feat: use a lighter quant
Browse files
app.py
CHANGED
@@ -8,7 +8,7 @@ from uuid import uuid4
|
|
8 |
# repo_id = "Qwen/Qwen2.5-0.5B-Instruct-GGUF"
|
9 |
# filename = "qwen2.5-0.5b-instruct-q8_0.gguf"
|
10 |
repo_id = "Qwen/Qwen2.5-3B-Instruct-GGUF"
|
11 |
-
filename = "qwen2.5-3b-instruct-
|
12 |
|
13 |
|
14 |
llm = load_llm(repo_id, filename)
|
|
|
8 |
# repo_id = "Qwen/Qwen2.5-0.5B-Instruct-GGUF"
|
9 |
# filename = "qwen2.5-0.5b-instruct-q8_0.gguf"
|
10 |
repo_id = "Qwen/Qwen2.5-3B-Instruct-GGUF"
|
11 |
+
filename = "qwen2.5-3b-instruct-q4_k_m.gguf"
|
12 |
|
13 |
|
14 |
llm = load_llm(repo_id, filename)
|