GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
Danielbrdz
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
Danielbrdz
Model Tree
Qwen/Qwen3.5-0.8B
Input Modalities
Output Modalities
Supported Functionality
| Pregunta | acc base → ft | Brier base → ft |
|---|---|---|
| tipo_contrato | 0.08 → 0.97 | 0.0593 → 0.0015 |
| requiere_abogado | 0.93 → 0.93 | 0.1544 → 0.0284 |
| riesgo | 0.17 → 0.73 | 0.0564 → 0.0319 |
| urgencia | 0.42 → 0.67 | 0.0529 → 0.0317 |
Global: accuracy 0.8325 · ECE 0.0466 · record accuracy 0.53 (las 4 bien por fila).
Barcenas Jev 0.8b
Based on Qwen 3.5 0.8b and trained on the Barcenas Contratos MX Decision dataset.
It is a "Jev" type model (decision model) in Mexican Spanish: it reads a contract and returns decisions with calibrated probabilities, rather than text. It specializes in contract triage: identifying the contract type, whether a lawyer is required, and the level of risk and urgency.
| Question | base acc → ft | base Brier → ft |
|---|---|---|
| contract_type | 0.08 → 0.97 | 0.0593 → 0.0015 |
| requires_lawyer | 0.93 → 0.93 | 0.1544 → 0.0284 |
| risk | 0.17 → 0.73 | 0.0564 → 0.0319 |
| urgency | 0.42 → 0.67 | 0.0529 → 0.0317 |
Overall: accuracy 0.8325 · ECE 0.0466 · record accuracy 0.53 (all 4 correct per row).
Made with ❤️ in Guadalupe, Nuevo Leon, Mexico 🇲🇽