LLaMA-7B
LLaMA-7B.
Meta Released 24 Feb 2023 13
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Meta Feb 2023 LLaMA-7B
Open weights
96.9
#250 of 271
Epoch Capabilities Index
- LAMBADA 73.3% #16 /26
- ScienceQA 36.2% #23 /23
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 | |
|---|---|---|---|---|
| OpenBookQA | 57.2% | #20 of 41 | Epoch AI | |
| HellaSwag | 76.2% | #42 of 74 | Epoch AI | |
| BoolQ | 76.5% | #45 of 75 | Epoch AI | |
| LAMBADA | 73.3% | #16 of 26 | Epoch AI | |
| WinoGrande | 70.1% | #49 of 76 | Epoch AI | |
| ARC-Challenge | 47.6% | #49 of 75 | Epoch AI | |
| PIQA | 79.8% | #38 of 58 | Epoch AI | |
| TriviaQA | 71% | #29 of 38 | Epoch AI | |
| MMLU | 35.6% | #117 of 127 | Epoch AI | |
| BIG-Bench Hard | 33.5% | #42 of 45 | Epoch AI |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GSM8K | 11% | #68 of 82 | Epoch AI |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| ScienceQA | 36.2% | #23 of 23 | Epoch AI |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HELM Classic | 53.3 | #11 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.