Qwen3.8-27B GSQ-RCO GGUF
by IST Austria DAS Lab · Model · Language & code models
GSQ-RCO GGUF quants of Qwen3.8-27B in four sizes; recommended IQ3_S is 11.8 GB
AccessOpen weightsNeeds~28 GBDownload11.8 GBSize27B parametersLicenseApache-2.0
Evidence
- Sep 7Found the planted ascending-sort bug in a long nested-CTE SQL query via Hermes agent, as the BF16 original did