Vik Korrapati: prediction

Transcript context

…Okay, easy enough. All right. Hi, everyone. My name is Vic. I've been working on Moondream for almost a year now. Like Shawn mentioned, I just went and looked and it turns out the first version I released December [00:41:00] 29, 2023. It's been a fascinating journey. So Moonbeam started off as a tiny vision language model. Since then, we've expanded scope a little bit to also try and build some tooling, client libraries, et cetera, to help people really deploy it. Unlike traditional large models that are focused at assistant type use cases, we're laser focused on building capabilities that developers can, sorry, it's yeah, we're basically focused on building capabilities that developers can use to build vision applications that can run anywhere. So, in a lot of cases for vision more so than for text, you really care about being able to run on the edge, run in real time, etc. So That's really important. We have we have different output modalities that we support. There's query where you can ask general English questions about an image and get back human like answers. There's captioning, which a lot of our users use for generating synthetic datasets to then train diffusion models and whatnot. We've done a lot of work to minimize those sessions there. [00:42:00] So that's. Use lot. We have open vocabulary object detection built in similar to a couple of more recent models like Palagem, et cetera, where rather than having to train a dedicated model, you can just say show me soccer balls in this image or show me if there are any deer in this image, it'll detect it. More recently, earlier this month, we released pointing capability where if all you're interested in is the center of an object you can just ask it to point out where that is. This is very useful when you're doing, you know, I automation type stuff. Let's see, LA we, we have two models out right now. There's a general purpose to be para model, which runs fair. Like it's, it's it's fine if you're running on server. It's good for our local Amma desktop friends and it can run on flagship, flagship mobile phones, but it never. so much for joining us today, and we'll see you in the [00:43:00] next one. Less memory even with our not yet fully optimized inference client. So the way we built our 0. 5b model was to start with the 2 billion parameter model and prune it while doing continual training to retain performance. We, our objective during the pruning was to preserve accuracy across a broad set of benchmarks. So the way we went about it was to estimate the importance of different components of the model, like attention heads, channels MLP rows and whatnot using basically a technique based on the gradient. So the way we went about it was to estimate the importance of different components of the model, like attention heads, channels MLP rows and whatnot using basically a technique based on the gradient. I'm not sure how much people want to know details. We'll be writing a paper about this, but feel free to grab me if you have more questions. Then we iteratively prune a small chunk that will minimize loss and performance retrain the model to recover performance and bring it back. The 0. 5b we released is more of a proof of concept that this is possible. I think the thing that's really exciting about this is it makes it possible for for developers to build using the 2B param [00:44:00] model and just explore, build their application, and then once they're ready to deploy figure out what exactly they need out of the model and prune those capabilities into a smaller form factor that makes sense for their deployment target. So yeah, very excited about that. Let me talk to you folks a little bit about another problem I've been working on recently, which is similar to the clocks example we've been talking about. We had a customer reach out who was talking about, like, who had a bunch of gauges out in the field. This is very common in manufacturing and oil and gas, where you have a bunch of analog devices that you need to monitor. It's expensive to. And I was like, okay, let's have humans look at that and monitor stuff and make sure that the system gets shut down when the temperature goes over 80 or something. So I was like, yeah, this seems easy enough. Happy to, happy to help you distill that. Let's, let's get it going. Turns out our model couldn't do it at all. I went and looked at other open source models to see if I could just generate a bunch of data and learn from that. Did not work either. So I was like, let's look at what the folks with [00:45:00] hundreds of billions of dollars in market cap have to offer. And yeah, that doesn't work either. My hypothesis is that like the, the way these models are trained are using a large amount of image text data scraped from the internet. And that can be biased. In the case of gauges, most gauge images aren't gauges in the wild, they're product images. Detail images like these, where it's always set to zero. It's paired with an alt text that says something like GIVTO, pressure sensor, PSI, zero to 30 or something. And so the models are fairly good at picking up those details. It'll tell you that it's a pressure gauge. It'll tell you what the brand is, but it doesn't really learn to pay attention to the needle over there. And so, yeah, that's a gap we need to address. So naturally my mind goes to like, let's use synthetic data to, Solve this problem. That works, but it's problematic because it turned out we needed millions of synthetic gauge images to get to reasonable performance. ind goes to like, let's use synthetic data to, Solve this problem. That works, but it's problematic because it turned out we needed millions of synthetic gauge images to get to reasonable performance. And thinking about it, reading a gauge is like [00:46:00] not a one, like it's not a zero short process in our minds, right? Like if you had to tell me the reading in Celsius for this, Real world gauge. There's two dials on there. So first you have to figure out which one you have to be paying attention to, like the inner one or the outer one. You look at the tip of the needle, you look at what labels it's between, and you count how many and do some math to figure out what that probably is. So what happens if we just add that as a Chain of thought to give the model better understanding of the different sub, to allow the model to better learn the subtasks it needs to perform to accomplish this goal. So you can see in this example, this was actually generated by the latest version of our model. It's like, okay, Celsius is the inner scale. It's between 50 and 60. There's 10 ticks. So the second tick, it's a little debatable here, like there's a weird shadow situation going on, the dial is off, so I don't know what the ground truth is, but it works okay. There's points on there that are, the points [00:47:00] over there are actually grounded. I don't know if this is easy to see, but when I click on those, there's a little red dot that moves around on the image. The model actually has to predict where this points are, I was already trying to do this with bounding boxes, but then Malmo came out with pointing capabilities. And it's like pointing is a much better paradigm to to represent this. We see pretty good results. This one's actually for clock reading. I couldn't find our chart for gauge reading at the last minute. So the light. Blue chart is with our rounded chain of thought. This measures, we have, we built a clock reading benchmark about 500 images. This measures accuracy on that. You can see it's a lot more sample efficient when you're using the chain of thought to model. Another big benefit from this approach is like, you can kind of understand how the model is. it and how it's failing. So in this example, the actual correct reading is 54 Celsius, the model output [00:48:00] 56, not too bad but you can actually go and see where it messed up. Like it got a lot of these right, except instead of saying it was on the 7th tick, it actually predicted that it was the 8th tick and that's why it went with 56. So now that you know that this. Failing in this way, you can adjust how you're doing the chain of thought to maybe say like, actually count out each tick from 40, instead of just trying to say it's the eighth tick. Or you might say like, okay, I see that there's that middle thing, I'll count from there instead of all the way from 40.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.