Enable array API testing for jax.experimental.array_api #29647
Conversation
This file contains hidden or 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
|
I forgot to add jax to one of the CI builds with |
|
We get 41 failed array API tests. Most (all?) of them seem caused by the inplace assignment problem. Also note that 140 array API tests pass with jax though. EDIT: here is an archive of the 42 failures observed on the CPU-only run (vs 41 for the CUDA run): I checked and the two The only other failure is an assertion error in
>>> a = xp.arange(2 ** 32 + 1, dtype=xp.uint8)
>>> xp.asarray(2 ** 32 + 1).dtype
Traceback (most recent call last):
Cell In[18], line 1
xp.asarray(2 ** 32 + 1).dtype
File ~/miniforge3/envs/dev/lib/python3.11/site-packages/jax/_src/numpy/lax_numpy.py:3568 in asarray
return array(a, dtype=dtype, copy=bool(copy), order=order, device=device)
File ~/miniforge3/envs/dev/lib/python3.11/site-packages/jax/_src/numpy/lax_numpy.py:3387 in array
_ = dtypes.coerce_to_array(object, dtype)
File ~/miniforge3/envs/dev/lib/python3.11/site-packages/jax/_src/dtypes.py:322 in coerce_to_array
dtype = _scalar_type_to_dtype(type(x), x)
File ~/miniforge3/envs/dev/lib/python3.11/site-packages/jax/_src/dtypes.py:311 in _scalar_type_to_dtype
raise OverflowError(f"Python int {value} too large to convert to {dtype}")
OverflowError: Python int 4294967297 too large to convert to int32
>>> xp.asarray([2 ** 32 + 1])
Array([1], dtype=int32) |
This was referenced Aug 9, 2024
Open
|
some progress at scipy/scipy#22070 |
1 task
|
For the record, #30340 was recently merged to EDIT: updating this draft PR to explore usage of |
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Outdated suggestions cannot be applied.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 draft PR to run the existing array API tests on jax inputs. This is not expected to work as jax does not support inplace updates via
__setitem__in particular as discussed in:.atfor simulating in-place ops data-apis/array-api#609Note that scipy started to run its array API tests against jax but maintains a list of tests to skip because of that design decision:
TODO before considering a review for merge