Enhance cache strategy with board-specific cache keys

[anilrajrimal1]

Summary:

This PR refines the caching strategy used in the Compile Library GitHub Actions workflow by making cache keys board-specific. Previously, caches for pip and PlatformIO were shared across all board configurations, which could lead to cache conflicts or inefficiencies. Now, each board (e.g., nodemcuv2, lolin32) has its own unique cache, improving cache hits and reducing build times.

Changes:

• Modified cache keys for pip and PlatformIO to incorporate the board matrix value.
• This change ensures better cache management when multiple boards are being built, preventing cache collisions.

Motivation:

This change improves the efficiency of the CI/CD pipeline by ensuring that the cache is appropriately scoped for each board, optimizing build times for different configurations.