Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / dgasmith/opt_einsum issues and pull requests

#235 - `contract_path` with ellipsis fails when `shapes=True`

Issue - State: open - Opened by nova77 15 days ago - 1 comment

#234 - opt_einsum.contract is slower than raw numpy.einsum

Issue - State: open - Opened by lin-cp about 1 month ago - 3 comments

#233 - Alternative Greedy Algorithms

Issue - State: open - Opened by dgasmith about 1 month ago - 1 comment

#232 - Finalizes typing and updates documentation

Pull Request - State: closed - Opened by dgasmith about 1 month ago - 1 comment

#231 - Document state of PyTorch upstreaming in the README

Issue - State: open - Opened by bmillwood about 1 month ago - 2 comments

#230 - Github identifier format in docs/source/conf.py

Issue - State: open - Opened by NGC2023 about 2 months ago - 1 comment

#229 - handle empty contraction list in PathInfo

Pull Request - State: closed - Opened by jcmgray about 2 months ago - 1 comment

#228 - add jaxlib alias

Pull Request - State: closed - Opened by jcmgray about 2 months ago - 1 comment

#227 - Environment and MyPy updates

Pull Request - State: closed - Opened by dgasmith about 2 months ago - 1 comment

#226 - Sevier county Tennessee

Issue - State: closed - Opened by BMOHARRIS 4 months ago

#225 - ARREST

Issue - State: closed - Opened by BMOHARRIS 4 months ago

#224 - `contract` failed to find required methods from `jax.numpy` backend

Issue - State: closed - Opened by zazbone 5 months ago - 4 comments

#223 - Setup.py says install requires Numpy >= 1.7

Issue - State: closed - Opened by chabazite 7 months ago - 1 comment

#221 - New release with Python 3.12 support?

Issue - State: open - Opened by singingwolfboy 8 months ago - 1 comment

#220 - Sub-optimal contraction path when using broadcasting

Issue - State: open - Opened by pimdh 9 months ago

#219 - contract_path with `optimize=bool` raises

Issue - State: open - Opened by ev-br 10 months ago - 3 comments

#218 - contract_expression inconsistent with contract

Issue - State: open - Opened by SebastianSemper 12 months ago

#216 - Suggestion: Multiple einsum steps

Issue - State: open - Opened by davystrong about 1 year ago - 2 comments

#215 - Multiple GPUs

Issue - State: closed - Opened by alex-orca about 1 year ago - 2 comments

#214 - Add Aesara backend

Pull Request - State: closed - Opened by rlouf over 1 year ago - 7 comments

#213 - Add Aesara as a backend

Issue - State: open - Opened by rlouf over 1 year ago - 4 comments

#212 - Return type of opt_einsum.contract -- linter errors

Issue - State: closed - Opened by danielkelshaw over 1 year ago - 2 comments

#210 - deduplicate backends.md

Pull Request - State: closed - Opened by loriab over 1 year ago

#209 - 'out' option in 'contract' makes a copy

Issue - State: closed - Opened by wxj6000 over 1 year ago - 6 comments

#208 - Use ConfigParser instead of SafeConfigParser

Pull Request - State: closed - Opened by hongxu-jia over 1 year ago - 1 comment

#207 - [WIP] add a preprocessor

Pull Request - State: closed - Opened by jcmgray over 1 year ago - 1 comment

#206 - Way to keep tensor Ids from being rearranged to cretae custom paths

Issue - State: closed - Opened by LionFrangoulis over 1 year ago - 2 comments

#205 - Disable torch.backends.opt_einsum to avoid duplicate work

Pull Request - State: closed - Opened by janeyx99 over 1 year ago - 19 comments

#204 - Remove NumPy as a hard dependency

Pull Request - State: open - Opened by dgasmith over 1 year ago - 7 comments

#203 - Thoughts on making the numpy dependency optional?

Issue - State: open - Opened by janeyx99 over 1 year ago - 5 comments

#201 - Update README.md

Pull Request - State: closed - Opened by s-mandra over 1 year ago - 2 comments

#200 - [BUG FIX] Add shapes support to parse_einsum_input

Pull Request - State: closed - Opened by janeyx99 almost 2 years ago - 10 comments

#199 - Fink optimal path for caching constant intermediates

Issue - State: open - Opened by fangzhangmnm almost 2 years ago - 3 comments

#198 - Zero dimension tensors (as zero entries) lead opt_einsum slower than einsum

Issue - State: open - Opened by Geositta2000 almost 2 years ago - 1 comment

#197 - Support for Python array API standard

Pull Request - State: open - Opened by IsaacBreen almost 2 years ago - 4 comments

#196 - allow backend=None

Pull Request - State: closed - Opened by jcmgray almost 2 years ago - 2 comments

#195 - Fix `dp` to work for contractions with no sized axes (scalars only).

Pull Request - State: closed - Opened by romanngg almost 2 years ago - 1 comment

#194 - Could `opt_einsum` understand repeated inputs?

Issue - State: open - Opened by romanngg almost 2 years ago

#193 - allow `backend=None` as shorthand for 'auto'

Issue - State: closed - Opened by jcmgray almost 2 years ago - 1 comment

#192 - Bug fixes and tweaks for v3.4.0

Pull Request - State: closed - Opened by dgasmith almost 2 years ago - 4 comments

#191 - Opt Einsum v3.5.0

Issue - State: open - Opened by dgasmith almost 2 years ago - 2 comments

#190 - Fixes typo in the new markdown docs

Pull Request - State: closed - Opened by dgasmith about 2 years ago - 1 comment

#189 - Slow contractions when scalars are present and output is large

Issue - State: open - Opened by Geositta2000 about 2 years ago - 3 comments

#188 - Fully qualify license trove classifier

Pull Request - State: closed - Opened by johnthagen about 2 years ago - 3 comments

#187 - Empty step in contractions with scalar by dp

Issue - State: open - Opened by Geositta2000 about 2 years ago - 1 comment

#186 - Reuse path including a scalar constant in the contraction

Issue - State: closed - Opened by Geositta2000 about 2 years ago - 2 comments

#185 - Possibilities to utilize symmetry in opt_einsum

Issue - State: closed - Opened by Lancashire3001 about 2 years ago - 1 comment

#184 - [question]: compute contraction order and provide order at future call

Issue - State: closed - Opened by rohany about 2 years ago - 2 comments

#183 - error in Test opt_enisum

Issue - State: closed - Opened by jiaminglei-lei over 2 years ago - 2 comments

#182 - UnicodeEncodeError in contract_path

Issue - State: closed - Opened by ChienKaiMa over 2 years ago

#181 - more custom dp scores for faster contractions

Pull Request - State: closed - Opened by jcmgray over 2 years ago - 1 comment

#180 - How to use tblis to improve TDOT?

Issue - State: closed - Opened by Lancashire3000 over 2 years ago - 7 comments

#179 - Optimizing a sum of einsums?

Issue - State: closed - Opened by yaroslavvb over 2 years ago - 2 comments

#178 - Use timeit to check the performance

Issue - State: closed - Opened by ghost over 2 years ago - 1 comment

#177 - Customize library in binary contractions

Issue - State: closed - Opened by Lancashire3000 over 2 years ago - 3 comments

#176 - "Finalize" MyPy

Pull Request - State: closed - Opened by dgasmith over 2 years ago - 2 comments

#175 - Fix typos

Pull Request - State: closed - Opened by kianmeng over 2 years ago - 1 comment

#174 - Prevents double contraction of floats in dynamic path, fixes #172

Pull Request - State: closed - Opened by dgasmith over 2 years ago - 1 comment

#173 - Is there anything like np.show_config() in opt_einsum

Issue - State: closed - Opened by ghost over 2 years ago - 1 comment

#172 - Bug: Dynamic_programming path solver may return illegal contraction path

Issue - State: closed - Opened by refraction-ray over 2 years ago - 5 comments

#171 - Customized cost functions for faster contractions

Issue - State: closed - Opened by jcmgray over 2 years ago - 4 comments
Labels: enhancement

#170 - Do you provide a way to compute the output shape of an einsum?

Issue - State: closed - Opened by mariogeiger over 2 years ago - 2 comments

#169 - paths.ssa_to_linear() might be buggy

Issue - State: closed - Opened by springer13 almost 3 years ago - 1 comment

#168 - when used with pytorch, is it possible to apply gradient checkpointing?

Issue - State: closed - Opened by seongwook-ham almost 3 years ago - 1 comment

#167 - Is opt_einsum.paths.optimal algorithm correct?

Issue - State: open - Opened by igtrnt almost 3 years ago - 7 comments

#166 - Update helpers.py(fix typos)

Pull Request - State: closed - Opened by ax7e almost 3 years ago - 1 comment

#165 - Create CITATION.cff

Pull Request - State: closed - Opened by dgasmith almost 3 years ago - 4 comments

#163 - Updates README

Pull Request - State: closed - Opened by dgasmith about 3 years ago - 1 comment

#162 - Taking strides into account

Issue - State: closed - Opened by danpovey about 3 years ago - 2 comments

#161 - Python Type Annotations

Pull Request - State: closed - Opened by dgasmith about 3 years ago - 2 comments

#160 - Docs: Mkdocs Material Overhaul

Pull Request - State: closed - Opened by dgasmith about 3 years ago - 2 comments

#159 - Switch CI to GHA from Travis

Pull Request - State: closed - Opened by dgasmith about 3 years ago - 2 comments

#158 - Question: optimal contractions and batch dimension

Issue - State: closed - Opened by Linux-cpp-lisp over 3 years ago - 6 comments

#157 - Contraction route "ipx, iqx -> pq" probably slow and not multi-threading most time

Issue - State: closed - Opened by ajz34 over 3 years ago - 5 comments

#155 - Add flake8 and spell check, whitelist dictionary, docstrings

Pull Request - State: closed - Opened by asears over 3 years ago - 1 comment

#154 - Small fixes: 'dp' and memory_limit + tensordot axes order

Pull Request - State: closed - Opened by jcmgray over 3 years ago - 1 comment

#153 - infinite loop in dynamic-programming with 'max_input' memory limit

Issue - State: closed - Opened by rc over 3 years ago - 7 comments

#152 - fix typo in readme

Pull Request - State: closed - Opened by samuelstjean almost 4 years ago - 1 comment

#151 - contract_expression with batch dimenstion

Issue - State: closed - Opened by PgLoLo almost 4 years ago - 2 comments

#150 - Docs: Adds v3.3.0 changelog

Pull Request - State: closed - Opened by dgasmith almost 4 years ago - 1 comment

#149 - contraction_list: only keep remaining for last 20 contractions

Pull Request - State: closed - Opened by jcmgray almost 4 years ago - 2 comments

#148 - Remove 'remaining' from PathInfo? (incurs n^2 memory overhead)

Issue - State: closed - Opened by jcmgray almost 4 years ago - 1 comment

#147 - Test fails with newer numpy

Issue - State: closed - Opened by avdv almost 4 years ago - 6 comments

#146 - Einsum output order

Issue - State: open - Opened by dgasmith almost 4 years ago - 2 comments

#145 - support contraction of arbitrary objects

Pull Request - State: closed - Opened by jcmgray almost 4 years ago - 2 comments

#144 - cuTensor with cupy

Issue - State: closed - Opened by cnk113 about 4 years ago - 1 comment

#142 - Symbolic einsum support?

Issue - State: closed - Opened by refraction-ray about 4 years ago - 8 comments

#141 - stateful path optimizers check args + minor doc tweaks

Pull Request - State: closed - Opened by jcmgray about 4 years ago - 1 comment

#140 - more detailed error when backend is missing a function e.g. einsum

Pull Request - State: closed - Opened by jcmgray about 4 years ago - 1 comment

#139 - Problem with sparse

Issue - State: closed - Opened by Tomohirohashizume about 4 years ago - 3 comments

#138 - questions about helpers.flop_count

Issue - State: closed - Opened by refraction-ray about 4 years ago - 4 comments

#137 - Possible bug in RandomGreedy path

Issue - State: closed - Opened by emprice about 4 years ago - 2 comments