led-large-book-summary
by pszemraj
460M params · summarization · 120 likes · 3.0k downloads
led-large-book-summary is a 460M parameter model. At Q4 quantization it requires 0GB of VRAM. It runs comfortably on GeForce RTX 4090 (2855 tok/s), GeForce RTX 5090 (4281 tok/s), M4 Max 128GB (1044 tok/s).
Inference providers
| Provider | $/1M in | $/1M out | Throughput |
|---|---|---|---|
| HF Inference |
GPU compatibility
| GPU | VRAM | Q4 Decode | Verdict |
|---|---|---|---|
| GeForce RTX 4090 | 24GB | 2855 tok/s | comfortable |
| GeForce RTX 5090 | 32GB | 4281 tok/s | comfortable |
| M4 Max 128GB | 128GB | 1044 tok/s | comfortable |
| M4 Pro 48GB | 48GB | 522 tok/s | comfortable |
| M4 Pro 24GB | 24GB | 522 tok/s | comfortable |
| A100 PCIe 80 GB | 80GB | 5231 tok/s | comfortable |
| H100 SXM5 80 GB | 80GB | 10521 tok/s | comfortable |
| GeForce RTX 3090 | 24GB | 2524 tok/s | comfortable |
| Radeon RX 7900 XTX | 24GB | 2087 tok/s | comfortable |
| GeForce RTX 4080 | 16GB | 2026 tok/s | comfortable |