Opus Mt Lt De Onnx Int8
encodermodelquantized.onnx decodermodelquantized.onnx decoderwithpastmodelquantized.onnx (KV cache, decoding O(1) por token)
Fonte do modelo
- Tipo
- Other
- Tamanho do arquivo
- 47,4–83,0 MB
- Versão
- vRevision 5e47835
- Atualizado
- 14 de set. de 2026
- Publicador
- @R4kSo1997
- Downloads
- 5
- Fonte
- Hugging Face
Descrição da fonte
opus-mt-lt-de ONNX int8 (movil offline)
Conversion movil-ready de Helsinki-NLP/opus-mt-lt-de.
ONNX
- encoder_model_quantized.onnx
- decoder_model_quantized.onnx
- decoder_with_past_model_quantized.onnx (KV cache, decoding O(1) por token)
Quant: int8 dynamic ARM64 + per_channel.
Tokenizer
- tokenizer.json (Fast / Xenova / transformers.js)
- source.spm, target.spm, vocab.json (Marian raw)
Source: Helsinki-NLP/opus-mt-lt-de
Repo: R4kSo1997/opus-mt-lt-de-onnx-int8
Fontes
1 fonte
Fontes
1 fonteVerificado 10 de ago.
Artefatos de modelo
3 artefatos
Artefatos de modelo
3 artefatosTrechos de fonte
3 trechos
Trechos de fonte
3 trechos--- library_name: transformers tags: - onnx - int8 - marian - opus-mt - translation - offline - lt - de --- # opus-mt-lt-de ONNX int8 (movil offline) Conversion movil-ready de `Helsinki-NLP/opus-mt-lt-de`. ## ONNX - encoder_model_quantized.onnx - decoder_model_quantized.onnx - decoder_with_past_model_quantized.onnx (KV cache, decoding O(1) por token) Quant: int8 dynamic ARM64 + per_channel. ## Tokenizer - tokenizer.json (Fast / Xenova / transformers.js) - source.spm, target.spm, vocab.json (Marian raw) Source: `Helsinki-NLP/opus-mt-lt-de` Repo: `R4kSo1997/opus-mt-lt-de-onnx-int8`
Source context: 5 downloads · 0 likes · Pipeline translation · Library transformers · Repo R4kSo1997/opus-mt-lt-de-onnx-int8
Source context: 2 downloads · 0 likes · Pipeline translation · Library transformers · Repo R4kSo1997/opus-mt-lt-de-onnx-int8
