Version 0.5.0 #82
Merged
Version 0.5.0 #82
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
* Updated filter_{jit,grad,value_and_grad}
* Documented that BatchNorm supports multiple axis_names
* Doc fix
* added tree_inference
* Added filter_vmap and filter_pmap
* Made filter_{vmap,pmp} non-experimental. Fixed doc issues.
* Added 'inference' call-time argument to 'MultiheadAttention', which was missed when switching over 'Dropout' to use 'inference'.
* Added test for tree_inference
* Doc improvements
* Doc tweaks
* Updated tests
* Two minor bugfixes for filter_{jit,grad,vmap,pmap,value_and_grad}.
1. Applying filter_{jit,vmap,pmap} to a function with *args and **kwargs should now work. Previously inspect.Signature.apply_defaults was not filling these in.
2. The output of all filtered transformations has become a Module. This means that e.g. calling filter_jit(filter_grad(f)) multiple times will not induce any recompilation, as all filter_grad(f)s are PyTrees of the same structure as each other.
* Crash and test fixes
* tidy up
* Fixes and test fixes
* Finished adding tests.
- Can now substitute arbitrary nodes, not just leaves - These substituted nodes can now have different structures to each other. - Will raise an error if `where` depends on the values of the leaves of the PyTree. In addition, `tree_equal` should now work under JIT.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is a big update.
Exciting new features!
Added
filter_vmap.filter_vmap,filter_pmap#65.)Added
filter_pmap.filter_vmap,filter_pmap#65.)Added pooling layers:
eqx.nn.Pooleqx.nn.AvgPool1deqx.nn.AvgPool2deqx.nn.AvgPool3deqx.nn.MaxPool1deqx.nn.MaxPool2deqx.nn.MaxPool3dAdded
tree_serialise_leavesandtree_deserialise_leaves.Added
tree_inference, as a convenience for toggling all inference flags through a model.Refactoring for nicer APIs
filter_{jit,grad,value_and_grad}now have an easier-to-use API for specifying which arguments have what behaviour.(args, kwargs)as a single PyTree, then you can specify adefault,args,kwargsseparately. In particular this avoids doing messy stuff likefilter_spec=((...), {})when you had no kwargs.argsandkwargsagainst their runtime values. Both the runtime values, and the filter specification, are matched up against the function signature.e.g. you can do
filter_jit(lambda x: x, kwargs=dict(x=True))(1), using a keyword argument at JIT-time and a positional argument at call time.filter_jit(fun)andfilter_jit(default=...)(fun)will work.tree_atcan now replace subtrees, and not just leaves.tree_at. #47.)filter,partitionnow support anis_leafargument.is_leafargument tofilterandpartition#68.)Miscellaneous
filter_jit(filter_grad(fun))twice will no longer lead to unnecessary recompilation: the secondfilter_grad(fun)instance will be a PyTree that looks like the firstfilter_grad(fun)instance, and thus we won't get any recompilation.