5 ms·
Well you most likely need a CUDA compatible GPU with a non-negligible amount of vram to load the model. And a few hours of your time to mash your head at the ke
by moffkalast 5y ago
Well you most likely need a CUDA compatible GPU with a non-negligible amount of vram to load the model. And a few hours of your time to mash your head at the keyboard in frustration because of the inevitable dependencies that won't want to resolve properly.
Most "open source" models don't actully upload the final models either, just the code they used to train them (not sure if this is the case here), so the next step after that is downloading 100G of data and running your workstation for a few days to train the actual model first.
I recall at one point I attempted to run BERT or GPT-J or one of those lighter language models locally, and it was going well until I realized I needed a 24G of vRAM to even load the model hah.
- hadlock 5y agoDo CUDA AI projects not containerize well? I understand that AI projects often have many, many dependencies, but this seems like something containers ("docker") would be ideally suited for. Something like `docker pull dall-e` and `docker run -ie -n dalle dall-e` probably volume mount in a local folder? I'm curious what the hurdles are here for packaging something like this.
- indiv0 5y agoThey do containerize but it’s not enough. You still need the right drivers and CUDA stuff installed on the docker host, which can be finicky to setup. Not to mention figuring out how to get docker to actually pass the GPU to the container.
- analyte123 5y agoI see this has not gotten any better in the last 5 years. For future reference, if you as a dev are interacting with a product that is this hard to use and there is no other option but to use it (CUDA), you should buy their stock.
- danuker 5y agoI'd much rather look out for competitors. Resting on laurels won't last. If a company stops improving its offer, a competitor might catch up.
- ceeplusplus 5y agoThe competition has even more broken software (AMD). At least Nvidia's works when you need it to.
- moffkalast 5y agoI think OpenCL support has been getting better, and it may be possible to run a lot of models with it, but that just doubles your already frustrating amount of hours trying to set the damn thing up.
- boolemancer 5y ago> Not to mention figuring out how to get docker to actually pass the GPU to the container. Should just be passing `--gpus all` these days, shouldn't it?
- sailingparrot 5y ago> I'm curious what the hurdles are here for packaging something like this. No hurdle, it's just that Docker is not part of the standard toolkit of ML people. ML people all use conda (or virtualenv etc.) which already solves most of the dependency problems, making learning Docker not especially appealing. But virtually all training/inference platform (including the ones used by OpenAI) are using docker. It's not a technical limitation.
- visarga 5y agoNo, ML people use Docker to deploy their models. Nobody got time or patience to replicate environments for inference.
- jaxrtech 5y agoWhy not have some ML-rendition of BitTorrent designed for sharing pre-trained models? At least then you're not paying for AWS egress ha
- jtvjan 5y agoThat's not the issue. You can already download weights of other models through BitTorrent or from large archive sites like Hugging Face or the-eye. OpenAI just doesn't want to release them, so that they can sell access to it through their cloud service.