- 26 4月, 2023 1 次提交
-
-
由 Alexander Jipa 提交于
Co-authored-by: NAlexander Jipa <azzhipa@amazon.com> Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com> Co-authored-by: NLogan Adams <114770087+loadams@users.noreply.github.com>
-
- 18 4月, 2023 1 次提交
-
-
由 Stas Bekman 提交于
-
- 12 4月, 2023 1 次提交
-
-
由 dawei-wang 提交于
Fix microsoft/DeepSpeed#3163
-
- 24 3月, 2023 1 次提交
-
-
由 Ma, Guokai 提交于
Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com>
-
- 09 1月, 2023 1 次提交
-
-
由 Stas Bekman 提交于
Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com>
-
- 04 1月, 2023 1 次提交
-
-
由 Stas Bekman 提交于
* [doc] fix `min_loss_scale` default * align
-
- 13 12月, 2022 1 次提交
-
-
由 Conglong Li 提交于
Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 28 11月, 2022 1 次提交
-
-
由 Joe Mayer 提交于
* Adding gradient accumulation dtype config. * Switching to new DtypeEnum * Adding standalone check function, and unit tests * Variable disambiguation * Adding checks for unsupported states. * Updating for PR comments. * Reorganizing unit test. Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com>
-
- 05 11月, 2022 1 次提交
-
-
由 Joe Mayer 提交于
* Updating autotune default in docs. * Running pre-commit.
-
- 25 10月, 2022 1 次提交
-
-
由 Joe Mayer 提交于
Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 24 9月, 2022 1 次提交
-
-
由 Jeff Rasley 提交于
-
- 04 8月, 2022 1 次提交
-
-
由 Jeff Rasley 提交于
Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com>
-
- 31 7月, 2022 1 次提交
-
-
由 Jeff Rasley 提交于
-
- 30 7月, 2022 1 次提交
-
-
由 Arpan Jain 提交于
Co-authored-by: NArpan Jain <t-arpanjain@microsoft.com> Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 26 7月, 2022 1 次提交
-
-
由 Quentin Anthony 提交于
Co-authored-by: NAmmar Ahmad Awan <ammar.awan@microsoft.com> Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 22 7月, 2022 1 次提交
-
-
由 Jeff Rasley 提交于
Co-authored-by: NAmmar Ahmad Awan <ammar.awan@microsoft.com> Co-authored-by: Nyaozhewei <zheweiy@berkeley.edu> Co-authored-by: NSamyam Rajbhandari <samyamr@microsoft.com>
-
- 20 7月, 2022 1 次提交
-
-
由 Zhewei Yao 提交于
Co-authored-by: Nyaozhewei <zheweiy@berkeley.edu> Co-authored-by: Nxiaoxiawu <yxiaoxiawu@microsoft.com> Co-authored-by: NConglong Li <conglong.li@gmail.com> Co-authored-by: NXiaoxia (Shirley) Wu <94406484+xiaoxiawu-microsoft@users.noreply.github.com> Co-authored-by: NJeff Rasley <jerasley@microsoft.com> Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com> Co-authored-by: NMichael Wyatt <michaelwyatt@microsoft.com>
-
- 14 7月, 2022 1 次提交
-
-
由 Manuel R. Ciosici 提交于
Co-authored-by: NConglong Li <conglong.li@gmail.com>
-
- 16 6月, 2022 1 次提交
-
-
由 Quentin Anthony 提交于
Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 23 3月, 2022 1 次提交
-
-
由 Sayed Hadi Hashemi 提交于
Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 11 3月, 2022 1 次提交
-
-
由 Yucheng Lu 提交于
Co-authored-by: NConglong Li <conglong.li@gmail.com> Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 23 1月, 2022 1 次提交
-
-
由 Alex Hedges 提交于
-
- 22 1月, 2022 1 次提交
-
-
由 Manuel R. Ciosici 提交于
-
- 21 1月, 2022 1 次提交
-
-
由 Justin Chiu 提交于
* Changes for bfloat16 Zero2 * ZeRO stage3 optimizations, with some bug fixes optimizations for stage3: - prefetching improvements - batching allgather calls to amortize fixed overhead and improve bandwidth utilization - batching reduce_scatter calls to amortize fixed overhead and improve bandwidth utilization - using *_base variants of allgather and reduce scatter to reduce memory allocations and data movement - more fine grained synchronization for communication that allows blocking on less work - precomputation of fetching code - using a fetch queue rather than deciding what to (pre)fetch at each iteration - limiting queued coalesced communication ops to reduce memory pressure on pytorch cuda caching allocator (not elegant solution) optimizations for stage3-offload: - made some host-device tensor copies async to improve performance bug fixes and qol improvements: - fix init context method when parent modules modify child weights - speed up model initialization by moving model to GPU before weight initialization - fixed unit test imports so that unit tests can be run from any directory - change performance logging to include memory consumption - add logging w/ model size when done partitioning model new features - bfloat16 support for ZeRO 3 * fix import in ut * ran yapf * improvements to cache flush warn log * backwards compatibility with older versions of pytorch * handle edge case where reduced tensor smaller than world size * moved event synchronization to allgather handle wait() call * removed unnecessary barrier call * formatting fix after resolving merge conflict * skip nvme prefetch when trace not complete * opportunistically avoid memory allocation in allgather coalesced where possible * fix indentation after merge * fixes to account for parameter offload * accounting for torch.cuda.memory_stats not being available * moved partition_all_params to optimizer step * allgathering on params before item gets called * fix param status checks needed after moving partition_all_parameters call to optimizer step * fix grad accumulation with optimizer offload * grad norm computation fix for optimizer offload * change post divide in reduce-scatter to pre divide * fix gradient race condition w/ optimizer offload * improve inf/nan gradient tracking * don't prefetch when not in training mode * format fix after merging * fix prefetching issue when using NVME offload * improved defragmentation for fp16 parameters * relative imports for bf16 tests * changes for bwd compatibility with pytorch 1.2 * remove buffered_reduce_fallback * removed unused parameter offset bookkeeping * fixed tracking for multiple param groups * unbroke bfloat16 config after merge conflict * using base allgather params when only 1 param * cleanup/fixes for fp16 partition defragmentation * switch to CRLF * convert to same new-line style as master * align new line with master * Fix merge issues * switch to CRLF * fix to LF line endings * minor merge fixes * remove extra bfloat16_enabled definition * asserting params inflight for AllGatherHandle * remove get_cuda_mem_allocated_str * Format fixes * fix bfloat16 zero stage check (broken after merge commit) * +self.communication_data_type, -self.allreduce_always_fp32; delete dead code * Add self.reduce_scatter * Format fix * Fix merge issues * iterate over params_to_fetch rather than make another iterator * add some TODOs * remove unnecessary division by micro_step_id * rename config keys "bfloat16" -> "bf16" * rename stage3_gather_fp16_weights_on_model_save -> stage3_gather_16bit_weights_on_model_save * add unit test to check backwards compatibility for gather_16bit_weights * added test to confirm bf16 key bwd compatibility * Format fixes Co-authored-by: NRana Ali Amjad <raamjad@amazon.com> Co-authored-by: NJustin Chiu <justchiu@amazon.com> Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com> Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 04 1月, 2022 1 次提交
-
-
由 Manuel R. Ciosici 提交于
Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 14 12月, 2021 1 次提交
-
-
由 Jeff Rasley 提交于
Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com>
-
- 27 11月, 2021 1 次提交
-
-
由 Mikhail Druzhinin 提交于
* fp16 allreduce * Undo sparse sum in nan check * communication_data_type instead of fp32_allreduce and fp16_allreduce * sparse_allreduce with fp32 or fp16 data type * FIx communication_data_type checks * Allow only torch data types for communication_data_type * Fix Zero assert messages Co-authored-by: NJeff Rasley <jerasley@microsoft.com> Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com>
-
- 23 11月, 2021 1 次提交
-
-
由 Manuel R. Ciosici 提交于
Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 13 11月, 2021 2 次提交
-
-
由 Cheng Li 提交于
* [squash] Staging autotuning v4 Co-authored-by: NCheng Li <pistasable@gmail.com> Co-authored-by: NMinjia Zhang <minjiaz@microsoft.com> Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com> Co-authored-by: NJeff Rasley <jerasley@microsoft.com> * add new extra, guard xgboost, cleanup dead files (#268) * Fix autotuning docs (#1553) * fix docs * rewording the goal * fix typos * fix typos (#1556) * fix typos * fix format * fix bug (#1557) * fix bug Co-authored-by: NJeff Rasley <jerasley@microsoft.com> Co-authored-by: NMinjia Zhang <minjiaz@microsoft.com> Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com>
-
由 Manuel R. Ciosici 提交于
Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 02 10月, 2021 1 次提交
-
-
由 Alex Hedges 提交于
* Fix typos in docs/ * Fix typos in code comments and output strings * Fix typos in the code itself * Fix typos in tests/ Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com>
-
- 01 10月, 2021 1 次提交
-
-
由 Jeff Rasley 提交于
-
- 17 8月, 2021 1 次提交
-
-
由 Conglong Li 提交于
Co-authored-by: NConglong Li <conglong.li@gmail.com> Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 30 7月, 2021 1 次提交
-
-
由 Olatunji Ruwase 提交于
* Fix docstring * Make screenshots clickable for easier viewing * Navigation menu in alphabetical order; More clicable screenshots * Rename 1Cycle doc * Tweak naming * Remove no longer used flag * ZeRO3 Offload release * Single GPU results * Rearrange figures * Single GPU text * tweak intro * zero3-offload section * Add asynchronous i/o docs * Fix print_per_steps doc * Document round_robin_gradients * Tweak description * Trigger CI
-
- 02 7月, 2021 1 次提交
-
-
由 Samyam Rajbhandari 提交于
* contiguous gradients should be set to True by default * Set contiguous gradients to True by default Features such as reduce_scatter depends on contiguous gradients being True. This is also the preferred default configuration.
-
- 17 6月, 2021 1 次提交
-
-
由 Olatunji Ruwase 提交于
* Fix docstring * Make screenshots clickable for easier viewing * Navigation menu in alphabetical order; More clicable screenshots * Rename 1Cycle doc * Tweak naming * Remove no longer used flag * ZeRO3 Offload release * Single GPU results * Rearrange figures * Single GPU text * tweak intro * zero3-offload section * Add asynchronous i/o docs * Fix print_per_steps doc
-
- 09 6月, 2021 1 次提交
-
-
由 Stas Bekman 提交于
Co-authored-by: NOlatunji Ruwase <olruwase@microsoft.com> Co-authored-by: NJeff Rasley <jerasley@microsoft.com>
-
- 20 5月, 2021 1 次提交
-
-
由 Jeff Rasley 提交于
-
- 14 5月, 2021 1 次提交
-
-
由 Olatunji Ruwase 提交于
* Fix docstring * Make screenshots clickable for easier viewing * Navigation menu in alphabetical order; More clicable screenshots * Rename 1Cycle doc * Tweak naming * Remove no longer used flag * ZeRO3 Offload release * Single GPU results * Rearrange figures * Single GPU text * tweak intro * zero3-offload section * Add asynchronous i/o docs
-
- 13 5月, 2021 1 次提交
-
-
由 Cheng Li 提交于
* use the original function's name as the key to old_functions dict * update profile output format * print at global rank 0 * add flops calculation in bwd pass using time from ds timers * improve aggregated profiling out to show all depth * print samples/second * update readme and examples * update docs * fix typo and reorder printing * fix format
-