请问在moderarts上怎么配置,里边预训练模型这些在哪儿下？

[marvellee1]

[zhangjiajin]

请给出更详细的日志。

[marvellee1]

(MindSpore) [ma-user examples]$vega ./nas/sp_nas/spnas_md.yml -b m -d NPU
2022-12-08 19:13:41.377 INFO ------------------------------------------------
2022-12-08 19:13:41.377 INFO task id: 1208.191334.031
2022-12-08 19:13:41.377 INFO ------------------------------------------------
2022-12-08 19:13:41.379 INFO configure: {
"general": {
"backend": "m",
"device_category": "NPU",
"cluster": {}
},
"pipeline": [
"serial"
],
"serial": {
"pipe_step": {
"type": "SearchPipeStep"
},
"search_algorithm": {
"type": "SpNasS",
"max_sample": 1,
"objective_keys": "AP50"
},
"search_space": {
"type": "SearchSpace",
"hyperparameters": [
{
"key": "network.backbone.code",
"type": "CATEGORY",
"range": [
"111-2111-211111-211"
]
}
]
},
"model": {
"model_desc": {
"type": "Faster_Rcnn_MD"
}
},
"trainer": {
"type": "SpNasTrainerCallback",
"epochs": 6,
"get_train_metric_after_epoch": false,
"model_statistics": false,
"is_detection_trainer": true,
"perfs_cmp_key": "AP50",
"optimizer": {
"type": "SGD",
"params": {
"lr": 0.03,
"momentum": 0.9,
"weight_decay": 0.0001
}
},
"lr_scheduler": {
"type": "WarmupScheduler",
"by_epoch": false,
"params": {
"warmup_type": "linear",
"warmup_iters": 2000,
"warmup_ratio": 0.001,
"after_scheduler_config": {
"type": "MultiStepLR",
"by_epoch": true,
"params": {
"milestones": [
10,
20
],
"gamma": 0.1
}
}
}
},
"loss": {
"type": "SumLoss"
},
"metric": {
"type": "coco",
"params": {
"anno_path": "/home/ma-user/work/cache/datasets/COCO2017/annotations/instances_val2017.json"
}
}
},
"dataset": {
"type": "CocoDataset",
"common": {
"batch_size": 16,
"num_parallel_workers": 1,
"flip_ratio": 0.5,
"expand_ratio": 1.0,
"img_width": 700,
"img_height": 512,
"keep_ratio": true,
"device_id": 0,
"device_num": 1,
"rank_id": 0,
"python_multiprocessing": true,
"coco_root": "/home/ma-user/work/cache/datasets/COCO2017",
"mindrecord_dir": "/home/ma-user/work/cache/MindRecord_COCO_TRAIN",
"instance_set": "annotations/instances_{}.json",
"coco_classes": [
"target",
"fault"
],
"num_classes": 3
},
"train": {
"train_data_type": "train2017"
},
"val": {
"val_data_type": "val2017",
"test_batch_size": 64
}
}
},
"parallel": {
"pipe_step": {
"type": "SearchPipeStep",
"models_folder": "{local_base_path}/output/serial/",
"pretrained_folder": "{local_base_path}/output/serial/"
},
"search_algorithm": {
"type": "SpNasP"
},
"search_space": {
"type": "SearchSpace",
"hyperparameters": [
{
"key": "network.neck.code",
"type": "CATEGORY",
"range": [
[
0,
1,
2,
3
]
]
}
]
},
"model": {
"model_desc": {
"type": "Faster_Rcnn_MD"
}
},
"trainer": {
"ref": "serial.trainer",
"type": "SpNasTrainerCallback",
"epochs": 6,
"get_train_metric_after_epoch": false,
"model_statistics": false,
"is_detection_trainer": true,
"perfs_cmp_key": "AP50",
"optimizer": {
"type": "SGD",
"params": {
"lr": 0.03,
"momentum": 0.9,
"weight_decay": 0.0001
}
},
"lr_scheduler": {
"type": "WarmupScheduler",
"by_epoch": false,
"params": {
"warmup_type": "linear",
"warmup_iters": 2000,
"warmup_ratio": 0.001,
"after_scheduler_config": {
"type": "MultiStepLR",
"by_epoch": true,
"params": {
"milestones": [
10,
20
],
"gamma": 0.1
}
}
}
},
"loss": {
"type": "SumLoss"
},
"metric": {
"type": "coco",
"params": {
"anno_path": "/home/ma-user/work/cache/datasets/COCO2017/annotations/instances_val2017.json"
}
}
},
"dataset": {
"ref": "serial.dataset",
"type": "CocoDataset",
"common": {
"batch_size": 16,
"num_parallel_workers": 1,
"flip_ratio": 0.5,
"expand_ratio": 1.0,
"img_width": 700,
"img_height": 512,
"keep_ratio": true,
"device_id": 0,
"device_num": 1,
"rank_id": 0,
"python_multiprocessing": true,
"coco_root": "/home/ma-user/work/cache/datasets/COCO2017",
"mindrecord_dir": "/home/ma-user/work/cache/MindRecord_COCO_TRAIN",
"instance_set": "annotations/instances_{}.json",
"coco_classes": [
"target",
"fault"
],
"num_classes": 3
},
"train": {
"train_data_type": "train2017"
},
"val": {
"val_data_type": "val2017",
"test_batch_size": 64
}
}
},
"fullytrain": {
"pipe_step": {
"type": "TrainPipeStep",
"models_folder": "{local_base_path}/output/parallel/",
"pretrained_folder": "{local_base_path}/output/parallel/"
},
"trainer": {
"ref": "serial.trainer",
"epochs": 24,
"type": "SpNasTrainerCallback",
"get_train_metric_after_epoch": false,
"model_statistics": false,
"is_detection_trainer": true,
"perfs_cmp_key": "AP50",
"optimizer": {
"type": "SGD",
"params": {
"lr": 0.03,
"momentum": 0.9,
"weight_decay": 0.0001
}
},
"lr_scheduler": {
"type": "WarmupScheduler",
"by_epoch": false,
"params": {
"warmup_type": "linear",
"warmup_iters": 2000,
"warmup_ratio": 0.001,
"after_scheduler_config": {
"type": "MultiStepLR",
"by_epoch": true,
"params": {
"milestones": [
10,
20
],
"gamma": 0.1
}
}
}
},
"loss": {
"type": "SumLoss"
},
"metric": {
"type": "coco",
"params": {
"anno_path": "/home/ma-user/work/cache/datasets/COCO2017/annotations/instances_val2017.json"
}
}
},
"dataset": {
"ref": "serial.dataset",
"type": "CocoDataset",
"common": {
"batch_size": 16,
"num_parallel_workers": 1,
"flip_ratio": 0.5,
"expand_ratio": 1.0,
"img_width": 700,
"img_height": 512,
"keep_ratio": true,
"device_id": 0,
"device_num": 1,
"rank_id": 0,
"python_multiprocessing": true,
"coco_root": "/home/ma-user/work/cache/datasets/COCO2017",
"mindrecord_dir": "/home/ma-user/work/cache/MindRecord_COCO_TRAIN",
"instance_set": "annotations/instances_{}.json",
"coco_classes": [
"target",
"fault"
],
"num_classes": 3
},
"train": {
"train_data_type": "train2017"
},
"val": {
"val_data_type": "val2017",
"test_batch_size": 64
}
}
},
"abs_path": true
}
2022-12-08 19:13:41.379 INFO ------------------------------------------------
{}
2022-12-08 19:13:42.534 INFO ------------------------------------------------
2022-12-08 19:13:42.534 INFO Step: serial
2022-12-08 19:13:42.534 INFO ------------------------------------------------
2022-12-08 19:13:42.547 INFO Serial-level Sample1: expend -> swap -> expend. Success.
2022-12-08 19:13:42.548 INFO desc:{'network.backbone.code': '1111-211-12111111112-11'}
2022-12-08 19:13:42.575 INFO submit trainer, id=1
2022-12-08 19:13:42.577 INFO Run train/val in mode: 0.
2022-12-08 19:13:42.578 INFO minspore context, mode: 0, target: Ascend, device_id: 0
2022-12-08 19:13:42.578 INFO DEVICE_ID: 0
2022-12-08 19:13:42.578 INFO Dataset_sink_mode:True.
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.413.059 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.416.528 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.432.797 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.436.952 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.453.311 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.457.830 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.472.690 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.478.160 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.498.580 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.511.148 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.511.647 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.513.333 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.513.772 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.515.253 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:43.942.775 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:44.232.465 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:44.301.238 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
[WARNING] ME(119171:281472820496960,MainProcess):2022-12-08-19:13:44.304.187 [mindspore/common/tensor.py:1637] WARN_DEPRECATED: The usage of to_tensor is deprecated. Please use init_data
2022-12-08 19:13:44.327 INFO Model was created.
CHECKING MINDRECORD FILES ...
[WARNING] PIPELINE(119171,ffff7f7aca40,python3.7):2022-12-08-19:13:46.765.392 [mindspore/ccsrc/pipeline/jit/pipeline.cc:173] CheckArgValid] The data types of Tensor:[[ True False False ... False False False]
[ True False False ... False False False]
[ True False False ... False False False]
...
[ True False False ... False False False]
[ True False False ... False False False]
[ True False False ... False False False]] is bool, which may cause SelectKernelInfo failure for operator [AddN]. For more details, please refer to the FAQ at https://www.mindspore.cn.
[ERROR] ANALYZER(119171,ffff7f7aca40,python3.7):2022-12-08-19:14:10.941.108 [mindspore/ccsrc/pipeline/jit/static_analysis/async_eval_result.cc:66] HandleException] Exception happened, check the information as below.

The function call stack (See file '/home/ma-user/work/vega-master/examples/rank_0/om/analyze_fail.dat' for more details):

0 In file /home/ma-user/.local/lib/python3.7/site-packages/vega/algorithms/nas/sp_nas/src/network_define.py(155)

    if self.reduce_flag:

1 In file /home/ma-user/.local/lib/python3.7/site-packages/vega/algorithms/nas/sp_nas/src/network_define.py(154)

    grads = self.grad(self.network, weights)(x, img_shape, gt_bboxe, gt_label, gt_num, self.sens)
            ^

2 In file /home/ma-user/.local/lib/python3.7/site-packages/vega/algorithms/nas/sp_nas/src/network_define.py(107)

    loss1, loss2, loss3, loss4, loss5, loss6 = self._backbone(x, img_shape, gt_bboxe, gt_label, gt_num)
                                               ^

3 In file /home/ma-user/.local/lib/python3.7/site-packages/vega/networks/mindspore/faster_rcnn/faster_rcnn_resnet.py(267)

    if self.training:

4 In file /home/ma-user/.local/lib/python3.7/site-packages/vega/networks/mindspore/faster_rcnn/faster_rcnn_resnet.py(330)

                                   self.cast(x[0], mstype.float32),
                                             ^

5 In file /home/ma-user/.local/lib/python3.7/site-packages/vega/networks/mindspore/faster_rcnn/fpn_neck.py(102)

    for i in range(self.fpn_layer):

6 In file /home/ma-user/.local/lib/python3.7/site-packages/vega/networks/mindspore/faster_rcnn/fpn_neck.py(111)

    for i in range(self.fpn_layer - 1, -1, -1):
    ^

7 In file /home/ma-user/.local/lib/python3.7/site-packages/vega/networks/mindspore/faster_rcnn/fpn_neck.py(106)

    y = y + (x[2] + self.interpolate1(y[self.fpn_layer - 4]),)
             ^

2022-12-08 19:14:11.8 ERROR Failed to run worker, id: 1, message: For 'Add', x.shape and y.shape are supposed to broadcast, where broadcast means that x.shape[i] = 1 or -1 or y.shape[i] = 1 or -1 or x.shape[i] = y.shape[i], but now x.shape and y.shape can not broadcast, got i: -2, x.shape: [16, 256, 32, 44], y.shape: [16, 256, 48, 80].
2022-12-08 19:14:11.59 INFO Update Success. step_name=serial, worker_id=1
2022-12-08 19:14:11.59 INFO waiting for the workers [1] to finish
2022-12-08 19:14:11.60 INFO Best values: []
2022-12-08 19:14:11.62 WARNING Failed to dump pareto front records, report is emplty.
2022-12-08 19:14:13.74 INFO ------------------------------------------------
2022-12-08 19:14:13.74 INFO Pipeline end.
2022-12-08 19:14:13.74 INFO
2022-12-08 19:14:13.75 INFO task id: 1208.191334.031
2022-12-08 19:14:13.75 INFO output folder: /home/ma-user/work/vega-master/examples/tasks/1208.191334.031/output
2022-12-08 19:14:13.75 INFO
2022-12-08 19:14:13.75 INFO running time:
2022-12-08 19:14:13.76 INFO serial: 0:00:28 [2022-12-08 19:13:42.540940 - 2022-12-08 19:14:11.072154]
2022-12-08 19:14:13.76 INFO
2022-12-08 19:14:13.80 INFO result file output.csv is not existed or empty
2022-12-08 19:14:13.80 INFO ------------------------------------------------
2022-12-08 19:14:16.178 INFO Shutdown urgently.

Exception ignored in: <function _PythonMultiprocessing.del at 0xffff1a6b6710>
Traceback (most recent call last):
File "/home/ma-user/anaconda3/envs/MindSpore/lib/python3.7/site-packages/mindspore/dataset/engine/datasets.py", line 3165, in del
File "/home/ma-user/anaconda3/envs/MindSpore/lib/python3.7/site-packages/mindspore/dataset/engine/datasets.py", line 2841, in terminate
TypeError: 'NoneType' object is not callable

[zhangjiajin]

mindspore的版本号是多少？

[marvellee1]

1.7.0

预训练模型有提供下载的地址么？

[zhangjiajin]

这个预训练模型在这：
https://download.pytorch.org/models/fasterrcnn_resnet50_fpn_coco-258fb6c6.pth