fix(llm): 分场景使用不同模型

- nvidia 激活、模型改为 step-3.7-flash → 用于日常生成/采集/优化等 - compliance_optimizer.py 显式指定 provider=sensenova → 合规审查走 deepseek-v4-flash - _FALLBACK 同步更新为 step-3.7-flash - 管理后台标签同步更新
2026-06-01 14:35:41 +08:00
parent 475945418e
commit f1dc29f439
3 changed files with 3 additions and 3 deletions
@@ -182,7 +182,7 @@ def polish_with_llm(html: str, platform: str, remaining_issues: Optional[List[Di
                prompt = get_prompt("compliance_fix", issues_desc=issues_desc, html=html)
            else:
                prompt = get_prompt("compliance_polish", html=html)
-            polished = call_llm(prompt, provider=None, temperature=temperature, max_tokens=max_tokens, system_prompt=system_prompt)
+            polished = call_llm(prompt, provider="sensenova", temperature=temperature, max_tokens=max_tokens, system_prompt=system_prompt)
            polished = clean_html_content(polished)
            polished = strip_ai_preface(polished)
            polished = strip_thinking_html(polished)