From 970844a59e452a58676099c205cb53093307e6cd Mon Sep 17 00:00:00 2001 From: wangmin Date: Mon, 7 Sep 2020 16:38:25 +0800 Subject: [PATCH] fix init interface --- ...rder_optimizer_for_resnet50_application.md | 28 +++++++++---------- 1 file changed, 13 insertions(+), 15 deletions(-) diff --git a/tutorials/source_zh_cn/advanced_use/second_order_optimizer_for_resnet50_application.md b/tutorials/source_zh_cn/advanced_use/second_order_optimizer_for_resnet50_application.md index 440e668f07..5aff6381f4 100644 --- a/tutorials/source_zh_cn/advanced_use/second_order_optimizer_for_resnet50_application.md +++ b/tutorials/source_zh_cn/advanced_use/second_order_optimizer_for_resnet50_application.md @@ -114,7 +114,7 @@ def create_dataset(dataset_path, do_train, repeat_num=1, batch_size=32, target=" if target == "Ascend": device_num, rank_id = _get_rank_info() else: - init("nccl") + init() rank_id = get_rank() device_num = get_group_size() if device_num == 1: @@ -249,7 +249,7 @@ if __name__ == "__main__": # learning rate setting lr = get_model_lr(0, config.lr_init, config.lr_decay, config.lr_end_epoch, step_size, decay_epochs=39) # define the optimizer - net_opt = opt = THOR(filter(lambda x: x.requires_grad, net.get_parameters()), Tensor(lr), config.momentum, + opt = THOR(filter(lambda x: x.requires_grad, net.get_parameters()), Tensor(lr), config.momentum, filter(lambda x: 'matrix_A' in x.name, net.get_parameters()), filter(lambda x: 'matrix_G' in x.name, net.get_parameters()), filter(lambda x: 'A_inv_max' in x.name, net.get_parameters()), @@ -360,20 +360,18 @@ sh run_distribute_train_gpu.sh [DATASET_PATH] [DEVICE_NUM] - `DATASET_PATH`：训练数据集路径。 - `DEVICE_NUM`: 实际的运行卡数。 -在GPU训练时，无需设置`DEVICE_ID`环境变量，因此在主训练脚本中不需要调用`int(os.getenv('DEVICE_ID'))`来获取卡的物理序号，同时`context`中也无需传入`device_id`。我们需要将device_target设置为GPU，并需要调用`init("nccl")`来使能NCCL。 +在GPU训练时，无需设置`DEVICE_ID`环境变量，因此在主训练脚本中不需要调用`int(os.getenv('DEVICE_ID'))`来获取卡的物理序号，同时`context`中也无需传入`device_id`。我们需要将device_target设置为GPU，并需要调用`init()`来使能NCCL。 训练过程中loss打印示例如下： ```bash ... -epoch: 1 step: 5004, loss is 4.3069 -epoch: 2 step: 5004, loss is 3.5695 -epoch: 3 step: 5004, loss is 3.5893 -epoch: 4 step: 5004, loss is 3.1987 -epoch: 5 step: 5004, loss is 3.3526 +epoch: 1 step: 5004, loss is 4.2546034 +epoch: 2 step: 5004, loss is 4.0819564 +epoch: 3 step: 5004, loss is 3.7005644 +epoch: 4 step: 5004, loss is 3.2668946 +epoch: 5 step: 5004, loss is 3.023509 ...... -epoch: 40 step: 5004, loss is 1.9482 -epoch: 41 step: 5004, loss is 1.8950 -epoch: 42 step: 5004, loss is 1.9023 +epoch: 36 step: 5004, loss is 1.645802 ... ``` @@ -385,13 +383,13 @@ epoch: 42 step: 5004, loss is 1.9023 ├─resnet-1_5004.ckpt ├─resnet-2_5004.ckpt │ ...... - ├─resnet-42_5004.ckpt + ├─resnet-36_5004.ckpt ...... ├─ckpt_7 ├─resnet-1_5004.ckpt ├─resnet-2_5004.ckpt │ ...... - ├─resnet-42_5004.ckpt + ├─resnet-36_5004.ckpt ``` @@ -434,7 +432,7 @@ if __name__ == "__main__": ``` ### 执行推理 -推理网络定义完成之后，调用/scripts目录下的shell脚本，进行推理。 +推理网络定义完成之后，调用`scripts`目录下的shell脚本，进行推理。 #### Ascend 910 在Ascend 910硬件平台上，推理的执行命令如下： ``` @@ -462,5 +460,5 @@ sh run_eval_gpu.sh [DATASET_PATH] [CHECKPOINT_PATH] 推理的结果如下： ``` -result: {'top_5_accuracy': 0.9281169974391805, 'top_1_accuracy': 0.7593830025608195} ckpt=train_parallel0/resnet-42_5004.ckpt +result: {'top_5_accuracy': 0.9287972151088348, 'top_1_accuracy': 0.7597031049935979} ckpt=train_parallel/resnet-36_5004.ckpt ``` -- Gitee