stt-en-fastconformer-ctc-large-GGUF

GGUF quantisations of nvidia/stt_en_fastconformer_ctc_large for CrispASR.

Quant	Size	Description
F16	222 MB	Full precision
Q8_0	132 MB	8-bit
Q5_0	95 MB	5-bit
Q4_K	83 MB	4-bit K-quant (recommended)

Architecture

18-layer NeMo FastConformer encoder + Conv1d CTC head. d_model=512, 8 heads, 1024 SentencePiece vocab, English only, 80 log-mel features, ~115M params.

Usage

crispasr --backend fastconformer-ctc -m stt-en-fastconformer-ctc-large-q4_k.gguf -f audio.wav

NeMo Family

Same --backend fastconformer-ctc supports large (18L/512d), xlarge (24L/1024d), and xxlarge (42L/1024d).

Provenance and EU AI Act Art. 53 note

• Upstream model: nvidia/stt_en_fastconformer_ctc_large — published by nvidia.
• Upstream licence: cc-by-4.0. This repository redistributes under the same terms; it grants no rights the upstream licence does not.
• What was done here: format conversion and/or quantisation only (GGUF). No training, no fine-tuning, no merging, no distillation, no change to architecture, vocabulary or capability. Only the numeric representation of the upstream weights differs.
• Training data: documented — where it is documented at all — by the upstream provider; see the upstream model card. No training data was used, added or selected by this repository.
• Provider status: under Regulation (EU) 2024/1689 the upstream authors remain the provider of this model. Converting the serialisation format does not make this repository the provider of a new general-purpose AI model, and no such claim is made. Questions about training content, copyright policy or model capability belong upstream.