LLMs are fundamentally transforming the paradigm of ML deployment and compilation. Simultaneously, the increasing complexity of ML optimization pipelines has rendered many legacy components inadequate for meeting rapidly evolving requirements.
On the other hand, the open-source community faces a shortage of volunteers willing to maintain these codebases consistently. Consequently, we must prioritize and concentrate our efforts on key strategic approaches to address these challenges effectively. For most common use cases, the Unity flow can effectively replace legacy components, incorporating features such as static shape auto-tuning and BYOC capabilities. While we acknowledge that some niche scenarios (e.g., microTVM) may not be fully supported initially, we can address these later if strong demand persists. In summary, I concur that the time has come to gradually phase out legacy components. This strategic move will serve two crucial purposes: 1. Cleanup the codebase: By removing outdated or redundant elements, we can significantly reduce complexity and improve maintainability. 2. Unify our focus: Concentrating our efforts on the new unity flow will allow for more efficient development and innovation. --- [Visit Topic](https://discuss.tvm.apache.org/t/phasing-out-legacy-components/17703/6) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/a470d6b4a9962507e1538c08c58931f059e369be34f5625ad647afe16c8d0294).
