Update predict_tr logger.debug

app.py

@@ -17,14 +17,31 @@ from ctransformers import AutoModelForCausalLM
 from huggingface_hub import hf_hub_download
 from loguru import logger
 
-filename_list = ["Wizard-Vicuna-7B-Uncensored.ggmlv3.q2_K.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q3_K_L.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q3_K_M.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q3_K_S.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q4_0.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q4_1.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q4_K_M.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q4_K_S.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q5_0.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q5_1.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q5_K_M.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q5_K_S.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q6_K.bin", "Wizard-Vicuna-7B-Uncensored.ggmlv3.q8_0.bin"]
+filename_list = [
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q2_K.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q3_K_L.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q3_K_M.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q3_K_S.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q4_0.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q4_1.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q4_K_M.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q4_K_S.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q5_0.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q5_1.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q5_K_M.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q5_K_S.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q6_K.bin",
+    "Wizard-Vicuna-7B-Uncensored.ggmlv3.q8_0.bin",
+]
 
 URL = "https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGML/raw/main/Wizard-Vicuna-7B-Uncensored.ggmlv3.q4_K_M.bin"  # 4.05G
 MODEL_FILENAME = Path(URL).name
-MODEL_FILENAME = filename_list[0]  # q2_K
-MODEL_FILENAME = filename_list[5]  # q4_1
+MODEL_FILENAME = filename_list[0]  # q2_K 4.05G
+MODEL_FILENAME = filename_list[5]  # q4_1 4.21
 
-REPO_ID = "/".join(urlparse(URL).path.strip('/').split('/')[:2])  # TheBloke/Wizard-Vicuna-7B-Uncensored-GGML
+REPO_ID = "/".join(
+    urlparse(URL).path.strip("/").split("/")[:2]
+)  # TheBloke/Wizard-Vicuna-7B-Uncensored-GGML
 
 DESTINATION_FOLDER = "models"
 
@@ -71,7 +88,10 @@ def predict_str(prompt, bot):  # bot is in fact bot_history
     # bot.append([prompt, f"{response} {_}"])
     # return prompt, bot
 
-    return prompt, bot + [[prompt, None]]
+    _ = bot + [[prompt, None]]
+    logger.debug(f"{prompt=}, {_=}")
+
+    return prompt, _
 
 
 def bot_str(bot):
@@ -242,6 +262,7 @@ def generate(
         **asdict(generation_config),
     )
 
+
 # if "mpt" in model_filename:
 #     config = AutoConfig.from_pretrained("mosaicml/mpt-30b-cha t", context_length=8192)
 #     llm = AutoModelForCausalLM.from_pretrained(
@@ -278,7 +299,7 @@ LLM = AutoModelForCausalLM.from_pretrained(
     # "TheBloke/WizardCoder-15B-1.0-GGML",
     REPO_ID,  # DESTINATION_FOLDER,  # model_path_or_repo_id: str required
     model_file=_,
-    model_type="llama",   # "starcoder",  AutoConfig.from_pretrained(REPO_ID)
+    model_type="llama",  # "starcoder",  AutoConfig.from_pretrained(REPO_ID)
     threads=cpu_count,
 )