Compare LLM Models
Select 2-5 models to compare side by side.
Select models to compare
0 / 5 selected
Top downloads
Select 2-5 models
| Attribute |
meta-llama
|
|---|---|
| Company | meta-llama |
| Release Date | 2024-07-23 |
| Parameters (B) | 405 |
| Architecture | Transformer |
| Context Window | 128000 |
| Input Modalities | text |
| Open Source | Yes |
| License | llama3.1 |
| Score | 8 |
| VRAM (GB) | 810 |
| Compute | Multi-node GPU cluster with 8+ NVIDIA H100 80GB GPUs (FP8) or 16+ H100 GPUs (FP16); ~810 GB VRAM for full-precision weights plus additional memory for KV cache at long context |
| Benchmark: BBH | 89.2 |
| Benchmark: HumanEval | 89 |
| Benchmark: MATH | 73.9 |
| Benchmark: MBPP | 88.6 |
| Benchmark: MMLU | 88.6 |
| Benchmark: MMLU-Pro | 73.3 |
| Benchmark: SWE-Bench-Pro | 11.2 |
| Benchmark: TruthfulQA | 0 |
| Pricing (Input) | — |
| Pricing (Output) | — |