Issues
- 8
- 3
- 4
[Question] Interface for generic logger
#114 opened by AwePhD - 0
On using multi-GPU workers with DistributedDataParallel (DDP) and PyTorch Lightning
#135 opened by simon-ging - 0
[Warning] FutureWarning: The behavior of DataFrame concatenation with empty or all-NA entries is deprecated.
#113 opened by eddiebergman - 0
- 1
[Choice] Remove usage of json
#79 opened by eddiebergman - 2
- 0
Road to Declarative Usage
#65 opened by danrgll - 0
- 0
[UX] Replace "cost" with "budget_cost" and "loss" with "optimization_loss" for clarity to users
#125 opened by eddiebergman - 0
[SearchSpace] Allow `None`/ `bool` as a valid value in Categorical/Constant
#124 opened by eddiebergman - 0
- 0
[SearchSpace] Un-intuative behaviour for a search space with a fidelity parameter when ignored, i.e. `RandomSearch`
#122 opened by eddiebergman - 0
[Optimizer] Multi-fidelity algorithms should explicitly check for a fidelity parameter
#120 opened by eddiebergman - 0
- 0
[Optimizers] Make all available optimizers through a "string" be instatiable without kwargs
#118 opened by eddiebergman - 0
[UX] Write to log search space that used and save the pipeline space used as a YAML
#97 opened by eddiebergman - 4
[Bug] `summarize_results_all_tasks_all_devs` only returns results for first seed found.
#80 opened by eddiebergman - 1
[Debugging Feature] Enhance information(e.g. default value, confidence level..) when printing parameter
#59 opened by danrgll - 1
DeepCAVE usage with NePS?
#58 opened by danrgll - 1
[Examples] Enhance examples with standalone clear usage comments for immediate understanding without prior docs familiarization
#57 opened by danrgll - 1
- 0
Custom searcher ignoring max_cost_total
#75 opened by danrgll - 0
- 0
[API] `is_fidelity` with `default` and `default_confidence`, what does it mean?
#89 opened by eddiebergman - 0
- 3
Priorband: Defining default value + default_confidence for fidelity parameter
#101 opened by danrgll - 2
- 0
[Optimizers] Programmatic access to Optimizer's capabilities/requirements
#119 opened by eddiebergman - 0
[Optim] Consider pickle for optimizer state file in run (with option to toggle)
#64 opened by eddiebergman - 0
Enable the use of dict for pipeline_space when the custom optimizer is defined via yaml
#87 opened by danrgll - 3
[Bug] Parallelism and seeding
#108 opened by Neeratyoy - 0
[CI] Ruff/Mypy
#82 opened by eddiebergman - 0
Enable argument modification for custom optimizer class that has been loaded via yaml
#69 opened by danrgll - 4
- 0
Debate on the choice of the name "run_args"
#86 opened by danrgll - 0
Change design for pre_load_hooks
#100 opened by danrgll - 0
Enhance flexibility of 'run_args' to allow mixed usage of direct and YAML-based arguments
#67 opened by danrgll - 4
- 1
- 3
- 1
- 1
[API] A categorical/constant fidelity??
#88 opened by eddiebergman - 1
Generate a pipeline_space.yaml as feedback on which parameter type your configurations have been converted to
#84 opened by danrgll - 0
- 0
- 1
[DOCS] Clarity about enabling parallelization
#85 opened by danrgll - 1
[CI] Migration to ruff
#77 opened by eddiebergman - 0
Concurrent Use of 'neps.run()' Arguments 'post_run_summary' and 'development_stage_id' Causes an Error
#52 opened by danrgll