Fine-tuning speech models for Algerian Darija: a team's four open questions
Fair_Bed_914 · reddit · 2026-10-13
A team building a phone voice agent for Algeria finds current models handle Algerian Darija poorly and wants to fine-tune their own, despite no prior speech-model experience. They have plenty of real call recordings and ask four questions:
- Which STT/TTS models with commercial licenses to pick
- Fine-tune vs LoRA, hardware needs, good guides/notebooks
- Data requirements: hours, formats, how perfect transcripts must be
- Dialect handling: Darija varies sharply across regions (east/west/capital/south), with French and Berber mixing — balanced per-region data or one big mixed dataset?
More from coding & agent
- AI Engineer New York opens with a print daily: 300K-dev telemetry says org, not model — jonmarkgo · 2026-10-13
- One-line decorator turns any combinatorial solver into a differentiable NN layer — PMinervini · 2026-10-13
- Google's enterprise Gemini agent: one memory across Workspace, M365 and Slack, routes work to Claude — BenBajarin · 2026-10-13
- Reverse-Engineering Elgato 4K Capture Card With T3 Code, Latency Down to ~24ms — talkaboutdesign · 2026-10-13
- Air Street Capital's Nathan Benaich: Codex predicted prompts are very good — nathanbenaich · 2026-10-13
- AI bot posting on Reddit for its user sparks debate on where agent approval lines should be — ChatGPTworkBot · 2026-10-13