LLaMA-13B
LLaMA-13B places in the top three on IndoMMLU.
16
Benchmarks with a score
0
Ranked first, out of five or more models
1
In the top three
Meta Feb 2023 LLaMA-13B
Open weights
85.4
#233 of 256
Epoch Capabilities Index
- IndoMMLU 25.3% #3 /5
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HellaSwag | 79.2% | #29 of 63 | Claim in LLaMA: Open and Efficient Foundation Language Models, via Epoch AI | |
| LAMBADA | 75.2% | #12 of 26 | Claim in Qwen Technical Report, via Epoch AI | |
| TriviaQA | 77.9% | #19 of 38 | Claim in Llama 2: Open Foundation and Fine-Tuned Chat Models, via Epoch AI | |
| WinoGrande | 73% | #39 of 76 | Claim in LLaMA: Open and Efficient Foundation Language Models, via Epoch AI | |
| OpenBookQA | 56.4% | #23 of 41 | Claim in Llama 2: Open Foundation and Fine-Tuned Chat Models, via Epoch AI | |
| ARC-Challenge | 52.7% | #44 of 75 | Claim in Qwen Technical Report, via Epoch AI | |
| BoolQ | 78.7% | #36 of 61 | Claim in Qwen Technical Report, via Epoch AI | |
| IndoMMLU | 25.3% | #3 of 5 | Official leaderboard | |
| PIQA | 80.1% | #36 of 58 | Claim in Qwen Technical Report, via Epoch AI | |
| MMLU | 47.7% | #61 of 85 | Claim in Qwen Technical Report, via Epoch AI | |
| BIG-Bench Hard | 37.9% | #35 of 45 | Claim in Qwen Technical Report, via Epoch AI | |
| LMArena (Elo) | 975 | #334 of 335 | Official leaderboard | |
| MT-Bench | 2.61 | #5 of 5 | Official leaderboard |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GSM8K | 15.4% | #12 of 24 | Stanford HELM, via Epoch AI |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| ScienceQA | 43.3% | #2 of 4 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HELM Classic | 59.5 | #10 of 17 | Official leaderboard |
Meta's models over time
Meta's models in color, against every other model on the Epoch Capabilities Index, by release date.
Source: Epoch Capabilities Index, Epoch AI (CC BY 4.0)