Release Tasks

Current Work-in-progress:

• Multi-shader kernels in a single command buffer
• Tutorial program

P0

• README
• Finish slide presentation
• Implement transformer block

P1

• Benchmarking suite
• Optimize Kernels to perform competitively
• Wrap current test code in a testing framework
• Implement Sibhoem matmul series in WGSL

P2

• separate logging instance in GPUContext so it doesn't step on the user's logging
• Implement an end-to-end architecture - gemma and/or gpt2
• Separate template project repo (based on examples/hello-world but without the relative directory pointers)

Done

• Naming convention?
• Refactor Shape to not be dynamically allocated
• Builds work for all examples
• Update deprecated SetDeviceLostCallback --> SetDeviceLostCallbackInfo