Update app.py

app.py

@@ -5,6 +5,7 @@ import os
 import time
 from dataclasses import asdict, dataclass
 from pathlib import Path
+from urllib.parse import urlparse
 from types import SimpleNamespace
 
 import gradio as gr
@@ -229,7 +230,7 @@ def generate(
         **asdict(generation_config),
     )
 
-
+_ = '''
 _ = """full url: https://huggingface.co/TheBloke/mpt-30B-chat-GGML/blob/main/mpt-30b-chat.ggmlv0.q4_1.bin"""
 
 # https://huggingface.co/TheBloke/mpt-30B-chat-GGML
@@ -247,15 +248,13 @@ MODEL_FILENAME = "WizardCoder-15B-1.0.ggmlv3.q4_1.bin"  # 11.9G
 
 # https://huggingface.co/TheBloke/WizardLM-13B-V1.0-Uncensored-GGML
 MODEL_FILENAME = "wizardlm-13b-v1.0-uncensored.ggmlv3.q4_1.bin"  # 8.4G
+# '''
 
-DESTINATION_FOLDER = "models"
-
-REPO_ID = "TheBloke/mpt-30B-chat-GGML"
-if "WizardCoder" in MODEL_FILENAME:
-    REPO_ID = "TheBloke/WizardCoder-15B-1.0-GGML"
+URL = "https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGML/raw/main/Wizard-Vicuna-7B-Uncensored.ggmlv3.q4_K_M.bin"
+MODEL_FILENAME = Path(URL).name
+REPO_ID = "/".join(urlparse(url).path.strip('/').split('/')[:2])  # TheBloke/Wizard-Vicuna-7B-Uncensored-GGML
 
-if "uncensored" in MODEL_FILENAME.lower():
-    REPO_ID = "TheBloke/WizardLM-13B-V1.0-Uncensored-GGML"
+DESTINATION_FOLDER = "models"
 
 logger.info(f"start dl, {REPO_ID=}, {MODEL_FILENAME=}, {DESTINATION_FOLDER=}")
 download_quant(DESTINATION_FOLDER, REPO_ID, MODEL_FILENAME)
@@ -365,7 +364,7 @@ with gr.Blocks(
         #     """<center><a href="https://huggingface.co/spaces/mikeee/mpt-30b-chat?duplicate=true"><img src="https://bit.ly/3gLdBN6" alt="Duplicate"></a> and spin a CPU UPGRADE to avoid the queue</center>"""
         # )
         gr.Markdown(
-            f"""<h4><center>{MODEL_FILENAME}</center></h4>
+            f"""<h4><center>{REPO_ID} {MODEL_FILENAME}</center></h4>
             It takes about 100 seconds for the initial reply
             message to appear. Average streaming rate ~1 sec/chat. The bot only speaks English.