Why I can't get a good result?

[chaiyixuan]

I run the Blood Cell Detection.ipynb, the training loss: 0.0035 ,but the val_loss is 0.2372. The predict result is:

I use the following setting: `LABELS = ['RBC']

IMAGE_H, IMAGE_W = 416, 416 GRID_H, GRID_W = 13 , 13 BOX = 5 CLASS = len(LABELS) CLASS_WEIGHTS = np.ones(CLASS, dtype='float32') OBJ_THRESHOLD = 0.3 NMS_THRESHOLD = 0.3 ANCHORS = [1.3221, 1.73145, 3.19275, 4.00944, 5.05587, 8.09892, 9.47112, 4.84053, 11.2364, 13.0071]

NO_OBJECT_SCALE = 1.0 OBJECT_SCALE = 5.0 COORD_SCALE = 1.0 CLASS_SCALE = 1.0

BATCH_SIZE = 10 WARM_UP_BATCHES = 100 TRUE_BOX_BUFFER = 10

wt_path = 'yolov3.weights'
`

[chaiyixuan]

optimizer = Adam(lr=1e-4, beta_1=0.9, beta_2=0.999, epsilon=1e-08, decay=0.0) with 100 epochs

[experiencor]

You use too many warmup batches. WARM_UP_BATCHES of 3 is enough.