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 |
google
|
|---|---|
| Company | |
| Release Date | 2024-07-31 |
| Parameters (B) | 2 |
| Architecture | Transformer |
| Context Window | 8192 |
| Input Modalities | text |
| Open Source | Yes |
| License | gemma |
| Score | 0 |
| VRAM (GB) | 5 |
| Compute | Runs on a single consumer GPU with ~5 GB VRAM in BF16 (about 3 GB with 4-bit quantization); trained on 512 TPUv5e chips. |
| Benchmark: AGIEval | 30.6 |
| Benchmark: ARC | 55.4 |
| Benchmark: BBH | 41.9 |
| Benchmark: DROP | 52 |
| Benchmark: GSM8K | 23.9 |
| Benchmark: HellaSwag | 73 |
| Benchmark: HumanEval | 17.7 |
| Benchmark: MATH | 15 |
| Benchmark: MBPP | 29.6 |
| Benchmark: MMLU | 51.3 |
| Pricing (Input) | — |
| Pricing (Output) | — |