modify some network scripts

official/cv/centerface/src/centerface.py

@@ -233,10 +233,8 @@ class TrainingWrapper(nn.TrainOneStepWithLossScaleCell):
         if self.reducer_flag:
             grads = self.grad_reducer(grads)
         cond = self.get_overflow_status(status, grads)
-        overflow = self.process_loss_scale(cond)
         ret = (loss, cond, sens)
-        if not overflow:
-            self.optimizer(grads)
+        self.optimizer(grads)
         return ret
 
 

official/cv/retinanet/src/lr_schedule.py

@@ -45,7 +45,7 @@ def get_lr(global_step, lr_init, lr_end, lr_max, warmup_epochs1, warmup_epochs2,
     warmup_steps4 = warmup_steps3 + steps_per_epoch * warmup_epochs4
     warmup_steps5 = warmup_steps4 + steps_per_epoch * warmup_epochs5
     step_radio = [1e-4, 1e-3, 1e-2, 0.1]
-    if config.finetune:
+    if hasattr(config, finetune) and config.finetune:
         step_radio = [1e-4, 1e-2, 0.1, 1]
     for i in range(total_steps):
         if i < warmup_steps1:

official/cv/retinanet/train.py

@@ -127,7 +127,8 @@ def main():
     config.lr_end_rate = ast.literal_eval(config.lr_end_rate)
     device_id = get_device_id()
     if config.device_target == "Ascend":
-        context.set_context(mempool_block_size="31GB")
+        if context.get_context("mode") == context.PYNATIVE_MODE:
+            context.set_context(mempool_block_size="31GB")
     elif config.device_target == "GPU":
         set_graph_kernel_context(config.device_target)
     elif config.device_target == "CPU":
@@ -138,7 +139,7 @@ def main():
     context.set_context(mode=context.GRAPH_MODE, device_target=config.device_target)
     if config.distribute:
         init()
-        device_num = config.device_num
+        device_num = get_device_num()
         rank = get_rank()
         context.reset_auto_parallel_context()
         context.set_auto_parallel_context(parallel_mode=ParallelMode.DATA_PARALLEL, gradients_mean=True)
@@ -163,7 +164,7 @@ def main():
     retinanet = retinanet50(backbone, config)
     net = retinanetWithLossCell(retinanet, config)
     init_net_param(net)
-    if config.finetune:
+    if hasattr(config, "finetune") and config.finetune:
         init_net_param(net, initialize_mode='XavierUniform')
     else:
         init_net_param(net)

research/cv/glore_res/train.py

@@ -46,7 +46,7 @@ from src.save_callback import SaveCallback
 
 if config.isModelArts:
     import moxing as mox
-if config.net == 'resnet200' or config.net == 'resnet101':
+if config.net == 'resnet200' or config.net == 'resnet101' or config.net == 'resnet50':
     if config.device_target == "GPU":
         config.cast_fp16 = False
 

research/cv/yolov3_tiny/scripts/run_distribute_train.sh

@@ -80,6 +80,7 @@ do
         --training_shape=640 \
         --weight_decay=0.016 \
         --loss_scale=1024 \
+        --num_parallel_workers=32 \
         --lr_scheduler=cosine_annealing  > log.txt 2>&1 &
     cd ..
 done

research/cv/yolov3_tiny/scripts/run_standalone_train.sh

@@ -67,5 +67,6 @@ python train.py \
     --training_shape=640 \
     --per_batch_size=32 \
     --weight_decay=0.016 \
+    --num_parallel_workers=32 \
     --lr_scheduler=cosine_annealing > log.txt 2>&1 &
 cd ..

research/cv/yolov3_tiny/src/yolo_dataset.py

@@ -316,9 +316,9 @@ def create_yolo_dataset(
                 CV.Normalize(mean, std),
                 hwc_to_chw
             ],
-            num_parallel_workers=num_parallel_workers
+            num_parallel_workers=8
         )
-        ds = ds.batch(batch_size, num_parallel_workers=num_parallel_workers, drop_remainder=True)
+        ds = ds.batch(batch_size, num_parallel_workers=8, drop_remainder=True)
     else:
         ds = de.GeneratorDataset(
             yolo_dataset,
@@ -331,9 +331,9 @@ def create_yolo_dataset(
             input_columns=["image", "img_id"],
             output_columns=["image", "image_shape", "img_id"],
             column_order=["image", "image_shape", "img_id"],
-            num_parallel_workers=num_parallel_workers
+            num_parallel_workers=8
         )
-        ds = ds.map(operations=hwc_to_chw, input_columns=["image"], num_parallel_workers=num_parallel_workers)
+        ds = ds.map(operations=hwc_to_chw, input_columns=["image"], num_parallel_workers=8)
         ds = ds.batch(batch_size, drop_remainder=True)
     ds = ds.repeat(max_epoch)
     return ds, len(yolo_dataset)