Skip to content

P
pytensor
提交 2b571a86 authored 作者: Frederic's avatar Frederic
浏览文件
操作

Update the file NEWS.txt

上级 94c2e959
隐藏空白字符变更
内嵌 并排
正在显示 包含 108 行增加 和 112 行删除
NEWS.txt
浏览文件 @ 2b571a86
TODO for final release: TODO for final release:
- test python 2.4
- test theano-cache with "pip install Theano": issue 101
- Re-write this NEWS.txt file! - Re-write this NEWS.txt file!
if time check issue: 98. Modifications in the trunk since the 0.4.1 release (12 August 2011) up to 5 Dec 2011
Modifications in the trunk since the 0.4.1 release (12 August 2011) up to 2 Dec 2011
Every body is recommended to update Theano to 0.5 when released after Every body is recommended to update Theano to 0.5 when released after
...@@ -18,160 +14,160 @@ Important changes: ...@@ -18,160 +14,160 @@ Important changes:
* Moved to github: https://github.com/Theano/Theano/ * Moved to github: https://github.com/Theano/Theano/
* Old trac ticket moved to assembla ticket: https://www.assembla.com/spaces/theano/tickets * Old trac ticket moved to assembla ticket: https://www.assembla.com/spaces/theano/tickets
* Theano vision: https://deeplearning.net/software/theano/introduction.html#theano-vision (Many people) * Theano vision: https://deeplearning.net/software/theano/introduction.html#theano-vision (Many people)
* * Theano with GPU works in some cases on Windows now. Still experimental. (Sebastian Urban)
* See the Interface changes.
Interface Behavior Change (was deprecated and generated a warning since Theano 0.3 released the 23 Nov 2010): Interface Behavior Change (was deprecated and generated a warning since Theano 0.3 released the 23 Nov 2010):
* The current default value of the parameter axis of * The current default value of the parameter axis of
theano.{max,min,argmax,argmin,max_and_argmax} is now the same as theano.{max,min,argmax,argmin,max_and_argmax} is now the same as
numpy: None. i.e. operate on all dimensions of the tensor. numpy: None. i.e. operate on all dimensions of the tensor. (Frédéric Bastien, Olivier Delalleau)
Interface Feature Removed (was deprecated): Interface Feature Removed (was deprecated):
* The string mode FAST_RUN_NOGC and STABILIZE are not accepted. It was accepted only by theano.function(). Use Mode(linker='c|py_nogc') or Mode(optimizer='stabilize') instead. * The string mode FAST_RUN_NOGC and STABILIZE are not accepted. It was accepted only by theano.function(). Use Mode(linker='c|py_nogc') or Mode(optimizer='stabilize') instead.
* tensor.grad(cost, wrt) now return an object of the "same type" as wrt * tensor.grad(cost, wrt) now return an object of the "same type" as wrt
(list/tuple/TensorVariable). (list/tuple/TensorVariable). (Ian Goodfellow, Olivier)
* a few tag.shape and Join.vec_length left. * a few tag.shape and Join.vec_length left have been removed. (Frederic)
* scan interface change: RP * scan interface change: (Razvan Pascanu)
* The use of `return_steps` for specifying how many entries of the output * The use of `return_steps` for specifying how many entries of the output
to return has been deprecated. Instead, apply a subtensor to the output to return has been removed. Instead, apply a subtensor to the output
returned by scan to select a certain slice returned by scan to select a certain slice.
* The inner function (that scan receives) should return its outputs and * The inner function (that scan receives) should return its outputs and
updates following this order: updates following this order:
[outputs], [updates], [condition]. One can skip any of the three if not [outputs], [updates], [condition]. One can skip any of the three if not
used, but the order has to stay unchanged. used, but the order has to stay unchanged.
* shared.value is moved, use shared.set_value() or shared.get_value() instead. * shared.value is moved, use shared.set_value() or shared.get_value() instead. (Olivier D.)
New deprecation (will be removed in Theano 0.6, warning generated if you use them): Interface bug fixes:
* tensor.shared() renamed to tensor._shared (Olivier D.) You probably want to call theano.shared()! * Rop in some case should have returned a list of one theano variable, but returned the variable itself. (Razvan)
* Theano flags "home" is not used anymore as it was a duplicate. If you use it, theano should raise an error. (Olivier D.)
Interface bug fixes: New deprecation (will be removed in Theano 0.6, warning generated if you use them):
* Rop in some case should have returned a list of one theano variable, but returned the variable itself. * tensor.shared() renamed to tensor._shared. You probably want to call theano.shared()! (Olivier D.)
* Theano flags "home" is not used anymore as it was a duplicate. If you use it, theano should raise an error.
New features: New features:
* adding 1d advanced indexing support to inc_subtensor and set_subtensor (James) * Adding 1d advanced indexing support to inc_subtensor and set_subtensor (James Bergstra)
* tensor.{zeros,ones}_like now support the dtype param as numpy (Fred) * tensor.{zeros,ones}_like now support the dtype param as numpy (Frederic)
* config flags "exception_verbosity" to control the verbosity of exceptions (Ian) * Added configuration flag "exception_verbosity" to control the verbosity of exceptions (Ian)
* theano-cache list: list the content of the theano cache (Fred) * theano-cache list: list the content of the theano cache (Frederic)
* tensor.ceil_int_div FB * theano-cache unlock: remove the Theano lock (Olivier)
* MaxAndArgMax.grad now work with any axis(The op support only 1 axis) FB * tensor.ceil_int_div (Frederic)
* MaxAndArgMax.grad now work with any axis(The op support only 1 axis) (Frederic)
* used by tensor.{max,min,max_and_argmax} * used by tensor.{max,min,max_and_argmax}
* tensor.{all,any} RP * tensor.{all,any} (Razvan)
* tensor.roll as numpy: (Matthew Rocklin, DWF) * tensor.roll as numpy: (Matthew Rocklin, David Warde-Farley)
* Theano works in some cases on Windows now. Still experimental. (Sebastian Urban) * Theano with GPU works in some cases on Windows now. Still experimental. (Sebastian Urban)
* IfElse now allow to have a list/tuple as the result of the if/else branches. * IfElse now allow to have a list/tuple as the result of the if/else branches.
* They must have the same length and correspondig type) RP * They must have the same length and corresponding type) (Razvan)
* argmax dtype as int64. OD * Argmax output dtype now int64. (Olivier)
* added arccos IG * Added the element-wise operation arccos. (Ian)
* sparse dot with full output. (Yann Dauphin) * sparse dot with full grad output. (Yann Dauphin)
* Optimized to Usmm and UsmmCscDense in some case (YD) * Optimized to Usmm and UsmmCscDense in some case (Yann)
* Note: theano.dot, sparse.dot return a structured_dot grad( * Note: theano.dot, sparse.dot return a structured_dot grad.
This mean that the old grad returned a grad value with the same sparsity pattern then the inputs.
* GpuAdvancedSubtensor1 support broadcasted dimensions. (Frederic)
New optimizations: New optimizations:
* AdvancedSubtensor1 reuse preallocated memory if available(scan, c|py_nogc linker)(Fred) * AdvancedSubtensor1 reuse preallocated memory if available(scan, c|py_nogc linker)(Frederic)
* tensor_variable.size (as numpy) product of the shape elements OD * tensor_variable.size (as numpy) product of the shape elements. (Olivier)
* sparse_variable.size (as scipy) the number of stored value.OD * sparse_variable.size (as scipy) the number of stored value. (Olivier)
* dot22, dot22scalar work with complex(Fred) * dot22, dot22scalar work with complex. (Frederic)
* Doc how to wrap in Theano an existing python function(in numpy, scipy, ...) Fred * Generate Gemv/Gemm more often. (James)
* Generate Gemv/Gemm more often JB * remove scan when all computations can be moved outside the loop. (Razvan)
* remove scan when all computations can be moved outside the loop RP * scan optimization done earlier. This allow other optimization to be applied. (Frederic, Guillaume, Razvan)
* scan optimization done earlier. This allow other optimization to be applied FB, RP, GD * exp(x) * sigmoid(-x) is now correctly optimized to a more stable form. (Olivier)
* exp(x) * sigmoid(-x) is now correctly optimized to a more stable form. * Added Subtensor(Rebroadcast(x)) => Rebroadcast(Subtensor(x)) optimization. (Guillaume)
* Make the optimization process faster. (James)
* Allow fusion of elemwise when the scalar op need support code. (James)
GPU: * Better opt that lift transpose around dot. (James)
* GpuAdvancedSubtensor1 support broadcasted dimensions
Bug fixes (the result change):
Bug fixes that change results: * On CPU, if the convolution had received explicit shape information, they where not checked at runtime.
* On cpu, if the convolution had received explicit shape information, they where not checked at runtime. This caused wrong result if the input shape was not the one expected. (Fred, reported by Sander Dieleman) This caused wrong result if the input shape was not the one expected. (Frederic, reported by Sander Dieleman)
* Scan grad when the input of scan has sequence of different length. (RP reported by Michael Forbes) * Scan grad when the input of scan has sequence of different length. (Razvan, reported by Michael Forbes)
* Scan.infer_shape now work correctly when working with a condition for the number of loop. In the past, it returned n_steps as the shape, which is not always true. RP * Scan.infer_shape now work correctly when working with a condition for the number of loop.
* Theoretical bug: in some case we could have GPUSum return bad value. Was not able to produce the error.. In the past, it returned n_steps as the shape, which is not always true. (Razvan)
* Theoretical bug: in some case we could have GPUSum return bad value. Was not able to produce the error.
* pattern affected({0,1}*nb dim, 0 no reduction on this dim, 1 reduction on this dim ) * pattern affected({0,1}*nb dim, 0 no reduction on this dim, 1 reduction on this dim )
01, 011, 0111, 010, 10, 001, 0011, 0101: FB 01, 011, 0111, 010, 10, 001, 0011, 0101: (Frederic)
* div by zeros in verify_grad. This hid a bug in the grad of Images2Neibs. (JB) * div by zeros in verify_grad. This hid a bug in the grad of Images2Neibs. (James)
* theano.sandbox.neighbors.Images2Neibs grad was returning wrong value. The grad is now disabled and return an error. FB * theano.sandbox.neighbors.Images2Neibs grad was returning wrong value.
The grad is now disabled and return an error. (Frederic)
Crash fixed: Crash fixed:
* T.mean crash at graph building time by Ian G. * T.mean crash at graph building time. (Ian)
* "Interactive debugger" crash fix (Ian, Fred) * "Interactive debugger" crash fix. (Ian, Frederic)
* Do not call gemm with strides 0, some blas refuse it. (PL) * Do not call gemm with strides 0, some blas refuse it. (Pascal Lamblin)
* optimization crash with gemm and complex. (Fred) * Optimization crash with gemm and complex. (Frederic)
* Gpu crash with elemwise Fred * GPU crash with elemwise. (Frederic)
* compilation crash with amdlibm and the gpu. (Fred) * Compilation crash with amdlibm and the GPU. (Frederic)
* IfElse crash (Fred) * IfElse crash. (Frederic)
* Execution crash fix in AdvancedSubtensor1 on 32 bit computers (PL) * Execution crash fix in AdvancedSubtensor1 on 32 bit computers. (Pascal)
* gpu compilation crash on MacOS X (OD) * GPU compilation crash on MacOS X. (Olivier)
* gpu compilation crash on MacOS X (Fred) * Support for OSX Enthought Python Distribution 7.x. (Graham Taylor, Olivier)
* Support for OSX Enthought Python Distribution 7.x (Graham Taylor, OD) * When the subtensor inputs had 0 dimensions and the outputs 0 dimensions. (Frederic)
* When the subtensor inputs had 0 dimensions and the outputs 0 dimensions * Crash when the step to subtensor was not 1 in conjunction with some optimization. (Frederic, reported by Olivier Chapelle)
* Crash when the step to subtensor was not 1 in conjunction with some optimization * fix dot22scalar cast (Justin Bayer, Frédéric, Olivier)
Optimization:
* Added Subtensor(Rebroadcast(x)) => Rebroadcast(Subtensor(x)) optimization (GD)
* Scan optimization are executed earlier. This make other optimization being applied(like blas optimization, gpu optimization...)(GD, Fred, RP)
* Make the optimization process faster (JB)
* Allow fusion of elemwise when the scalar op need support code. (JB)
Know bug: Know bug:
* CAReduce with nan in inputs don't return the good output (`Ticket <http://trac-hg.assembla.com/theano/ticket/763>`_). * CAReduce with nan in inputs don't return the good output (`Ticket <https://www.assembla.com/spaces/theano/tickets/763>`_).
* This is used in tensor.{max,mean,prod,sum} and in the grad of PermuteRowElements. * This is used in tensor.{max,mean,prod,sum} and in the grad of PermuteRowElements.
* If you do grad of grad of scan you can have wrong number in some case. * If you do grad of grad of scan you can have wrong number in some case.
Sandbox: Sandbox:
* cvm, interface more consistent with current linker (James) * cvm, interface more consistent with current linker. (James)
* vm linker have a callback parameter (JB) * vm linker have a callback parameter. (James)
* review/finish/doc: diag/extract_diag AB,FB,GD * review/finish/doc: diag/extract_diag. (Arnaud Bergeron, Frederic, Olibier)
* review/finish/doc: AllocDiag/diag AB,FB,GD * review/finish/doc: AllocDiag/diag. (Arnaud, Frederic, Guillaume)
* review/finish/doc: MatrixInverse, matrix_inverse RP * review/finish/doc: MatrixInverse, matrix_inverse. (Razvan)
* review/finish/doc: matrix_dot RP * review/finish/doc: matrix_dot. (Razvan)
* review/finish/doc: det PH determinent op * review/finish/doc: det (determinent) op. (Philippe Hamel)
* review/finish/doc: Cholesky David determinent op * review/finish/doc: Cholesky determinent op. (David)
* review/finish/doc: ensure_sorted_indices Li Yao * review/finish/doc: ensure_sorted_indices. (Li Yao)
* review/finish/doc: spectral_radius_boud Xavier Glorot * review/finish/doc: spectral_radius_boud. (Xavier Glorot)
* review/finish/doc: sparse sum Valentin Bisson * review/finish/doc: sparse sum. (Valentin Bisson)
Sandbox New features(not enabled by default): Sandbox New features(not enabled by default):
* CURAND_RandomStreams for uniform and normal(not pickable, gpu only)(James) * CURAND_RandomStreams for uniform and normal. (not pickable, GPU only)(James)
Documentation: Documentation:
* Many update by many people: Olivier Delalleau, Fred, RP, David, * Many updates. (Many people)
* Updates to install doc on MacOS (OD) * Updates to install doc on MacOS. (Olivier)
* Updates to install doc on Windows(DWF, OD) * Updates to install doc on Windows. (David, Olivier)
* Doc how to use scan to loop with a condition as the number of iteration RP * Added how to use scan to loop with a condition as the number of iteration. (Razvan)
* Added how to wrap in Theano an existing python function .(in numpy, scipy, ...) (Frederic)
* Refactored GPU insatalltion of Theano. (Olivier)
Others: Others:
* Better error message at many places: David Warde-Farley, Ian, Fred, Olivier D. * Better error message at many places. (David, Ian, Frederic, Olivier)
* pep8: James, * pep8 fix. (Many people)
* min_informative_str to print graph: Ian G. * New min_informative_str() function to print graph. (Ian)
* Fix catching of exception. (Sometimes we catched interupt): Fred, David, Ian, OD, * Fix catching of exception. (Sometimes we catched interupt) (Frederic, David, Ian, Olivier)
* Better support for uft string(David WF) * Better support for uft string. (David)
* Fix pydotprint with a function compiled with a ProfileMode (Fred) * Fix pydotprint with a function compiled with a ProfileMode (Frederic)
* Was broken with change to the profiler. * Was broken with change to the profiler.
* warning when people have old cache entry (OD) * warning when people have old cache entry. (Olivier)
* More test for join on the gpu and cpu. * More test for join on the GPU and cpu. (Frederic)
* Don't request to load the gpu module by default in scan module. RP * Don't request to load the GPU module by default in scan module. (Razvan)
* Better opt that lift transpose around dot JB * Fix some import problem.
* Fix some import problem * Filtering update. (James)
* Filtering update JB * The buidbot raise optimization error instead of printing a warning. (Frederic)
* On Windows, the default compiledir changed to be local to the computer/user and not transfered with roaming profile. (Sebastian Urban)
Reviewers: Reviewers (alphabetical order):
James, David, Ian, Fred, Razvan, delallea * David, Frederic, Ian, James, Olivier, Razvan
Markdown 格式
0% 或
您添加了 0 人 到此讨论。请谨慎行事。
请先完成此评论的编辑！
请 注册 或者 后发表评论