# Multi-domain adaptation

FlexOlmo is a new type of large language model that supports a flexible paradigm for data collaboration, allowing data owners to contribute data without relinquishing control.

Hermes-3-Llama-3.1-405B is a large language model based on the Llama architecture, focusing on text generation tasks.

This is a large language model with 12B parameters merged using the mergekit tool, combining features from multiple excellent models.

XYZ-embedding-zh is a Chinese embedding model based on sentence-transformers, which can map sentences and paragraphs to a 1792-dimensional dense vector space and is suitable for tasks such as clustering and semantic search.

This is a merged model based on Mistral-Small-24B-Instruct-2501, utilizing the DARE TIES fusion method to integrate multiple 24B parameter-scale models.

Turkish speech recognition model fine-tuned based on OpenAI Whisper-small

German speech recognition model optimized based on Whisper Large v3 Turbo, fine-tuned on a 13-hour curated dataset, significantly improving German recognition accuracy

A Latvian automatic speech recognition model fine-tuned based on whisper-large-v3, trained by AiLab.lv, supporting Latvian speech-to-text tasks.

An English speech recognition model jointly developed by NVIDIA NeMo and Suno.ai, supporting punctuation and capitalization conversion, based on FastConformer-TDT-CTC architecture

A Transformer model specifically developed for bidirectional translation between Swahili and English, fine-tuned on 210,000 sentence pairs

This is the ctranslate2 converted version of bond005/whisper-large-v3-ru-podlodka, optimized for Russian speech recognition tasks.

Open-source model based on MIT license, specific functionality depends on the actual model

Turkish-Llama-8b-v0.1 is a fully fine-tuned Turkish text generation model based on LLaMA-3 8B, trained on a 30GB Turkish dataset.

Russian spelling correction model trained on FRED-T5-1.7B, capable of correcting spelling, punctuation, and capitalization errors

A Dutch-adapted version based on Meta Llama-2-7b, fine-tuned using LoRa technology with 32B Dutch Llama-2 tokens as the base model

A text generation model trained on the SlimPajama-627B dataset, utilizing the retina network architecture.

Distilled model based on FRED-T5-1.7B architecture for automatic correction of Russian spelling, punctuation, and capitalization errors

A Russian and English spelling correction model based on the mT5-large architecture, normalizing words to correct spelling and typographical errors.

Russian spelling correction model based on FRED-T5-large, capable of correcting spelling, punctuation, and capitalization errors

OpenSearch Neural Sparse Encoding Model v1, used to encode queries and documents into 30,522-dimensional sparse vectors for efficient search relevance and retrieval.

Carballo-bloom-1.3B is a 1.3B-parameter Transformer-based Galician causal language model, continuously pre-trained based on FLOR-1.3B.

A bilingual and cross-language reranking model optimized for RAG, supporting Chinese, English, Japanese, and Korean, providing explainable absolute scores

Parakeet CTC 1.1B is an automatic speech recognition model jointly developed by NVIDIA NeMo and Suno.ai, based on the FastConformer architecture with approximately 1.1 billion parameters, supporting English speech transcription.

Parakeet RNNT 1.1B is an automatic speech recognition model jointly developed by NVIDIA NeMo and Suno.ai, based on the FastConformer Transducer architecture with approximately 1.1 billion parameters, supporting English speech transcription.

A text-to-video generation model fine-tuned using the Hugging Face Transformers library, capable of generating video content based on text input.

Document classification model based on the Donut framework, trained on a small-scale RVL-CDIP dataset

DeBERTa-v3 model specifically designed for zero-shot classification, trained on 27 tasks and 310 categories, supporting multi-domain text classification

A Russian spelling check model trained based on M2M100-1.2B, capable of correcting spelling and typing errors

This is a translation model trained via AutoTrain, supporting translation tasks between unknown languages.

The NVIDIA FastConformer-Transducer is a high-performance model for English automatic speech recognition (ASR), utilizing an optimized FastConformer architecture and Transducer decoder with approximately 618 million parameters.

NVIDIA FastConformer-CTC XLarge is an Automatic Speech Recognition (ASR) model with approximately 600 million parameters, designed specifically for English speech transcription and trained using the FastConformer architecture and CTC loss.

This is a large automatic speech recognition (ASR) model based on the FastConformer architecture, specifically designed for transcribing English speech into text.

This is a large automatic speech recognition (ASR) model based on the FastConformer architecture, specifically designed for transcribing English speech into text.

This is a FastConformer hybrid model for Russian automatic speech recognition, combining Transducer and CTC decoders with approximately 115 million parameters.

This is a quantized version of the Helsinki-NLP/opus-mt-de-en model, enabling fast inference through CTranslate2, supporting German to English machine translation.

An English-to-Korean translation model fine-tuned from KETI-AIR/long-ke-t5-base, trained on multiple AI Hub datasets with a BLEU score of 42.463.

AI model released under MIT license, specific functions depend on the actual model

Top-tier English-Vietnamese and Vietnamese-English bidirectional translation model trained on MTet and PhoMT datasets, supporting high-quality multi-domain text translation

A Croatian automatic speech recognition model fine-tuned from facebook/wav2vec2-xls-r-300m, trained on the ParlaSpeech-HR v1.0 dataset

French-to-English neural machine translation model under the OPUS-MT project, based on transformer-big architecture, using SentencePiece tokenization