Datasets:
correct typo
Browse files
README.md
CHANGED
|
@@ -55,7 +55,7 @@ configs:
|
|
| 55 |
`LOOPerSet` was originally created to train the cost model for the [LOOPer autoscheduler](https://ieeexplore.ieee.org/document/11282943) (PACT '25). For a full description of the generation process and a diversity analysis, please see our [companion paper on arXiv](https://arxiv.org/abs/2510.10209).
|
| 56 |
|
| 57 |
### What is inside?
|
| 58 |
-
Each data point represents a **(Program, Schedule)
|
| 59 |
* **Source Code & IR:** Raw Tiramisu (C++) generator code, lowered Halide IR, and ISL ASTs.
|
| 60 |
* **Structured Features:** JSON-based representation of the program structure (loop hierarchy, memory access patterns, arithmetic expressions) for feature engineering.
|
| 61 |
* **Optimization Schedules:** Sequences of code transformations (tiling, skewing, fusion, interchange, unrolling, parallelization, etc) and the specific API commands used to apply them.
|
|
|
|
| 55 |
`LOOPerSet` was originally created to train the cost model for the [LOOPer autoscheduler](https://ieeexplore.ieee.org/document/11282943) (PACT '25). For a full description of the generation process and a diversity analysis, please see our [companion paper on arXiv](https://arxiv.org/abs/2510.10209).
|
| 56 |
|
| 57 |
### What is inside?
|
| 58 |
+
Each data point represents a **(Program, Schedule) → Performance** tuple containing:
|
| 59 |
* **Source Code & IR:** Raw Tiramisu (C++) generator code, lowered Halide IR, and ISL ASTs.
|
| 60 |
* **Structured Features:** JSON-based representation of the program structure (loop hierarchy, memory access patterns, arithmetic expressions) for feature engineering.
|
| 61 |
* **Optimization Schedules:** Sequences of code transformations (tiling, skewing, fusion, interchange, unrolling, parallelization, etc) and the specific API commands used to apply them.
|