Upgraded model to Qwen-30B-Q3-K-XL

app.py

@@ -3,6 +3,9 @@
 
 To do
 - create the custom gradio look
+- explore making it look better
+- get a better model — Qwen 30b coder
+- use zerogpu
 
 """
 from huggingface_hub import hf_hub_download
@@ -17,14 +20,14 @@ import re # remove thinking tag from response
 # ----- Get Model ----- #
 # Download Q4_K_M GGUF file from the repo
 model_path = hf_hub_download(
-    repo_id="Qwen/Qwen2.5-Coder-7B-Instruct-GGUF",
-    filename="qwen2.5-coder-7b-instruct-q4_k_m.gguf"
+    repo_id="unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF",
+    filename="Qwen3-Coder-30B-A3B-Instruct-UD-Q3_K_XL.gguf" # fallback: Q2_K_XL
 )
 
 # Initialize llama.cpp with the local cached path
 llm = Llama(
     model_path=model_path,
-    n_ctx=2048,
+    n_ctx=4096,
     n_threads=2
 )