Split torch ops into separate files

Fixes sbrunk#29

Bonus additions:

- Add `torch.multinomial`
- Add `torch.oneHot`
- Add `torch.gradient`

core/src/main/scala/torch/nn/functional/sparse.scala

@@ -0,0 +1,28 @@
+/*
+ * Copyright 2022 storch.dev
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package torch
+package nn
+package functional
+
+import org.bytedeco.pytorch.global.torch as torchNative
+
+/** Takes LongTensor with index values of shape `(*)` and returns a tensor of shape `(*,
+  * numClasses)` that have zeros everywhere except where the index of last dimension matches the
+  * corresponding value of the input tensor, in which case it will be 1.
+  */
+def oneHot(input: Tensor[Int64], numClasses: Long = -1): Tensor[Int64] =
+  Tensor(torchNative.one_hot(input.native, numClasses))

core/src/main/scala/torch/ops/BLASOps.scala

@@ -0,0 +1,25 @@
+/*
+ * Copyright 2022 storch.dev
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package torch
+
+/** BLAS and LAPACK Operations
+  *
+  * https://pytorch.org/docs/stable/torch.html#blas-and-lapack-operations
+  */
+
+def matmul[D1 <: DType, D2 <: DType](t1: Tensor[D1], t2: Tensor[D2]): Tensor[Promoted[D1, D2]] =
+  t1.matmul(t2)

core/src/main/scala/torch/ops/ComparisonOps.scala

@@ -0,0 +1,33 @@
+/*
+ * Copyright 2022 storch.dev
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package torch
+
+import org.bytedeco.pytorch.global.torch as torchNative
+
+/** Comparison Ops
+  *
+  * https://pytorch.org/docs/stable/torch.html#comparison-ops
+  */
+
+def allclose(
+    input: Tensor[?],
+    other: Tensor[?],
+    rtol: Double = 1e-05,
+    atol: Double = 1e-08,
+    equalNan: Boolean = false
+) =
+  torchNative.allclose(input.native, other.native, rtol, atol, equalNan)