Adding vision to a language model is more than one extra modality. With large images or long video, the vision encoder can use more compute than the language model itself, so efficient models need a different design. Our VLM team, including ryan, Tianshu Yu, Sina Semnani, and Sam Stevens,... explains how they teach LFMs to see, from mid-training to the custom vision encoder built for LFM3.show more

@rshubert98 @TianshuYu6 This VLM is so good!

@rshubert98 @TianshuYu6 @Presidentlin

@rshubert98 @TianshuYu6 ran d1 on a plain cpu box and loved it, vision next please

@rshubert98 @TianshuYu6 You guys are a gang to watch

@rshubert98 @TianshuYu6 The Liquid team did a really good job. The model's capability far exceeds what previous 1.7B and 2B vision models could do. Huge success! You guys are on the frontier of making Artificial Intelligence free for everyone.

@rshubert98 @TianshuYu6 Tested and created this video

@rshubert98 @TianshuYu6 @grok so does this latest lfm3 latest vision encoder can stream higher than 30fps or real time video understanding and able to identify events in industrial applications than general vlms ? General vlms down sample native rate and lose context and how come this can be different?

@rshubert98 @TianshuYu6 Teaching AI to truly see smarter, faster vision for the next generation!

@rshubert98 @TianshuYu6 Wow, as someone in the field, I can say @liquidai is gonna be a huge protagonist in the future of AI.

@rshubert98 @TianshuYu6 love this! how much of the compute does the encoder take on long video?

@rshubert98 @TianshuYu6 The compute flip is the key insight — once the vision encoder eats more FLOPs than the LM, bolting on a standard ViT stops being free. Makes me wonder how much of LFM3's efficiency actually comes from the encoder design vs. the language backbone.

@rshubert98 @TianshuYu6 🔥

@rshubert98 @TianshuYu6 Vision encoder outspending the LLM is the bit nobody budgets for. At what image size does that flip?

@rshubert98 @TianshuYu6 goats

@rshubert98 @TianshuYu6 Mid-training into a custom vision encoder is the move to watch. When vision tokens cost more than language, the mix has to change with resolution, not just with epoch. 🎛️

@rshubert98 @TianshuYu6 an encoder that out-spends the language model on long video quietly changes how you size everything

@rshubert98 @TianshuYu6 Mid-training to a custom encoder for LFM3 is the jump I'm watching — long video is where most bolted-on vision setups start to hurt.

@rshubert98 @TianshuYu6 encoder out-costing the llm on long video

@rshubert98 @TianshuYu6 We spent years moving everything to the cloud. Now the biggest flex is that nothing leaves your phone.