Are you planning to add support for Canadian banks anytime soon? I checked out Monarch and ended up going with Mint just because they were one of the only ones supporting Canadian banks
Literally just made his point. Every single person keeps repeating it’s like email, as if email is a fantastic place to go get your news and interact with others.
But that's not email's use case at all. It's not being compared to email because they serve the same function otherwise there'd be no need for Lemmy. It's being compared as a distributed and open protocol, and where the comparison is important is in adoption, and email proves out that an open distributed system can achieve not only strong, but universal adoption. We pulled it off in the 80s on ridiculously weak hardware and snail pace internet, I'm sure it can be done again.
It's not like e-mail. It already fixes the things you're complaining about, while the things OP was complaining about will be fixed eventually. The point of bringing up e-mail is that a) federation does work and b) that usability is not a hindrance by itself.
First part is correct, the second part is not. GPT Neo is a 2.7B param model, the largest GPT is 175B (they have various flavours, up to 175B). I appreciate the sentiment and what ElutherAI is doing with GPT Neo, but there is no open source equivlenet of the full GPT-3 available for the public to use. Hopefuly it's just a matter of time.
GPT-J is 6B and comes pretty close. Also practically I haven’t noticed a difference.
Keep in mind there are also closed source alternatives: for example, AI21’s Jurassic-1 models are comparable, cheaper, and technically larger (albeit somewhat comically, 178B instead of 175B parameters).
This is a good question. I want to write a more detailed post about this in the future, but here are a few points for now:
- Neuropod is an abstraction layer so it can do useful things on top of just running models locally. For example, we can transparently proxy model execution to remote machines. This can be super useful for running large scale jobs with compute intensive models. Including GPUs in all our cluster machines doesn’t make sense from a resource efficiency perspective so instead, if we proxy model execution to a smaller cluster of GPU-enabled servers, we can get higher GPU utilization while using fewer GPUs. The "Model serving" section of the blog post ([1]) goes into more detail on this. We can also do interesting things with model isolation (see the "Out-of-process execution" section of the post).
- ONNX converts models while Neuropod wraps them. We use TensorFlow, TorchScript, etc. under the hood to run a model. This is important because we have several models that use custom ops, TensorRT, etc. We can use the same custom ops that we use at training time during inference. One of the goals of Neuropod is to make experimentation, deployment, and iteration easier so not having to do additional "conversion" work is useful.
- When we started building Neuropod, ONNX could only do trace-based conversions of PyTorch models. We've generally had lots of trouble with correctness of trace-based conversions for non-trivial models (even with TorchScript). Removing intermediate conversion steps (and their corresponding verification steps) can save a lot of time and make the experimentation process more efficient.
- Being able to define a "problem" interface was important to us (e.g. "this is the interface of a model that does 2d object detection"). This lets us have multiple implementations that we can easily swap out because we concretely defined an interface. This capability is useful for comparing models across frameworks without doing a lot of work. The blog post ([1]) talks about this in more detail.
The blog post ([1]) goes into a lot more detail about our motivations and use cases so it's worth a read.
