add sd3.5-large and turbo gguf

xorbitsai · Dec 27, 2024 · 57b4181 · 57b4181
1 parent 7d7fb15
commit 57b4181
Show file tree

Hide file tree

Showing 2 changed files with 44 additions and 4 deletions.
diff --git a/xinference/model/image/model_spec.json b/xinference/model/image/model_spec.json
@@ -124,7 +124,17 @@
       "quantize_text_encoder": "text_encoder_3",
       "torch_dtype": "bfloat16",
       "transformer_nf4": true
-    }
+    },
+    "gguf_model_id": "city96/stable-diffusion-3.5-large-gguf",
+    "gguf_quantizations": [
+      "F16",
+      "Q4_0",
+      "Q4_1",
+      "Q5_0",
+      "Q5_1",
+      "Q8_0"
+    ],
+    "gguf_model_file_name_template": "sd3.5_large-{quantization}.gguf"
   },
   {
     "model_name": "sd3.5-large-turbo",
@@ -145,7 +155,17 @@
     "default_generate_config": {
       "guidance_scale": 1.0,
       "num_inference_steps": 4
-    }
+    },
+    "gguf_model_id": "city96/stable-diffusion-3.5-large-turbo-gguf",
+    "gguf_quantizations": [
+      "F16",
+      "Q4_0",
+      "Q4_1",
+      "Q5_0",
+      "Q5_1",
+      "Q8_0"
+    ],
+    "gguf_model_file_name_template": "sd3.5_large_turbo-{quantization}.gguf"
   },
   {
     "model_name": "sd-turbo",

diff --git a/xinference/model/image/model_spec_modelscope.json b/xinference/model/image/model_spec_modelscope.json
@@ -129,7 +129,17 @@
       "quantize_text_encoder": "text_encoder_3",
       "torch_dtype": "bfloat16",
       "transformer_nf4": true
-    }
+    },
+    "gguf_model_id": "Xorbits/stable-diffusion-3.5-large-gguf",
+    "gguf_quantizations": [
+      "F16",
+      "Q4_0",
+      "Q4_1",
+      "Q5_0",
+      "Q5_1",
+      "Q8_0"
+    ],
+    "gguf_model_file_name_template": "sd3.5_large-{quantization}.gguf"
   },
   {
     "model_name": "sd3.5-large-turbo",
@@ -151,7 +161,17 @@
     "default_generate_config": {
       "guidance_scale": 1.0,
       "num_inference_steps": 4
-    }
+    },
+    "gguf_model_id": "Xorbits/stable-diffusion-3.5-large-turbo-gguf",
+    "gguf_quantizations": [
+      "F16",
+      "Q4_0",
+      "Q4_1",
+      "Q5_0",
+      "Q5_1",
+      "Q8_0"
+    ],
+    "gguf_model_file_name_template": "sd3.5_large_turbo-{quantization}.gguf"
   },
   {
     "model_name": "sd-turbo",