vasanth009's picture
MacWispr polish training + eval datasets (structure-v2/v3 era)
a39a7d3 verified
|
Raw
History Blame Contribute Delete
2.18 kB
metadata
license: mit
language:
  - en
task_categories:
  - text-generation
tags:
  - dictation
  - speech-to-text
  - text-cleanup
  - on-device
  - synthetic
pretty_name: MacWispr Polish Training Data
size_categories:
  - 1K<n<10K

MacWispr Polish — training & eval datasets

The complete open dataset behind MacWispr's on-device dictation polish model (Qwen3.5-0.8B post-trained to turn raw speech-to-text into clean, structured writing). Training pipeline and verifier live in the MacWispr repo.

Contents

Path Rows What it is
sft/train.jsonl (+valid/test) 3,011 / 276 / 173 Main SFT pool. {"text": "### Input:\n<raw>\n\n### Output:\n<gold>"}
synthetic/synth_hard.jsonl 311 Synthetic hard-category examples (multi-list, numbered, mixed styles, checklist) generated with Grok, validated by the rule-based polish_verifier, deduped vs all other pools. {"raw", "gold", "tags", "source"}
eval/ood_eval_set.jsonl 40 Out-of-distribution eval suite (held out from all training). {"id", "raw", "tags", ...}
dpo/dpo_prompts.jsonl 220 Prompts + golds used to build DPO preference pairs
results/ Benchmark vs Claude Sonnet (same suite, same scorer), incl. per-case outputs

Task

Input: raw ASR transcript with disfluencies and a spoken formatting request. Output: cleaned text with the requested structure (bullets / 1. numbered / - [ ] checklists / multiple labelled lists / email), fillers removed, self-corrections applied, questions preserved as questions (never answered).

Benchmark snapshot (2026-07-21)

40-case OOD suite, shared rule-based scorer:

System Pass Mean latency
MacWispr local 0.8B (4-bit, on-device) 23/40 (57.5%) 191 ms
Claude Sonnet (cloud) 25/40 (62.5%) 4,407 ms

Provenance & license

Synthetic and curated data created for this project (no user dictations — MacWispr never collects transcripts). Grok was used to draft the synthetic examples; every gold is machine-validated by the open verifier. MIT.