pytensor
pytensor copied to clipboard
PyTensor allows you to define, optimize, and efficiently evaluate mathematical expressions involving multi-dimensional arrays.
## Description This PR adds mlx dispatches for the following Ops: - `nlinalg.kron` - `nlinalg.svd` - `nlinalg.inv` - `nlinalg.pinv` - `slinalg.cholesky` - `slinalg.solve` - `slinalg.solve_triangular` - `slinalg.lu` Most (all?) of...
### Description Missing: * `A@B + A@C = A@(B+C)` (one less matmul) * `s*A @ B = s*(A@B)` (which can be done by a single gemm routine) ```python from pytensor.graph...
Bumps [actions/upload-artifact](https://github.com/actions/upload-artifact) from 4.6.2 to 5.0.0. Release notes Sourced from actions/upload-artifact's releases. v5.0.0 What's Changed BREAKING CHANGE: this update supports Node v24.x. This is not a breaking change per-se but...
Bumps [actions/download-artifact](https://github.com/actions/download-artifact) from 5.0.0 to 6.0.0. Release notes Sourced from actions/download-artifact's releases. v6.0.0 What's Changed BREAKING CHANGE: this update supports Node v24.x. This is not a breaking change per-se but...
Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3 to 4. Release notes Sourced from github/codeql-action's releases. v3.30.8 CodeQL Action Changelog See the releases page for the relevant changes to the CodeQL CLI and language...
### Describe the issue: Reported in https://github.com/pymc-devs/pytensor/pull/1667#issuecomment-3447731251 If MKL detection fails, then we see the following new error: > Mingw-w64 runtime failure: > 32 bit pseudo relocation at 00007FF6E32314BC out...
I noticed cache clear wasn't working and always used purge instead. It was a mistake introduced by 671a821d5d6ce33eb8ff3baa05d392c99da4133f (downstream of https://github.com/aesara-devs/aesara/pull/1312). This changed many time.time() to to time.perf_counter(). But here...
Natural follow up to #1318 is to also get `convolve2d` in. This PR is just to get the ball rolling, it's just a super dumb wrapper around `scipy.signal.convolve2d`. Remainings TODO:...
### Description The following raises: ```py import pytensor.tensor as pt import numpy as np x = pt.dmatrix('x') x_val = np.arange(9).reshape((3, 3)) out = pt.subtensor.inc_subtensor(x[:, :2], 10) out.eval({x:x_val}, mode='MLX') ``` Traceback...
### Description ```python import pytensor import pytensor.tensor as pt x0 = pt.scalar("x") xs, _ = pytensor.scan(lambda x: x+1, outputs_info=[x0], n_steps=4) fn = pytensor.function([x0], xs) fn.dprint() # Scan still in the...