Dedicated Endpoints
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Get help setting up a custom Dedicated Endpoints.
Talk with our engineer to get a quote for reserved GPU instances with discounts.
README
License: apache-2.0Training Setup
- Base model: KasuleTrevor/cdli-qwen3-asr-lg-textpre-0p6b-v2
- Datasets: KasuleTrevor/lg_100hrs + dmusingu/yogera-dataset (Luganda split)
- Training language tag:
Luganda - Forced inference language: disabled
- Epochs: 5
- Batch size: 8
- Gradient accumulation: 2
- Learning rate: 2e-05
- Scheduler: cosine
- Warmup ratio: 0.03
- Save steps: 500
- Selected checkpoint: checkpoint-20500
- Selection reason: best validation normalized WER (0.218026, CER 0.048474)
Final Test Metrics
- Corpus WER (normalized): 0.265987
- Corpus CER (normalized): 0.069419
- Average utterance WER (normalized): 0.270043
- Average utterance CER (normalized): 0.070111
Checkpoint Selection Evidence
| checkpoint | wer_normalized | cer_normalized | avg_wer_normalized | avg_cer_normalized | eval_loss |
|---|---|---|---|---|---|
| checkpoint-20500 | 0.2180256376371085 | 0.04847391182601662 | 0.24485461110063012 | 0.05152957450996647 | 0.3063507378101349 |
| checkpoint-21000 | 0.2181049292982688 | 0.048431131576491086 | 0.24545219959764403 | 0.051482337430740925 | 0.30632004141807556 |
| checkpoint-20000 | 0.2182106515131492 | 0.04872670420957655 | 0.24445939342968798 | 0.051514217055101334 | 0.30641767382621765 |
| checkpoint-19000 | 0.21842209594291 | 0.04854002675710152 | 0.2448993132712972 | 0.051402762833701474 | 0.3062317371368408 |
| checkpoint-21095 | 0.2184749570503502 | 0.048847266730966676 | 0.24567316623026003 | 0.05165522814465006 | 0.30632004141807556 |
| checkpoint-19500 | 0.21947931809171403 | 0.049080613546560466 | 0.24568551036129177 | 0.05167133029624753 | 0.306130051612854 |
Source Dataset Breakdown
| source_dataset | n_samples | mean_wer | mean_cer | median_wer | median_cer |
|---|---|---|---|---|---|
| lg100 | 2828 | 0.27 | 0.0701 | 0.2222 | 0.0379 |
Artifacts
- Result folder:
results/checkpoint-20500/ - Includes checkpoint validation summaries, final test predictions, scored outputs, and grouped analyses.
Model provider
KasuleTrevor
Model tree
Base
KasuleTrevor/cdli-qwen3-asr-lg-textpre-0p6b-v2
Fine-tuned
this model
Modalities
Input
Audio
Output
Text
Pricing
Dedicated Endpoints
View detailsSupported Functionality
Model APIs
Dedicated Endpoints
Container
More information