- 10 8月, 2020 1 次提交
-
-
由 tangwei12 提交于
* add paddle.fleet.AsyncOptimizer Co-authored-by: Ndongdaxiang <dongdaxiang@baidu.com>
-
- 05 8月, 2020 1 次提交
-
-
由 WangXi 提交于
Add dgc to fleet meta optimizer, rm dgc from optimizer all
-
- 03 8月, 2020 3 次提交
-
-
由 WangXi 提交于
-
由 JZ-LIANG 提交于
-
由 Dong Daxiang 提交于
* split meta optimizer files * add graph execution in execution, update two properties in DistributedStrategy, unit tests for these features
-
- 29 7月, 2020 1 次提交
-
-
由 Dong Daxiang 提交于
* refine strategy compiler and meta optimizers make async as a_sync
-
- 28 7月, 2020 1 次提交
-
-
由 Dong Daxiang 提交于
* add more settings for distributed strategy Basically, DistributedStrategy has several parts of configurations: - BuildStrategy: the same as paddle.fluid.BuildStrategy, but the distributed arguments are moved out of BuildStrategy - ExecutionStrategy: the same as paddle.fluid.ExecutionStrategy - collective communication configs: nccl_comm_num, hierarchical allreduce and so on - distributed algorithms: async_update(mainly used in PS), lars, lamb and so on
-
- 20 7月, 2020 1 次提交
-
-
由 Dong Daxiang 提交于
refactor fleet api under paddle.fleet update DistributedStrategy
-
- 06 7月, 2020 1 次提交
-
-
由 Dong Daxiang 提交于
* add paddle.fleet.DistributedStrategy for 2.0
-