-
Notifications
You must be signed in to change notification settings - Fork 1
Pull requests: davidbenjamin/permutect
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
New learned aggregation for sets, permutation invariant but much more expressive than the mean
#172
by davidbenjamin
was merged Jan 17, 2025
RaggedSet class to encapsulate operations on batches of different-size sets; no change in output
#171
by davidbenjamin
was merged Jan 17, 2025
New label-balancing weight scheme; option to use specific samples for calibration
#170
by davidbenjamin
was merged Jan 14, 2025
fixed remnant of constant alt count in BaseBatch that affected metrics
#167
by davidbenjamin
was merged Dec 5, 2024
Loading…
Update requirements for compatibility with GATK, particularly move to pytorch 2.1
#166
by davidbenjamin
was merged Dec 4, 2024
Loading…
better balance of false positives, filtered false negatives, true positives in filtering tensorboard projector
#164
by davidbenjamin
was merged Dec 1, 2024
Loading…
WDL always uses GPU and SSD with google cloud backend; rename self-attention arguments
#163
by davidbenjamin
was merged Nov 29, 2024
Loading…
Posterior model updates priors and somatic spectrum with EM step between epochs
#162
by davidbenjamin
was merged Nov 29, 2024
Loading…
save base data as big tensors, not as Python lists
#161
by davidbenjamin
was merged Nov 25, 2024
Loading…
quick bug fix for when het_beta is None in filter_variants
#159
by davidbenjamin
was merged Nov 21, 2024
Loading…
Option to broaden germline allele fraction range for linseq evaluation
#157
by davidbenjamin
was merged Nov 14, 2024
Loading…
GPU support and optimizations for posterior model
#156
by davidbenjamin
was merged Nov 7, 2024
Loading…
Artifact Model and Posterior Model now use GPU, although not very well yet
#154
by davidbenjamin
was merged Nov 1, 2024
Loading…
Adversarial base model training forcing learned features to forget alt count
#150
by davidbenjamin
was merged Oct 14, 2024
Loading…
Base model and artifact model are saved together after artifact model training
#149
by davidbenjamin
was merged Oct 10, 2024
Loading…
Improvements and refactoring of loss weighting
#148
by davidbenjamin
was merged Oct 9, 2024
Loading…
Previous Next
ProTip!
Find all pull requests that aren't related to any open issues with -linked:issue.