Hotfix paralle nccl args (#903)

* fix parallel

* typo

* fix data best_map parameter bug

fluid/object_detection/README.md

@@ -52,22 +52,22 @@ Declaration: the MobileNet-v1 SSD model is converted by [TensorFlow model](https
 #### Train on PASCAL VOC
   - Train on one device (/GPU).
   ```python
-  env CUDA_VISIABLE_DEVICES=0 python -u train.py --parallel=False --dataset='pascalvoc' --pretrained_model='pretrained/ssd_mobilenet_v1_coco/'
+  env CUDA_VISIBLE_DEVICES=0 python -u train.py --parallel=False --dataset='pascalvoc' --pretrained_model='pretrained/ssd_mobilenet_v1_coco/'
   ```
   - Train on multi devices (/GPUs).
 
   ```python
-  env CUDA_VISIABLE_DEVICES=0,1 python -u train.py --batch_size=64 --dataset='pascalvoc' --pretrained_model='pretrained/ssd_mobilenet_v1_coco/'
+  env CUDA_VISIBLE_DEVICES=0,1 python -u train.py --batch_size=64 --dataset='pascalvoc' --pretrained_model='pretrained/ssd_mobilenet_v1_coco/'
   ```
 
 #### Train on MS-COCO
   - Train on one device (/GPU).
   ```python
-  env CUDA_VISIABLE_DEVICES=0 python -u train.py --parallel=False --dataset='coco2014' --pretrained_model='pretrained/mobilenet_imagenet/'
+  env CUDA_VISIBLE_DEVICES=0 python -u train.py --parallel=False --dataset='coco2014' --pretrained_model='pretrained/mobilenet_imagenet/'
   ```
   - Train on multi devices (/GPUs).
   ```python
-  env CUDA_VISIABLE_DEVICES=0,1 python -u train.py --batch_size=64 --dataset='coco2014' --pretrained_model='pretrained/mobilenet_imagenet/'
+  env CUDA_VISIBLE_DEVICES=0,1 python -u train.py --batch_size=64 --dataset='coco2014' --pretrained_model='pretrained/mobilenet_imagenet/'
   ```
 
 TBD
@@ -90,13 +90,13 @@ Note we set the defualt test list to the dataset's test/val list, you can use yo
 
 #### Evaluate on PASCAL VOC
 ```python
-env CUDA_VISIABLE_DEVICES=0 python eval.py --dataset='pascalvoc' --model_dir='train_pascal_model/90' --data_dir='data/pascalvoc' --test_list='test.txt' --ap_version='11point'
+env CUDA_VISIBLE_DEVICES=0 python eval.py --dataset='pascalvoc' --model_dir='train_pascal_model/90' --data_dir='data/pascalvoc' --test_list='test.txt' --ap_version='11point'
 ```
 
 #### Evaluate on MS-COCO
 ```python
-env CUDA_VISIABLE_DEVICES=0 python eval.py --dataset='coco2014' --nms_threshold=0.5 --model_dir='train_coco_model/40' --test_list='annotations/instances_minival2014.json' --ap_version='integral'
-env CUDA_VISIABLE_DEVICES=0 python eval_coco_map.py --dataset='coco2017' --nms_threshold=0.5 --model_dir='train_coco_model/40' --test_list='annotations/instances_minival2017.json'
+env CUDA_VISIBLE_DEVICES=0 python eval.py --dataset='coco2014' --nms_threshold=0.5 --model_dir='train_coco_model/40' --test_list='annotations/instances_minival2014.json' --ap_version='integral'
+env CUDA_VISIBLE_DEVICES=0 python eval_coco_map.py --dataset='coco2017' --nms_threshold=0.5 --model_dir='train_coco_model/40' --test_list='annotations/instances_minival2017.json'
 ```
 
 TBD
@@ -104,7 +104,7 @@ TBD
 ### Infer and Visualize
 
 ```python
-env CUDA_VISIABLE_DEVICES=0 python infer.py --model_dir='train_coco_model/20' --image_path='./data/coco/val2014/COCO_val2014_000000000139.jpg'
+env CUDA_VISIBLE_DEVICES=0 python infer.py --model_dir='train_coco_model/20' --image_path='./data/coco/val2014/COCO_val2014_000000000139.jpg'
 ```
 Below is the examples after running python infer.py to inference and visualize the model result.
 <p align="center">

fluid/object_detection/eval_coco_map.py

@@ -69,7 +69,7 @@ def eval(args, data_args, test_list, batch_size, model_dir=None):
         place=place,
         feed_list=[image, gt_box, gt_label, gt_iscrowd, gt_image_info])
 
-    def get_dt_res(nmsed_out_v):
+    def get_dt_res(nmsed_out_v, data):
         dts_res = []
         lod = nmsed_out_v[0].lod()[0]
         nmsed_out_v = np.array(nmsed_out_v[0])
@@ -100,6 +100,7 @@ def eval(args, data_args, test_list, batch_size, model_dir=None):
                     'score': score
                 }
                 dts_res.append(dt_res)
+        return dts_res
 
     def test():
         dts_res = []
@@ -111,7 +112,7 @@ def eval(args, data_args, test_list, batch_size, model_dir=None):
                                   return_numpy=False)
             if batch_id % 20 == 0:
                 print("Batch {0}".format(batch_id))
-            dts_res += get_dt_res(nmsed_out_v)
+            dts_res += get_dt_res(nmsed_out_v, data)
 
         with open("detection_result.json", 'w') as outfile:
             json.dump(dts_res, outfile)

fluid/object_detection/train.py

@@ -18,6 +18,8 @@ add_arg('learning_rate',    float, 0.001,     "Learning rate.")
 add_arg('batch_size',       int,   32,        "Minibatch size.")
 add_arg('num_passes',       int,   120,       "Epoch number.")
 add_arg('use_gpu',          bool,  True,      "Whether use GPU.")
+add_arg('parallel',         bool,  True,      "Parallel.")
+add_arg('use_nccl',         bool,  True,      "NCCL.")
 add_arg('dataset',          str,   'pascalvoc', "coco2014, coco2017, and pascalvoc.")
 add_arg('model_save_dir',   str,   'model',     "The path to save model.")
 add_arg('pretrained_model', str,   'pretrained/ssd_mobilenet_v1_coco/', "The init model path.")
@@ -274,6 +276,7 @@ def parallel_exe(args,
             best_map = test_map[0]
             save_model('best_model')
         print("Pass {0}, test map {1}".format(pass_id, test_map[0]))
+        return best_map
 
     for pass_id in range(num_passes):
         start_time = time.time()
@@ -295,7 +298,7 @@ def parallel_exe(args,
             if batch_id % 20 == 0:
                 print("Pass {0}, batch {1}, loss {2}, time {3}".format(
                     pass_id, batch_id, loss_v, start_time - prev_start_time))
-        test(pass_id, best_map)
+        best_map = test(pass_id, best_map)
         if pass_id % 10 == 0 or pass_id == num_passes - 1:
             save_model(str(pass_id))
     print("Best test map {0}".format(best_map))