GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
caiovicentino1
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
caiovicentino1
Model Tree
caiovicentino1/Ekbasis-Skopos-27B
Input Modalities
Output Modalities
Supported Functionality
The release protocol on these exact files, served by vLLM and the System One API, as for the bf16 model:
| INT4 | Skopos bf16 | Ekbasis-27B | |
|---|---|---|---|
| git v3, known and held-out types, all questions | 96.78% | 96.94% | 96.62% |
| Never-trained worlds, changed answers (one prompt / read once) | 81.51% / 77.91% | 81.51% / 78.68% | 81.16% / 78.03% |
ftest_family, changed answers (629) | 80.45% | 80.29% | 81.88% |
| ECE on git v3 | 0.025 | 0.023 | 0.024 |
| Guard, known types: flagged / false alarms (of 863) | 836 / 25 | 837 / 27 | 828 / 37 |
| Guard, held-out types: flagged / false alarms (of 799) | 778 / 25 | 776 / 26 | 767 / 49 |
On the 3,014 changed answers of the fresh never-trained-family confirmation (5,385 items), this build (served) scores 83.5% against 81.8% for Ekbasis-27B in bf16 by trainer readout (the confirmation's reference; +1.7 points, 95% CI [+0.8, +2.7]). Only the summary of this run was kept, not its per-item answers.
The quantization gate of the Ekbasis repository (PREREG_quantized.md: accuracy within 1 point of bf16 on git, never-trained worlds and ftest_family, ECE within 0.01, the guard within 1 point, the same answer as bf16 on at least 97% of questions and 99% of its confident ones; applied here with the guard read on the larger guard sets above instead of git v3, and without the real-repository scenarios, T3c, which were not run on this build) is passed: same answer as bf16 on 97.44% of the questions and 99.87% of the answers bf16 gives at 0.9 or more.
decision_config.json sets this build's calibrated_temperature, 1.67, fit on validation (fval_git3 + m3val) for this build: serve.py then adds calibrated to every answer, as the hosted API does (remove the key to turn it off).
bash
hf download caiovicentino1/Ekbasis-Skopos-27B-INT4 --local-dir Ekbasis-Skopos-27B-INT4bash Ekbasis-Skopos-27B-INT4/serve_vllm.sh Ekbasis-Skopos-27B-INT4 8001python Ekbasis-Skopos-27B-INT4/serve.py --model Ekbasis-Skopos-27B-INT4 --vllm-url http://127.0.0.1:8001 --port 8000pip install ekbasis && EKBASIS_URL=http://127.0.0.1:8000 ekbasis git-check -- "git reset --hard"Apache-2.0, as Ekbasis-Skopos-27B (LICENSE, LICENSE-Qwen, LICENSE-Eikos, NOTICE). Its training data contains data
derived from ShIO-bash-26.1 (Ragsdale and Boppana, 2026) and swe-rebench-v2-CodeWorldModeling (Marin Community, 2026),
both CC BY 4.0: NOTICE gives the attributions and the changes made.
hf download caiovicentino1/Ekbasis-Skopos-27B-INT4 --local-dir Ekbasis-Skopos-27B-INT4bash Ekbasis-Skopos-27B-INT4/serve_vllm.sh Ekbasis-Skopos-27B-INT4 8001python Ekbasis-Skopos-27B-INT4/serve.py --model Ekbasis-Skopos-27B-INT4 --vllm-url http://127.0.0.1:8001 --port 8000pip install ekbasis && EKBASIS_URL=http://127.0.0.1:8000 ekbasis git-check -- "git reset --hard"