v4.0.6

TensorOperations v4.0.6

Diff since v4.0.5

Merged pull requests:

• Automatic Differentiation test rewrite (#150) (@lkdvos)

v4.0.5

TensorOperations v4.0.5

Diff since v4.0.4

v4.0.4

TensorOperations v4.0.4

Diff since v4.0.3

v4.0.3

TensorOperations v4.0.3

Diff since v4.0.2

Merged pull requests:

• Refactor branching on α and β (#146) (@lkdvos)
• Bump actions/checkout from 1 to 4 (#147) (@dependabot[bot])

Closed issues:

• Floating Point Accuracy of @tensor results with CUDA (#144)
• Enable multithreads when doing the permutedims in the TTGT algorithms (#145)

v4.0.2

TensorOperations v4.0.2

Diff since v4.0.1

v4.0.1

TensorOperations v4.0.1

Diff since v4.0.0

Closed issues:

• Why drop caching Tensors? (#140)
• Is TensorOperations able to take advantage of symmetry in the output? (#142)

v4.0.0

TensorOperations v4.0.0

Diff since v3.2.5

Closed issues:

• Consider splitting out CuArrays? (#105)
• Cryptic IndexError messages (#127)
• Specify output type (#129)
• Update Cuda.jl to v4? (#130)
• Sums are not commutative (#136)
• Question: @cutensor not defined (#138)

Merged pull requests:

• fix scalar multplication with multiple factors (#131) (@fatteneder)
• Version 4 (WIP) (#137) (@Jutho)
• Split backend for operations and allocators (#141) (@lkdvos)

v3.2.5

TensorOperations v3.2.5

Diff since v3.2.4

Closed issues:

• Unintended contraction throws error. (#118)
• Certain contractions with matrices resulting in zeros (#120)
• Uncontracted indices error when summing two contracted objects (#121)
• Does TensorOperations.jl not support Zygote.jl? (#126)

Merged pull requests:

• Update link (#122) (@davidhbrann)

v3.2.4

TensorOperations v3.2.4

Diff since v3.2.3

support for CUDA 3.3 and above
other small fixes

Closed issues:

• Unsupported assignment operator ":=" (#106)
• TensorOperations and CUDA 3.3, 3.4 (#107)
• Difficulties using GPU: @cutensor not defined (#112)
• similar_from_structure always call fill! which is not need for bitstype (#114)
• Citation to the package (#117)

Merged pull requests:

• update for CUDA 3.3 (@workspace -> with_workspace) (#109) (@jfeist)
• in-place operations no longer have to be in-place (#110) (@maartenvd)
• #114 remove undesired fill! (#115) (@jemiryguo)
• Fix link to LRUCache in docs (#116) (@jishnub)

v3.2.3

TensorOperations v3.2.3

Diff since v3.2.2