-
Notifications
You must be signed in to change notification settings - Fork 19.4k
Pull requests: keras-team/keras
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Remove non-used param & add initial value for callback.py
size:XS
#20043
by shashaka
was merged Jul 25, 2024
Loading…
Support torch
convert_to_numpy
for all devices
size:XS
#20042
by aboubezari
was merged Jul 25, 2024
Loading…
Truncate msg in
Progbar
if it exceeds the terminal column size
size:XS
#20038
by james77777778
was closed Jul 25, 2024
Loading…
Add clear deprecation error message for add_metric()
size:XS
#20034
by grasskin
was merged Jul 23, 2024
Loading…
Fix jax's
in_top_k
and numpy's top_k
size:S
#20033
by james77777778
was merged Jul 23, 2024
Loading…
Fix QLoRA saving/loading
Gemma
Gemma model specific issues
size:S
#20029
by james77777778
was merged Jul 23, 2024
Loading…
Fix test_softmax_correctness_with_axis_tuple test case failed on torch gpu ci.
size:XS
#20025
by shashaka
was merged Jul 22, 2024
Loading…
Bring back loss information for multiple outputs
size:M
#20023
by james77777778
was merged Jul 24, 2024
Loading…
Fixing issue #20021: Unexpected result on keras.ops.nn.softmax when running with tuple axis
size:M
#20022
by shashaka
was merged Jul 22, 2024
Loading…
Faster host to device transfer on Jax backend
size:S
#20018
by Hilly12
was merged Jul 23, 2024
Loading…
Fixing issue #19754: len is not well defined for a symbolic Tensor
size:XS
#20016
by shashaka
was closed Jul 28, 2024
Loading…
Revert to the original default behavior of
dtype
in Loss
and Metric
size:M
#20014
by james77777778
was merged Jul 19, 2024
Loading…
Previous Next
ProTip!
Exclude everything labeled
bug
with -label:bug.