| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| GSM8k | FLARE-2B | Throughput (tok/s)2,087 | 27 | 23h ago | |
| HumanEval | FLARE-2B | Throughput (tok/s)1,763.9 | 27 | 23h ago | |
| Llama 2 7B inference v1.0 | QTIP | Decoding Throughput (TOK/s)188 | 6 | 3mo ago | |
| Llama 3.1 8B | LAQuant | Throughput (TOK/s)225.01 | 5 | 22d ago | |
| Qwen3-8B | LAQuant | Throughput (tokens/sec)196.8 | 5 | 22d ago | |
| Qwen3-4B | LAQuant | Throughput (tokens/sec)231.3 | 5 | 22d ago | |
| Qwen3 1.7B | LAQuant | Decoding Throughput (tokens/sec)316.9 | 5 | 22d ago | |
| Llama 2 70B v1.0 (inference) | QTIP | Throughput (TOK/s)23.5 | 5 | 3mo ago | |
| Alpaca | AR Throughput (tok/s)88.8 | 2 | 3mo ago | ||
| Ultrafeedback | AR Throughput (tok/s)88.8 | 2 | 3mo ago |
