-
## Context
`at::cpp_custom_type_hack` is a deprecated utility to shepherd arbitrary C++ classes through Python by embedding an owning pointer inside a byte Tensor's storage. It's usage is already hea…
-
**Describe the bug 🐞**
I'm getting the error:
```
ERROR: LoadError: MethodError: no method matching ldiv!(::Vector{Float64}, ::LinearAlgebra.QRCompactWY{Float64, Matrix{Float64}, Matrix{Float64…
-
## 🚀 Feature
Currently (as of 1.8.1) torch.nn.quantized.functional.conv1d/2d/3d/linear require the output to always be requantized to 8b. Conv operators ask for an output scale and zero point, while …
-
### Feature request
Currently, when loading a model in quantized form, the `HFQuantizer` is created based on other kwargs passed into the `from_pretrained` function. See current implementation below:…
-
### 🚀 The feature, motivation and pitch
# Motivation
Intel GPUs could significantly improve the workload performance. As described in [[RFC] Intel GPU Upstreaming](https://github.com/pytorch/pytor…
-
The following Linear Operator Library has recently been released, https://cola.readthedocs.io/en/latest/index.html, which allows for lazy evaluation in the following manner,
`lazy_A = cola.fns.lazi…
-
- GC: make sure we are able to GC everything that's GC'able in theory
- [x] GC for hopping windows #1889
- [x] GC for tumbling windows
- [ ] GC for group operators (LAG/LEAD/topk)
- [x] G…
-
Currently, we only take advantage of savings for computing the kernel gram matrix, and computing a gram solve in GP predictions.
We neglect:
- Efficiency of matrix solves and log-determinants…
-
Linear operators should adhere by similar broadcasting rules as ``RandomVariable``s and ``np.ndarray``s. For example
```python
from probnum.linalg.linear_operators import Identity
A = Identity(5)…
-
Currently the only supported format for reading/writing matrices is the matrix market format (see PRs #94 and #98).
There are other, more general, data iterchange formats (JSON, YAML, etc.) which cou…