Minor fixes.

Author: souradipp76

readme_ready/main.py

@@ -56,7 +56,7 @@ def url_validator(x):
     if mode.lower() == "readme":
         headings = questionary.text(
             message="List of Readme Headings?(comma separated)"
-            + "[Example: # Introduction,## Usage]"
+            + "[Example: Introduction,Usage]"
         ).ask()
 
     model_name = questionary.select(

readme_ready/query/query.py

@@ -122,7 +122,7 @@ def generate_readme(
         for heading in headings:
             question = (
                 "Provide the README content for the section with "
-                + f"heading \"{heading}\" starting with ## {heading}."
+                + f'heading "{heading}" starting with ## {heading}.'
             )
             try:
                 response = chain.invoke({"input": question})

readme_ready/utils/llm_utils.py

@@ -22,7 +22,9 @@ def get_gemma_chat_model(model_name: str, streaming=False, model_kwargs=None):
         gguf_file = model_kwargs["gguf_file"]
         _ = hf_hub_download(model_name, gguf_file)
     tokenizer = get_tokenizer(model_name, gguf_file)
-    if sys.platform == "linux" or sys.platform == "linux2":
+    if (
+        sys.platform == "linux" or sys.platform == "linux2"
+    ) and "gptq" not in model_name.lower():
         from transformers import BitsAndBytesConfig
 
         bnb_config = BitsAndBytesConfig(
@@ -76,7 +78,9 @@ def get_llama_chat_model(model_name: str, streaming=False, model_kwargs=None):
         _ = hf_hub_download(model_name, gguf_file)
     tokenizer = get_tokenizer(model_name, gguf_file)
     tokenizer.pad_token = tokenizer.eos_token
-    if sys.platform == "linux" or sys.platform == "linux2":
+    if (
+        sys.platform == "linux" or sys.platform == "linux2"
+    ) and "gptq" not in model_name.lower():
         from transformers import BitsAndBytesConfig
 
         bnb_config = BitsAndBytesConfig(