模型
具体的模型版本，带来源的结构和权重文件信息。
| 模型 | 总参数量 | 估算占用 |
|---|---|---|
| Qwen2.5-7B-Instruct | 7.61B | 4.86 GiB |
| Qwen2.5-0.5B-Instruct | 0.49B | 1.08 GiB |
| Qwen2.5-1.5B-Instruct | 1.54B | 3.16 GiB |
| Qwen2.5-3B-Instruct | 3.09B | 2.30 GiB |
| Qwen2.5-14B-Instruct | 14.70B | 9.93 GiB |
| Qwen2.5-32B-Instruct | 32.50B | 20.55 GiB |
| Qwen2.5-72B-Instruct | 72.70B | 43.55 GiB |
| Qwen3-1.7B | 1.70B | 4.72 GiB |
| Qwen3-4B | 4.00B | 3.51 GiB |
| Qwen3-8B | 8.20B | 5.87 GiB |
| Qwen3-14B | 14.80B | 9.70 GiB |
| Qwen3-32B | 32.80B | 20.47 GiB |
| Qwen3-30B-A3B | 30.50B | 18.09 GiB |
| Phi-3.5-mini-instruct | 3.82B | 10.18 GiB |
| Mistral-7B-Instruct-v0.3 | 7.25B | 14.56 GiB |
| Mixtral-8x7B-Instruct-v0.1 | 46.70B | 88.05 GiB |
| DeepSeek-R1-Distill-Llama-8B | 8.03B | 16.02 GiB |
| DeepSeek-R1-Distill-Qwen-32B | 32.50B | 63.09 GiB |
| DeepSeek-V2-Lite-Chat | 15.70B | 29.26 GiB |