Update README.md
#2
by JUNLLM - opened
README.md
CHANGED
|
@@ -1,3 +1,45 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
---
|
| 4 |
+
|
| 5 |
+
# CircuitNet 3.0 Dataset
|
| 6 |
+
|
| 7 |
+
This is the public dataset release for [CircuitNet 3.0: A Multi-Modal Dataset with Task-Oriented Augmentation for AI-Driven Circuit Design](https://openreview.net/forum?id=lEDb4gQ4dB).
|
| 8 |
+
|
| 9 |
+
The code and documentation companion lives in [sklp-eda-lab/iclr-circuitnet_3.0](https://github.com/sklp-eda-lab/iclr-circuitnet_3.0).
|
| 10 |
+
|
| 11 |
+
CircuitNet 3.0 is a multi-stage, multi-modal dataset for AI-driven circuit design. It is intended to support early-stage timing and power prediction, along with cross-stage representation learning over RTL, netlist, and layout-oriented artifacts.
|
| 12 |
+
|
| 13 |
+
## Current release
|
| 14 |
+
|
| 15 |
+
At the time of writing, the public archive is packaged as `circuitNetv3.zip`. The archive currently exposes:
|
| 16 |
+
|
| 17 |
+
- `dataset/RTL/` source RTL modules
|
| 18 |
+
- `dataset/Netlist/` synthesized netlist artifacts
|
| 19 |
+
- `dataset/Final/*/feature.json` gate-level feature metadata
|
| 20 |
+
- `dataset/Final/*/final_netlist.v` final synthesized netlists
|
| 21 |
+
- `dataset/Final/*/power_summary.txt` power summaries
|
| 22 |
+
|
| 23 |
+
Layout artifacts are large and are being uploaded incrementally. If a layout subset you need is not yet available, or if you would like to request a specific large layout package, please contact [wangmingjun000613@gmail.com](mailto:wangmingjun000613@gmail.com).
|
| 24 |
+
|
| 25 |
+
## Dataset summary
|
| 26 |
+
|
| 27 |
+
- 8,659 validated open-source RTL designs
|
| 28 |
+
- 15,863 total design instances after task-oriented augmentation
|
| 29 |
+
- Source-level train/validation/test partitioning, with augmented variants of test sources excluded from train/val to avoid leakage
|
| 30 |
+
- Timing labels: AT, WNS, and TNS
|
| 31 |
+
- Power labels: total power summaries
|
| 32 |
+
|
| 33 |
+
## What is intentionally excluded
|
| 34 |
+
|
| 35 |
+
- Commercial PDK files, standard-cell libraries, and other restricted technology collateral
|
| 36 |
+
- Proprietary EDA binaries or license-dependent artifacts
|
| 37 |
+
- Redundant examples that do not improve downstream use or reproducibility
|
| 38 |
+
|
| 39 |
+
## Update policy
|
| 40 |
+
|
| 41 |
+
This dataset card will be updated as new artifacts are uploaded. The Hugging Face hub is the source of truth for the latest available files.
|
| 42 |
+
|
| 43 |
+
## Citation
|
| 44 |
+
|
| 45 |
+
If you use this dataset, please cite the paper above.
|