fix eval script (#464)

* fix eval script

* fix dali shell

ppcls/modeling/architectures/hrnet.py

@@ -179,7 +179,8 @@ class Branches(nn.Layer):
         outs = []
         for idx, input in enumerate(inputs):
             conv = input
-            for basic_block_func in self.basic_block_list[idx]:
+            basic_block_list = self.basic_block_list[idx]
+            for basic_block_func in basic_block_list:
                 conv = basic_block_func(conv)
             outs.append(conv)
         return outs

ppcls/modeling/architectures/resnest.py

@@ -165,6 +165,7 @@ class SplatConv(nn.Layer):
 
         atten = self.conv3(gap)
         atten = self.rsoftmax(atten)
+        atten = paddle.reshape(x=atten, shape=[-1, atten.shape[1], 1, 1])
 
         if self.radix > 1:
             attens = paddle.split(atten, num_or_sections=self.radix, axis=1)

ppcls/utils/misc.py

@@ -48,12 +48,12 @@ class AverageMeter(object):
 
     @property
     def total_minute(self):
-        return '{self.name}_sum: {s:{self.fmt}}{self.postfix} min'.format(
+        return '{self.name} {s:{self.fmt}}{self.postfix} min'.format(
             s=self.sum / 60, self=self)
 
     @property
     def mean(self):
-        return '{self.name}_avg: {self.avg:{self.fmt}}{self.postfix}'.format(
+        return '{self.name}: {self.avg:{self.fmt}}{self.postfix}'.format(
             self=self) if self.need_avg else ''
 
     @property

tools/eval.py

@@ -67,9 +67,9 @@ def main(args, return_dict={}):
     init_model(config, net, optimizer=None)
     valid_dataloader = Reader(config, 'valid', places=place)()
     net.eval()
-
-    top1_acc = program.run(valid_dataloader, config, net, None, None, 0,
-                           'valid')
+    with paddle.no_grad():
+        top1_acc = program.run(valid_dataloader, config, net, None, None, 0,
+                               'valid')
     return_dict["top1_acc"] = top1_acc
     return top1_acc
 

tools/eval.sh

-python -m paddle.distributed.launch \
-    --selected_gpus="0" \
+python3.7 -m paddle.distributed.launch \
+    --gpus="0,1,2,3" \
     tools/eval.py \
-        -c ./configs/eval.yaml \
-        -o load_static_weights=True \
-        -o use_gpu=False
+        -c ./configs/ResNet/ResNet50.yaml \
+        -o pretrained_model="./ResNet50_pretrained" \
+        -o use_gpu=True

tools/program.py

@@ -298,6 +298,11 @@ def run(dataloader,
 
     tic = time.time()
     for idx, batch in enumerate(dataloader()):
+        # avoid statistics from warmup time
+        if idx == 10:
+            metric_list["batch_time"].reset()
+            metric_list["reader_time"].reset()
+
         metric_list['reader_time'].update(time.time() - tic)
         batch_size = len(batch[0])
         feeds = create_feeds(batch, use_mix)
@@ -327,11 +332,15 @@ def run(dataloader,
         metric_list["batch_time"].update(time.time() - tic)
         tic = time.time()
 
-        fetchs_str = ' '.join([str(m.value) for m in metric_list.values()])
+        fetchs_str = ' '.join([
+            str(metric_list[key].mean)
+            if "time" in key else str(metric_list[key].value)
+            for key in metric_list
+        ])
 
         if idx % print_interval == 0:
             ips_info = "ips: {:.5f} images/sec.".format(
-                batch_size / metric_list["batch_time"].val)
+                batch_size / metric_list["batch_time"].avg)
             if mode == 'eval':
                 logger.info("{:s} step:{:<4d}, {:s} {:s}".format(
                     mode, idx, fetchs_str, ips_info))

tools/run.sh

 #!/usr/bin/env bash
 
-python -m paddle.distributed.launch \
-    --selected_gpus="0,1,2,3" \
+python3.7 -m paddle.distributed.launch \
+    --gpus="0,1,2,3" \
     tools/train.py \
         -c ./configs/ResNet/ResNet50.yaml \
         -o print_interval=10

tools/static/run_dali.sh

@@ -4,7 +4,7 @@ export CUDA_VISIBLE_DEVICES="0,1,2,3"
 export FLAGS_fraction_of_gpu_memory_to_use=0.80
 
 python3.7 -m paddle.distributed.launch \
-    --selected_gpus="0,1,2,3" \
+    --gpus="0,1,2,3" \
     tools/static/train.py \
         -c ./configs/ResNet/ResNet50.yaml \
         -o print_interval=10 \

tools/train.py

@@ -92,8 +92,9 @@ def main(args):
         # 2. validate with validate dataset
         if config.validate and epoch_id % config.valid_interval == 0:
             net.eval()
-            top1_acc = program.run(valid_dataloader, config, net, None, None,
-                                   epoch_id, 'valid')
+            with paddle.no_grad():
+                top1_acc = program.run(valid_dataloader, config, net, None,
+                                       None, epoch_id, 'valid')
             if top1_acc > best_top1_acc:
                 best_top1_acc = top1_acc
                 best_top1_epoch = epoch_id