Alpaca
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Alpaca instruction-following 52K
61.67TPR
5
Alpaca 15% retention (train)
134.25Total Sum
5
Alpaca 7B (test)
1.292HV Score
5
Alpaca
2.98Speedup
5
Alpaca 800 samples
62.1BLEU
5
Alpaca R1
0.61Kendall's Tau (τb)
4
Alpaca Llama
0.75Kendall's Tau
4
Alpaca
98Monitor Accuracy
4
Alpaca (test)
84.8MaJ
4
Alpaca
2.55Acceptance Length (SGLang)
4
Alpaca
102.44Latency (ms/tok)
4
Alpaca
0KL Divergence (Specificity)
4
Alpaca-52K (train)
65.88Average Score
4
Alpaca 52K instruction LLaMA-7B (test)
32.9BBH
4
Alpaca-Data (10% sub-sampled)
59.13Win Rate (vs. Clust Rand)
4
Alpaca GPT-4
0.96Kendall's Tau
3
Alpaca Llama 3.3 70B
74.75Memory Usage (GB/GPU)
3
Alpaca
0.665Spearman Rho (AdamW Influence)
3
Alpaca-65B Weak-to-strong Safety Alignment (test)
131.81HV Score
3
Alpaca (train)
1VSR
3
Alpaca
88.8AR Throughput (tok/s)
2
Alpaca avg prompt instance
15.225Inference Time (s)
2
Alpaca-style instruction-tuning 5K samples (held-out)
1.342Cross-Entropy Loss (Std)
1