VAMB Benchmark Results

Output of VAMB on the benchmark datasets: CAMI2: Airways, GI, Oral, Skin, Urog and the MetaHIT dataset. Each folder contains: - clusters.tsv: Clustername and contigname- tnf.npz: Tetranucleotide Frequency information for each contig in Numpy npz format- latent.npz: Latent representation of each contig in Numpy npz format- mask.npz: Contig mask, ie. if contigs should not be used in encoding in Numpy npz format- model.pt: Final VAE model in PyTorch- log.txt: Logfile of the run containing information on settings, loss and clustering- lengths.npz: Length of each contig in Numpy npz format