Semibin

* logging: Use 'SemiBin2' as logger name
* SemiBin: Always log to file in DEBUG level and log command-line arguments
* coverage: Better error messages (168)
* SemiBin: Print version number in logs
* SemiBin: Better method to save the model which is more compatible with
newer versions of PyTorch
* update_model: Add subcommand to convert old models to new format
* SemiBin: Respect the number of threads requested better (140)
* SemiBin: Switch to pixi
* check_install: Prints out information on the GPU
* check_install: No error if MMseqs2 is not installed
* SemiBin: Remove `SemiBin` command. Only `SemiBin1` and `SemiBin2` are
available (and `SemiBin1` is deprecated)
* SemiBin: Deprecate `--prodigal-output-faa` argument
* SemiBin: Do not fail if no bins are produced (170 & 173)

* SemiBin: Support running SemiBin with strobealign-aemb
(--abundance/-a)
* citation: Add citation subcommand
* SemiBin1: Introduce separate SemiBin1 command
* internal: Code simplification and refactor
* deprecation: Deprecate --orf-finder=fraggenescan option
* Update abundance normalization
* SemiBin: do not use more processes than can be taken advantage of (155)

* multi_easy_bin: Fix multi_easy_bin with --write-pre-recluster (128)

* train_self: Fix bug with --mode
* concatenate_fasta: Fix bug with compression
* bin_short: Make alias work

* SemiBin: Better error checking throughout
* SemiBin: Write a log file
* concatenate_fasta: support compression
* concatenate_fasta: slightly better error message when contig ID already
contains separator
* SemiBin: add `bin_short` as alias for `bin`

* Bugfix: using --no-recluster with multi_easy_bin (128)