onednnl 0.0.1

high-level bindings to oneDNN Deep Learning library
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37

# To be done

## Primitives completed / to be done

| Primitive         | Forward | Backward | Test | Bench |
| ------------------ | :------: | :------: | :--: | :---: |
| `au_gru`          |    ✅    |    ❌    |  ❌  |  ❌   |
| `batch_norm`      |    ✅    |    ❌    |  ❌  |  ❌   |
| `binary`          |    ✅    |    ⬜    |  ✅  |  ✅   |
| `concat`          |    ❌    |    ⬜    |  ❌  |  ❌   | 
| `convolution`     |    ❌    |    ❌    |  ❌  |  ❌   |
| `deconvolution`   |    ❌    |    ❌    |  ❌  |  ❌   |
| `eltwise`         |    ✅    |    ✅    |  ✅  |  ❌   |
| `gemm`            |    ❌    |    ⬜    |  ❌  |  ❌   | 
| `group_norm`      |    ❌    |    ❌    |  ❌  |  ❌   |
| `gru`             |    ❌    |    ❌    |  ❌  |  ❌   |
| `inner_product`   |    ✅    |    ✅    |  ✅  |  ❌   |
| `layer_norm`      |    ❌    |    ❌    |  ❌  |  ❌   |
| `lbr_augru`       |    ❌    |    ❌    |  ❌  |  ❌   | 
| `lbr_gru`         |    ❌    |    ❌    |  ❌  |  ❌   | 
| `lrn`             |    ❌    |    ❌    |  ❌  |  ❌   |
| `lstm`            |    ❌    |    ❌    |  ❌  |  ❌   |
| `matmul`          |    ✅    |    ⬜    |  ✅  |  ❌   |
| `pooling`         |    ❌    |    ❌    |  ❌  |  ❌   |
| `prelu`           |    ✅    |    ❌    |  ❌  |  ❌   |
| `reduction`       |    ✅    |    ⬜    |  ✅  |  ❌   | 
| `reorder`         |    ❌    |    ⬜    |  ❌  |  ❌   | 
| `resampling`      |    ❌    |    ❌    |  ❌  |  ❌   |
| `shuffle`         |    ❌    |    ❌    |  ❌  |  ❌   |
| `softmax`         |    ❌    |    ❌    |  ❌  |  ❌   |
| `sum`             |    ❌    |    ⬜    |  ❌  |  ❌   | 
| `vanilla_rnn`     |    ❌    |    ❌    |  ❌  |  ❌   |

## Known Issues

- Missing support for GPU (SYCL and OpenCL)
- Missing postops