reduce max tokens; refresh requirements.txt

2026-07-06 13:07:52 +02:00 · 2025-05-21 10:25:11 -06:00
parent ffacc738b7
commit 7960d225c6
2 changed files with 2 additions and 205 deletions
@@ -47,7 +47,7 @@ class Phi3LanguageModel:
            local_files_only=True
        )
        model = ORTModelForCausalLM.from_pretrained(
-            model_path,  # Change model_id to just model_path
+            model_path, 
            provider="CPUExecutionProvider",
            trust_remote_code=True,
            local_files_only=True
@@ -59,7 +59,7 @@ class Phi3LanguageModel:
            "text-generation",
            model=model,
            tokenizer=tokenizer,
-            max_new_tokens=512,
+            max_new_tokens=256,
            temperature=0.7,
            top_p=0.9,
            repetition_penalty=1.1,