rosetta-cortex VS CortexTheseus

Machine-Learning(ML), CortexVirtual Machine(CVM) & Model RepresentationTool(MRT) R&D 1. Refactor the internal implementation of MRT (yamrt), add Fake-Quant Forward & Backward Pass, and support Quant-Aware Training; 2. yamrt adds various Post Training Quantization algorithms on the basis of the new architecture, maintains the old static quantization methods, and expands the latest quantization research methods; Cortex ZkRollup 1. ZkRollup fixes the logic problem of contract code withdraw; 2. Investigate and test ZkProver’s optimization solutions for issues such as proof generation and memory usage, such as Prover multi-instance parallel calculation, replacement of pairing calculation methods, etc.; 3. ZkProver updated and tested dependent libraries such as libsnark, optimized and tested the proof generation; 4. ZkNode python interface adjustment, fix the state transition problem of the state execution library, and update the code document; Cortex Full Node 1. Fix the crash problem of full node log; 2. New safe mode startup option for all nodes; 3. rosseta interface https://github.com/CortexFoundation/rosetta-cortex;

I have been eyeing https://www.cortexlabs.ai/ - Their reddit isn't active but their Medium/Github is. They just released a new whitepaper.

Running a Cortex Node (Cortex Github)