Llama-2-7b-hf
by meta-llama
6.7B params · text-generation · 2.3k likes · 686.3k downloads
Llama-2-7b-hf is a 6.7B parameter model. At Q4 quantization it requires 3GB of VRAM. It runs comfortably on GeForce RTX 4090 (194 tok/s), GeForce RTX 5090 (292 tok/s), M4 Max 128GB (71 tok/s).
GPU compatibility
| GPU | VRAM | Q4 Decode | Verdict |
|---|---|---|---|
| GeForce RTX 4090 | 24GB | 194 tok/s | comfortable |
| GeForce RTX 5090 | 32GB | 292 tok/s | comfortable |
| M4 Max 128GB | 128GB | 71 tok/s | comfortable |
| M4 Pro 48GB | 48GB | 35 tok/s | comfortable |
| M4 Pro 24GB | 24GB | 35 tok/s | comfortable |
| A100 PCIe 80 GB | 80GB | 356 tok/s | comfortable |
| H100 SXM5 80 GB | 80GB | 718 tok/s | comfortable |
| GeForce RTX 3090 | 24GB | 172 tok/s | comfortable |
| Radeon RX 7900 XTX | 24GB | 142 tok/s | comfortable |
| GeForce RTX 4080 | 16GB | 138 tok/s | comfortable |