Transcript: Man, wow, I had a great call with Jordan and I'll speak on that more later after I get more things done. But I did not expect to be quantizing lava and having it work. And my script is really hacky, but at least the first version of the quantization, as far as I can tell, gave me pretty good results. And it might be possible to run more or less like GPT-4 vision style captioning locally, which would be pretty wild. Like that would be actually crazy. So that is super, super interesting. Wow, just really want a fast internet connection now. But yeah, I'm looking forward to uploading these quantized models and sharing them and seeing what happens. But yeah, I've got to get the first one out first and then I'll go to Jordan's house and upload the rest of the models. Yeah, it'll be interesting.
I invested the day delving into quantization, writing a script, and successfully running it. Afterwards, I visited Jordan's house to upload a massive 100 gigabytes of files. I'm hopeful that people will find these files useful and that there will be at least a few downloads. My aim is to further test the script on more images and present the outcomes.
I recently discovered that because my upload speed is only two megabytes per second, I had to go over to Jordan's house with two flash drives to upload our quantized model. I find it amusing that I had to resort to such an old-school method. It's not clear how frequently this will happen, but if it becomes a common occurrence, I'll need to find a better solution than using Jordan's internet.
84.68% similar
The speaker is focused on their API project and mentioned the stable diffusion model. They have also worked on running and testing various local language models, including Whisper and Orca 7 billion. They are curious to wire the models as a pipeline step and compare the output with GPT. The speaker is unsure of the success of their API project and the effectiveness of the language models, but they express eagerness to explore and experiment further.
84.67% similar
Today marked a significant advancement in the burrito project, where the image pipeline, established the previous day, became fully functional and integrated into a webpage, complete with an effective querying system. The visual aspect of the project, particularly the image embeddings, was both intriguing and aesthetically pleasing, although its effectiveness is still under review. The project is now at a stage where the creator is keen to move beyond personal experiments to sharing the results with others, with the immediate goal being to encourage a small group of individuals to test the developments. The focus for the week has shifted to actual user engagement through getting people to sign up and provide feedback, driven by the enthusiasm of witnessing the project's imagery features come to life.
84.66% similar
The user provided context discusses a phone call with someone named Pascal, during which they had a FaceTime call in Vision Pro. They discussed screen sharing, catching up, and some work-related matters. The conversation explored the idea of being comfortable in a remote environment and collapsing space in an interesting way, possibly through VR or AR. The user expressed thoughts on priorities related to Vision Pro and live transformations, highlighting the importance of adding reactivity and real-time updates to the Vision Pro scenario. They also discussed local considerations and the potential for running Lava 1.6 locally. The user expressed curiosity and a desire to continue pushing forward with these ideas.