v1.4.4

Lux v1.4.4

Diff since v1.4.3

Merged pull requests:

• feat: update ConvMixer to support reactant (#1063) (@avik-pal)
• feat: more nested AD rules (#1151) (@avik-pal)
• fix: update default rng for reactant (#1152) (@avik-pal)
• Change update step to thousands in PINN 2D PDE (#1153) (@abhro)

Closed issues:

• Random Numbers & Reactant (#1131)
• CUDA 2nd order AD with MaxPool and logsoftmax (#1150)

MLDataDevices-v1.6.6

MLDataDevices MLDataDevices-v1.6.6

Diff since MLDataDevices-v1.6.5

Merged pull requests:

• fix: remove old patches around reactant bug (#1135) (@avik-pal)
• CompatHelper: bump compat for Flux in [weakdeps] to 0.16, (keep existing compat) (#1136) (@github-actions[bot])
• chore: bump crate-ci/typos from 1.28.2 to 1.28.3 (#1137) (@dependabot[bot])
• fix: update to new reactant changes (#1140) (@avik-pal)
• chore: bump crate-ci/typos from 1.28.3 to 1.28.4 (#1144) (@dependabot[bot])
• don't declare implicitly exported functions public (#1147) (@simeonschaub)
• use return_type instead of _return_type (#1148) (@simeonschaub)
• feat: more nested AD rules (#1151) (@avik-pal)
• fix: update default rng for reactant (#1152) (@avik-pal)

Closed issues:

• Random Numbers & Reactant (#1131)
• Problem with Lux & SymbolicsLuxExt (#1132)
• How to implement a detach operation similar to Pytorch? (#1138)
• Unexpected handling of LR Schedulers in TrainState (#1143)
• Directly construct Optimiser state on Reactant buffers (#1145)
• (AbstractDevice)(x) should respect Adapt.adapt_structure (#1149)
• CUDA 2nd order AD with MaxPool and logsoftmax (#1150)

LuxLib-v1.4.0

LuxLib LuxLib-v1.4.0

Diff since LuxLib-v1.3.11

Merged pull requests:

• feat: more nested AD rules (#1151) (@avik-pal)

Closed issues:

• CUDA 2nd order AD with MaxPool and logsoftmax (#1150)

v1.4.3

Lux v1.4.3

Diff since v1.4.2

Merged pull requests:

• CompatHelper: bump compat for Flux in [weakdeps] to 0.16, (keep existing compat) (#1136) (@github-actions[bot])
• chore: bump crate-ci/typos from 1.28.2 to 1.28.3 (#1137) (@dependabot[bot])
• fix: update to new reactant changes (#1140) (@avik-pal)
• chore: bump crate-ci/typos from 1.28.3 to 1.28.4 (#1144) (@dependabot[bot])
• don't declare implicitly exported functions public (#1147) (@simeonschaub)
• use return_type instead of _return_type (#1148) (@simeonschaub)

Closed issues:

• Problem with Lux & SymbolicsLuxExt (#1132)
• How to implement a detach operation similar to Pytorch? (#1138)
• Unexpected handling of LR Schedulers in TrainState (#1143)
• Directly construct Optimiser state on Reactant buffers (#1145)
• (AbstractDevice)(x) should respect Adapt.adapt_structure (#1149)

LuxLib-v1.3.11

LuxLib LuxLib-v1.3.11

Diff since LuxLib-v1.3.10

Merged pull requests:

• Update exporting_to_jax.md (#1107) (@wsmoses)
• CompatHelper: bump compat for LossFunctions in [weakdeps] to 1, (keep existing compat) (#1108) (@github-actions[bot])
• Add TrainState docstring with Optimisers API (#1110) (@abhro)
• Fix markdown list in docstring (#1111) (@abhro)
• chore: bump crate-ci/typos from 1.27.3 to 1.28.1 (#1113) (@dependabot[bot])
• fix: handle debug leafs with dispatch (#1115) (@avik-pal)
• test: allow the latest AMDGPU to be installed (#1116) (@avik-pal)
• test: add unsafe_free to skip list (#1117) (@avik-pal)
• fix: use the correct dispatches for device overloads (#1118) (@avik-pal)
• test: try fixing enzyme test (#1119) (@avik-pal)
• ci(github-actions): use julia-actions/cache (#1122) (@avik-pal)
• test: re-enable flux testing (#1123) (@avik-pal)
• chore: bump minimum Reactant version (#1125) (@avik-pal)
• fix: try fixing cuda install in tests (#1126) (@avik-pal)
• docs: run partial dataset only on CI (#1128) (@avik-pal)
• chore: bump crate-ci/typos from 1.28.1 to 1.28.2 (#1130) (@dependabot[bot])
• fix: preserve object when device is same (#1133) (@avik-pal)
• fix: use functors for testing wrapped arrays (#1134) (@avik-pal)
• fix: remove old patches around reactant bug (#1135) (@avik-pal)
• CompatHelper: bump compat for Flux in [weakdeps] to 0.16, (keep existing compat) (#1136) (@github-actions[bot])
• chore: bump crate-ci/typos from 1.28.2 to 1.28.3 (#1137) (@dependabot[bot])
• fix: update to new reactant changes (#1140) (@avik-pal)
• chore: bump crate-ci/typos from 1.28.3 to 1.28.4 (#1144) (@dependabot[bot])
• don't declare implicitly exported functions public (#1147) (@simeonschaub)
• use return_type instead of _return_type (#1148) (@simeonschaub)

Closed issues:

• Immutable Arrays (#8)
• Downstream Compat Updates (#880)
• Enzyme 0.13 fails with batched matrix multiply (#1024)
• getkeypath and layer_map not fully working with model with Parallel layers (#1068)
• Re-enable Flux compatibility testing (#1070)
• [AMDGPU CI] Circular dependencies disabling precompilation (#1095)
• LuxTestUtils.Constant clashes with DifferentiationInterface.Constant (#1103)
• Error in trying to use Optimization.jl for LSTM training based on Lux.jl (#1114)
• Documentation Build Stalls (#1120)
• CUDA Test CI is broken (#1121)
• [MLDataDevices] devices don't preserve identity (#1129)
• Problem with Lux & SymbolicsLuxExt (#1132)
• How to implement a detach operation similar to Pytorch? (#1138)
• Unexpected handling of LR Schedulers in TrainState (#1143)
• Directly construct Optimiser state on Reactant buffers (#1145)
• (AbstractDevice)(x) should respect Adapt.adapt_structure (#1149)

v1.4.2

Lux v1.4.2

Diff since v1.4.1

Merged pull requests:

• docs: run partial dataset only on CI (#1128) (@avik-pal)
• chore: bump crate-ci/typos from 1.28.1 to 1.28.2 (#1130) (@dependabot[bot])
• fix: preserve object when device is same (#1133) (@avik-pal)
• fix: use functors for testing wrapped arrays (#1134) (@avik-pal)
• fix: remove old patches around reactant bug (#1135) (@avik-pal)

Closed issues:

• Immutable Arrays (#8)
• Downstream Compat Updates (#880)
• [MLDataDevices] devices don't preserve identity (#1129)

MLDataDevices-v1.6.5

MLDataDevices MLDataDevices-v1.6.5

Diff since MLDataDevices-v1.6.4

Merged pull requests:

• fix: use functors for testing wrapped arrays (#1134) (@avik-pal)

MLDataDevices-v1.6.4

MLDataDevices MLDataDevices-v1.6.4

Diff since MLDataDevices-v1.6.3

Merged pull requests:

• test: re-enable flux testing (#1123) (@avik-pal)
• chore: bump minimum Reactant version (#1125) (@avik-pal)
• fix: try fixing cuda install in tests (#1126) (@avik-pal)
• docs: run partial dataset only on CI (#1128) (@avik-pal)
• chore: bump crate-ci/typos from 1.28.1 to 1.28.2 (#1130) (@dependabot[bot])
• fix: preserve object when device is same (#1133) (@avik-pal)

Closed issues:

• Immutable Arrays (#8)
• Downstream Compat Updates (#880)
• Re-enable Flux compatibility testing (#1070)
• Documentation Build Stalls (#1120)
• CUDA Test CI is broken (#1121)
• [MLDataDevices] devices don't preserve identity (#1129)

v1.4.1

Lux v1.4.1

Diff since v1.4.0

Merged pull requests:

• Update exporting_to_jax.md (#1107) (@wsmoses)
• CompatHelper: bump compat for LossFunctions in [weakdeps] to 1, (keep existing compat) (#1108) (@github-actions[bot])
• Add TrainState docstring with Optimisers API (#1110) (@abhro)
• Fix markdown list in docstring (#1111) (@abhro)
• chore: bump crate-ci/typos from 1.27.3 to 1.28.1 (#1113) (@dependabot[bot])
• fix: handle debug leafs with dispatch (#1115) (@avik-pal)
• test: allow the latest AMDGPU to be installed (#1116) (@avik-pal)
• test: add unsafe_free to skip list (#1117) (@avik-pal)
• fix: use the correct dispatches for device overloads (#1118) (@avik-pal)
• test: try fixing enzyme test (#1119) (@avik-pal)
• ci(github-actions): use julia-actions/cache (#1122) (@avik-pal)
• test: re-enable flux testing (#1123) (@avik-pal)
• chore: bump minimum Reactant version (#1125) (@avik-pal)
• fix: try fixing cuda install in tests (#1126) (@avik-pal)

Closed issues:

• Enzyme 0.13 fails with batched matrix multiply (#1024)
• getkeypath and layer_map not fully working with model with Parallel layers (#1068)
• Re-enable Flux compatibility testing (#1070)
• [AMDGPU CI] Circular dependencies disabling precompilation (#1095)
• LuxTestUtils.Constant clashes with DifferentiationInterface.Constant (#1103)
• Error in trying to use Optimization.jl for LSTM training based on Lux.jl (#1114)
• Documentation Build Stalls (#1120)
• CUDA Test CI is broken (#1121)

MLDataDevices-v1.6.3

MLDataDevices MLDataDevices-v1.6.3

Diff since MLDataDevices-v1.6.2

Merged pull requests:

• test: try re-enabling enzyme testing on 0.13.16 (#1042) (@avik-pal)
• chore: bump codecov/codecov-action from 4 to 5 (#1093) (@dependabot[bot])
• ci: install specific AMDGPU version (#1096) (@avik-pal)
• ci: use sources for docs (#1100) (@avik-pal)
• Add Reactant and TPU to autodiff.md (#1101) (@wsmoses)
• refactor: cleanup some old pre-1.0 hacks (#1102) (@avik-pal)
• feat: add bf16 function (#1104) (@avik-pal)
• docs: add CUDA.CURAND.default_rng() to docs (#1105) (@avik-pal)
• fix: use generic broadcasting for complex numbers (#1106) (@avik-pal)
• Update exporting_to_jax.md (#1107) (@wsmoses)
• CompatHelper: bump compat for LossFunctions in [weakdeps] to 1, (keep existing compat) (#1108) (@github-actions[bot])
• Add TrainState docstring with Optimisers API (#1110) (@abhro)
• Fix markdown list in docstring (#1111) (@abhro)
• chore: bump crate-ci/typos from 1.27.3 to 1.28.1 (#1113) (@dependabot[bot])
• fix: handle debug leafs with dispatch (#1115) (@avik-pal)
• test: allow the latest AMDGPU to be installed (#1116) (@avik-pal)
• test: add unsafe_free to skip list (#1117) (@avik-pal)
• fix: use the correct dispatches for device overloads (#1118) (@avik-pal)
• test: try fixing enzyme test (#1119) (@avik-pal)
• ci(github-actions): use julia-actions/cache (#1122) (@avik-pal)

Closed issues:

• Zygote + ForwardDiff support for complex differentiation (#977)
• Add CUDA.CURAND.default_rng() to the table (#1003)
• getkeypath and layer_map not fully working with model with Parallel layers (#1068)
• [AMDGPU CI] Circular dependencies disabling precompilation (#1095)
• LuxTestUtils.Constant clashes with DifferentiationInterface.Constant (#1103)
• Error in trying to use Optimization.jl for LSTM training based on Lux.jl (#1114)