Uploaded my collected dataset that used mouse inputs.

Roughly 50k image and action pairs, sampled every third frame and smoothed steering with moving average.

center.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef7ae378e6f74b4095b970daa3f5a95d41018df7b3de1d8d55d7931888cdaff1
+size 4052428928

left.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:116e428c19d592defe4c65124b50547c5720f01da6e57412559c349ac458ef00
+size 4052428928

right.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d78d41dae2def02b07baf5461bc3d544bc1c67e8a2cb52cb143fa5efaedc3537
+size 4052428928

steer.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10be7a0b4bf8e1977b2059d1e96fa8d124634944ad8807a1cf1d8a8fcf2f5aa2
+size 211192