PaddleFleetX
PaddleFleetX copied to clipboard
AttributeError: 'Fleet' object has no attribute 'DistributedStrategy'
当使用示例代码进行调试时报错:
Traceback (most recent call last):
File "test.py", line 17, in
版本信息: paddlepaddle-gpu==2.0rc1,fleet-x==0.0.7
请问这是什么原因呢?
试了一下没有问题,请检查一下环境配置和paddle2.0.0rc1安装路径是否正确。
import paddle paddle.version '2.0.0-rc1' import paddle.distributed.fleet as fleet dist_strategy = fleet.DistributedStrategy() print(dist_strategy)
+==============================================================================+
| |
| DistributedStrategy Overview |
| |
+==============================================================================+
| a_sync=True <-> a_sync_configs |
+------------------------------------------------------------------------------+
| k_steps -1 |
| max_merge_var_num 1 |
| send_queue_size 16 |
| independent_recv_thread False |
| min_send_grad_num_before_recv 1 |
| thread_pool_size 1 |
| send_wait_times 1 |
| runtime_split_send_recv False |
| launch_barrier True |
| heter_worker_device_guard cpu |
+==============================================================================+
| Environment Flags, Communication Flags |
+------------------------------------------------------------------------------+
| mode 1 |
| elastic False |
| auto False |
| sync_nccl_allreduce True |
| nccl_comm_num 1 |
| use_hierarchical_allreduce False |
| hierarchical_allreduce_inter_nranks 1 |
| sync_batch_norm False |
| fuse_all_reduce_ops True |
| fuse_grad_size_in_MB 32 |
| fuse_grad_size_in_TFLOPS 50.0 |
| cudnn_exhaustive_search True |
| conv_workspace_size_limit 4000 |
| cudnn_batchnorm_spatial_persistent True |
| fp16_allreduce False |
| last_comm_group_size_MB 1.0 |
+==============================================================================+