LM Studio 0.2.22下载Deepseek-R1-distill-Qwen-7b GGUF，加载模型报错
llama.cpp error: ‘error loading model vocabulary: unknown pre-tokenizer type: ‘deepseek-r1-qwen’
错误内容
1 | json { "title": "Failed to load model", "cause": "llama.cpp error: 'error loading model vocabulary: unknown pre-tokenizer type: 'deepseek-r1-qwen''", "errorData": { "n_ctx": 2048, "n_batch": 512, "n_gpu_layers": 10 }, "data": { "memory": { "ram_capacity": "16.00 GB", "ram_unused": "3.40 GB" }, "gpu": { "type": "AppleMetal", "vram_recommended_capacity": "10.67 GB", "vram_unused": "2.85 GB" }, "os": { "platform": "darwin", "version": "15.2" }, "app": { "version": "0.2.22", "downloadsDir": "/Users/as/.cache/lm-studio/models" }, "model": {} } }``` |
解决方法
去LM Studio官网下载最新版DMG软件替换现有版本。亲测0.3.9 版本可用。