fix flash attention test (#2263)

address comments

Add Mixtral,Qwen-MoE presets and Update conversion script. (#2248)

Create pull_request_template.md

Bump the python group with 2 updates

Update requirements-jax-cuda.txt (#2252)

update of kaggle handle for moonshine (#2253)

Add Moonshine to KerasHub (#2093)

Update the int8 quant logic of ReversibleEmbedding. (#2250)

Add new CSPNet preset and add manual padding. (#2212)

fix the layer name (#2247)

Made label data optional for inference and adopted other required ch…

Add Mixtral (#2196)

Add Qwen Moe (#2163)

Sharded weights support (#2218)

script for converting retinanet weights from trochvision (#2233)

Add qwen presets (#2241)

Bump the python group with 2 updates

Fix up package build scripts with better error checking (#2230)

Cast a few more input to tensors in SD3 (#2234)

Fix up package build scripts with better error checking

Fix up package build scripts with better error checking

Fix up package build scripts with better error checking

Another fix for nightly builds (#2229)

Fix nightly (#2227)

Set GPU timeouts to 2 hours (#2226)

Add utility to map COCO IDs to class names (#2219)

Fix the keras_hub package for typecheckers and IDEs (#2222)

fix: apply torchfix weights_only suggestion (#2215)