mirror of
https://github.com/microsoft/SkillOpt.git
synced 2026-07-03 14:02:58 +08:00
244e346b83
- Skill optimization framework with training loop analogy - 11 benchmarks, 4 model backends (Azure OpenAI, Claude, Codex, Qwen) - WebUI for browser-based training control - Pluggable architecture for extending benchmarks and backends
9 lines
283 B
Python
9 lines
283 B
Python
"""ReflACT Scheduler -- edit budget and learning rate scheduling.
|
|
|
|
Analogous to learning rate schedulers (cosine annealing, step decay, warmup)
|
|
in neural network training. Controls how the edit_budget evolves over the
|
|
course of training.
|
|
|
|
Placeholder for future implementations.
|
|
"""
|