AI News Google Introduces T5Gemma 2: Encoder Decoder Models with Multimodal Inputs via SigLIP and 128K Context December 19, 2025 0 200 FacebookXPinterestWhatsAppLinkedinReddItEmailPrintTumblrTelegramMix