2080 Ti
8 records across 1 models (quants: EXL2 4.0bpw).
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2080 Ti | Llama 3 8B (body says llama3; linked gist script loads mistral-7b-instruct-exl2) | EXL2 4.0bpw | ExLlamaV2 | 195 | – | – | – | – | 2024-05 | community | – | source |
| 2080 Ti | Llama 3 8B (body says llama3; linked gist script loads mistral-7b-instruct-exl2) | EXL2 4.0bpw | ExLlamaV2 | 177 | – | – | – | – | 2024-05 | community | – | source |
| 2080 Ti | Llama 3 8B (body says llama3; linked gist script loads mistral-7b-instruct-exl2) | EXL2 4.0bpw | ExLlamaV2 | 121 | – | – | – | – | 2024-05 | community | – | source |
| 2080 Ti | Llama 3 8B (body says llama3; linked gist script loads mistral-7b-instruct-exl2) | EXL2 4.0bpw | ExLlamaV2 | 119 | – | – | – | – | 2024-05 | community | – | source |
| 2080 Ti | Llama 3 8B (body says llama3; linked gist script loads mistral-7b-instruct-exl2) | EXL2 4.0bpw | ExLlamaV2 | 113 | – | – | – | – | 2024-05 | community | – | source |
| 2080 Ti | Llama 3 8B (body says llama3; linked gist script loads mistral-7b-instruct-exl2) | EXL2 4.0bpw | ExLlamaV2 | 101 | – | – | – | – | 2024-05 | community | – | source |
| 2080 Ti | Llama 3 8B (body says llama3; linked gist script loads mistral-7b-instruct-exl2) | EXL2 4.0bpw | ExLlamaV2 | 82 | – | – | – | – | 2024-05 | community | – | source |
| 2080 Ti | Llama 3 8B (body says llama3; linked gist script loads mistral-7b-instruct-exl2) | EXL2 4.0bpw | ExLlamaV2 | 80 | – | – | – | – | 2024-05 | community | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085