# syhw/wer_are_we

Attempt at tracking states of the arts and recent results (bibliography) on speech recognition.

Repository: https://github.com/syhw/wer_are_we
Canonical: https://ross.abutalabs.com/products/wer_are_we
License Family: other
Topics: deep-neural-network, wer, speech-recognition
Last push: 2022-06-27T08:24:34+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 4045, "days_push": 1528, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1864, forks 224 (observed 2026-08-28T04:05:46.170600+00:00)

## What it is
A curated, community-maintained tracker of state-of-the-art word error rate (WER) results on speech recognition benchmarks like LibriSpeech, with links to the source papers. It serves as a bibliography and leaderboard for automatic speech recognition research progress.

## Use cases
- find the current state-of-the-art WER on LibriSpeech
- look up papers achieving low word error rates in speech recognition
- compare ASR model benchmark results over time
- find a bibliography of speech recognition research papers
- check how close speech recognition is to human-level performance
- track recent results in automatic speech recognition

## When to choose
- you need an up-to-date reference of SOTA speech recognition benchmark numbers
- you want paper citations for ASR results to compare your own model against
- you are researching progress in speech recognition over the years

## When to avoid
- you need runnable speech recognition code or models rather than benchmark tables
- you need benchmarks for languages or datasets other than those covered (e.g., beyond LibriSpeech)
- you need a maintained software library with a license

## Facets
- artifact type: dataset
- maturity: maintenance
- function: speech-recognition, benchmarking, documentation
- domain: speech-processing, machine-learning
- platform: cross-platform
- tags: word-error-rate, state-of-the-art-tracking, librispeech, asr-benchmark, bibliography, leaderboard, research

## Member repositories
- syhw/wer_are_we (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:46.170600+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:15:37.042324+00:00, confidence not recorded.
  - readme: https://github.com/syhw/wer_are_we (fetched 2026-08-28T04:05:46.170600+00:00, sha 00a80e967563)
- Data as of 2026-08-30T08:39:29.467469+00:00.