hearbenchmark hear-eval-kit issues

hearbenchmark / hear-eval-kit

Evaluation kit for the HEAR Benchmark

https://hearbenchmark.com

Apache License 2.0

56 stars 17 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

change json schema of the end score in the task_prediction log_metric function in pl model

#336 khumairraj opened 3 years ago
0
Save Predictions + Add prediction logging

#335 khumairraj closed 3 years ago
0
Allow metrics to return tuple of scores

#334 khumairraj closed 3 years ago
0
f+p+r from sed eval should be returned as a tuple and consumed in training, with the first entry as primary

#333 khumairraj opened 3 years ago
0
dcase remove generic grid from prediction

#332 khumairraj opened 3 years ago
0
Add event based metrics

#331 khumairraj closed 3 years ago
0
Store test output

#330 turian opened 3 years ago
0
GPU device name output

#329 turian closed 3 years ago
0
Event labels as strigns

#328 jorshi closed 3 years ago
0
Prediction output

#327 jorshi closed 2 years ago
0
[WIP] Leaderboard tweaks

#326 turian closed 3 years ago
0
Variable Audio Embedding Generation | ADD LWLRAP, mAP, D_prime

#325 khumairraj closed 3 years ago
0
Add nsynth kfold CI

#324 khumairraj closed 3 years ago
0
Spoken digit integration test

#323 jorshi closed 3 years ago
0
Update version numbers is README

#322 turian opened 3 years ago
0
CI should test on a task with kfold

#321 turian opened 3 years ago
0
Break task_prediction function to smaller functions

#320 khumairraj opened 3 years ago
0
Refactor to make the `data_splits` variable in task predictions.py more clearer and less ambiguous

#319 khumairraj opened 3 years ago
0
'=0.14.2' in Docker / ?

#318 turian opened 3 years ago
0
Memory profiling mode, on random audio?

#317 turian opened 3 years ago
0
Pinning memory can cause errors for smaller GPUs

#316 jorshi opened 3 years ago
1
K-fold v2

#315 jorshi closed 3 years ago
0
Minimum pytorch lightning version

#314 jorshi opened 3 years ago
2
Embeddings runner should skip hidden dirs

#313 jorshi opened 3 years ago
0
[Snyk] Fix for 2 vulnerabilities

#312 snyk-bot closed 3 years ago
0
Train prediction model end2end

#311 faroit closed 3 years ago
5
[WIP] Add Training for Data with Folds; Do full training on test and valid before final testing

#310 khumairraj closed 3 years ago
3
[WIP] Add Leave One Out Cross Validation and Grid Search on Dev Set; Train on Full

#309 khumairraj closed 3 years ago
0
Delete TF model in GPU memory (predictions.runner)

#308 seungheondoh closed 3 years ago
3
v2021.0.5

#307 turian closed 3 years ago
0
V2021.0.3

#306 turian closed 3 years ago
0
Sanity check we have all the embeddings we need (ESP with no labels for multilabel)

#305 turian opened 3 years ago
0
Device count

#304 jorshi closed 3 years ago
0
Better multi-GPU support

#303 turian closed 3 years ago
0
Validate that models run on the GPU they are supposed to

#302 turian closed 3 years ago
0
Tighter postprocessing grid

#301 turian closed 3 years ago
0
Explain metadata for events in prediction dataloader

#300 turian opened 3 years ago
0
CI/CD from github

#299 turian closed 3 years ago
0
Prediction output JSON should be unified and cleaned up

#298 turian opened 3 years ago
0
Log validation score

#297 turian closed 3 years ago
0
Put small datasets in github not GS

#296 turian closed 3 years ago
0
Downstream CI

#295 turian closed 3 years ago
0
Don't include full metadata in instances

#294 turian opened 3 years ago
0
Data loader optimizations

#293 turian closed 3 years ago
0
Use hash for partitioning for larger sets?

#292 turian closed 3 years ago
0
Add download URLs with -LATEST tar

#291 turian closed 3 years ago
0
Add hear version to tasks/ canonical filename?

#290 turian closed 3 years ago
0
Check that changing the base path doesn't affect the sorting

#289 turian closed 3 years ago
0
Rename dcase2016 to office_events in sampler, etc.

#288 turian closed 3 years ago
0
Repackaging

#287 turian closed 3 years ago
0

Previous Next