增加RecModel配合WideResNet代码以及参数转换成RecModel

docs/zh_CN/training/semi_supervised_learning/FixMatchCCSSL.md

@@ -71,7 +71,7 @@ cifar10数据在训练过程中会自动下载到默认缓存路径 `~/.cache/pa
 
 单卡训练执行以下命令
 ```
-python tools/train.py -c ppcls/configs/ssl/FixMatch_CCSSL_cifar10_4000.yaml
+python tools/train.py -c ppcls/configs/ssl/FixMatchCCSSL/FixMatchCCSSL_cifar10_4000.yaml
 ```
 
 4卡训练执行以下操作
@@ -86,7 +86,7 @@ python -m paddle.distributed.launch --gpus='0,1,2,3' tools/train.py -c ppcls/con
 准备用于评估的 `*.pdparams` 模型参数文件，可以使用训练好的模型，也可以使用 *4. 模型训练* 中保存的模型。
 * 以训练过程中保存的 `best_model_ema.ema.pdparams`为例，执行如下命令即可进行评估。
 ```
-python3.7 tools/eval.py -c ppcls/configs/ssl/FixMatch_CCSSL_cifar10_4000.yaml -o Global.pretrained_model="./output/WideResNetCCSSL/best_model_ema.ema"
+python3.7 tools/eval.py -c ppcls/configs/ssl/FixMatchCCSSL/FixMatchCCSSL_cifar10_4000_4gpu.yaml -o Global.pretrained_model="./output/WideResNet/best_model_ema.ema"
 ```
 
 * 以训练好的模型为例，下载提供的已经训练好的模型，到 `PaddleClas/pretrained_models` 文件夹中，执行如下命令即可进行评估。

ppcls/arch/__init__.py

@@ -85,10 +85,18 @@ class RecModel(TheseusLayer):
             self.head = build_gear(config["Head"])
         else:
             self.head = None
+        
+        if "Decoup" in config:
+            self.decoup = build_gear(config['Decoup'])
+        else:
+            self.decoup = None
 
     def forward(self, x, label=None):
+        
         out = dict()
         x = self.backbone(x)
+        if self.decoup is not None:
+            return self.decoup(x)
         out["backbone"] = x
         if self.neck is not None:
             x = self.neck(x)

ppcls/arch/backbone/model_zoo/wideresnet.py

@@ -201,8 +201,6 @@ class Wide_ResNet(nn.Layer):
         feat = self.relu(self.bn1(feat))
         feat = F.adaptive_avg_pool2d(feat, 1)
         feat = paddle.reshape(feat, [-1, self.channels])
-        if not self.training:
-            return self.fc(feat)
             
         if self.proj:
             pfeat = self.fc1(feat)

ppcls/arch/gears/__init__.py

@@ -20,6 +20,7 @@ from .vehicle_neck import VehicleNeck
 from paddle.nn import Tanh
 from .bnneck import BNNeck
 from .adamargin import AdaMargin
+from .decoup import Decoup
 
 __all__ = ['build_gear']
 
@@ -27,7 +28,7 @@ __all__ = ['build_gear']
 def build_gear(config):
     support_dict = [
         'ArcMargin', 'CosMargin', 'CircleMargin', 'FC', 'VehicleNeck', 'Tanh',
-        'BNNeck', 'AdaMargin'
+        'BNNeck', 'AdaMargin', 'FRFBNeck', 'Decoup'
     ]
     module_name = config.pop('name')
     assert module_name in support_dict, Exception(

ppcls/arch/gears/decoup.py

+import paddle
+import paddle.nn as nn
+
+
+class Decoup(nn.Layer):
+    def __init__(self, logits_index, features_index, **kwargs):
+        super(Decoup, self).__init__()
+        self.logits_index = logits_index
+        self.features_index = features_index
+
+
+    def forward(self, out, **kwargs):
+        assert isinstance(out, (list, tuple)), 'out must  be list or tuple'
+        out = {'logits': out[self.logits_index], 'features':out[self.features_index]}
+        return out
+

ppcls/configs/ssl/FixMatchCCSSL/FixMatchCCSSL_cifar10_4000_4gpu.yaml

@@ -23,14 +23,23 @@ EMA:
   decay: 0.999
 
 Arch:
-  name: WideResNet
-  widen_factor: 2
-  depth: 28
-  dropout: 0     # CCSSL为 drop_rate
-  num_classes: &sign_num_classes 10
-  low_dim: 64
-  proj: true
-  proj_after: false
+  name: RecModel
+  infer_output_key: logits
+  infer_add_softmax: false
+  Backbone:
+    name: WideResNet
+    widen_factor: 2
+    depth: 28
+    dropout: 0     # CCSSL为 drop_rate
+    num_classes: &sign_num_classes 10
+    low_dim: 64
+    proj: true
+    proj_after: false
+
+  Decoup:
+    name: Decoup
+    logits_index: 0
+    features_index: 1 
 
   use_sync_bn: true
 

ppcls/engine/train/train_fixmatch_ccssl.py

@@ -10,6 +10,7 @@ from ppcls.engine.train.utils import update_loss, update_metric, log_info
 from ppcls.utils import profiler
 from paddle.nn import functional as F
 import numpy as np
+import paddle
 # from reprod_log import ReprodLogger
 
 
@@ -20,6 +21,9 @@ def train_epoch_fixmatch_ccssl(engine, epoch_id, print_batch_step):
     # epoch = 0
     ##############################################################
 
+    paddle.save(engine.model.state_dict(), '../recmodel.pdparams')
+
+    assert 1==0
     tic = time.time()
     if not hasattr(engine, 'train_dataloader_iter'):
         engine.train_dataloader_iter = iter(engine.train_dataloader)
@@ -81,7 +85,7 @@ def train_epoch_fixmatch_ccssl(engine, epoch_id, print_batch_step):
         inputs_w, inputs_s1, inputs_s2 = unlabel_data_batch
 
         batch_size_label = inputs_x.shape[0]
-        inputs = paddle.concat([inputs_x, inputs_w, inputs_s1, inputs_s2])
+        inputs = paddle.concat([inputs_x, inputs_w, inputs_s1, inputs_s2], axis=0)
 
         loss_dict, logits_label = get_loss(engine, inputs, batch_size_label, 
                                            temperture, threshold, targets_x,
@@ -134,8 +138,9 @@ def get_loss(engine,
              targets_x,
              **kwargs
              ):
-
-    logits, feats = engine.model(inputs)
+    out = engine.model(inputs)
+    logits, feats = out['logits'], out['features']
+    # logits, feats = engine.model(inputs)
     feat_w, feat_s1, feat_s2 = feats[batch_size_label:].chunk(3)
     feat_x = feats[:batch_size_label]
     logits_x = logits[:batch_size_label]