Metal backend: Add MPSGraph caching (pytorch#15346)

This pull request introduces a new caching infrastructure for compiled
MPSGraph objects in the Metal backend, significantly improving
performance for repeated matrix multiplication and convolution
operations by reusing previously compiled graphs and their associated
tensors. The changes also include cache statistics tracking and
refactoring of the relevant code paths to leverage the cache. Note that
caching for the attention operation is not yet implemented.

### MPSGraph Caching Infrastructure

* Added a new global cache (`graph_cache`) for compiled `MPSGraph`
objects and their input/output tensors, keyed by operation type and
parameters, enabling reuse and reducing graph compilation overhead.
* Introduced `GraphCacheKey` and `CachedGraph` structures, along with a
custom hash function, to uniquely identify and store cached graphs for
matrix multiplication and convolution operations.

### Matrix Multiplication and Convolution Refactoring

* Refactored the matrix multiplication (`aoti_torch_mps_mm_out`) and
convolution (`aoti_torch_mps_convolution`) functions to check the cache
for an existing compiled graph before creating a new one, and to store
newly compiled graphs in the cache for future reuse.
[[1]](diffhunk://#diff-8b30270bea48d11579e41d09a082cfad335aa3bbb302302c2320d4e0da6b4680R261-R303)
[[2]](diffhunk://#diff-8b30270bea48d11579e41d09a082cfad335aa3bbb302302c2320d4e0da6b4680L505-R654)
* Removed manual release of `MPSGraph` objects after execution, as
cached graphs are now retained for reuse, preventing unnecessary
recompilation and memory leaks.
[[1]](diffhunk://#diff-8b30270bea48d11579e41d09a082cfad335aa3bbb302302c2320d4e0da6b4680L282-L285)
[[2]](diffhunk://#diff-8b30270bea48d11579e41d09a082cfad335aa3bbb302302c2320d4e0da6b4680R895-L759)

### Cache Statistics and Monitoring

* Added `CacheStats` structure to track cache hits and misses.

Author: manuelcandales