Testing of GGUF Llama3.2 3B

README.md

@@ -5,7 +5,7 @@ colorFrom: green
 colorTo: indigo
 sdk: gradio
 app_file: app.py
-pinned: true
+pinned: flase
 license: cc-by-sa-4.0
 short_description: SmallZOO runnigng SLMs directly on CPU with Llama.cpp&Python
 ---

app.py

@@ -2,10 +2,9 @@ import gradio as gr
 from llama_cpp import Llama
 import requests
 
-
 llm = Llama.from_pretrained(
-    repo_id="cognitivecomputations/dolphin-2.9.2-qwen2-7b-gguf",
-    filename="*Q4_K_S.gguf",
+    repo_id="lmstudio-community/Llama-3.2-3B-Instruct-GGUF",
+    filename="*Q4_K_M.gguf",
     verbose=True,
     n_ctx=32768,
     n_threads=2,

models.lst

@@ -0,0 +1,3 @@
+Stack of modesl to try:
+
+https://huggingface.co/lmstudio-community/Llama-3.2-3B-Instruct-GGUF