| | Card: first downstream check, MMLU-Pro AP-IQ2_S 77.1% vs ISTA GSQ-RCO-IQ2_S 75.7% (70 questions) | agentionai | Oct 1, 2026 |
| | Charts: AP-IQ3_XXS/IQ3_XS/IQ3_S/Q3_K_XL rebuilt on the extended calibration (from Fieldwork as-shipped records) | agentionai |
| | AP-IQ3_XXS, AP-IQ3_XS, AP-IQ3_S, AP-Q3_K_XL: rebuilt with the extended (mixed-v4) calibration | agentionai |
| | Update README.md | agentionai |
| | Card: new AP-IQ2_S numbers (clean held-out, overlap note), pagoda first-try test, regenerated charts | agentionai |
| | AP-IQ2_S: rebuild with extended calibration (agentic transcripts); same 8.95 GiB map | agentionai |
| | Update README.md | agentionai |
| | charts: AP-IQ2_S re-encode | agentionai |
| | AP-IQ2_S: updated numbers after the IQ1-coverage re-encode | agentionai |
| | AP-IQ2_S: re-encode with full IQ1 coverage | agentionai |
| | Upload folder using huggingface_hub | agentionai |
| | Upload README.md with huggingface_hub | agentionai |
| | Upload Qwen3.8-27B-AP-IQ2_S.gguf with huggingface_hub | agentionai |
| | Upload README.md with huggingface_hub | agentionai |
| | Upload folder using huggingface_hub | agentionai |
| | Upload Qwen3.8-27B-AP-IQ3_S.gguf with huggingface_hub | agentionai |
| | Upload README.md with huggingface_hub | agentionai |
| | Update README.md | agentionai |
| | Upload README.md with huggingface_hub | agentionai |
| | Upload Qwen3.8-27B-AP-IQ3_XS.gguf with huggingface_hub | agentionai |