Offline English ASR model based on the Icefall/K2 pruned transducer Zipformer-CR-CTC recipe. The model was trained with a transducer objective and CR-CTC (Consistency-Regularized CTC) auxiliary loss, and supports Icefall transducer and CTC decoding.

Files

model.pt: Zipformer CR-CTC transducer checkpoint
bpe.model: SentencePiece BPE model
tokens.txt: Icefall token table exported from bpe.model
config.yaml: model architecture, feature extraction, tokenizer, Zipformer model, decoding settings, and Hugging Face Hub download-metrics query file

Evaluation

Open ASR Leaderboard English short-form result, decoded with modified_beam_search and beam size 6:

Metric	Value
Average WER	5.56
Parameters	288M

Dataset WERs:

Dataset	WER
AMI	11.30
Earnings22	7.64
GigaSpeech	8.41
LibriSpeech Clean	1.31
LibriSpeech Other	3.01
SPGISpeech	1.64
VoxPopuli	5.60
Average	5.56

Training Data

The model was trained on a combined English training mixture built from the training portions of the datasets below.

Dataset	Train Hours	Source
LibriSpeech	960.0	ESB datasets
Earnings-22	105.0	ESB datasets
AMI Meeting Corpus	78.0	ESB datasets
Common Voice Scripted Speech 25.0 - English	~1,679.0	Mozilla Data Collective
Common Voice Spontaneous Speech 3.0 - English	~3.6	Mozilla Data Collective
GigaSpeech XL	10,000.0	ESB datasets
SPGISpeech	4,900.0	ESB datasets
TED-LIUM Release 3	454.0	ESB datasets
VoxPopuli	523.0	ESB datasets
Total	~18,700	ESB datasets and Mozilla Data Collective

Common Voice Scripted Speech 25.0 - English and Common Voice Spontaneous Speech 3.0 - English are the English scripted and spontaneous Common Voice releases checked on May 12, 2026.

For data normalization, the training transcripts were processed with direct LLM normalization using a self-hosted GLM-4.7 setup, plus custom agentic workflows built in collaboration with Claude Code 4.7 and Codex 5.4.

Training

The model was trained with the k2-fsa/Icefall framework, PyTorch 2.11, and CUDA 13.0 on 8 NVIDIA B300 GPUs using bf16 automatic mixed precision.

Usage With Icefall

From an Icefall checkout, download this model repo locally:

bash

huggingface-cli download soundsgoodai/Zipformer-cr-ctc-transducer-XL-290M \
  --local-dir Zipformer-cr-ctc-transducer-XL-290M

Run offline decoding through Icefall. Input audio must already be 16 kHz:

bash

cd icefall/egs/librispeech/ASR

PYTHONPATH=../../.. python zipformer/pretrained.py \
  --checkpoint /path/to/Zipformer-cr-ctc-transducer-XL-290M/model.pt \
  --tokens /path/to/Zipformer-cr-ctc-transducer-XL-290M/tokens.txt \
  --method modified_beam_search \
  --beam-size 6 \
  --num-encoder-layers "2,2,4,5,4,2" \
  --downsampling-factor "1,2,4,8,4,2" \
  --feedforward-dim "512,1024,2048,3072,2048,1024" \
  --num-heads "4,4,4,8,4,4" \
  --encoder-dim "192,384,768,1024,768,384" \
  --encoder-unmasked-dim "192,192,320,384,320,192" \
  --query-head-dim "32" \
  --value-head-dim "12" \
  --pos-head-dim "4" \
  --pos-dim 48 \
  --cnn-module-kernel "31,31,15,15,15,31" \
  --decoder-dim 512 \
  --joiner-dim 512 \
  --context-size 2 \
  --causal false \
  --chunk-size "16,32,64,-1" \
  --left-context-frames "64,128,256,-1" \
  --use-transducer true \
  --use-ctc true \
  --use-attention-decoder false \
  --use-cr-ctc true \
  /path/to/audio_1.wav \
  /path/to/audio_2.wav

For CTC decoding, use Icefall's CTC entrypoint with the same architecture settings:

bash

cd icefall/egs/librispeech/ASR

PYTHONPATH=../../.. python zipformer/pretrained_ctc.py \
  --checkpoint /path/to/Zipformer-cr-ctc-transducer-XL-290M/model.pt \
  --tokens /path/to/Zipformer-cr-ctc-transducer-XL-290M/tokens.txt \
  --method ctc-decoding \
  --sample-rate 16000 \
  --num-encoder-layers "2,2,4,5,4,2" \
  --downsampling-factor "1,2,4,8,4,2" \
  --feedforward-dim "512,1024,2048,3072,2048,1024" \
  --num-heads "4,4,4,8,4,4" \
  --encoder-dim "192,384,768,1024,768,384" \
  --encoder-unmasked-dim "192,192,320,384,320,192" \
  --query-head-dim "32" \
  --value-head-dim "12" \
  --pos-head-dim "4" \
  --pos-dim 48 \
  --cnn-module-kernel "31,31,15,15,15,31" \
  --decoder-dim 512 \
  --joiner-dim 512 \
  --context-size 2 \
  --causal false \
  --chunk-size "16,32,64,-1" \
  --left-context-frames "64,128,256,-1" \
  --use-transducer true \
  --use-ctc true \
  --use-attention-decoder false \
  --use-cr-ctc true \
  /path/to/audio_1.wav \
  /path/to/audio_2.wav

The architecture and decoding values above are also recorded in config.yaml.

Decoding Methods

The model supports the following Icefall decoding methods:

Transducer:

greedy_search
modified_beam_search
fast_beam_search

CTC:

ctc-decoding

The reported Open ASR Leaderboard result uses modified_beam_search with beam size 6. CTC decoding is supported, but the table above does not report CTC WERs.

Feature Extraction And Resampling

For this model, use Kaldi-style fbank features. kaldifeat and kaldi-native-fbank are the recommended feature extraction backends.

Audio should be mono 16 kHz before feature extraction. For sample-rate conversion, audioop.ratecv is recommended because it matches the resampling path used for evaluation. On Python 3.13 and newer, use an audioop-compatible package such as audioop-lts.

Output Formatting

The model emits normalized English text with punctuation and capitalization. It does not automatically capitalize the first word of every sentence unless that word is normally capitalized, such as a proper noun, honorific, acronym, or similar named expression.

The model also normalizes common written forms, including numbers, dates, and currency, to digit-based forms.

Open ASR Leaderboard Evaluation

The open_asr_leaderboard/soundsgoodai runner downloads this model repo, resamples audio to 16 kHz with audioop, computes fbank features with kaldi-native-fbank, and decodes with Icefall modified_beam_search.

bash

cd open_asr_leaderboard/soundsgoodai
bash run_zipformer.sh

Notes

The model expects 16 kHz mono audio.
The packaged leaderboard decoding configuration uses transducer modified_beam_search with beam size 6.
The checkpoint is an offline ASR model and is not intended for streaming decoding.
The checkpoint includes both transducer and CR-CTC heads, so Icefall loading should use --use-ctc true and --use-cr-ctc true.

References

Zipformer paper: Zengwei Yao, Liyong Guo, Xiaoyu Yang, Wei Kang, Fangjun Kuang, Yifan Yang, Zengrui Jin, Long Lin, and Daniel Povey. "Zipformer: A faster and better encoder for automatic speech recognition." https://arxiv.org/abs/2310.11230
CR-CTC paper: Zengwei Yao, Wei Kang, Xiaoyu Yang, Fangjun Kuang, Liyong Guo, Han Zhu, Zengrui Jin, Zhaoqing Li, Long Lin, and Daniel Povey. "CR-CTC: Consistency regularization on CTC for improved speech recognition." https://arxiv.org/abs/2410.05101
Icefall: https://github.com/k2-fsa/icefall
k2: https://github.com/k2-fsa/k2
kaldifeat: https://csukuangfj.github.io/kaldifeat/intro.html
kaldi-native-fbank: https://github.com/csukuangfj/kaldi-native-fbank
Python audioop.ratecv: https://docs.python.org/3.11/library/audioop.html#audioop.ratecv
audioop-lts for Python 3.13+: https://pypi.org/project/audioop-lts/
ESB datasets: https://huggingface.co/datasets/esb/datasets
LibriSpeech: https://www.openslr.org/12/
Earnings22: https://github.com/revdotcom/speech-datasets/tree/main/earnings22
AMI Meeting Corpus: https://www.idiap.ch/webarchives/sites/www.amiproject.org/ami-scientific-portal/meeting-corpus/
Common Voice Scripted Speech 25.0 - English: https://mozilladatacollective.com/datasets/cmndapwry02jnmh07dyo46mot
Common Voice Spontaneous Speech 3.0 - English: https://datacollective.mozillafoundation.org/datasets/cmn1pv5hi00uto1072y1074y7
GigaSpeech: https://github.com/SpeechColab/GigaSpeech
SPGISpeech: https://huggingface.co/datasets/kensho/spgispeech
TED-LIUM Release 3: https://lium.univ-lemans.fr/en/ted-lium3/
VoxPopuli: https://github.com/facebookresearch/voxpopuli