Learning Rate Schedules
Linear Warmup With Linear Decay
Linear Warmup With Linear Decay is a learning rate schedule in which we increase the learning rate linearly for $n$ updates and then linearly decay afterwards.
Papers
| Paper | Code | Results | Date | Stars |
|---|
Tasks
| Task | Papers | Share |
|---|---|---|
| Retrieval | 152 | 15.82% |
| Language Modelling | 88 | 9.16% |
| Question Answering | 69 | 7.18% |
| Large Language Model | 42 | 4.37% |
| Sentence | 29 | 3.02% |
| Information Retrieval | 26 | 2.71% |
| Text Classification | 25 | 2.60% |
| Text Generation | 24 | 2.50% |
| Sentiment Analysis | 20 | 2.08% |
Usage Over Time
Components
| Component | Type |
|
|---|---|---|
| 🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |
