Fix RoIFeatureTransform Methods with ArgMax

Summary:
For RoIFeatureTransform Methods with ArgMax (RoIPoolF/RoIPoolLateQuantization), the current pipeline will report
```
TypeError: I/O blob #0 of unsupported type: (BlobReference("gpu_0/roi_features"), BlobReference("gpu_0/_argmax_roi_features")) of type <type 'tuple'>
```
as in the example training: https://our.intern.facebook.com/intern/fblearner/details/51149202/

Reviewed By: viswanathgs

Differential Revision: D7907518

fbshipit-source-id: 38caa51c955f5a577d11c04688b9061cb6dd3c22

detectron/modeling/detector.py

@@ -298,7 +298,7 @@ class DetectionModelHelper(cnn.CNNModelHelper):
                 sampling_ratio=sampling_ratio
             )
         # Only return the first blob (the transformed features)
-        return xform_out
+        return xform_out[0] if isinstance(xform_out, tuple) else xform_out
 
     def ConvShared(
         self,