tanhevg
tanhevg
## Expected Behavior I am expecting hhblits to produce the same alignments when executed on the same input multiple times ## Current Behavior When default settings are used, alignments are...
When trying to build an HHBlits database from linclust output of 250mio clusters over 0.5bn sequences, `cstranslate_mpi` runs for a few days and does not produce an output ## Your...
The vast majority of labels from TIMIT corpus - all except 49, i.e. ~98.5% - have all 8 samples parsed as having the true bit in the first position, i.e....
Hi, thanks for the excellent example of using DistributedDataParallel in PyTorch; it is very easy to understand and is much better that Pytorch docs. One important bit that is missing...
Currently, the RNN cells are initialised as `param`'s. (e.g. [here](https://github.com/FluxML/Flux.jl/blob/3d5b76c0dfe9d9e885aa5ea62b312833a8d0759a/src/layers/recurrent.jl#L86) and [here](https://github.com/FluxML/Flux.jl/blob/3d5b76c0dfe9d9e885aa5ea62b312833a8d0759a/src/layers/recurrent.jl#L125-L126)). This causes the initial state to be modified during the backprop, which can in turn affect the model...
Hello, Thanks for a very useful and well documented package. Great effort! I have tried to train the model with flash attention enabled, and ran into the following error: ```...
Consider the following use case: * Reference data and output of simulator function are square matrices * We want to use trace norm of the difference between the matrices as...
Hello. Thank you for the great package! I am trying to run `diamond blastp` for detecting exact protein sequence matches. I am using a custom matrix that has 1's on...