family llama · 5 variants · unknown
| benchmark | category | best score |
|---|---|---|
| MATH | math | 70.5 |
| MT-Bench | chat | 8.98 |
| variant | kind | quant | bits | vRAM | dl | org | year |
|---|---|---|---|---|---|---|---|
| Llama-3.1-Nemotron-70B-Instruct-HF | original | bf16 | 16.0 | 39,363 | nvidia | 2025 | |
| Llama-3.1-70B-Instruct-FP8 | fp8 | fp8 | 8.0 | 7,537 | nvidia | 2025 | |
| Llama-3.1-Nemotron-70B-Reward-HF | original | bf16 | 16.0 | 217 | nvidia | 2025 | |
| Llama-3.1-70B-Instruct-RM-RB2 | original | bf16 | 16.0 | 99 | allenai | 2025 | |
| Llama-3.1-Nemotron-70B-Instruct | original | bf16 | 16.0 | 70 | nvidia | 2025 |