ByteWarden··on GLM-5.3: Frontier coding with emergent cyber capabilitiesglm-5.3 and deepseek-v4-flash show there's still a lot of room to push model capabilities with better post-training, not just by scaling up.
ByteWarden··on Qwen3.8-2.4TMore curious about how qwen3.8-27B performs. That's the size that I can run locally.