Bruno Cabral
Bruno Cabral
@usptact , could you please provide an example of sequence tagging in Vowpal ? What command line and input format ?
+1 that is really frustrating that something such as the used config can´t be shared if you are not a customer. They should advertise llm-foundry as open-core and not open-source.
It is possible to test it here: https://huggingface.co/spaces/mosaicml/mpt-30b-chat The results are impressive!
@rohan-varma just a +1 for this idea. Please help us GPU poor !! A 70B full finetuning with paged/low-precision optimizer, FSDP wrapping, activation checkpoint, full bf16 + an option for...
Great, I will try over the weekend and will report the results. What the range of the input ? Seems like is from -7.0 to 7.0 in the Suture example...
@winglian please don´t let it die
Super excited about this!
@robertgshaw2-neuralmagic any luck with this patch? I benchmarked and those kernels are really something. Great boost on my internal tests!
@andrewkho This is my config ( I used this torchdata dataloader but there is no difference for me in the main branch. It is using only 1 cpu core) ```...
@andrewkho thank you so much. I´m using Python 3.11 in Ubuntu 22.04 Unfortunately I cannot share the jsonl file, as it has sensitive information. But is like that with 4...