OpenCompass/configs/datasets/nq/README.md

70 lines
2.4 KiB
Markdown
Raw Permalink Normal View History

2024-05-28 23:09:59 +08:00
# NQ
```bash
python3 run.py --models hf_internlm2_7b --datasets nq_open_1shot_gen_20a989 --debug
python3 run.py --models hf_internlm2_chat_7b --datasets nq_open_1shot_gen_01cf41 --debug
```
2024-05-28 23:09:59 +08:00
## Base Models
| model | nq |
|:------------------------:|------:|
| llama-7b-turbomind | 15.10 |
| llama-13b-turbomind | 16.43 |
| llama-30b-turbomind | 22.11 |
| llama-65b-turbomind | 26.09 |
| llama-2-7b-turbomind | 14.35 |
| llama-2-13b-turbomind | 21.69 |
| llama-2-70b-turbomind | 23.27 |
| llama-3-8b-turbomind | 18.78 |
| llama-3-70b-turbomind | 31.88 |
| internlm2-1.8b-turbomind | 20.66 |
| internlm2-7b-turbomind | 41.05 |
| internlm2-20b-turbomind | 43.55 |
| qwen-1.8b-turbomind | 5.68 |
| qwen-7b-turbomind | 17.87 |
| qwen-14b-turbomind | 13.77 |
| qwen-72b-turbomind | 18.20 |
| qwen1.5-0.5b-hf | 6.01 |
| qwen1.5-1.8b-hf | 10.28 |
| qwen1.5-4b-hf | 15.73 |
| qwen1.5-7b-hf | 18.61 |
| qwen1.5-14b-hf | 16.07 |
| qwen1.5-32b-hf | 21.75 |
| qwen1.5-72b-hf | 20.53 |
| qwen1.5-moe-a2-7b-hf | 16.62 |
| mistral-7b-v0.1-hf | 20.66 |
| mistral-7b-v0.2-hf | 20.78 |
| mixtral-8x7b-v0.1-hf | 24.85 |
| mixtral-8x22b-v0.1-hf | 34.43 |
| yi-6b-hf | 10.08 |
| yi-34b-hf | 13.96 |
| deepseek-7b-base-hf | 8.45 |
| deepseek-67b-base-hf | 17.59 |
## Chat Models
| model | nq |
|:-----------------------------:|------:|
| qwen1.5-0.5b-chat-hf | 7.42 |
| qwen1.5-1.8b-chat-hf | 10.22 |
| qwen1.5-4b-chat-hf | 19.31 |
| qwen1.5-7b-chat-hf | 16.87 |
| qwen1.5-14b-chat-hf | 20.53 |
| qwen1.5-32b-chat-hf | 25.26 |
| qwen1.5-72b-chat-hf | 35.21 |
| qwen1.5-110b-chat-hf | 36.98 |
| internlm2-chat-1.8b-hf | 19.09 |
| internlm2-chat-1.8b-sft-hf | 18.14 |
| internlm2-chat-7b-hf | 28.73 |
| internlm2-chat-7b-sft-hf | 30.78 |
| internlm2-chat-20b-hf | 28.75 |
| internlm2-chat-20b-sft-hf | 34.10 |
| llama-3-8b-instruct-hf | 30.17 |
| llama-3-70b-instruct-hf | 40.25 |
| llama-3-8b-instruct-lmdeploy | 28.28 |
| llama-3-70b-instruct-lmdeploy | 39.14 |
| mistral-7b-instruct-v0.1-hf | 22.47 |
| mistral-7b-instruct-v0.2-hf | 25.18 |
| mixtral-8x7b-instruct-v0.1-hf | 32.08 |