Search results

MLX backend POC

backend compatibility

enhancementNew feature or requestNew feature or request

Status: Open (in progress).

pymc-devs/pytensornumber 1365

#1365 In pymc-devs/pytensor;

· williambdean opened on Apr 11, 2025

Add <code>C</code> and <code>Python</code> modes

backend compatibility

Status: Open.

#1274 In pymc-devs/pytensor;

· ricardoV94 opened on Mar 7, 2025

 ·  

3.0 release

Port SymbolicRandomVariable Op to PyTensor and get rid of many RVs in PyTensor

backend compatibility

OpFromGraph

random variables

refactor

Status: Open.

#1221 In pymc-devs/pytensor;

· ricardoV94 opened on Feb 19, 2025

Add numba overload for <code>scipy.linalg.svd</code>

backend compatibility

linalgLinear algebraLinear algebra

numba

SciPy compatibility

Status: Open.

#1197 In pymc-devs/pytensor;

· jessegrabowski opened on Feb 6, 2025

Don't force <code>.cpu()</code> on all PyTorch outputs

backend compatibility

enhancementNew feature or requestNew feature or request

torchPyTorch backendPyTorch backend

Status: Open.

#1052 In pymc-devs/pytensor;

· ricardoV94 opened on Oct 29, 2024

Proposal: Make compiler optional and fall-back to numba or JAX

backend compatibility

C-backend

installation

needs infoAdditional information requiredAdditional information required

Status: Open.

#1027 In pymc-devs/pytensor;

· twiecki opened on Oct 10, 2024

AdvancedSubtensor1 and AdvancedIncSubtensor1 are not needed in JAX and Numba backends

backend compatibility

graph rewriting

Status: Open.

#838 In pymc-devs/pytensor;

· ricardoV94 opened on Jun 21, 2024

Default typify in JITLinkers expects tensors/arrays, why?

backend compatibility

Status: Open.

#819 In pymc-devs/pytensor;

· ricardoV94 opened on Jun 14, 2024

Implement <code>lu_factor</code> and <code>lu_solve</code> for Blockwise <code>Solve</code> rewrites

backend compatibility

feature request

graph rewriting

linalgLinear algebraLinear algebra

performance

SciPy compatibility

vectorization

Status: Open.

#582 In pymc-devs/pytensor;

· jessegrabowski opened on Jan 10, 2024

Implement helper <code>@as_jax_op</code> to wrap JAX functions in PyTensor

backend compatibility

feature request

jax

torchPyTorch backendPyTorch backend

Status: Open.

#537 In pymc-devs/pytensor;

· ricardoV94 opened on Dec 7, 2023

No JAX dispatch for <code>mul_without_zeros</code>

backend compatibility

enhancementNew feature or requestNew feature or request

jax

Status: Open.

#526 In pymc-devs/pytensor;

· jessegrabowski opened on Nov 30, 2023

Implement proper JAX/Numba dispatch for IfElse

backend compatibility

jax

numba

Status: Open.

#501 In pymc-devs/pytensor;

· ricardoV94 opened on Nov 15, 2023