Lines Matching full:backward

20 (options) -> (inputs, params, forward, backward_setup, backward)
27     Then, we pass backward_inputs to backward. If None, then it is assumed to
29 backward: Given `output = backward_setup(*forward(*inputs))`, performs
32 fastrnns.bench times the forward and backward invocations.
37     "ModelDef", ["inputs", "params", "forward", "backward_setup", "backward"]
55     return output.backward(grad_output, **kwargs)
65         backward=simple_backward,
77         backward=simple_backward,
106         backward=simple_backward,
134         backward=simple_backward,
146         backward=simple_backward,
158         backward=simple_backward,
170         backward=simple_backward,
182         backward=simple_backward,
197             backward=simple_backward,
253         # Un-padding the output also makes the backward pass 2x slower...
262         backward=simple_backward,
314         backward=simple_backward,
322 # is invariant), the lowerbound of backward pass is hard to get since we lose the
352         backward=None,
470 …r all timesteps instead of going step-by-step, results in a single reduction kernel in the backward
471 …# FIXME matmul(x,y) + bias currently goes through jit AD, and backward formula in AD is not optimi…