add "Q4_K_M" precision

app.py

@@ -21,7 +21,7 @@ MODELS = {
         "clip_repo":  "ggml-org/SmolVLM2-256M-Video-Instruct-GGUF",
         "model_prefix": "SmolVLM2-256M-Video-Instruct",
         "clip_prefix":  "mmproj-SmolVLM2-256M-Video-Instruct",
-        "model_variants": ["f16", "Q8_0", "Q2_K", ],
+        "model_variants": ["Q4_K_M", "f16", "Q8_0", "Q2_K", ],
         "clip_variants":  ["f16", "Q8_0", ],
     },
     "500M": {
@@ -29,7 +29,7 @@ MODELS = {
         "clip_repo":  "ggml-org/SmolVLM2-500M-Video-Instruct-GGUF",
         "model_prefix": "SmolVLM2-500M-Video-Instruct",
         "clip_prefix":  "mmproj-SmolVLM2-500M-Video-Instruct",
-        "model_variants": ["Q8_0", "Q2_K", "f16"],
+        "model_variants": ["Q4_K_M", "Q8_0", "Q2_K", "f16"],
         "clip_variants":  ["Q8_0", "f16"],
     },
     "2.2B": {
@@ -37,7 +37,7 @@ MODELS = {
         "clip_repo":  "ggml-org/SmolVLM2-2.2B-Instruct-GGUF",
         "model_prefix": "SmolVLM2-2.2B-Instruct",
         "clip_prefix":  "mmproj-SmolVLM2-2.2B-Instruct",
-        "model_variants": ["Q8_0", "Q2_K","Q4_K_M", "f16"],
+        "model_variants": ["Q4_K_M", "Q8_0", "Q2_K","Q4_K_M", "f16"],
         "clip_variants":  ["Q8_0", "f16"],
     },
 }