-
Heyo, in NePS, when trying out EI instead of PI, I got an issue where the shape output by `ei` was actually just duplicated vstacked columns, where as for `pi` and `ucb`, it was a single column output…
-
Hi,
I have implemented the forward pass using a convolution + sigmoid_fwd activation and am now working on the backpropagation of the graph. However, according to the [document](https://docs.nvidia.…
-
There seem to be inconsistencies in TTNN IR op layout inputs and outputs. Currently all op inputs are set to TILE layout, while all outputs are ROW_MAJOR. Example:
```mlir
#device = #tt.device, l1Ma…
-
**Describe the bug**
Currently I use BFLOAT4_B to emulate the quantization type Q4_0 in my GGML backend. However currently it seems I cannot cast BF16 into BFLOAT4_B. Is this intended?
https://git…
-
### Description
Using `Tensor.ReverseDimension(slice, dimension)` on a `TensorSpan` created via `TensorSpan.Slice` creates a tensor with partially invalid values - in this case, the source tensor has…
-
### Search before asking
- [X] I have searched the Ultralytics YOLO [issues](https://github.com/ultralytics/ultralytics/issues) and [discussions](https://github.com/ultralytics/ultralytics/discussi…
-
### 📚 The doc issue
[The doc](https://pytorch.org/docs/stable/generated/torch.linalg.norm.html) of `linalg.norm()` says there is `A` parameter for `linalg.norm()` as shown below:
> torch.linalg.no…
-
Is there a reason this isn't present in linear-base? I know you can't make & (co)datatypes directly in Haskell and have to encode them, but it seems like it's worth having. I think I read somewhere th…
-
RuntimeError: The size of tensor a (96) must match the size of tensor b (144) at non-singleton dimension 1
-
Following the notebook file, I tried to construct a DMPNN model for tox21.
Training for some reason takes way longer than the pytorch implementation onn chemprop.
So I tried to parallelize it in mul…