Update forindo branch

app.py

@@ -17,8 +17,6 @@ from dl_hf_model import dl_hf_model
 from loguru import logger
 
 url = "https://huggingface.co/TheBloke/llama-2-13B-Guanaco-QLoRA-GGML/blob/main/llama-2-13b-guanaco-qlora.ggmlv3.q4_K_S.bin"  # 8.14G
-if "forindo" in platform.node():
-    url = "https://huggingface.co/TheBloke/llama-2-70b-Guanaco-QLoRA-GGML/blob/main/llama-2-70b-guanaco-qlora.ggmlv3.q3_K_S.bin"  # 29.7G
 
 # Prompt template: Guanaco
 prompt_template = """You are a helpful assistant. Let's think step by step.
@@ -36,19 +34,24 @@ logger.debug(f"{cpu_count=}")
 
 LLM = None
 
-try:
-    model_loc, file_size = dl_hf_model(url)
-except Exception as exc_:
-    logger.error(exc_)
-    raise SystemExit(1) from exc_
+if "forindo" in platform.node():
+    # url = "https://huggingface.co/TheBloke/llama-2-70b-Guanaco-QLoRA-GGML/blob/main/llama-2-70b-guanaco-qlora.ggmlv3.q3_K_S.bin"  # 29.7G
+    model_loc = "/home/mu2018/github/langchain-llama-2-70b-guanaco-qlora-ggml/llama-2-70b-guanaco-qlora.ggmlv3.q3_K_S.bin"
+else:
+    try:
+        model_loc, file_size = dl_hf_model(url)
+        logger.info(f"done load llm {model_loc=} {file_size=}G")
+    except Exception as exc_:
+        logger.error(exc_)
+        raise SystemExit(1) from exc_
 
+logger.debug(f"{model_loc=}")
 LLM = AutoModelForCausalLM.from_pretrained(
     model_loc,
     model_type="llama",
     threads=cpu_count,
 )
 
-logger.info(f"done load llm {model_loc=} {file_size=}G")
 
 os.environ["TZ"] = "Asia/Shanghai"
 try: