AI provider

rinna

rinna is a Japanese AI developer that took over Microsoft's Rinna chatbot business in 2020. It later published Japanese language, image-generation and image-text models.

Models & tools

  • Bilingual GPT-NeoX 4B PPO

    rinna · Bilingual GPT-NeoX

    This release adds PPO alignment to the bilingual 4B SFT model using Japanese-translated Anthropic HH-RLHF data.

    Launch date Downloadable
  • Japanese Stable Diffusion

    rinna · Japanese Stable Diffusion

    A Stable Diffusion adaptation for Japanese prompts and image concepts.

    Launch date Restricted accessDownloadable
  • Japanese CLIP ViT-B/16

    rinna · Japanese image-text encoders

    Japanese CLIP ViT-B/16 learns shared image and Japanese-text representations through contrastive image-text training.

    Launch date Downloadable
  • Japanese CLOOB ViT-B/16

    rinna · Japanese image-text encoders

    Japanese CLOOB ViT-B/16 learns shared image and Japanese-text representations through Contrastive Leave One Out Boost training.

    Launch date Downloadable
  • Japanese RoBERTa Base

    rinna · rinna Japanese RoBERTa

    A Japanese masked-language foundation encoder with specific tokenizer and position-ID requirements.

    Launch date Research releaseDownloadable
  • Bilingual GPT-NeoX 4B

    rinna · Bilingual GPT-NeoX

    An English-Japanese GPT-NeoX language model with 3.8B parameters.

    Launch date not confirmedResearch releaseDownloadable
  • Bilingual GPT-NeoX 4B 8K

    rinna · Bilingual GPT-NeoX

    An English-Japanese 3.8B language model with an 8,192-token sequence length.

    Launch date not confirmedResearch releaseDownloadable
  • This visual-language release joins rinna's 3.8B bilingual GPT-NeoX model with BLIP-2 components and a projection layer.

    Launch date not confirmedResearch releaseDownloadable
  • Bilingual GPT-NeoX 4B SFT

    rinna · Bilingual GPT-NeoX

    The supervised instruction release adapts Bilingual GPT-NeoX 4B using English and Japanese instruction data.

    Launch date not confirmedDownloadable
  • A Japanese reasoning model combining Bakeneko with a DeepSeek-R1-Distill-Qwen-32B chat vector and ORPO.

    Launch date not confirmedDownloadable
  • Baku 2B continues pretraining Google's Gemma 2 2B on about 80B Japanese and English tokens.

    Launch date not confirmedResearch releaseDownloadable
  • Baku 2B Instruct adds Google's Gemma instruction chat vector to rinna's Japanese-adapted base, then applies odds-ratio preference optimization.

    Launch date not confirmedDownloadable
  • Japanese GPT 1B

    rinna · Japanese GPT

    A 1.3B-parameter Japanese GPT model for text completion.

    Launch date not confirmedResearch releaseDownloadable
  • A Japanese GPT-2 language model trained to predict continuations of text.

    Launch date not confirmedResearch releaseDownloadable
  • A Japanese GPT-2 completion model with a larger decoder than the XSmall release.

    Launch date not confirmedResearch releaseDownloadable
  • A compact Japanese text-completion model trained on Japanese CC-100 and Wikipedia.

    Launch date not confirmedResearch releaseDownloadable
  • Japanese GPT-NeoX 3.6B PPO

    rinna · Japanese GPT-NeoX 3.6B

    A Japanese GPT-NeoX 3.6B conversation model aligned with PPO after supervised fine-tuning.

    Launch date not confirmedDownloadable
  • Japanese GPT-NeoX 3.6B SFT v2

    rinna · Japanese GPT-NeoX 3.6B

    A second supervised conversation fine-tune of Japanese GPT-NeoX 3.6B.

    Launch date not confirmedDownloadable
  • A Japanese GPT-NeoX model trained with EleutherAI's training code.

    Launch date not confirmedResearch releaseDownloadable
  • Youko 70B continues pretraining Meta Llama 3 70B on 5B Japanese and English tokens.

    Launch date not confirmedResearch releaseDownloadable
  • Youko 70B Instruct combines supervised fine-tuning with a chat vector from Meta's Llama 3 instruction model.

    Launch date not confirmedDownloadable
  • Youko 8B continues pretraining Meta Llama 3 8B on 22B Japanese and English tokens.

    Launch date not confirmedResearch releaseDownloadable
  • Youko 8B Instruct combines supervised fine-tuning with a chat vector from Meta's Llama 3 instruction model, followed by DPO.

    Launch date not confirmedDownloadable
  • Nekomata 14B

    rinna · Nekomata

    Nekomata 14B continues pretraining Qwen-14B on about 66B tokens of Japanese and English text.

    Launch date not confirmedResearch releaseDownloadable
  • This model instruction-tunes Nekomata 14B, retaining its Qwen ancestry and using the Alpaca task format.

    Launch date not confirmedDownloadable
  • Nekomata 7B

    rinna · Nekomata

    Nekomata 7B continues pretraining Qwen-7B on about 30B tokens of Japanese and English text.

    Launch date not confirmedResearch releaseDownloadable
  • This model instruction-tunes Nekomata 7B, retaining its Qwen ancestry and using the Alpaca task format.

    Launch date not confirmedDownloadable
  • Bakeneko 32B continues pretraining Qwen2.5-32B on about 18B tokens from Japanese and English corpora.

    Launch date not confirmedResearch releaseDownloadable
  • Bakeneko Instruct combines the Japanese-adapted base with Qwen2.5's instruction chat vector and Simple Preference Optimization, or SimPO.

    Launch date not confirmedDownloadable
  • A Japanese instruction model that merges Bakeneko with QwQ Bakeneko.

    Launch date not confirmedDownloadable
  • QwQ Bakeneko 32B combines the Japanese-adapted Bakeneko base with a reasoning chat vector from QwQ-32B, followed by ORPO.

    Launch date not confirmedDownloadable
  • A Vicuna 13B fine-tune for the LangChain MRKL action format, distributed as a weight delta.

    Launch date not confirmedResearch releaseDownloadable
  • Youri 7B

    rinna · Youri

    Youri 7B continues pretraining Meta's Llama 2 7B on about 40B Japanese and English tokens.

    Launch date not confirmedResearch releaseDownloadable
  • Youri 7B Chat is an instruction-tuned Youri release that uses a conversation-style input format.

    Launch date not confirmedDownloadable
  • Youri 7B Instruction tunes the Llama 2-derived Japanese model for the Alpaca instruction format.

    Launch date not confirmedDownloadable