| import subprocess |
| |
| class LocalLLM: |
| """ |
| Thin wrapper around Ollama (or LM Studio) to run a local model. |
| |
| Parameters |
| ---------- |
| model : str |
| Ollama model tag, e.g. "deepseek-coder:6.7b-instruct". |
| """ |
| |
| def __init__(self, model: str = "codellama:7b-instruct"): |
| self.model = model |
| |
| # ------------- PUBLIC API ------------------------------------------------ |
| def generate( |
| self, |
| prompt: str, |
| system_prompt: str = "", |
| temperature: float = 0.0, |
| num_tokens: int | None = None, # optional n-token limit |
| ) -> str: |
| """ |
| Call the local LLM with a prompt. |
| |
| Uses the Ollama chat command: |
| /set parameter temperature <value> |
| |
| Notes |
| ----- |
| * Works even on older Ollama builds that don’t support --temp. |
| * You can still change top-p, top-k, etc. the same way. |
| """ |
| # prepend the /set command, then optional system prompt |
| header_lines = [f"/set parameter temperature {temperature}"] |
| if system_prompt.strip(): |
| header_lines.append(system_prompt.strip()) |
| header = "\n\n".join(header_lines) |
| |
| full_prompt = f"{header}\n\n{prompt.strip()}" |
| |
| cmd = ["ollama", "run", self.model, full_prompt] |
| if num_tokens is not None: |
| cmd += ["-n", str(num_tokens), "--no-cache"] |
| |
| result = subprocess.run(cmd, capture_output=True, text=True) |
| if result.returncode != 0: |
| raise RuntimeError(f"Ollama stderr:\n{result.stderr}") |
| if not result.stdout.strip(): |
| raise RuntimeError("Ollama returned an empty response.") |
| return result.stdout.strip() |
