22 14:45:57 +08:00
ac88de6dfc
[WIP] Applying ColoTensor on TP-1D-row Linear. ( #831 )
...
* revert zero tensors back
* [tensor] init row 1d linear
2022-04-22 14:03:26 +08:00
294a6060d0
[tensor] ZeRO use ColoTensor as the base class. ( #828 )
...
* [refactor] moving InsertPostInitMethodToModuleSubClasses to utils.
* [tensor] ZeRO use ColoTensor as the base class.
* polish
2022-04-22 12:00:48 +08:00
8e6fdb4f29
[tensor]fix test_linear ( #826 )
2022-04-21 17:18:56 +08:00
1a9e2c2dff
[tensor] fix kwargs in colo_tensor torch_funtion ( #825 )
2022-04-21 16:47:35 +08:00
2ecc3d7a55
[tensor] lazy init ( #823 )
2022-04-21 15:40:23 +08:00
660d2d1f1b
[Tensor] apply ColoTensor on Torch functions ( #821 )
...
* Revert "[zero] add ZeroTensorShardStrategy (#793 )"
This reverts commit 88759e289e
2022-04-21 14:21:10 +08:00
0ce8924ceb
[tensor] reorganize files ( #820 )
2022-04-21 14:15:48 +08:00
ab962b9735
[gemini] a new tensor structure ( #818 )
...
* Revert "[zero] add ZeroTensorShardStrategy (#793 )"
This reverts commit 88759e289e
2022-04-21 11:42:37 +08:00
e761ad2cd7
Revert "[zero] add ZeroTensorShardStrategy ( #793 )" ( #806 )
2022-04-19 14:40:02 +08:00
88759e289e
[zero] add ZeroTensorShardStrategy ( #793 )
2022-04-19 14:32:45 +08:00
681addb512
[refactor] moving grad acc logic to engine ( #804 )
2022-04-19 14:03:21 +08:00
4d9332b4c5
[refactor] moving memtracer to gemini ( #801 )
2022-04-19 10:13:08 +08:00
4c4388c46e
[hotfix] fix memory leak in zero ( #781 )
2022-04-18 13:57:03 +08:00
5a1a095b92
[test] refactored with the new rerun decorator ( #763 )
...
* [test] refactored with the new rerun decorator
* polish test case
2022-04-15 00:33:04 +08:00
10ef8afdd2
[gemini] init genimi individual directory ( #754 )
2022-04-14 16:40:26 +08:00
dcca614eee
[hotfix] fix test_stateful_tensor_mgr ( #762 )
2022-04-14 15:50:09 +08:00
a93a7d7364
[hotfix] fix reuse_fp16_shard of sharded model ( #756 )
...
* fix reuse_fp16_shard
* disable test stm
* polish code
2022-04-14 14:56:46 +08:00
84c6700b2a
[zero] refactor memstats_collector ( #746 )
2022-04-14 12:01:12 +08:00
e396bb71f2
[zero] add tensor placement policies ( #743 )
...
* add tensor placement policies
* polish comments
* polish comments
* update moe unit tests
2022-04-13 15:00:48 +08:00
22c4b88d56
[zero] refactor ShardedParamV2 for convenience ( #742 )
2022-04-13 14:54:26 +08:00
f4f42d4c3c
[bug] fixed DDP compatibility with torch 1.8 ( #739 )
2022-04-13 00:08:46 +08:00
53cb584808
[utils] correct cpu memory used and capacity in the context of multi-process ( #726 )
2022-04-12 14:57:54 +08:00
b9b469ea50
[moe] add checkpoint for moe zero test ( #729 )
2022-04-12 12:11:54 +08:00
e88a498c9c
[test] removed trivial outdated test
2022-04-12 11:08:15 +08:00
62b4ce7326
[test] added missing decorators to model checkpointing tests
2022-04-12 11:08:15 +08:00
4d90a7b513
[refactor] zero directory ( #724 )
2022-04-11 23:13:02 +08:00
20ab1f5520
[bug] fixed broken test_found_inf ( #725 )
2022-04-11 22:00:27 +08:00
193dc8dacb
[refactor] refactor the memory utils ( #715 )
2022-04-11 16:47:57 +08:00
dbd96fe90a
[zero] check whether gradients have inf and nan in gpu ( #712 )
2022-04-11 15:40:13 +08:00
a9b8300d54
[zero] improve adaptability for not-shard parameters ( #708 )
...
* adapt post grad hooks for not-shard parameters
* adapt optimizer for not-shard parameters
* offload gradients for not-replicated parameters
2022-04-11 13:38:51 +08:00
ab8c6b4a0e
[zero] refactor memstats collector ( #706 )
...
* refactor memstats collector
* fix disposable
* polish code
2022-04-11 10:46:08 +08:00
ee112fe1da
[zero] adapt zero hooks for unsharded module ( #699 )
2022-04-08 20:23:26 +08:00
3c9cd5bb5e
[zero] stateful tensor manager ( #687 )
...
* [WIP] stateful tensor manager
* add eviction strategy
* polish code
* polish code
* polish comment
* add unit test
* fix sampler bug
* polish code
* fix max sampling cnt resetting bug
* fix sampler bug
* polish code
* fix bug
* fix unit test
Co-authored-by: jiaruifang <fangjiarui123@gmail.com>
2022-04-08 17:51:34 +08:00
d7ecaf362b
[zero] fix init bugs in zero context ( #686 )
...
* adapt model weight initialization for methods in Pytorch nn.init
2022-04-07 17:38:45 +08:00
0aab52301e
[hotfix] fix a bug in model data stats tracing ( #655 )
2022-04-03 21:48:06 +08:00
ade05a5d83
[refactor] pipeline, put runtime schedule into engine. ( #627 )
2022-04-03 20:46:45 +08:00
e5d615aeee
[hotfix] fix bugs in testing ( #659 )
...
* remove hybrid adam in test_moe_zero_optim
* fix activation checkpointing and its unitest
2022-04-02 21:58:47 +08:00
b31daed4cf
fix bugs in CPU adam ( #633 )
...
* add cpu adam counter for all cpu adam
* fixed updating error in adam kernel
2022-04-02 17:04:05 +08:00
055fbf5be6
[zero] adapt zero for unsharded paramters (Optimizer part) ( #601 )
2022-04-01 20:10:47 +08:00
354b7954d1
[model checkpoint] added unit tests for checkpoint save/load ( #599 )
2022-04-01 16:53:32 +08:00
93f14d2a33
[zero] test zero tensor utils ( #609 )
2022-04-01 15:16:59 +08:00
e956d93ac2
[refactor] memory utils ( #577 )
2022-04-01 09:22:33 +08:00
e6d50ec107
[zero] adapt zero for unsharded parameters ( #561 )
...
* support existing sharded and unsharded parameters in zero
* add unitest for moe-zero model init
* polish moe gradient handler
2022-03-31 18:34:11 +08:00
7c6c427db1
[zero] trace states of fp16/32 grad and fp32 param ( #571 )
2022-03-31 16:26:54 +08:00
7675366fce
[polish] rename col_attr -> colo_attr ( #558 )
2022-03-31 12:25:45 +08:00
014bac0c49
[zero] hijack p.grad in sharded model ( #554 )
...
* hijack p.grad in sharded model
* polish comments
* polish comments
2022-03-30 18:14:50 +08:00
f552b11294
[zero] label state for param fp16 and grad ( #551 )
2022-03-30 15:57:46 +08:00
214da761d4
[zero] add stateful tensor ( #549 )
2022-03-30 13:51:37 +08:00
8c90d4df54
[zero] add zero context manager to change config during initialization ( #546 )
2022-03-29 17:57:59 +08:00
ec5086c49c
Refactored docstring to google style
2022-03-29 17:17:47 +08:00
53b1b6e340
[zero] non model data tracing ( #545 )
2022-03-29 15:45:48 +08:00
1f90a3b129
[zero] polish ZeroInitContext ( #540 )
2022-03-29 09:09:04 +08:00
c11ff81b15
[zero] get memory usage of sharded optim v2. ( #542 )
2022-03-29 09:08:18 +08:00
a30e2b4c24
[zero] adapt for no-leaf module in zero ( #535 )
...
only process module's own parameters in Zero context
add zero hooks for all modules that contrain parameters
gather parameters only belonging to module itself
2022-03-28 17:42:18 +08:00
705f56107c
[zero] refactor model data tracing ( #537 )
2022-03-28 16:38:18 +08:00
a590ed0ba3
[zero] improve the accuracy of get_memory_usage of sharded param ( #538 )
2022-03-28 16:19:19 +08:00
37cb70feec
[zero] get memory usage for sharded param ( #536 )
2022-03-28 15:01:21 +08:00
105c5301c3
[zero]added hybrid adam, removed loss scale in adam ( #527 )
...
* [zero]added hybrid adam, removed loss scale of adam
* remove useless code
2022-03-25 18:03:54 +08:00
8d8c5407c0
[zero] refactor model data tracing ( #522 )
2022-03-25 18:03:32 +08:00
3601b2bad0
[test] fixed rerun_on_exception and adapted test cases ( #487 )
2022-03-25 17:25:12 +08:00
4d322b79da
[refactor] remove old zero code ( #517 )
2022-03-25 14:54:39 +08:00
6a3f9fda83
[cuda] modify the fused adam, support hybrid of fp16 and fp32 ( #497 )
2022-03-25 14:15:53 +08:00
920c5889a7
[zero] add colo move inline ( #521 )
2022-03-25 14:02:55 +08:00
0bebda6ea5
[zero] fix init device bug in zero init context unittest ( #516 )
2022-03-25 12:24:18 +08:00
7ef3507ace
[zero] show model data cuda memory usage after zero context init. ( #515 )
2022-03-25 11:23:35 +08:00
9330be0f3c
[memory] set cuda mem frac ( #506 )
2022-03-24 16:57:13 +08:00
0035b7be07
[memory] add model data tensor moving api ( #503 )
2022-03-24 14:29:41 +08:00
a445e118cf
[polish] polish singleton and global context ( #500 )
2022-03-23 18:03:39 +08:00
9ec1ce6ab1
[zero] sharded model support the reuse of fp16 shard ( #495 )
...
* sharded model supports reuse fp16 shard
* rename variable
* polish code
* polish code
* polish code
2022-03-23 14:59:59 +08:00
62b0a8d644
[zero] sharded optim support hybrid cpu adam ( #486 )
...
* sharded optim support hybrid cpu adam
* update unit test
* polish docstring
2022-03-22 14:56:59 +08:00
b334822163
[zero] polish sharded param name ( #484 )
...
* [zero] polish sharded param name
* polish code
* polish
* polish code
* polish
* polsih
* polish
2022-03-22 14:36:16 +08:00
65c0f380c2
[format] polish name format for MOE ( #481 )
2022-03-21 23:19:47 +08:00
7544347145
[MOE] add unitest for MOE experts layout, gradient handler and kernel ( #469 )
2022-03-21 13:35:04 +08:00
84fd7c1d4d
add moe context, moe utilities and refactor gradient handler ( #455 )
2022-03-18 16:38:32 +08:00
af185b5519
[test] fixed amp convergence comparison test ( #454 )
2022-03-18 16:28:16 +08:00
a241f61b34
[zero] Update initialize for ZeRO ( #458 )
...
* polish code
* shard strategy receive pg in shard() / gather()
* update zero engine
* polish code
2022-03-18 16:18:31 +08:00
642846d6f9
update sharded optim and fix zero init ctx ( #457 )
2022-03-18 15:44:47 +08:00
e2e9f82588
Revert "[zero] update sharded optim and fix zero init ctx" ( #456 )
...
* Revert "polish code"
This reverts commit 8cf7ff08cfe99af94ab846add4a5c557567ee768
2022-03-18 15:22:43 +08:00
8cf7ff08cf
polish code
2022-03-18 14:25:25 +08:00
46add4a5c5
remove surplus imports
2022-03-18 14:25:25 +08:00
57567ee768
update sharded optim and fix zero init ctx
2022-03-18 14:25:25 +08:00
f27d801a13
[test] optimized zero data parallel test ( #452 )
2022-03-18 11:35:54 +08:00
0fcfb1e00d
[test] make zero engine test really work ( #447 )
2022-03-17 17:24:25 +08:00
bb2790cf0b
optimize engine and trainer test ( #448 )
2022-03-17 15:44:17 +08:00
b72b8445c6
optimized context test time consumption ( #446 )
2022-03-17 14:40:52 +08:00
496cbb0760
[hotfix] fix initialize bug with zero ( #442 )
2022-03-17 13:16:22 +08:00
17b8274f8a
[unitest] polish zero config in unittest ( #438 )
2022-03-17 10:20:53 +08:00
640a6cd304
[refactory] refactory the initialize method for new zero design ( #431 )
2022-03-16 19:29:37 +08:00
fce9432f08
sync before creating empty grad
2022-03-16 14:24:09 +08:00
f9c762df85
[test] merge zero optim tests ( #428 )
2022-03-16 12:22:45 +08:00
5d7dc3525b
[hotfix] run cpu adam unittest in pytest ( #424 )
2022-03-16 10:39:55 +08:00
adebb3e041
[zero] cuda margin space for OS ( #418 )
2022-03-15 12:02:19 +08:00
56bb412e72
[polish] use GLOBAL_MODEL_DATA_TRACER ( #417 )
2022-03-15 11:29:46 +08:00
23ba3fc450
[zero] refactory ShardedOptimV2 init method ( #416 )
2022-03-15 10:45:55 +08:00
e79ea44247
[fp16] refactored fp16 optimizer ( #392 )
2022-03-15 10:05:38 +08:00
21dc54e019
[zero] memtracer to record cuda memory usage of model data and overall system ( #395 )
2022-03-14 22:05:30 +08:00
a37bf1bc42
[hotfix] rm test_tensor_detector.py ( #413 )
2022-03-14 21:39:48 +08:00
370f567e7d
[zero] new interface for ShardedOptimv2 ( #406 )
2022-03-14 20:48:41 +08:00
a9c27be42e
Added tensor detector ( #393 )
...
* Added tensor detector
* Added the - states
* Allowed change include_cpu when detect()
2022-03-14 18:01:46 +08:00
54fd37f0e0
polish unit test
2022-03-14 15:06:02 +08:00
1e4bf85cdb
fixed bug in activation checkpointing test ( #387 )
2022-03-11 15:50:28 +08:00
3af13a2c3e
[zero] polish ShardedOptimV2 unittest ( #385 )
...
* place params on cpu after zero init context
* polish code
* bucketzed cpu gpu tensor transter
* find a bug in sharded optim unittest
* add offload unittest for ShardedOptimV2.
* polish code and make it more robust
2022-03-11 15:50:28 +08:00
526a318032
[unit test] Refactored test cases with component func ( #339 )
...
* refactored test with component func
* fixed bug
2022-03-11 15:50:28 +08:00
de46450461
Added activation offload ( #331 )
...
* Added activation offload
* Fixed the import bug, used the pytest
2022-03-11 15:50:28 +08:00
b5f43acee3
[zero] find miss code ( #378 )
2022-03-11 15:50:28 +08:00
6b6002962a
[zero] zero init context collect numel of model ( #375 )
2022-03-11 15:50:28 +08:00
d9217e1960
Revert "[zero] bucketized tensor cpu gpu copy ( #368 )"
...
This reverts commit bef05489b6
2022-03-11 15:50:28 +08:00
00670c870e
[zero] bucketized tensor cpu gpu copy ( #368 )
2022-03-11 15:50:28 +08:00
44e4891f57
[zero] able to place params on cpu after zero init context ( #365 )
...
* place params on cpu after zero init context
* polish code
2022-03-11 15:50:28 +08:00
ea2872073f
[zero] global model data memory tracer ( #360 )
2022-03-11 15:50:28 +08:00
cb34cd384d
[test] polish zero related unitest ( #351 )
2022-03-11 15:50:28 +08:00
532ae79cb0
add test sharded optim with cpu adam ( #347 )
2022-03-11 15:50:28 +08:00
425bb0df3f
Added Profiler Context to manage all profilers ( #340 )
2022-03-11 15:50:28 +08:00
d0ae0f2215
[zero] update sharded optim v2 ( #334 )
2022-03-11 15:50:28 +08:00
2b8cddd40e
skip bert in test engine
2022-03-11 15:50:28 +08:00
f5f0ad266e
fix bert unit test
2022-03-11 15:50:28 +08:00
d271f2596b
polish engine unitest
2022-03-11 15:50:28 +08:00
354c0f9047
polish code
2022-03-11 15:50:28 +08:00
4d94cd513e
adapting bert unitest interface
2022-03-11 15:50:28 +08:00
7977422aeb
add bert for unitest and sharded model is not able to pass the bert case
2022-03-11 15:50:28 +08:00
1388671699
[zero] Update sharded model v2 using sharded param v2 ( #323 )
2022-03-11 15:50:28 +08:00
799d105bb4
using pytest parametrize
2022-03-11 15:50:28 +08:00
dec24561cf
show pytest parameterize
2022-03-11 15:50:28 +08:00
11bddb6e55
[zero] update zero context init with the updated test utils ( #327 )
2022-03-11 15:50:28 +08:00
6268446b81
[test] refactored testing components ( #324 )
2022-03-11 15:50:28 +08:00
de0468c7a8
[zero] zero init context ( #321 )
...
* add zero init context
* add more flags for zero init context
fix bug of repeated converting param to ShardedParamV2
* polish code
2022-03-11 15:50:28 +08:00
73bff11288
Added profiler communication operations
...
Fixed bug for learning rate scheduler
2022-03-11 15:50:28 +08:00
a3269de5c9
[zero] cpu adam kernel ( #288 )
...
* Added CPU Adam
* finished the cpu adam
* updated the license
* delete useless parameters, removed resnet
* modified the method off cpu adam unittest
* deleted some useless codes
* removed useless codes
Co-authored-by: ver217 <lhx0217@gmail.com>
Co-authored-by: Frank Lee <somerlee.9@gmail.com>
Co-authored-by: jiaruifang <fangjiarui123@gmail.com>
2022-03-11 15:50:28 +08:00
90d3aef62c
[zero] yet an improved sharded param ( #311 )
2022-03-11 15:50:28 +08:00
c9e7d9582d
[zero] polish shard strategy ( #310 )
...
* init shard param from shape tuple
* add more unitest for shard param
* add set_payload method for ShardedParam
* [zero] add shareded tensor class
* polish code
* add shard stratgy
* move shard and gather logic to shard strategy from shard tensor.
* polish code
2022-03-11 15:50:28 +08:00
36f9a74ab2
fix sharded param hook and unit test
2022-03-11 15:50:28 +08:00
001ca624dd
impl shard optim v2 and add unit test
2022-03-11 15:50:28 +08:00
74f77e314b
[zero] a shard strategy in granularity of tensor ( #307 )
2022-03-11 15:50:28 +08:00
80364c7686
[zero] sharded tensor ( #305 )
...
* init shard param from shape tuple
* add more unitest for shard param
* add set_payload method for ShardedParam
* [zero] add shareded tensor class
* polish code
2022-03-11 15:50:28 +08:00
d344689274
[profiler] primary memory tracer
2022-03-11 15:50:28 +08:00
e17e92c54d
Polish sharded parameter ( #297 )
...
* init shard param from shape tuple
* add more unitest for shard param
* add more unittests to shareded param
2022-03-11 15:50:28 +08:00
7aef75ca42
[zero] add sharded grad and refactor grad hooks for ShardedModel ( #287 )
2022-03-11 15:50:28 +08:00
27155b8513
added unit test for sharded optimizer ( #293 )
...
* added unit test for sharded optimizer
* refactor for elegance
2022-03-11 15:50:28 +08:00
e17e54e32a
added buffer sync to naive amp model wrapper ( #291 )
2022-03-11 15:50:28 +08:00
8d653af408
add a common util for hooks registered on parameter. ( #292 )
2022-03-11 15:50:28 +08:00
5a560a060a
Feature/zero ( #279 )
...
* add zero1 (#209 )
* add zero1
* add test zero1
* update zero stage 1 develop (#212 )
* Implement naive zero3 (#240 )
* naive zero3 works well
* add zero3 param manager
* add TODOs in comments
* add gather full param ctx
* fix sub module streams
* add offload
* fix bugs of hook and add unit tests
* fix bugs of hook and add unit tests (#252 )
* add gather full param ctx
* fix sub module streams
* add offload
* fix bugs of hook and add unit tests
* polish code and add state dict hook
* fix bug
* update unit test
* refactor reconstructed zero code
* clip_grad support zero3 and add unit test
* add unit test for Zero3ParameterManager
* [WIP] initialize the shard param class
* [WIP] Yet another sharded model implementation (#274 )
* [WIP] initialize the shard param class
* [WIP] Yes another implementation of shardModel. Using a better hook method.
* torch.concat -> torch.cat
* fix test_zero_level_1.py::test_zero_level_1 unitest
* remove deepspeed implementation and refactor for the reconstructed zero module
* polish zero dp unittests
Co-authored-by: ver217 <lhx0217@gmail.com>
Co-authored-by: Frank Lee <somerlee.9@gmail.com>
2022-03-11 15:50:28 +08:00
82023779bb
Added TPExpert for special situation
2022-03-11 15:50:28 +08:00
219df6e685
Optimized MoE layer and fixed some bugs;
...
Decreased moe tests;
Added FFNExperts and ViTMoE model
2022-03-11 15:50:28 +08:00
3dba070580
fixed padding index issue for vocab parallel embedding layers; updated 3D linear to be compatible with examples in the tutorial
2022-03-11 15:50:28 +08:00
9ee197d0e9
moved env variables to global variables; ( #215 )
...
added branch context;
added vocab parallel layers;
moved split_batch from load_batch to tensor parallel embedding layers;
updated gpt model;
updated unit test cases;
fixed few collective communicator bugs
2022-02-15 11:31:13 +08:00
569357fea0
add pytorch hooks ( #179 )
...
* add pytorch hooks
fix #175
* remove licenses in src code
* add gpu memory tracer
* replacing print with logger in ophooks.
2022-01-25 22:20:54 +08:00
e2089c5c15
adapted for sequence parallel ( #163 )
2022-01-20 13:44:51 +08:00
7bf1e98b97
pipeline last stage supports multi output ( #151 )
2022-01-17 15:57:47 +08:00
96780e6ee4
Optimize pipeline schedule ( #94 )
...
* add pipeline shared module wrapper and update load batch
* added model parallel process group for amp and clip grad (#86 )
* added model parallel process group for amp and clip grad
* update amp and clip with model parallel process group
* remove pipeline_prev/next group (#88 )
* micro batch offload
* optimize pipeline gpu memory usage
* pipeline can receive tensor shape (#93 )
* optimize pipeline gpu memory usage
* fix grad accumulation step counter
* rename classes and functions
Co-authored-by: Frank Lee <somerlee.9@gmail.com>
2021-12-30 15:56:46 +08:00
01a80cd86d
Hotfix/Colossalai layers ( #92 )
...
* optimized 1d layer apis; reorganized nn.layer modules; fixed tests
* fixed 2.5d runtime issue
* reworked split batch, now called in trainer.schedule.load_batch
Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-29 23:32:10 +08:00
0fedef4f3c
Layer integration ( #83 )
...
* integrated parallel layers for ease of building models
* integrated 2.5d layers
* cleaned codes and unit tests
* added log metric by step hook; updated imagenet benchmark; fixed some bugs
* reworked initialization; cleaned codes
Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-27 15:04:32 +08:00
8f02a88db2
add interleaved pipeline, fix naive amp and update pipeline model initializer ( #80 )
2021-12-20 23:26:19 +08:00
91c327cb44
fixed zero level 3 dtype bug ( #76 )
2021-12-20 17:00:53 +08:00
cd9c28e055
added CI for unit testing ( #69 )
2021-12-16 10:32:08 +08:00
da01c234e1
Develop/experiments ( #59 )
...
* Add gradient accumulation, fix lr scheduler
* fix FP16 optimizer and adapted torch amp with tensor parallel (#18 )
* fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes
* fixed trainer
* Revert "fixed trainer"
This reverts commit 2e0b0b7699#23 )
Co-authored-by: 1SAA <c2h214748@gmail.com>
* Split conv2d, class token, positional embedding in 2d, Fix random number in ddp
Fix convergence in cifar10, Imagenet1000
* Integrate 1d tensor parallel in Colossal-AI (#39 )
* fixed 1D and 2D convergence (#38 )
* optimized 2D operations
* fixed 1D ViT convergence problem
* Feature/ddp (#49 )
* remove redundancy func in setup (#19 ) (#20 )
* use env to control the language of doc (#24 ) (#25 )
* Support TP-compatible Torch AMP and Update trainer API (#27 )
* Add gradient accumulation, fix lr scheduler
* fix FP16 optimizer and adapted torch amp with tensor parallel (#18 )
* fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes
* fixed trainer
* Revert "fixed trainer"
This reverts commit 2e0b0b7699#23 )
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: ver217 <lhx0217@gmail.com>
* add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29 )
* add explanation for ViT example (#35 ) (#36 )
* support torch ddp
* fix loss accumulation
* add log for ddp
* change seed
* modify timing hook
Co-authored-by: Frank Lee <somerlee.9@gmail.com>
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: binmakeswell <binmakeswell@gmail.com>
* Feature/pipeline (#40 )
* remove redundancy func in setup (#19 ) (#20 )
* use env to control the language of doc (#24 ) (#25 )
* Support TP-compatible Torch AMP and Update trainer API (#27 )
* Add gradient accumulation, fix lr scheduler
* fix FP16 optimizer and adapted torch amp with tensor parallel (#18 )
* fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes
* fixed trainer
* Revert "fixed trainer"
This reverts commit 2e0b0b7699#23 )
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: ver217 <lhx0217@gmail.com>
* add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29 )
* add explanation for ViT example (#35 ) (#36 )
* optimize communication of pipeline parallel
* fix grad clip for pipeline
Co-authored-by: Frank Lee <somerlee.9@gmail.com>
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: binmakeswell <binmakeswell@gmail.com>
* optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51 )
* Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset
* update api for better usability (#58 )
update api for better usability
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: ver217 <lhx0217@gmail.com>
Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com>
Co-authored-by: binmakeswell <binmakeswell@gmail.com>
Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com>
Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 15:08:29 +08:00
3defa32aee
Support TP-compatible Torch AMP and Update trainer API ( #27 )
...
* Add gradient accumulation, fix lr scheduler
* fix FP16 optimizer and adapted torch amp with tensor parallel (#18 )
* fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes
* fixed trainer
* Revert "fixed trainer"
This reverts commit 2e0b0b7699#23 )
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: ver217 <lhx0217@gmail.com>
2021-11-18 19:45:06 +08:00
3245a69fc2
cleaned test scripts
2021-10-29 00:48:14 +08:00
404ecbdcc6
Migrated project
2021-10-28 18:21:23 +02:00
Jiarui Fang