SHRDFU-7B-overbaked-lora [Adapter]: benchmark results
maldv SHRDFU 7B Overbaked LoRA is an adapter release for cgato/Thespis-CurtainCall-7b-v0.3 trained using crabcanon. Its adapter identity remains separate from the full SHRDFU checkpoints. Provider: maldv. Access: Open.
Aggregate Rating 1406 ± 33, rank #5790 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - ARC Challenge | 64.33 | Normalized accuracy (%) (25-shot) | 66.2 |
| Open LLM Leaderboard v1 - WinoGrande | 78.69 | Accuracy (%) (5-shot) | 64.9 |
| Open LLM Leaderboard v1 - HellaSwag | 83.46 | Normalized accuracy (%) (10-shot) | 58.8 |
| Open LLM Leaderboard v1 - MMLU | 62.42 | Accuracy (%) (5-shot) | 55.2 |
| Open LLM Leaderboard v1 - GSM8K | 38.44 | Accuracy (%) (5-shot) | 53.4 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 50.98 | MC2 (%) (0-shot) | 50.2 |
Interactive version: theaggregate.ai/model?slug=shrdfu-7b-overbaked-lora-adapter · How It Works · Data refreshed daily, snapshot 2026-11-08.