模型的地址： https://huggingface.co/Undi95/LewdMistral-7B-0.2-GGUF 


source .venv/bin/activate



huggingface-cli download \
  Undi95/LewdMistral-7B-0.2-GGUF \
  LewdMistral-7B-0.2.q8_0.gguf \
  --local-dir downloads \
  --local-dir-use-symlinks False




vim Modelfile-Lewd



FROM "./downloads/LewdMistral-7B-0.2.q8_0.gguf"
PARAMETER stop "<|im_start|>"
PARAMETER stop "<|im_end|>"
TEMPLATE """
<|im_start|>system
{{ .System }}<|im_end|>
<|im_start|>user
{{ .Prompt }}<|im_end|>
<|im_start|>assistant
"""



https://sarinsuriyakoon.medium.com/unsloth-lora-with-ollama-lightweight-solution-to-full-cycle-llm-development-edadb6d9e0f0

参考这一篇：



FROM "./downloads/LewdMistral-7B-0.2.q8_0.gguf"
PARAMETER stop "### Response:"
PARAMETER stop "### Instruction:"
PARAMETER stop "### Input:"
TEMPLATE """
Below is an instruction that describes a task. Write a response that appropriately completes the request.

### Instruction:
{{ .System }}

### Input:
{{ .Prompt }}

### Response:
"""


然后命令行：


ollama create LewdMistral7BQ8v1 -f Modelfile-Lewd


ollama run LewdMistral7BQ8v1