Intended Use
This model is intended for Japanese ASR in speech with galgame or anime-like delivery, including:
- visual novel and game voice transcription
- subtitle generation workflows
- Japanese character dialogue with expressive voice acting
- research on domain adaptation from general ASR models to anime-style speech
It is not yet validated as a general-purpose Japanese ASR model. For broad Japanese speech, compare against the original base model before production use.
Training
The eval loss above comes from the training run's internal evaluation split. It should not be treated as an external benchmark score.
Evaluation
Fixed 800-Clip Benchmark
The following numbers are from a fixed 800-clip Japanese ASR evaluation set sampled with seed 20260531. The set contains 200 clips from each source:
Table with columns: source, dataset, split, clips, duration| source | dataset | split | clips | duration |
|---|
| Nekopara | grider-transwithai/nekopara-speech | train | 200 | 991.0s |
| Anime Speech | joujiboi/japanese-anime-speech | train | 200 | 1053.5s |
| JSUT Basic5000 | japanese-asr/ja_asr.jsut_basic5000 |
Total: 800 clips, 4108.3s audio, 17354 reference characters.
Metric: strict character error rate (CER) after removing whitespace and common Japanese/ASCII punctuation. S, I, and D are substitution, insertion, and deletion rates divided by reference characters. The same decoding and normalization were used for all models.
Table with columns: model, rows, CER, S, I, D| model | rows | CER | S | I | D |
|---|
Qwen/Qwen3-ASR-0.6B | 800 | 0.1673 | 0.1025 | 0.0214 | 0.0434 |
jaykwok/Qwen3-ASR-0.6B-JA-Galgame | 800 | 0.1438 | 0.0962 | 0.0228 | 0.0249 |
|
CER by source:
Table with columns: model, Nekopara, Anime Speech, JSUT, Common Voice| model | Nekopara | Anime Speech | JSUT | Common Voice |
|---|
Qwen/Qwen3-ASR-0.6B | 0.2900 | 0.1244 | 0.1297 | 0.1552 |
jaykwok/Qwen3-ASR-0.6B-JA-Galgame | 0.2392 | 0.0811 | 0.1207 | 0.1568 |
Qwen/Qwen3-ASR-1.7B | 0.2803 |
For this 1.7B checkpoint, full SFT improves overall CER from 0.1437 to 0.1285, a 10.6% relative reduction. The largest improvement is deletion reduction, from 0.0418 to 0.0242. In-domain gains are stronger: Nekopara CER improves by 18.8% relative, and Anime Speech CER improves by 26.8% relative. JSUT and Common Voice are slightly worse than the 1.7B base in this small sample, so this checkpoint should still be treated primarily as a galgame/anime-domain model rather than a general Japanese ASR upgrade.
These numbers are a small reproducible sanity benchmark, not a comprehensive public leaderboard. Strict character CER can over-penalize kana/kanji variants, long-vowel spelling, expressive writing, and transcript style differences.
Additional Evaluation Candidates
Recommended additional evaluation sets:
For a larger follow-up benchmark, use a fixed sample instead of evaluating every available hour. A practical next pass would be:
Table with columns: dataset, domain, suggested subset, reason| dataset | domain | suggested subset | reason |
|---|
ntaquan0125/steinsgate-voice | visual novel | 500-2000 clips | small, strongly in-domain, but check access/license first |
grider-transwithai/nekopara-speech | visual novel/game voice | 500-2000 fixed random clips | relevant character voice with metadata; use the full distribution unless you need content filtering |
joujiboi/japanese-anime-speech | anime/VN dialogue | 1000-3000 fixed random clips |
Report CER plus substitution, insertion, and deletion rates, with the exact normalization and decoding settings.
Repository Contents
This repository intentionally includes training recovery artifacts:
model.safetensors- tokenizer and processor files
optimizer.ptscheduler.ptrng_state.pthtrainer_state.jsontraining_args.bin
For inference-only use, the optimizer and scheduler files are not required.
Inference
Use the same inference stack as the upstream Qwen3-ASR models, replacing the model id with:
jaykwok/Qwen3-ASR-1.7B-JA-Galgame
Refer to the upstream Qwen3-ASR documentation for the latest supported inference commands and runtime requirements.
Limitations
- The model is specialized for galgame/anime-style Japanese speech and may be less reliable on news, meetings, lectures, or spontaneous conversation.
- The training data may contain adult or NSFW source material. Downstream users should account for domain and content bias.
- The published benchmark is small and should be treated as a sanity check rather than a full leaderboard result.
- Transcriptions may still contain hallucinations, punctuation differences, or style-specific handling of non-speech vocalizations.
License and Use
The base model Qwen/Qwen3-ASR-1.7B is released under Apache-2.0.
This fine-tuned checkpoint was trained on litagin/Galgame_Speech_ASR_16kHz. Users must review and comply with the dataset license and upstream terms before redistribution, commercial use, or further fine-tuning. This model card does not grant rights beyond the upstream model and dataset licenses.