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 |
ibm-granite
|
|---|---|
| Company | ibm-granite |
| Release Date | 2024-10-21 |
| Parameters (B) | 1.3 |
| Architecture | MoE Transformer |
| Context Window | 4096 |
| Input Modalities | text |
| Open Source | Yes |
| License | apache-2.0 |
| Score | 0 |
| VRAM (GB) | 4 |
| Compute | Runs in BF16 on GPUs with ~4GB VRAM; 400M active parameters per token make it suitable for CPU servers, edge devices, and low-latency on-device inference |
| Benchmark: AGIEval | 20.4 |
| Benchmark: ARC | 41.2 |
| Benchmark: BBH | 31.8 |
| Benchmark: GPQA | 23.1 |
| Benchmark: GSM8K | 26.3 |
| Benchmark: HellaSwag | 65.6 |
| Benchmark: HumanEval | 30.2 |
| Benchmark: IFEval | 32.4 |
| Benchmark: MATH | 10.9 |
| Benchmark: MBPP | 15.4 |
| Benchmark: MMLU | 32 |
| Benchmark: MMLU-Pro | 12.2 |
| Benchmark: TruthfulQA | 49.1 |
| Pricing (Input) | — |
| Pricing (Output) | — |