eeva/tensorflow-haskell
1
0
Fork 0
mirror of https://github.com/tensorflow/haskell.git synced 2024-11-27 05:19:45 +01:00
Code Issues Releases Wiki Activity
tensorflow-haskell/tensorflow-ops/tensorflow-ops.cabal

308 lines
8.4 KiB
Text
Raw Normal View History

Initial commit
2016-10-24 21:26:42 +02:00
name: tensorflow-ops
version: 0.1.0.0
synopsis: Friendly layer around TensorFlow bindings.
description: Please see README.md
homepage: https://github.com/tensorflow/haskell#readme
license: Apache
Fix .cabal files so 'stack check' passes. (#110) - Add LICENSE files for all packages. - Add descriptions for packages that were missing one. - Work around google/proto-lens#69 by symlinking third_party into tensorflow-proto.
2017-05-10 20:37:00 +02:00
license-file: LICENSE
Initial commit
2016-10-24 21:26:42 +02:00
author: TensorFlow authors
maintainer: tensorflow-haskell@googlegroups.com
copyright: Google Inc.
category: Machine Learning
build-type: Simple
cabal-version: >=1.22
library
hs-source-dirs: src
exposed-modules: TensorFlow.Gradient
, TensorFlow.Ops
, TensorFlow.EmbeddingOps
Add Minimize module with gradient descent and adam implementations (#125)
2017-05-26 04:19:22 +02:00
, TensorFlow.Minimize
Consolidate some packages. (#111) - Merge tensorflow-nn and tensorflow-queue into tensorflow-ops. They don't add extra dependencies and each contain a single module, so I don't think it's worth separating them at the package level. - Remove google-shim in favor of direct use of test-framework.
2017-05-11 00:26:03 +02:00
, TensorFlow.NN
, TensorFlow.Queue
Add resource-based variable ops. (#98) The main difference between these and the `Ref`-bases ops is the explicit `readValue` op. I'm not sure how this should interact with gradients and save/restore, so I'm keeping it as a separate module for now. Once we figure out the details, we can merge it into `TensorFlow.Ops` and replace all uses of the old `Ref`-based ops. (That would also fix #92.) Also replaces our special case newtype `ResourceHandle` to `Tensor Value ResourceHandle`, where `ResourceHandle` is the TF proto corresponding to `DT_RESOURCE`.
2017-04-16 18:24:02 +02:00
, TensorFlow.Variable
Adapt to lts-8.6 and use proto-lens-0.2.0.1 (#97)
2017-04-11 23:09:01 +02:00
build-depends: proto-lens == 0.2.*
Initial commit
2016-10-24 21:26:42 +02:00
, base >= 4.7 && < 5
, bytestring
, fgl
, mtl
, data-default
, lens-family
, containers
, tensorflow == 0.1.*
, tensorflow-proto == 0.1.*
, tensorflow-core-ops == 0.1.*
, text
default-language: Haskell2010
Add example to README + make haddock link more prominent (#60)
2017-01-17 05:44:45 +01:00
Test-Suite RegressionTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: RegressionTest.hs
hs-source-dirs: tests
build-depends: base
, HUnit
Add resource-based variable ops. (#98) The main difference between these and the `Ref`-bases ops is the explicit `readValue` op. I'm not sure how this should interact with gradients and save/restore, so I'm keeping it as a separate module for now. Once we figure out the details, we can merge it into `TensorFlow.Ops` and replace all uses of the old `Ref`-based ops. (That would also fix #92.) Also replaces our special case newtype `ResourceHandle` to `Tensor Value ResourceHandle`, where `ResourceHandle` is the TF proto corresponding to `DT_RESOURCE`.
2017-04-16 18:24:02 +02:00
, lens-family
, transformers
Add example to README + make haddock link more prominent (#60)
2017-01-17 05:44:45 +01:00
, random
, tensorflow
, tensorflow-core-ops
, tensorflow-ops
added matrix factorization test (#101)
2017-04-28 02:05:34 +02:00
Test-Suite MatrixTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: MatrixTest.hs
hs-source-dirs: tests
build-depends: base
, HUnit
, random
, tensorflow
, tensorflow-core-ops
, tensorflow-ops
, tensorflow-test
, test-framework
, test-framework-hunit
, transformers
, vector
Whitespace
2017-05-04 09:31:25 +02:00
Initial commit
2016-10-24 21:26:42 +02:00
Test-Suite BuildTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: BuildTest.hs
hs-source-dirs: tests
build-depends: HUnit
, base
, proto-lens
, lens-family
, tensorflow
, tensorflow-ops
, tensorflow-proto
, test-framework
, test-framework-hunit
, transformers
, vector
Test-Suite EmbeddingOpsTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: EmbeddingOpsTest.hs
hs-source-dirs: tests
build-depends: HUnit
, QuickCheck
, base
, proto-lens
, lens-family
, tensorflow
test case to show can't calculate grad for embedding (and associated fix) (#23) * Fix for embedding gradient calculation - Passes vectors instead of scalars to slice - converts the numRows to a scalar - add `toScalar` utility function - minor change to test case so that it actually works * added lib for testing helper functions * add flatSlice function
2016-11-17 22:54:36 +01:00
, tensorflow-test
Initial commit
2016-10-24 21:26:42 +02:00
, tensorflow-core-ops
, tensorflow-ops
, tensorflow-proto
, test-framework
, test-framework-hunit
, test-framework-quickcheck2
Distinguish between "rendered" and "unrendered" Tensors. (#88) Distinguish between "rendered" and "unrendered" Tensors. There are now three types of `Tensor`: - `Tensor Value a`: rendered value - `Tensor Ref a`: rendered reference - `Tensor Build a` : unrendered value The extra bookkeeping makes it easier to track (and enforce) which tensors are rendered or not. For examples where this has been confusing in the past, see With this change, pure ops look similar to before, returning `Tensor Build` instead of `Tensor Value`. "Stateful" (monadic) ops are unchanged. For example: add :: OneOf [..] t => Tensor v'1 t -> Tensor v'2 t -> Tensor Build t assign :: (MonadBuild m, TensorType t) => Tensor Ref t -> Tensor v'2 t -> m (Tensor Ref t) The `gradients` function now requires that the variables over which it's differentiating are pre-rendered: gradients :: (..., Rendered v2) => Tensor v1 a -> [Tensor v2 a] -> m [Tensor Value a] (`Rendered v2` means that `v2` is either a `Ref` or a `Value`.) Additionally, the implementation of `gradients` now takes care to render every intermediate value when performing the reverse accumulation. I suspect this fixes an exponential blowup for complicated expressions.
2017-04-07 00:10:33 +02:00
, transformers
Initial commit
2016-10-24 21:26:42 +02:00
, vector
Test-Suite ArrayOpsTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: ArrayOpsTest.hs
hs-source-dirs: tests
build-depends: HUnit
, QuickCheck
, base
, proto-lens
, lens-family
, tensorflow
, tensorflow-core-ops
, tensorflow-ops
, tensorflow-proto
, test-framework
, test-framework-hunit
, test-framework-quickcheck2
, transformers
, vector
Test-Suite OpsTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: OpsTest.hs
hs-source-dirs: tests
build-depends: HUnit
, QuickCheck
, base
, bytestring
, proto-lens
, lens-family
, temporary
, tensorflow
, tensorflow-core-ops
, tensorflow-ops
, tensorflow-proto
, test-framework
, test-framework-hunit
, test-framework-quickcheck2
, transformers
, vector
Add resource-based variable ops. (#98) The main difference between these and the `Ref`-bases ops is the explicit `readValue` op. I'm not sure how this should interact with gradients and save/restore, so I'm keeping it as a separate module for now. Once we figure out the details, we can merge it into `TensorFlow.Ops` and replace all uses of the old `Ref`-based ops. (That would also fix #92.) Also replaces our special case newtype `ResourceHandle` to `Tensor Value ResourceHandle`, where `ResourceHandle` is the TF proto corresponding to `DT_RESOURCE`.
2017-04-16 18:24:02 +02:00
Test-Suite VariableTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: VariableTest.hs
hs-source-dirs: tests
build-depends: HUnit
, base
, tensorflow
, tensorflow-core-ops
, tensorflow-ops
, test-framework
, test-framework-hunit
, transformers
, vector
Initial commit
2016-10-24 21:26:42 +02:00
Test-Suite DataFlowOpsTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: DataFlowOpsTest.hs
hs-source-dirs: tests
build-depends: HUnit
, QuickCheck
, base
, proto-lens
, lens-family
, tensorflow
, tensorflow-core-ops
, tensorflow-ops
, tensorflow-proto
, test-framework
, test-framework-hunit
, test-framework-quickcheck2
, vector
Test-Suite GradientTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: GradientTest.hs
hs-source-dirs: tests
build-depends: HUnit
, base
, proto-lens
, lens-family
, tensorflow
Fix gradients calculation for min and max (#48)
2016-12-12 18:47:02 +01:00
, tensorflow-core-ops
Initial commit
2016-10-24 21:26:42 +02:00
, tensorflow-ops
, tensorflow-proto
, test-framework
, test-framework-hunit
Fixed matMul gradients for transposed arguments
2017-05-04 09:39:15 +02:00
, transformers
Fix gradients calculation for min and max (#48)
2016-12-12 18:47:02 +01:00
, vector
Initial commit
2016-10-24 21:26:42 +02:00
Test-Suite MiscTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: MiscTest.hs
hs-source-dirs: tests
build-depends: HUnit
, base
, bytestring
, vector
, transformers
, tensorflow
Replaced topK with topKV2. (#21) topK is obsolete and generating warnings.
2016-11-09 05:57:22 +01:00
, tensorflow-core-ops
Initial commit
2016-10-24 21:26:42 +02:00
, tensorflow-ops
, tensorflow-proto
, test-framework
, test-framework-hunit
Consolidate some packages. (#111) - Merge tensorflow-nn and tensorflow-queue into tensorflow-ops. They don't add extra dependencies and each contain a single module, so I don't think it's worth separating them at the package level. - Remove google-shim in favor of direct use of test-framework.
2017-05-11 00:26:03 +02:00
Test-Suite NNTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: NNTest.hs
hs-source-dirs: tests
build-depends: HUnit
, QuickCheck
, base
, tensorflow
, tensorflow-test
, tensorflow-ops
, test-framework
, test-framework-hunit
, test-framework-quickcheck2
, vector
Test-Suite QueueTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: QueueTest.hs
hs-source-dirs: tests
-- Uses multiple threads and blocks without this option.
ghc-options: -threaded
build-depends: HUnit
, base
, bytestring
, proto-lens
, lens-family
, tensorflow
, tensorflow-core-ops
, tensorflow-ops
, test-framework
, test-framework-hunit
, transformers
, vector
Added sessionTracer to log graph operations. (#26) * Added TracingTest.
2016-11-15 00:14:51 +01:00
Test-Suite TracingTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: TracingTest.hs
hs-source-dirs: tests
build-depends: HUnit
, base
, bytestring
, data-default
, lens-family
, tensorflow
, tensorflow-ops
, test-framework
, test-framework-hunit
Initial commit
2016-10-24 21:26:42 +02:00
Test-Suite TypesTest
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: TypesTest.hs
hs-source-dirs: tests
build-depends: HUnit
, QuickCheck
, base
, bytestring
, proto-lens
, lens-family
, tensorflow
Fix TensorData encode and decode for Bool (#49)
2016-12-13 04:40:32 +01:00
, tensorflow-core-ops
Initial commit
2016-10-24 21:26:42 +02:00
, tensorflow-ops
, tensorflow-proto
, transformers
, test-framework
, test-framework-hunit
, test-framework-quickcheck2
, vector
Optimize fetching (#27) * Add MNIST data to gitignore * Add simple tensor round-trip benchmark * Use deepseq + cleaner imports * Use safe version of fromIntegral in FFI code * Don't copy data when fetching tensors BEFORE benchmarking feedFetch/4 byte time 55.79 μs (54.88 μs .. 56.62 μs) 0.998 R² (0.997 R² .. 0.999 R²) mean 55.61 μs (55.09 μs .. 56.11 μs) std dev 1.828 μs (1.424 μs .. 2.518 μs) variance introduced by outliers: 34% (moderately inflated) benchmarking feedFetch/4 KiB time 231.4 μs (221.9 μs .. 247.3 μs) 0.988 R² (0.974 R² .. 1.000 R²) mean 226.6 μs (224.1 μs .. 236.2 μs) std dev 13.45 μs (7.115 μs .. 27.14 μs) variance introduced by outliers: 57% (severely inflated) benchmarking feedFetch/4 MiB time 485.8 ms (424.6 ms .. 526.7 ms) 0.998 R² (0.994 R² .. 1.000 R²) mean 515.7 ms (512.5 ms .. 517.9 ms) std dev 3.320 ms (0.0 s .. 3.822 ms) variance introduced by outliers: 19% (moderately inflated) AFTER benchmarking feedFetch/4 byte time 53.11 μs (52.12 μs .. 54.22 μs) 0.996 R² (0.995 R² .. 0.998 R²) mean 54.64 μs (53.59 μs .. 56.18 μs) std dev 4.249 μs (2.910 μs .. 6.076 μs) variance introduced by outliers: 75% (severely inflated) benchmarking feedFetch/4 KiB time 83.83 μs (82.72 μs .. 84.92 μs) 0.999 R² (0.998 R² .. 0.999 R²) mean 83.82 μs (83.20 μs .. 84.35 μs) std dev 1.943 μs (1.557 μs .. 2.614 μs) variance introduced by outliers: 20% (moderately inflated) benchmarking feedFetch/4 MiB time 95.54 ms (93.62 ms .. 97.82 ms) 0.999 R² (0.998 R² .. 1.000 R²) mean 96.61 ms (95.76 ms .. 97.51 ms) std dev 1.408 ms (1.005 ms .. 1.889 ms)
2016-11-17 19:41:49 +01:00
Benchmark FeedFetchBench
default-language: Haskell2010
type: exitcode-stdio-1.0
main-is: FeedFetchBench.hs
hs-source-dirs: tests
build-depends: base
, criterion
, deepseq
, tensorflow
, tensorflow-ops
, transformers
, vector
ghc-options: -O2 -threaded
Initial commit
2016-10-24 21:26:42 +02:00
source-repository head
type: git
location: https://github.com/tensorflow/haskell
Reference in a new issue Copy permalink