Technical details

The full technical story.

The product blog tells the story in one pass. These 12 pages expand each part of it, in the same order. The indented pages go one level further into the method.

01

Data collection

Where the speech comes from and how it was written.

Blog chapter
02

Tokenizer

Why the vocabulary sets decoding speed.

Blog chapter
03

Evaluation

Why WER misleads, and how the judge was validated.

Blog chapter
04

Results

Every system, every judge, every metric.

Blog chapter
05

DarijaCS benchmark

External validation, error breakdowns, and quantization.

Blog chapter
06

Local inference

Running the 0.6B model on a laptop CPU.

Blog chapter