Takuma Seno

[WIP] Implement nnabla.experimental.distributions

1

Squashed version of #392 . cc. @TE-AkioHayakawa

Hi, @TE-AkioHayakawa san , @TE-TakuyaNarihira san. I've added `to_grad` option to `F.assign` function. This will be useful to build original your own gradient manipulation with use of `nn.grad`. ```py x...

Fix F.clip_grad_by_norm in the case of small norm

1

The `clip_by_norm` operation takes the maximum value between the L2 norm and the clipping value. For reference, see [tensorflow](https://github.com/tensorflow/tensorflow/blob/r2.1/tensorflow/python/ops/clip_ops.py#L179). cc @TE-AkioHayakawa

Implement nnabla.experimental.distributions modules

1

I implemented distribution modules such as [chainer](https://docs.chainer.org/en/stable/reference/generated/chainer.Distribution.html) and [tensorflow](https://www.tensorflow.org/api_docs/python/tf/distributions). ```py import nnabla as nn from nnabla.experimental.distributions import Normal loc = nn.Variable((2, 3)) scale = nn.Variable((2, 3)) inpt = nn.Variable((2, 3))...

[WIP] Implement backward_all as a counterpart of forward_all

1

Hi, @TE-AkioHayakawa san, @TE-TakuyaNarihira san! I've implemented `nn.backward_all` as a counterpart of `nn.forward_all`. It seems to work correctly. But I have some concerns about this. - implementation design looks a...

SinGAN implemenation

1

Hi, @TE-AkioHayakawa san , @TE-TakuyaNarihira san, @TE-KazukiYoshiyama san. I've made SinGAN with NNabla. If you consider this to be good to this repository, I'll clean up codes and add more...

seed is not actually used in the script

1

Hello, @aviralkumar2907 . Thanks for sharing the CQL code. For the MuJoCo script, I've noticed that `seed` value is not actually used anywhere in the script. https://github.com/aviralkumar2907/CQL/blob/master/d4rl/examples/cql_mujoco_new.py Is this the...

Fix F.clip_grad_by_norm in the case of small norm

1

The operation of `clip_by_norm` must take the maximum value between the L2 norm and the clipping value. See https://github.com/sony/nnabla/pull/572 . cc. @TE-AkioHayakawa