Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -24,7 +24,7 @@ about two hours.
|
|
| 24 |
| eval set | exact-solution accuracy |
|
| 25 |
|---|---|
|
| 26 |
| standard test (200 puzzles, same difficulty as training) | 100% |
|
| 27 |
-
| hard set
|
| 28 |
|
| 29 |
Per-stratum on the hard set: raw 98.7%, hard24 99.3%, anneal150 100%,
|
| 30 |
anneal400 94.7%, anneal400x2 96.0%. All 17 failures occur on puzzles where the
|
|
|
|
| 24 |
| eval set | exact-solution accuracy |
|
| 25 |
|---|---|
|
| 26 |
| standard test (200 puzzles, same difficulty as training) | 100% |
|
| 27 |
+
| hard set (750 stratified puzzles, stratum medians up to ~40× the training median difficulty) | **97.7%** (733/750) |
|
| 28 |
|
| 29 |
Per-stratum on the hard set: raw 98.7%, hard24 99.3%, anneal150 100%,
|
| 30 |
anneal400 94.7%, anneal400x2 96.0%. All 17 failures occur on puzzles where the
|