sr-latch

7 ms·

31.
by sr-latch 3y ago
I find Docker running a full Linux userspace a little bloated. Thankfully there are distroless base images( https://github.com/GoogleContainerTools/distroless ). Haven't done service dev in a while, so I don't
32.
by sr-latch 3y ago
Apple's opinions seem to be agreeable...
33.
by sr-latch 3y ago
My hunch is that the next progression in AI will incorporate forward forward training
34.
by sr-latch 4y ago
I have a universal benchmark for judging how much knowledge a language model stores, and it's asking about the G-FOLD paper ( https://www.lpi.usra.edu/meetings/marsconcepts2012/pdf/4193.... ), because I no
35.
by sr-latch 4y ago
Are you the GGML dev?
36.
by sr-latch 4y ago
Not a single page, but almost all large language models with open weights are published on this website: https://huggingface.co/models
37.
by sr-latch 4y ago
Would be cool to try to incorporate the previous token's confidence embedding into this process, but that would make training with a triangular attention mask not possible.
38.
by sr-latch 4y ago
This looks similar to the WebGPT paper, is that referenced in any of langchain or haystack's publications? Introducing the mechanism of internal thought is very interesting, I wonder if there's a way to make it implicit in the mod
39.
by sr-latch 4y ago
Have you tried running it against a quantized model on HuggingFace with identical inputs and deterministic sampling to check if the outputs you're getting are identical? I think that should confirm/eliminate any concern of the mod