Upload 5 files

README.md

@@ -1,165 +1,166 @@
----
-title: 🧩 DiffuseCraft Mod (SDXL/SD1.5 Models Text-to-Image)
-emoji: 🧩🖼️📦
-colorFrom: red
-colorTo: pink
-sdk: gradio
-sdk_version: 5.45.0
-app_file: app.py
-pinned: true
-header: mini
-license: mit
-duplicated_from: r3gm/DiffuseCraft
-short_description: Stunning images using stable diffusion.
-preload_from_hub:
-  - madebyollin/sdxl-vae-fp16-fix config.json,diffusion_pytorch_model.safetensors
----
-
-## Using this Space programmatically
-
-You can call this Space from Python (via `gradio_client`) or from plain `curl`.
-
-> ⚠️ Note: This README may lag behind the actual API definition shown in the Space’s “View API” page.
-> If something does not work, always double-check the latest argument list and endpoint names there.
-
-Assumptions:
-
-- Space ID: `John6666/DiffuseCraftMod`
-- You have a valid Hugging Face access token: `hf_xxx...` (read access is enough)
-- Replace `hf_xxx...` with your own token
-
----
-
-### 1. Python examples (`gradio_client`)
-
-Install:
-
-```bash
-pip install gradio_client
-````
-
-#### 1.1 Synchronous API – `generate_image`
-
-```python
-from gradio_client import Client
-
-client = Client("John6666/DiffuseCraftMod", hf_token="hf_xxx...")
-
-status, images, info = client.predict(
-    # Core text controls
-    prompt="Hello!!",
-    negative_prompt=(
-        "lowres, bad anatomy, bad hands, missing fingers, extra digit, "
-        "fewer digits, worst quality, low quality"
-    ),
-
-    # Basic generation controls
-    num_images=1,
-    num_inference_steps=28,
-    guidance_scale=7.0,
-    clip_skip=0,
-    seed=-1,
-
-    # Canvas / model / task (optional, server has defaults)
-    height=1024,
-    width=1024,
-    model_name="votepurchase/animagine-xl-3.1",
-    vae_model="None",
-    task="txt2img",
-
-    # All other arguments are optional; defaults match the UI
-    api_name="/generate_image",
-)
-
-print(status)   # e.g. "COMPLETE"
-print(images)   # list of image paths / URLs
-print(info)     # generation metadata (seed, model, etc.)
-```
-
-#### 1.2 Streaming API – `generate_image_stream`
-
-```python
-from gradio_client import Client
-
-client = Client("John6666/DiffuseCraftMod", hf_token="hf_xxx...")
-
-job = client.submit(
-    prompt="Hello!!",
-    negative_prompt=(
-        "lowres, bad anatomy, bad hands, missing fingers, extra digit, "
-        "fewer digits, worst quality, low quality"
-    ),
-    num_images=1,
-    num_inference_steps=28,
-    guidance_scale=7.0,
-    clip_skip=0,
-    seed=-1,
-    height=1024,
-    width=1024,
-    model_name="votepurchase/animagine-xl-3.1",
-    vae_model="None",
-    task="txt2img",
-    api_name="/generate_image_stream",
-)
-
-for status, images, info in job:
-    # You will see progress messages, intermediate previews, and the final result.
-    print(status, images, info)
-```
-
-You can stop iterating once you see a `"COMPLETE"` status if you only care about the final output.
-
----
-
-### 2. `curl` examples
-
-When calling from `curl`, include your HF token; anonymous calls may be rate-limited or rejected.
-
-```bash
-export HF_TOKEN="hf_xxx..."  # your Hugging Face access token
-```
-
-The `data` field is a positional array. The order must match the function signature.
-For simplicity, the examples below only send the first few arguments and rely on server defaults for the rest.
-
-#### 2.1 Synchronous API – `generate_image`
-
-```bash
-curl -X POST "https://john6666-diffusecraftmod.hf.space/call/generate_image" \
-  -H "Authorization: Bearer $HF_TOKEN" \
-  -H "Content-Type: application/json" \
-  -d '{
-    "data": [
-      "Hello!!",  // prompt
-      "lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, worst quality, low quality", // negative_prompt
-      1,          // num_images
-      28,         // num_inference_steps
-      7.0,        // guidance_scale
-      0,          // clip_skip
-      -1          // seed
-      // All subsequent parameters will use their default values
-    ]
-  }'
-```
-
-#### 2.2 Streaming API – `generate_image_stream`
-
-```bash
-curl -X POST "https://john6666-diffusecraftmod.hf.space/call/generate_image_stream" \
-  -H "Authorization: Bearer $HF_TOKEN" \
-  -H "Content-Type: application/json" \
-  -d '{
-    "data": [
-      "Hello!!",
-      "lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, worst quality, low quality",
-      1,
-      28,
-      7.0,
-      0,
-      -1
-    ]
-  }'
-```
-
-For full parameter coverage (all advanced options such as LoRAs, ControlNet, IP-Adapter, etc.),
-refer to the Space’s “View API” page and adapt the examples above accordingly.
+---
+title: 🧩 DiffuseCraft Mod (SDXL/SD1.5 Models Text-to-Image)
+emoji: 🧩🖼️📦
+colorFrom: red
+colorTo: pink
+sdk: gradio
+sdk_version: 6.17.3
+python_version: "3.12"
+app_file: app.py
+pinned: true
+header: mini
+license: mit
+duplicated_from: r3gm/DiffuseCraft
+short_description: Stunning images using stable diffusion.
+preload_from_hub:
+  - madebyollin/sdxl-vae-fp16-fix config.json,diffusion_pytorch_model.safetensors
+---
+
+## Using this Space programmatically
+
+You can call this Space from Python (via `gradio_client`) or from plain `curl`.
+
+> ⚠️ Note: This README may lag behind the actual API definition shown in the Space’s “View API” page.
+> If something does not work, always double-check the latest argument list and endpoint names there.
+
+Assumptions:
+
+- Space ID: `John6666/DiffuseCraftMod`
+- You have a valid Hugging Face access token: `hf_xxx...` (read access is enough)
+- Replace `hf_xxx...` with your own token
+
+---
+
+### 1. Python examples (`gradio_client`)
+
+Install:
+
+```bash
+pip install gradio_client
+````
+
+#### 1.1 Synchronous API – `generate_image`
+
+```python
+from gradio_client import Client
+
+client = Client("John6666/DiffuseCraftMod", hf_token="hf_xxx...")
+
+status, images, info = client.predict(
+    # Core text controls
+    prompt="Hello!!",
+    negative_prompt=(
+        "lowres, bad anatomy, bad hands, missing fingers, extra digit, "
+        "fewer digits, worst quality, low quality"
+    ),
+
+    # Basic generation controls
+    num_images=1,
+    num_inference_steps=28,
+    guidance_scale=7.0,
+    clip_skip=0,
+    seed=-1,
+
+    # Canvas / model / task (optional, server has defaults)
+    height=1024,
+    width=1024,
+    model_name="votepurchase/animagine-xl-3.1",
+    vae_model="None",
+    task="txt2img",
+
+    # All other arguments are optional; defaults match the UI
+    api_name="/generate_image",
+)
+
+print(status)   # e.g. "COMPLETE"
+print(images)   # list of image paths / URLs
+print(info)     # generation metadata (seed, model, etc.)
+```
+
+#### 1.2 Streaming API – `generate_image_stream`
+
+```python
+from gradio_client import Client
+
+client = Client("John6666/DiffuseCraftMod", hf_token="hf_xxx...")
+
+job = client.submit(
+    prompt="Hello!!",
+    negative_prompt=(
+        "lowres, bad anatomy, bad hands, missing fingers, extra digit, "
+        "fewer digits, worst quality, low quality"
+    ),
+    num_images=1,
+    num_inference_steps=28,
+    guidance_scale=7.0,
+    clip_skip=0,
+    seed=-1,
+    height=1024,
+    width=1024,
+    model_name="votepurchase/animagine-xl-3.1",
+    vae_model="None",
+    task="txt2img",
+    api_name="/generate_image_stream",
+)
+
+for status, images, info in job:
+    # You will see progress messages, intermediate previews, and the final result.
+    print(status, images, info)
+```
+
+You can stop iterating once you see a `"COMPLETE"` status if you only care about the final output.
+
+---
+
+### 2. `curl` examples
+
+When calling from `curl`, include your HF token; anonymous calls may be rate-limited or rejected.
+
+```bash
+export HF_TOKEN="hf_xxx..."  # your Hugging Face access token
+```
+
+The `data` field is a positional array. The order must match the function signature.
+For simplicity, the examples below only send the first few arguments and rely on server defaults for the rest.
+
+#### 2.1 Synchronous API – `generate_image`
+
+```bash
+curl -X POST "https://john6666-diffusecraftmod.hf.space/call/generate_image" \
+  -H "Authorization: Bearer $HF_TOKEN" \
+  -H "Content-Type: application/json" \
+  -d '{
+    "data": [
+      "Hello!!",  // prompt
+      "lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, worst quality, low quality", // negative_prompt
+      1,          // num_images
+      28,         // num_inference_steps
+      7.0,        // guidance_scale
+      0,          // clip_skip
+      -1          // seed
+      // All subsequent parameters will use their default values
+    ]
+  }'
+```
+
+#### 2.2 Streaming API – `generate_image_stream`
+
+```bash
+curl -X POST "https://john6666-diffusecraftmod.hf.space/call/generate_image_stream" \
+  -H "Authorization: Bearer $HF_TOKEN" \
+  -H "Content-Type: application/json" \
+  -d '{
+    "data": [
+      "Hello!!",
+      "lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, worst quality, low quality",
+      1,
+      28,
+      7.0,
+      0,
+      -1
+    ]
+  }'
+```
+
+For full parameter coverage (all advanced options such as LoRAs, ControlNet, IP-Adapter, etc.),
+refer to the Space’s “View API” page and adapt the examples above accordingly.

app.py

@@ -878,7 +878,7 @@ CSS ="""
 .desc [src$='#float'] { float: right; margin: 20px; }
 """
 
-with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=False, css=CSS) as app:
+with gr.Blocks(elem_id="main", fill_width=True, fill_height=False) as app:
     gr.Markdown("# 🧩 DiffuseCraft Mod", elem_classes="title")
     gr.Markdown("This space is a modification of [r3gm's DiffuseCraft](https://huggingface.co/spaces/r3gm/DiffuseCraft).", elem_classes="info")
     with gr.Column():
@@ -923,9 +923,9 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                                     keep_tags_gui = gr.Radio(label="Remove tags leaving only the following", choices=["body", "dress", "all"], value="all")
                                 image_algorithms = gr.CheckboxGroup(["Use WD Tagger"], label="Algorithms", value=["Use WD Tagger"], visible=False)
                             generate_from_image_btn_gui = gr.Button(value="GENERATE TAGS FROM IMAGE")
-                        prompt_gui = gr.Textbox(lines=6, placeholder="1girl, solo, ...", label="Prompt", show_copy_button=True)
+                        prompt_gui = gr.Textbox(lines=6, placeholder="1girl, solo, ...", label="Prompt", buttons=["copy"])
                         with gr.Accordion("Negative prompt, etc.", open=False) as menu_negative:
-                            neg_prompt_gui = gr.Textbox(lines=3, placeholder="Enter Neg prompt", label="Negative prompt", value="lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, worst quality, low quality, very displeasing, (bad)", show_copy_button=True)
+                            neg_prompt_gui = gr.Textbox(lines=3, placeholder="Enter Neg prompt", label="Negative prompt", value="lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, worst quality, low quality, very displeasing, (bad)", buttons=["copy"])
                             translate_prompt_button = gr.Button(value="Translate prompt to English", size="sm", variant="secondary")
                             with gr.Row():
                                 insert_prompt_gui = gr.Radio(label="Insert reccomended positive / negative prompt", choices=["None", "Auto", "Animagine", "Pony"], value="Auto", interactive=True)                          
@@ -955,7 +955,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                         update_task_options,
                         [model_name_gui, task_gui],
                         [task_gui],
-                        show_api=False,
+                        api_visibility="undocumented",
                     )
 
                     load_model_gui = gr.HTML(elem_id="load_model", elem_classes="contain")
@@ -972,8 +972,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                         # height="auto",
                         interactive=False,
                         preview=False,
-                        show_share_button=False,
-                        show_download_button=True,
+                        buttons=["download", "fullscreen"],
                         selected_index=50,
                         format="png",
                     )
@@ -984,9 +983,9 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
 
                     with gr.Accordion("History", open=False):
                         history_files = gr.Files(interactive=False, visible=False)
-                        history_gallery = gr.Gallery(label="History", columns=6, object_fit="contain", format="png", interactive=False, show_share_button=False, show_download_button=True)
+                        history_gallery = gr.Gallery(label="History", columns=6, object_fit="contain", format="png", interactive=False, buttons=["download", "fullscreen"])
                         history_clear_button = gr.Button(value="Clear History", variant="secondary")
-                        history_clear_button.click(lambda: ([], []), None, [history_gallery, history_files], queue=False, show_api=False)
+                        history_clear_button.click(lambda: ([], []), None, [history_gallery, history_files], queue=False, api_visibility="undocumented")
 
                     with gr.Row(equal_height=False, variant="default"):
                         gpu_duration_gui = gr.Number(minimum=5, maximum=240, value=20, show_label=False, container=False, info="GPU time duration (seconds)")
@@ -1166,7 +1165,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                             return gr.Slider(minimum=-val_lora, maximum=val_lora, step=0.01, value=1.0, label=label, visible=visible)
                     
                         def lora_textbox(label):
-                            return gr.Textbox(label=label, info="Example of prompt:", value="None", show_copy_button=True, interactive=False, visible=False)
+                            return gr.Textbox(label=label, info="Example of prompt:", value="None", buttons=["copy"], interactive=False, visible=False)
                         
                         with gr.Row():
                             with gr.Column():
@@ -1237,7 +1236,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                             search_civitai_button_lora = gr.Button("Search on Civitai")
                             search_civitai_desc_lora = gr.Markdown(value="", visible=False, elem_classes="desc")
                             with gr.Accordion("Select from Gallery", open=False):
-                                search_civitai_gallery_lora = gr.Gallery([], label="Results", allow_preview=False, columns=5, show_share_button=False, interactive=False)
+                                search_civitai_gallery_lora = gr.Gallery([], label="Results", allow_preview=False, columns=5, buttons=["download", "fullscreen"], interactive=False)
                             search_civitai_result_lora = gr.Dropdown(label="Search Results", choices=[("", "")], value="", allow_custom_value=True, visible=False)
                             with gr.Row():
                                 text_lora = gr.Textbox(label="LoRA's download URL", placeholder="https://civitai.com/api/download/models/28907", info="It has to be .safetensors files, and you can also download them from Hugging Face.", lines=1, scale=4)
@@ -1321,8 +1320,8 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                             use_textual_inversion_gui = gr.CheckboxGroup(choices=get_embed_list(get_model_pipeline(model_name_gui.value)) if active_textual_inversion_gui.value else [], value=None, label="Use Textual Invertion in prompt")
                             def update_textual_inversion_gui(active_textual_inversion_gui, model_name_gui):
                                 return gr.update(choices=get_embed_list(get_model_pipeline(model_name_gui)) if active_textual_inversion_gui else [])
-                            active_textual_inversion_gui.change(update_textual_inversion_gui, [active_textual_inversion_gui, model_name_gui], [use_textual_inversion_gui], show_api=False)
-                            model_name_gui.change(update_textual_inversion_gui, [active_textual_inversion_gui, model_name_gui], [use_textual_inversion_gui], show_api=False)
+                            active_textual_inversion_gui.change(update_textual_inversion_gui, [active_textual_inversion_gui, model_name_gui], [use_textual_inversion_gui], api_visibility="undocumented")
+                            model_name_gui.change(update_textual_inversion_gui, [active_textual_inversion_gui, model_name_gui], [use_textual_inversion_gui], api_visibility="undocumented")
 
                         with gr.Accordion("ControlNet / Img2img / Inpaint", open=False, visible=True) as menu_i2i:
                             with gr.Row():
@@ -1360,7 +1359,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                                     change_preprocessor_choices,
                                     [task_gui],
                                     [preprocessor_name_gui],
-                                    show_api=False,
+                                    api_visibility="undocumented",
                                 )
 
                             with gr.Row():
@@ -1420,7 +1419,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                                 gr.Info(f"{len(sd_gen.model.STYLE_NAMES)} styles loaded")
                                 return gr.update(value=None, choices=sd_gen.model.STYLE_NAMES)
 
-                            style_button.click(load_json_style_file, [style_json_gui], [style_prompt_gui], show_api=False)
+                            style_button.click(load_json_style_file, [style_json_gui], [style_prompt_gui], api_visibility="undocumented")
 
                         with gr.Accordion("Other settings", open=False, visible=True) as menu_other:
                             with gr.Row():
@@ -1542,7 +1541,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
 
                     def change_visibility_canvas():
                         return gr.update(visible=True, interactive=True), gr.update(visible=False)
-                    show_canvas.click(change_visibility_canvas, [], [image_base, show_canvas], show_api=False)
+                    show_canvas.click(change_visibility_canvas, [], [image_base, show_canvas], api_visibility="undocumented")
 
                     invert_mask = gr.Checkbox(value=False, label="Invert mask")
                     btn = gr.Button("Create mask")
@@ -1556,7 +1555,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
 
                 def send_img(img_source, img_result):
                     return img_source, img_result
-                btn_send.click(send_img, [img_source, img_result], [image_control, image_mask_gui], show_api=False)
+                btn_send.click(send_img, [img_source, img_result], [image_control, image_mask_gui], api_visibility="undocumented")
 
         with gr.Tab("PNG Info"):
             with gr.Row():
@@ -1564,7 +1563,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                     image_metadata = gr.Image(label="Image with metadata", type="pil", sources=["upload"])
 
                 with gr.Column():
-                    result_metadata = gr.Textbox(label="Metadata", show_label=True, show_copy_button=True, interactive=False, container=True, max_lines=99)
+                    result_metadata = gr.Textbox(label="Metadata", show_label=True, buttons=["copy"], interactive=False, container=True, max_lines=99)
 
                     image_metadata.change(
                         fn=extract_exif_data,
@@ -1602,11 +1601,11 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
             [menu_model, menu_from_image, menu_negative, menu_gen, menu_hires, menu_lora, menu_advanced,
               menu_example, task_gui, quick_speed_gui],
             queue=False,
-            show_api=False,
+            api_visibility="undocumented",
         )
-        model_name_gui.change(get_t2i_model_info, [model_name_gui], [model_info_gui], queue=False, show_api=False)
-        translate_prompt_gui.click(translate_to_en, [prompt_gui], [prompt_gui], queue=False, show_api=False)\
-        .then(translate_to_en, [neg_prompt_gui], [neg_prompt_gui], queue=False, show_api=False)
+        model_name_gui.change(get_t2i_model_info, [model_name_gui], [model_info_gui], queue=False, api_visibility="undocumented")
+        translate_prompt_gui.click(translate_to_en, [prompt_gui], [prompt_gui], queue=False, api_visibility="undocumented")\
+        .then(translate_to_en, [neg_prompt_gui], [neg_prompt_gui], queue=False, api_visibility="undocumented")
 
         gr.on(
             triggers=[quick_model_type_gui.change, quick_genre_gui.change, quick_speed_gui.change, quick_aspect_gui.change],
@@ -1615,7 +1614,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
             outputs=[quality_selector_gui, style_selector_gui, sampler_selector_gui, optimization_gui, insert_prompt_gui],
             queue=False,
             trigger_mode="once",
-            show_api=False,
+            api_visibility="undocumented",
         )
         gr.on(
             triggers=[quality_selector_gui.change, style_selector_gui.change, insert_prompt_gui.change],
@@ -1624,7 +1623,7 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
             outputs=[prompt_gui, neg_prompt_gui, quick_model_type_gui],
             queue=False,
             trigger_mode="once",
-            show_api=False,
+            api_visibility="undocumented",
         )
         sampler_selector_gui.change(set_sampler_settings, [sampler_selector_gui], [sampler_gui, steps_gui, cfg_gui, clip_skip_gui, img_width_gui, img_height_gui, optimization_gui], queue=False)
         optimization_gui.change(set_optimization, [optimization_gui, steps_gui, cfg_gui, sampler_gui, clip_skip_gui, lora5_gui, lora_scale_5_gui], [steps_gui, cfg_gui, sampler_gui, clip_skip_gui, lora5_gui, lora_scale_5_gui], queue=False)
@@ -1647,15 +1646,15 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                     lora7_gui, lora_scale_7_gui, lora7_info_gui, lora7_copy_gui, lora7_desc_gui],
             queue=False,
             trigger_mode="once",
-            show_api=False,
+            api_visibility="undocumented",
         )
-        lora1_copy_gui.click(apply_lora_prompt, [prompt_gui, lora1_info_gui], [prompt_gui], queue=False, show_api=False)
-        lora2_copy_gui.click(apply_lora_prompt, [prompt_gui, lora2_info_gui], [prompt_gui], queue=False, show_api=False)
-        lora3_copy_gui.click(apply_lora_prompt, [prompt_gui, lora3_info_gui], [prompt_gui], queue=False, show_api=False)
-        lora4_copy_gui.click(apply_lora_prompt, [prompt_gui, lora4_info_gui], [prompt_gui], queue=False, show_api=False)
-        lora5_copy_gui.click(apply_lora_prompt, [prompt_gui, lora5_info_gui], [prompt_gui], queue=False, show_api=False)
-        lora6_copy_gui.click(apply_lora_prompt, [prompt_gui, lora6_info_gui], [prompt_gui], queue=False, show_api=False)
-        lora7_copy_gui.click(apply_lora_prompt, [prompt_gui, lora7_info_gui], [prompt_gui], queue=False, show_api=False)
+        lora1_copy_gui.click(apply_lora_prompt, [prompt_gui, lora1_info_gui], [prompt_gui], queue=False, api_visibility="undocumented")
+        lora2_copy_gui.click(apply_lora_prompt, [prompt_gui, lora2_info_gui], [prompt_gui], queue=False, api_visibility="undocumented")
+        lora3_copy_gui.click(apply_lora_prompt, [prompt_gui, lora3_info_gui], [prompt_gui], queue=False, api_visibility="undocumented")
+        lora4_copy_gui.click(apply_lora_prompt, [prompt_gui, lora4_info_gui], [prompt_gui], queue=False, api_visibility="undocumented")
+        lora5_copy_gui.click(apply_lora_prompt, [prompt_gui, lora5_info_gui], [prompt_gui], queue=False, api_visibility="undocumented")
+        lora6_copy_gui.click(apply_lora_prompt, [prompt_gui, lora6_info_gui], [prompt_gui], queue=False, api_visibility="undocumented")
+        lora7_copy_gui.click(apply_lora_prompt, [prompt_gui, lora7_info_gui], [prompt_gui], queue=False, api_visibility="undocumented")
         gr.on(
             triggers=[search_civitai_button_lora.click, search_civitai_query_lora.submit],
             fn=search_civitai_lora,
@@ -1664,54 +1663,54 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
             outputs=[search_civitai_result_lora, search_civitai_desc_lora, search_civitai_button_lora, search_civitai_query_lora, search_civitai_gallery_lora],
             queue=True,
             scroll_to_output=True,
-            show_api=False,
+            api_visibility="undocumented",
         )
-        search_civitai_result_lora.change(select_civitai_lora, [search_civitai_result_lora], [text_lora, search_civitai_desc_lora], queue=False, scroll_to_output=True, show_api=False)
-        search_civitai_gallery_lora.select(update_civitai_selection, None, [search_civitai_result_lora], queue=False, show_api=False)
-        button_lora.click(get_my_lora, [text_lora, romanize_text], [lora1_gui, lora2_gui, lora3_gui, lora4_gui, lora5_gui, lora6_gui, lora7_gui, new_lora_status], scroll_to_output=True, show_api=False)
-        upload_button_lora.upload(upload_file_lora, [upload_button_lora], [file_output_lora, upload_button_lora], show_api=False).success(
-            move_file_lora, [file_output_lora], [lora1_gui, lora2_gui, lora3_gui, lora4_gui, lora5_gui, lora6_gui, lora7_gui], scroll_to_output=True, show_api=False)
+        search_civitai_result_lora.change(select_civitai_lora, [search_civitai_result_lora], [text_lora, search_civitai_desc_lora], queue=False, scroll_to_output=True, api_visibility="undocumented")
+        search_civitai_gallery_lora.select(update_civitai_selection, None, [search_civitai_result_lora], queue=False, api_visibility="undocumented")
+        button_lora.click(get_my_lora, [text_lora, romanize_text], [lora1_gui, lora2_gui, lora3_gui, lora4_gui, lora5_gui, lora6_gui, lora7_gui, new_lora_status], scroll_to_output=True, api_visibility="undocumented")
+        upload_button_lora.upload(upload_file_lora, [upload_button_lora], [file_output_lora, upload_button_lora], api_visibility="undocumented").success(
+            move_file_lora, [file_output_lora], [lora1_gui, lora2_gui, lora3_gui, lora4_gui, lora5_gui, lora6_gui, lora7_gui], scroll_to_output=True, api_visibility="undocumented")
 
-        use_textual_inversion_gui.change(set_textual_inversion_prompt, [use_textual_inversion_gui, prompt_gui, neg_prompt_gui, prompt_syntax_gui], [prompt_gui, neg_prompt_gui], show_api=False)
+        use_textual_inversion_gui.change(set_textual_inversion_prompt, [use_textual_inversion_gui, prompt_gui, neg_prompt_gui, prompt_syntax_gui], [prompt_gui, neg_prompt_gui], api_visibility="undocumented")
 
         generate_from_image_btn_gui.click(
-            lambda: ("", "", ""), None, [series_dbt, character_dbt, prompt_gui], queue=False, show_api=False,
+            lambda: ("", "", ""), None, [series_dbt, character_dbt, prompt_gui], queue=False, api_visibility="undocumented",
         ).success(
             predict_tags_wd,
             [input_image_gui, prompt_gui, image_algorithms, general_threshold_gui, character_threshold_gui],
             [series_dbt, character_dbt, prompt_gui, copy_button_dbt],
-            show_api=False,
+            api_visibility="undocumented",
         ).success(
-            compose_prompt_to_copy, [character_dbt, series_dbt, prompt_gui], [prompt_gui], queue=False, show_api=False,
+            compose_prompt_to_copy, [character_dbt, series_dbt, prompt_gui], [prompt_gui], queue=False, api_visibility="undocumented",
         ).success(
-            remove_specific_prompt, [prompt_gui, keep_tags_gui], [prompt_gui], queue=False, show_api=False,
+            remove_specific_prompt, [prompt_gui, keep_tags_gui], [prompt_gui], queue=False, api_visibility="undocumented",
         ).success(
-            convert_danbooru_to_e621_prompt, [prompt_gui, tag_type_gui], [prompt_gui], queue=False, show_api=False,
+            convert_danbooru_to_e621_prompt, [prompt_gui, tag_type_gui], [prompt_gui], queue=False, api_visibility="undocumented",
         ).success(
-            insert_recom_prompt, [prompt_gui, neg_prompt_gui, recom_prompt_gui], [prompt_gui, neg_prompt_gui], queue=False, show_api=False,
+            insert_recom_prompt, [prompt_gui, neg_prompt_gui, recom_prompt_gui], [prompt_gui, neg_prompt_gui], queue=False, api_visibility="undocumented",
         )
         
-        prompt_type_button.click(convert_danbooru_to_e621_prompt, [prompt_gui, prompt_type_gui], [prompt_gui], queue=False, show_api=False)
-        random_character_gui.click(select_random_character, [series_dbt, character_dbt], [series_dbt, character_dbt], queue=False, show_api=False)
+        prompt_type_button.click(convert_danbooru_to_e621_prompt, [prompt_gui, prompt_type_gui], [prompt_gui], queue=False, api_visibility="undocumented")
+        random_character_gui.click(select_random_character, [series_dbt, character_dbt], [series_dbt, character_dbt], queue=False, api_visibility="undocumented")
         generate_db_random_button.click(
             v2_random_prompt,
             [prompt_gui, series_dbt, character_dbt,
             rating_dbt, aspect_ratio_dbt, length_dbt, identity_dbt, ban_tags_dbt, model_name_dbt],
             [prompt_gui, series_dbt, character_dbt],
-            show_api=False,
+            api_visibility="undocumented",
         ).success(
-            convert_danbooru_to_e621_prompt, [prompt_gui, tag_type_gui], [prompt_gui], queue=False, show_api=False,
+            convert_danbooru_to_e621_prompt, [prompt_gui, tag_type_gui], [prompt_gui], queue=False, api_visibility="undocumented",
         )
 
-        translate_prompt_button.click(translate_prompt, [prompt_gui], [prompt_gui], queue=False, show_api=False)
-        translate_prompt_button.click(translate_prompt, [character_dbt], [character_dbt], queue=False, show_api=False)
-        translate_prompt_button.click(translate_prompt, [series_dbt], [series_dbt], queue=False, show_api=False)
+        translate_prompt_button.click(translate_prompt, [prompt_gui], [prompt_gui], queue=False, api_visibility="undocumented")
+        translate_prompt_button.click(translate_prompt, [character_dbt], [character_dbt], queue=False, api_visibility="undocumented")
+        translate_prompt_button.click(translate_prompt, [series_dbt], [series_dbt], queue=False, api_visibility="undocumented")
 
         generate_button.click(
             fn=insert_model_recom_prompt,
             inputs=[prompt_gui, neg_prompt_gui, model_name_gui, recom_prompt_gui],
             outputs=[prompt_gui, neg_prompt_gui],
-            api_name=False,
+            api_visibility="private",
             queue=False,
         ).success(
             fn=sd_gen.load_new_model,
@@ -1854,8 +1853,8 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
             api_name="sd_gen_generate_pipeline",
             queue=True,
             show_progress="full",
-        ).success(save_gallery_images, [result_images, model_name_gui], [result_images, result_images_files], queue=False, show_api=False)\
-        .success(save_gallery_history, [result_images, result_images_files, history_gallery, history_files], [history_gallery, history_files], queue=False, show_api=False)
+        ).success(save_gallery_images, [result_images, model_name_gui], [result_images, result_images_files], queue=False, api_visibility="undocumented")\
+        .success(save_gallery_history, [result_images, result_images_files, history_gallery, history_files], [history_gallery, history_files], queue=False, api_visibility="undocumented")
 
         with gr.Tab("Danbooru Tags Transformer with WD Tagger", render=True):
             with gr.Column(scale=2):
@@ -1894,60 +1893,60 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
                     generate_btn = gr.Button(value="GENERATE TAGS", size="lg", variant="primary")
                 with gr.Row():
                     with gr.Group():
-                        output_text = gr.TextArea(label="Output tags", interactive=False, show_copy_button=True)
+                        output_text = gr.TextArea(label="Output tags", interactive=False, buttons=["copy"])
                         with gr.Row():
                             copy_btn = gr.Button(value="Copy to clipboard", size="sm", interactive=False)
                             copy_prompt_btn = gr.Button(value="Copy to primary prompt", size="sm", interactive=False)
                     with gr.Group():
-                        output_text_pony = gr.TextArea(label="Output tags (Pony e621 style)", interactive=False, show_copy_button=True)
+                        output_text_pony = gr.TextArea(label="Output tags (Pony e621 style)", interactive=False, buttons=["copy"])
                         with gr.Row():
                             copy_btn_pony = gr.Button(value="Copy to clipboard", size="sm", interactive=False)
                             copy_prompt_btn_pony = gr.Button(value="Copy to primary prompt", size="sm", interactive=False)
                 description_ui()
 
-        translate_input_prompt_button.click(translate_prompt, inputs=[input_general], outputs=[input_general], queue=False, show_api=False)
-        translate_input_prompt_button.click(translate_prompt, inputs=[input_character], outputs=[input_character], queue=False, show_api=False)
-        translate_input_prompt_button.click(translate_prompt, inputs=[input_copyright], outputs=[input_copyright], queue=False, show_api=False)
+        translate_input_prompt_button.click(translate_prompt, inputs=[input_general], outputs=[input_general], queue=False, api_visibility="undocumented")
+        translate_input_prompt_button.click(translate_prompt, inputs=[input_character], outputs=[input_character], queue=False, api_visibility="undocumented")
+        translate_input_prompt_button.click(translate_prompt, inputs=[input_copyright], outputs=[input_copyright], queue=False, api_visibility="undocumented")
 
         generate_from_image_btn.click(
-            lambda: ("", "", ""), None, [input_copyright, input_character, input_general], queue=False, show_api=False,
+            lambda: ("", "", ""), None, [input_copyright, input_character, input_general], queue=False, api_visibility="undocumented",
         ).success(
             predict_tags_wd,
             [input_image, input_general, image_algorithms, general_threshold, character_threshold],
             [input_copyright, input_character, input_general, copy_input_btn],
-            show_api=False,
+            api_visibility="undocumented",
         ).success(
-            remove_specific_prompt, inputs=[input_general, keep_tags], outputs=[input_general], queue=False, show_api=False,
+            remove_specific_prompt, inputs=[input_general, keep_tags], outputs=[input_general], queue=False, api_visibility="undocumented",
         ).success(
-            convert_danbooru_to_e621_prompt, inputs=[input_general, input_tag_type], outputs=[input_general], queue=False, show_api=False,
+            convert_danbooru_to_e621_prompt, inputs=[input_general, input_tag_type], outputs=[input_general], queue=False, api_visibility="undocumented",
         ).success(
-            insert_recom_prompt, inputs=[input_general, dummy_np, recom_prompt], outputs=[input_general, dummy_np], queue=False, show_api=False,
+            insert_recom_prompt, inputs=[input_general, dummy_np, recom_prompt], outputs=[input_general, dummy_np], queue=False, api_visibility="undocumented",
         ).success(lambda: gr.update(interactive=True), None, [copy_prompt_btn_input], queue=False)
-        copy_input_btn.click(compose_prompt_to_copy, inputs=[input_character, input_copyright, input_general], outputs=[input_tags_to_copy], show_api=False)\
-            .success(gradio_copy_text, inputs=[input_tags_to_copy], js=COPY_ACTION_JS, show_api=False)
-        copy_prompt_btn_input.click(compose_prompt_to_copy, inputs=[input_character, input_copyright, input_general], outputs=[input_tags_to_copy], show_api=False)\
-            .success(gradio_copy_prompt, inputs=[input_tags_to_copy], outputs=[prompt_gui], show_api=False)
+        copy_input_btn.click(compose_prompt_to_copy, inputs=[input_character, input_copyright, input_general], outputs=[input_tags_to_copy], api_visibility="undocumented")\
+            .success(gradio_copy_text, inputs=[input_tags_to_copy], js=COPY_ACTION_JS, api_visibility="undocumented")
+        copy_prompt_btn_input.click(compose_prompt_to_copy, inputs=[input_character, input_copyright, input_general], outputs=[input_tags_to_copy], api_visibility="undocumented")\
+            .success(gradio_copy_prompt, inputs=[input_tags_to_copy], outputs=[prompt_gui], api_visibility="undocumented")
         
-        pick_random_character.click(select_random_character, [input_copyright, input_character], [input_copyright, input_character], show_api=False)
+        pick_random_character.click(select_random_character, [input_copyright, input_character], [input_copyright, input_character], api_visibility="undocumented")
 
         generate_btn.click(
             v2_upsampling_prompt,
             [model_name, input_copyright, input_character, input_general,
             input_rating, input_aspect_ratio, input_length, input_identity, input_ban_tags],
             [output_text],
-            show_api=False,
+            api_visibility="undocumented",
         ).success(
-            convert_danbooru_to_e621_prompt, inputs=[output_text, tag_type], outputs=[output_text_pony], queue=False, show_api=False,
+            convert_danbooru_to_e621_prompt, inputs=[output_text, tag_type], outputs=[output_text_pony], queue=False, api_visibility="undocumented",
         ).success(
-            insert_recom_prompt, inputs=[output_text, dummy_np, recom_animagine], outputs=[output_text, dummy_np], queue=False, show_api=False,
+            insert_recom_prompt, inputs=[output_text, dummy_np, recom_animagine], outputs=[output_text, dummy_np], queue=False, api_visibility="undocumented",
         ).success(
-            insert_recom_prompt, inputs=[output_text_pony, dummy_np, recom_pony], outputs=[output_text_pony, dummy_np], queue=False, show_api=False,
+            insert_recom_prompt, inputs=[output_text_pony, dummy_np, recom_pony], outputs=[output_text_pony, dummy_np], queue=False, api_visibility="undocumented",
         ).success(lambda: (gr.update(interactive=True), gr.update(interactive=True), gr.update(interactive=True), gr.update(interactive=True)),
-                   None, [copy_btn, copy_btn_pony, copy_prompt_btn, copy_prompt_btn_pony], queue=False, show_api=False)
-        copy_btn.click(gradio_copy_text, inputs=[output_text], js=COPY_ACTION_JS, show_api=False)
-        copy_btn_pony.click(gradio_copy_text, inputs=[output_text_pony], js=COPY_ACTION_JS, show_api=False)
-        copy_prompt_btn.click(gradio_copy_prompt, inputs=[output_text], outputs=[prompt_gui], show_api=False)
-        copy_prompt_btn_pony.click(gradio_copy_prompt, inputs=[output_text_pony], outputs=[prompt_gui], show_api=False)
+                   None, [copy_btn, copy_btn_pony, copy_prompt_btn, copy_prompt_btn_pony], queue=False, api_visibility="undocumented")
+        copy_btn.click(gradio_copy_text, inputs=[output_text], js=COPY_ACTION_JS, api_visibility="undocumented")
+        copy_btn_pony.click(gradio_copy_text, inputs=[output_text_pony], js=COPY_ACTION_JS, api_visibility="undocumented")
+        copy_prompt_btn.click(gradio_copy_prompt, inputs=[output_text], outputs=[prompt_gui], api_visibility="undocumented")
+        copy_prompt_btn_pony.click(gradio_copy_prompt, inputs=[output_text_pony], outputs=[prompt_gui], api_visibility="undocumented")
 
         from typing import Any, Dict, List, Optional, Tuple, Generator
         # 1) Helper: model loader (keeps existing behavior)
@@ -2285,8 +2284,8 @@ with gr.Blocks(theme=args.theme, elem_id="main", fill_width=True, fill_height=Fa
             yield from _generate_image(argv)
 
     # 5) Register two APIs with explicit signatures
-    gr.api(generate_image, api_name="generate_image", show_api=True, queue=True, concurrency_id="gpu")
-    gr.api(generate_image_stream, api_name="generate_image_stream", show_api=True, queue=True, concurrency_id="gpu")
+    gr.api(generate_image, api_name="generate_image", api_visibility="public", queue=True, concurrency_id="gpu")
+    gr.api(generate_image_stream, api_name="generate_image_stream", api_visibility="public", queue=True, concurrency_id="gpu")
 
     gr.DuplicateButton(value="Duplicate Space for private use (This demo does not work on CPU. Requires GPU Space)")
 
@@ -2300,5 +2299,7 @@ if __name__ == "__main__":
         ssr_mode=args.ssr,
         mcp_server=False,
         allowed_paths=[allowed_path],
+        theme=args.theme,
+        css=CSS,
     )
 ## END MOD

env.py

@@ -50,6 +50,8 @@ LOAD_DIFFUSERS_FORMAT_MODEL = [
     'Raelina/Raehoshi-illust-XL-8.1',
     'Raelina/Raehoshi-illust-XL-9',
     'Raelina/Raehoshi-illust-XL-9.1',
+    'Raelina/Raehoshi-illust-XL-10',
+    'Raelina/Raehoshi-illust-XL-11',
     'Raelina/Raehoshi-illust-vpred',
     'camenduru/FLUX.1-dev-diffusers',
     'black-forest-labs/FLUX.1-schnell',

modutils.py

@@ -218,6 +218,8 @@ CIVITAI_RESOLVE_NEGATIVE_CACHE: dict[str, str] = {}
 CIVITAI_VERSION_JSON_CACHE: dict[str, dict] = {}
 CIVITAI_VERSION_NEGATIVE_CACHE: dict[str, str] = {}
 CIVITAI_WGET_FRESH_RETRY_LIMIT = 1
+CIVITAI_METADATA_RECONNECT_ATTEMPTS = 3
+CIVITAI_METADATA_RECONNECT_BACKOFF = 0.8
 CIVITAI_API_PROBE_TIMEOUT = (3.0, 8.0)
 CIVITAI_API_RETRYABLE_STATUSES = frozenset([404, 405, 429, 500, 502, 503, 504])
 CIVITAI_ACTIVE_API_ORIGIN = ""
@@ -492,6 +494,20 @@ def extract_civitai_model_version_id(url: str):
         return ""
     return ""
 
+def extract_civitai_file_id(url: str):
+    try:
+        parts = get_civitai_url_parts(url)
+        qs = urllib.parse.parse_qs(parts.query)
+        for key, values in qs.items():
+            if str(key).casefold() != "fileid" or not values:
+                continue
+            value = str(values[0] or "").strip()
+            if value.isdigit():
+                return value
+    except Exception:
+        return ""
+    return ""
+
 def get_civitai_query_filters(url: str):
     try:
         parts = get_civitai_url_parts(url)
@@ -706,6 +722,16 @@ def pick_civitai_file_from_version_json(json_data, source_url: str = ""):
     files = json_data.get("files", []) if isinstance(json_data, dict) else []
     if not isinstance(files, list) or not files:
         return {}
+    explicit_file_id = extract_civitai_file_id(source_url)
+    if explicit_file_id:
+        for file_info in files:
+            if not isinstance(file_info, dict):
+                continue
+            candidate_id = str(file_info.get("id") or file_info.get("fileId") or "").strip()
+            if candidate_id == explicit_file_id:
+                return dict(file_info)
+        print(f"[civitai] explicit fileId={explicit_file_id} not present in model version metadata")
+        return {}
     version_id = str((json_data or {}).get("id") or "")
     filters = get_civitai_query_filters(source_url)
     candidates = []
@@ -786,41 +812,61 @@ def request_json_data(url, api_key: str = ""):
         return None
 
     endpoint_path = f"/model-versions/{model_version_id}"
-    headers = get_civitai_headers(effective_api_key)
-    session = create_retry_session()
+    last_error = None
+    for attempt in range(1, CIVITAI_METADATA_RECONNECT_ATTEMPTS + 1):
+        session = create_retry_session()
+        headers = get_civitai_headers(effective_api_key)
+        if attempt > 1:
+            headers["Connection"] = "close"
+        endpoint_url = ""
+        try:
+            json_data, endpoint_url, result = request_civitai_api_json(
+                endpoint_path,
+                headers=headers,
+                timeout=CIVITAI_METADATA_TIMEOUT,
+                api_key=effective_api_key,
+                session=session,
+                stream=True,
+                allow_not_found=True,
+            )
+            if result.status_code == 404:
+                print(f"Civitai metadata lookup status=404: {endpoint_url}")
+                cache_put(CIVITAI_VERSION_NEGATIVE_CACHE, model_version_id, "status=404")
+                return None
+            if not json_data:
+                print(f"Civitai metadata lookup returned empty JSON: {endpoint_url}")
+                cache_put(CIVITAI_VERSION_NEGATIVE_CACHE, model_version_id, "empty_json")
+                return None
+            cache_put(CIVITAI_VERSION_JSON_CACHE, model_version_id, copy.deepcopy(json_data))
+            if normalized_url and normalized_url != raw_url:
+                cache_put(CIVITAI_RESOLVE_CACHE, raw_url, normalized_url)
+            return json_data
+        except Exception as e:
+            last_error = e
+            print(
+                f"[civitai] metadata reconnect attempt={attempt}/{CIVITAI_METADATA_RECONNECT_ATTEMPTS} "
+                f"url={sanitize_url_for_log(endpoint_url or endpoint_path)} "
+                f"error={type(e).__name__}: {sanitize_sensitive_log_text(e)}"
+            )
+        finally:
+            try:
+                session.close()
+            except Exception:
+                pass
+        if attempt < CIVITAI_METADATA_RECONNECT_ATTEMPTS:
+            time.sleep(min(2.5, CIVITAI_METADATA_RECONNECT_BACKOFF * attempt))
 
-    try:
-        json_data, endpoint_url, result = request_civitai_api_json(
-            endpoint_path,
-            headers=headers,
-            timeout=CIVITAI_METADATA_TIMEOUT,
-            api_key=effective_api_key,
-            session=session,
-            stream=True,
-            allow_not_found=True,
-        )
-        if result.status_code == 404:
-            print(f"Civitai metadata lookup status=404: {endpoint_url}")
-            cache_put(CIVITAI_VERSION_NEGATIVE_CACHE, model_version_id, "status=404")
-            return None
-        if not json_data:
-            print(f"Civitai metadata lookup returned empty JSON: {endpoint_url}")
-            cache_put(CIVITAI_VERSION_NEGATIVE_CACHE, model_version_id, "empty_json")
-            return None
-        cache_put(CIVITAI_VERSION_JSON_CACHE, model_version_id, copy.deepcopy(json_data))
-        if normalized_url and normalized_url != raw_url:
-            cache_put(CIVITAI_RESOLVE_CACHE, raw_url, normalized_url)
-        return json_data
-    except Exception as e:
-        print(f"Civitai metadata lookup failed: {endpoint_url} {type(e).__name__}: {sanitize_sensitive_log_text(e)}")
-        return None
+    if last_error is not None:
+        print(f"Civitai metadata lookup failed after reconnects: {type(last_error).__name__}: {sanitize_sensitive_log_text(last_error)}")
+    return None
 
 class ModelInformation:
     def __init__(self, json_data, source_url: str = ""):
         selected_file = pick_civitai_file_from_version_json(json_data, source_url=source_url)
+        explicit_file_id = extract_civitai_file_id(source_url)
         self.model_version_id = json_data.get("id", "")
         self.model_id = json_data.get("modelId", "")
-        self.download_url = selected_file.get("downloadUrl", "") or json_data.get("downloadUrl", "")
+        self.download_url = selected_file.get("downloadUrl", "") or ("" if explicit_file_id else json_data.get("downloadUrl", ""))
         self.model_url = f"{get_civitai_canonical_web_origin()}/models/{self.model_id}?modelVersionId={self.model_version_id}"
         self.filename_url = selected_file.get("name", "") or ""
         self.description = json_data.get("description", "")
@@ -1005,7 +1051,41 @@ def guess_downloaded_file_path(directory, before_files, expected_filename=""):
 
     return None
 
-def get_existing_completed_download_path(directory, expected_filename=""):
+def get_civitai_expected_size_bytes(file_info):
+    if not isinstance(file_info, dict):
+        return 0
+    raw_size_kb = file_info.get("sizeKB")
+    if raw_size_kb is None:
+        raw_size_kb = file_info.get("sizeKb")
+    try:
+        size_kb = float(raw_size_kb)
+    except (TypeError, ValueError):
+        return 0
+    if size_kb <= 0:
+        return 0
+    return max(1, int(round(size_kb * 1024.0)))
+
+def is_civitai_file_complete(path, file_info=None, *, log_mismatch=False):
+    candidate = Path(path)
+    if not candidate.exists() or not candidate.is_file():
+        return False
+    expected_size = get_civitai_expected_size_bytes(file_info)
+    if expected_size <= 0:
+        return True
+    try:
+        actual_size = int(candidate.stat().st_size)
+    except OSError:
+        return False
+    tolerance = 4096
+    complete = abs(actual_size - expected_size) <= tolerance
+    if not complete and log_mismatch:
+        print(
+            f"[civitai] existing file size mismatch; treating as incomplete "
+            f"path={candidate} actual={actual_size} expected={expected_size}"
+        )
+    return complete
+
+def get_existing_completed_download_path(directory, expected_filename="", file_info=None):
     expected_name = str(expected_filename or "").strip()
     if not expected_name:
         return ""
@@ -1021,7 +1101,7 @@ def get_existing_completed_download_path(directory, expected_filename=""):
             candidate_paths.append(legacy_nested_path)
 
     for candidate_path in candidate_paths:
-        if candidate_path.exists() and candidate_path.is_file():
+        if is_civitai_file_complete(candidate_path, file_info=file_info, log_mismatch=True):
             return str(candidate_path)
 
     return ""
@@ -1064,6 +1144,7 @@ def download_things(directory, url, hf_token="", civitai_api_key="", romanize=Fa
         if normalized_url != url:
             print(f"Civitai download URL normalized: {sanitize_url_for_log(url)} -> {sanitize_url_for_log(normalized_url)}")
         model_profile = retrieve_model_info(normalized_url, api_key=civitai_api_key)
+        selected_file = model_profile.selected_file if model_profile else {}
         if model_profile and model_profile.download_url:
             url = model_profile.download_url
             filename = model_profile.filename_url or ""
@@ -1087,7 +1168,7 @@ def download_things(directory, url, hf_token="", civitai_api_key="", romanize=Fa
         print(f"Filename: {filename}")
         print(f"[civitai] resolved signed host={signed_host or '-'} url={sanitize_url_for_log(url)}")
 
-        existing_completed_path = get_existing_completed_download_path(directory, expected_filename=filename)
+        existing_completed_path = get_existing_completed_download_path(directory, expected_filename=filename, file_info=selected_file)
         if existing_completed_path:
             print(f"[civitai] using existing completed file path={existing_completed_path}")
             downloaded_file_path = existing_completed_path
@@ -1139,10 +1220,14 @@ def download_things(directory, url, hf_token="", civitai_api_key="", romanize=Fa
             if download_status != 0:
                 log_download_error("civitai", "command_failed", url=url, status=download_status)
 
+        if not downloaded_file_path and download_status == 0:
+            candidate_path = guess_downloaded_file_path(directory, before_files, expected_filename=filename)
+            if candidate_path and is_civitai_file_complete(candidate_path, file_info=selected_file, log_mismatch=True):
+                downloaded_file_path = candidate_path
+            elif candidate_path:
+                log_download_error("civitai", "size_mismatch", url=url, detail=f"path={candidate_path}")
         if not downloaded_file_path:
-            downloaded_file_path = guess_downloaded_file_path(directory, before_files, expected_filename=filename)
-        if not downloaded_file_path:
-            existing_completed_path = get_existing_completed_download_path(directory, expected_filename=filename)
+            existing_completed_path = get_existing_completed_download_path(directory, expected_filename=filename, file_info=selected_file)
             if existing_completed_path:
                 print(f"[civitai] using existing completed file path={existing_completed_path}")
                 downloaded_file_path = existing_completed_path
@@ -1773,7 +1858,6 @@ def get_lora_info(lora_path: str):
     label = ""
     md = "None"
     if not lora_path or lora_path == "None":
-        print("LoRA file not found.")
         return is_valid, label, tag, md
     path = Path(lora_path)
     new_path = Path(f'{path.parent.name}/{escape_lora_basename(path.stem)}{path.suffix}')

requirements.txt

@@ -1,22 +1,23 @@
-stablepy==0.6.5
-diffusers
-transformers
-accelerate
-huggingface_hub
-spaces
-torch==2.8.0
-numpy<2
-gdown
-opencv-python
-optimum[onnxruntime]
-#dartrs
-git+https://github.com/John6666cat/dartrs
-translatepy
-timm
-rapidfuzz
-sentencepiece
-unidecode
-matplotlib-inline
-mediapipe==0.10.5
-einops
-# pydantic==2.10.6
+stablepy==0.6.5
+diffusers
+transformers>=4.47.1,<5,!=4.57.0
+accelerate
+huggingface_hub
+spaces
+torch==2.8.0
+numpy<2
+gdown
+opencv-python
+#dartrs
+git+https://github.com/John6666cat/dartrs
+translatepy
+timm
+rapidfuzz
+pandas
+safetensors
+sentencepiece
+unidecode
+matplotlib-inline
+mediapipe==0.10.13
+einops
+# pydantic==2.10.6