GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
teckedd
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
teckedd
Model Tree
teckedd/gha-whisper-small-twi-v6
Input Modalities
Output Modalities
Supported Functionality
| Metric | Value |
|---|---|
val_wer | 0.2970 |
val_cer | 0.1031 |
google/WaxalNLP (aka_asr)local (ghana-health-ai-recorder)fsicoli/common_voice_22_0 (tw)fsicoli/common_voice_22_0 (en)ghananlpcommunity/twi-speech-text-multispeaker-16k (default)python
from transformers import pipeline # or AutoModel + processor per task# repo: teckedd/gha-whisper-small-twi-v6
If you use this checkpoint, please credit Ghana Health AI / Serendepify and the upstream base model authors plus any listed datasets.
v6-local-holdoutFalse5e-06 / 600 / 8x4True · manifest /root/gha_local_asr/manifest.train32.jsonl · weight 0.080.31490.3044): FalseProduction decode uses num_beams=5 in modal/asr_service.py.
0.3283