Upload 5 files

app.py

@@ -138,9 +138,13 @@ with gr.Blocks(fill_width=True, elem_id="container", css=css, delete_cache=(60,
                                 lora5_copy = gr.Button(value="Copy example to prompt", visible=False)
                             lora5_md = gr.Markdown(value="", visible=False)
                     with gr.Accordion("From URL", open=True, visible=True):
+                        with gr.Row():
+                            lora_search_civitai_basemodel = gr.CheckboxGroup(label="Search LoRA for", choices=["Pony", "SD 1.5", "SDXL 1.0", "Flux.1 D", "Flux.1 S"], value=["Pony", "SDXL 1.0"])
+                            lora_search_civitai_sort = gr.Radio(label="Sort", choices=["Highest Rated", "Most Downloaded", "Newest"], value="Highest Rated")
+                            lora_search_civitai_period = gr.Radio(label="Period", choices=["AllTime", "Year", "Month", "Week", "Day"], value="AllTime")
                         with gr.Row():
                             lora_search_civitai_query = gr.Textbox(label="Query", placeholder="oomuro sakurako...", lines=1)
-                            lora_search_civitai_basemodel = gr.CheckboxGroup(label="Search LoRA for", choices=["Pony", "SD 1.5", "SDXL 1.0"], value=["Pony", "SDXL 1.0"])
+                            lora_search_civitai_tag = gr.Textbox(label="Tag", lines=1)
                             lora_search_civitai_submit = gr.Button("Search on Civitai")
                         with gr.Row():
                             lora_search_civitai_result = gr.Dropdown(label="Search Results", choices=[("", "")], value="", allow_custom_value=True, visible=False)
@@ -247,9 +251,9 @@ with gr.Blocks(fill_width=True, elem_id="container", css=css, delete_cache=(60,
     lora5_copy.click(apply_lora_prompt, [prompt, lora5_info], [prompt], queue=False, show_api=False)
 
     gr.on(
-        triggers=[lora_search_civitai_submit.click, lora_search_civitai_query.submit],
+        triggers=[lora_search_civitai_submit.click, lora_search_civitai_query.submit, lora_search_civitai_tag.submit],
         fn=search_civitai_lora,
-        inputs=[lora_search_civitai_query, lora_search_civitai_basemodel],
+        inputs=[lora_search_civitai_query, lora_search_civitai_basemodel, lora_search_civitai_sort, lora_search_civitai_period, lora_search_civitai_tag],
         outputs=[lora_search_civitai_result, lora_search_civitai_desc, lora_search_civitai_submit, lora_search_civitai_query],
         scroll_to_output=True,
         queue=True,

dc.py

@@ -21,12 +21,9 @@ from stablepy import (
     SD15_TASKS,
     SDXL_TASKS,
 )
+import time
 #import urllib.parse
 import gradio as gr
-from PIL import Image
-import IPython.display
-import time, json
-from IPython.utils import capture
 import logging
 logging.getLogger("diffusers").setLevel(logging.ERROR)
 import diffusers
@@ -381,7 +378,7 @@ class GuiSD:
         if vae_model:
             vae_type = "SDXL" if "sdxl" in vae_model.lower() else "SD 1.5"
             if model_type != vae_type:
-                gr.Info(msg_inc_vae)
+                gr.Warning(msg_inc_vae)
 
         self.model.device = torch.device("cpu")
         dtype_model = torch.bfloat16 if model_type == "FLUX" else torch.float16
@@ -395,7 +392,7 @@ class GuiSD:
         )
         #yield f"Model loaded: {model_name}"
 
-    @spaces.GPU
+    #@spaces.GPU
     @torch.inference_mode()
     def generate_pipeline(
         self,
@@ -508,7 +505,7 @@ class GuiSD:
         vae_model = vae_model if vae_model != "None" else None
         loras_list = [lora1, lora2, lora3, lora4, lora5]
         vae_msg = f"VAE: {vae_model}" if vae_model else ""
-        msg_lora = []
+        msg_lora = ""
 
         print("Config model:", model_name, vae_model, loras_list)
 
@@ -679,35 +676,94 @@ class GuiSD:
         return self.infer_short(self.model, pipe_params, progress), info_state
 ## END MOD
 
-# def sd_gen_generate_pipeline(*args):
+def dynamic_gpu_duration(func, duration, *args):
+
+    @spaces.GPU(duration=duration)
+    def wrapped_func():
+        yield from func(*args)
+
+    return wrapped_func()
+
+
+@spaces.GPU
+def dummy_gpu():
+    return None
+
+
+def sd_gen_generate_pipeline(*args):
+
+    gpu_duration_arg = int(args[-1]) if args[-1] else 59
+    verbose_arg = int(args[-2])
+    load_lora_cpu = args[-3]
+    generation_args = args[:-3]
+    lora_list = [
+        None if item == "None" or item == "" else item
+        for item in [args[7], args[9], args[11], args[13], args[15]]
+    ]
+    lora_status = [None] * 5
+
+    msg_load_lora = "Updating LoRAs in GPU..."
+    if load_lora_cpu:
+        msg_load_lora = "Updating LoRAs in CPU (Slow but saves GPU usage)..."
+
+    if lora_list != sd_gen.model.lora_memory and lora_list != [None] * 5:
+        yield None, msg_load_lora
+
+    # Load lora in CPU
+    if load_lora_cpu:
+        lora_status = sd_gen.model.lora_merge(
+            lora_A=lora_list[0], lora_scale_A=args[8],
+            lora_B=lora_list[1], lora_scale_B=args[10],
+            lora_C=lora_list[2], lora_scale_C=args[12],
+            lora_D=lora_list[3], lora_scale_D=args[14],
+            lora_E=lora_list[4], lora_scale_E=args[16],
+        )
+        print(lora_status)
+
+    if verbose_arg:
+        for status, lora in zip(lora_status, lora_list):
+            if status:
+                gr.Info(f"LoRA loaded in CPU: {lora}")
+            elif status is not None:
+                gr.Warning(f"Failed to load LoRA: {lora}")
+
+        if lora_status == [None] * 5 and sd_gen.model.lora_memory != [None] * 5 and load_lora_cpu:
+            lora_cache_msg = ", ".join(
+                str(x) for x in sd_gen.model.lora_memory if x is not None
+            )
+            gr.Info(f"LoRAs in cache: {lora_cache_msg}")
 
-#     # Load lora in CPU
-#     status_lora = sd_gen.model.lora_merge(
-#         lora_A=args[7] if args[7] != "None" else None, lora_scale_A=args[8],
-#         lora_B=args[9] if args[9] != "None" else None, lora_scale_B=args[10],
-#         lora_C=args[11] if args[11] != "None" else None, lora_scale_C=args[12],
-#         lora_D=args[13] if args[13] != "None" else None, lora_scale_D=args[14],
-#         lora_E=args[15] if args[15] != "None" else None, lora_scale_E=args[16],
-#     )
+        msg_request = f"Requesting {gpu_duration_arg}s. of GPU time"
+        gr.Info(msg_request)
+        print(msg_request)
 
-#     lora_list = [args[7], args[9], args[11], args[13], args[15]]
-#     print(status_lora)
-#     for status, lora in zip(status_lora, lora_list):
-#         if status:
-#             gr.Info(f"LoRA loaded: {lora}")
-#         elif status is not None:
-#             gr.Warning(f"Failed to load LoRA: {lora}")
+    # yield from sd_gen.generate_pipeline(*generation_args)
 
-#     # if status_lora == [None] * 5 and self.model.lora_memory != [None] * 5:
-#     #     gr.Info(f"LoRAs in cache: {", ".join(str(x) for x in self.model.lora_memory if x is not None)}")
+    start_time = time.time()
 
-#     yield from sd_gen.generate_pipeline(*args)
+    yield from dynamic_gpu_duration(
+        sd_gen.generate_pipeline,
+        gpu_duration_arg,
+        *generation_args,
+    )
+
+    end_time = time.time()
 
+    if verbose_arg:
+        execution_time = end_time - start_time
+        msg_task_complete = (
+            f"GPU task complete in: {round(execution_time, 0) + 1} seconds"
+        )
+        gr.Info(msg_task_complete)
+        print(msg_task_complete)
 
-# sd_gen_generate_pipeline.zerogpu = True
+
+dynamic_gpu_duration.zerogpu = True
+sd_gen_generate_pipeline.zerogpu = True
 
 from pathlib import Path
-import random
+from PIL import Image
+import random, json
 from modutils import (safe_float, escape_lora_basename, to_lora_key, to_lora_path,
     get_local_model_list, get_private_lora_model_lists, get_valid_lora_name,
     get_valid_lora_path, get_valid_lora_wt, get_lora_info,
@@ -723,6 +779,10 @@ def infer(prompt, negative_prompt, seed, randomize_seed, width, height, guidance
     import numpy as np
     MAX_SEED = np.iinfo(np.int32).max
 
+    load_lora_cpu = False
+    verbose_info = False
+    gpu_duration = 59
+
     images: list[tuple[PIL.Image.Image, str | None]] = []
     info: str = ""
     progress(0, desc="Preparing...")
@@ -739,7 +799,7 @@ def infer(prompt, negative_prompt, seed, randomize_seed, width, height, guidance
     prompt, negative_prompt = insert_model_recom_prompt(prompt, negative_prompt, model_name)
     progress(0.5, desc="Preparing...")
     lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora4, lora4_wt, lora5, lora5_wt = \
-        set_prompt_loras(prompt, lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora4, lora4_wt, lora5, lora5_wt)
+        set_prompt_loras(prompt, model_name, lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora4, lora4_wt, lora5, lora5_wt)
     lora1 = get_valid_lora_path(lora1)
     lora2 = get_valid_lora_path(lora2)
     lora3 = get_valid_lora_path(lora3)
@@ -748,7 +808,7 @@ def infer(prompt, negative_prompt, seed, randomize_seed, width, height, guidance
     progress(1, desc="Preparation completed. Starting inference preparation...")
 
     sd_gen.load_new_model(model_name, vae, TASK_MODEL_LIST[0], progress)
-    images, info = sd_gen.generate_pipeline(prompt, negative_prompt, 1, num_inference_steps,
+    images, info = sd_gen_generate_pipeline(prompt, negative_prompt, 1, num_inference_steps,
         guidance_scale, True, generator, lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt,
         lora4, lora4_wt, lora5, lora5_wt, sampler,
         height, width, model_name, vae, TASK_MODEL_LIST[0], None, "Canny", 512, 1024,
@@ -757,7 +817,7 @@ def infer(prompt, negative_prompt, seed, randomize_seed, width, height, guidance
         False, True, 1, True, False, False, False, False, "./images", False, False, False, True, 1, 0.55,
         False, False, False, True, False, "Use same sampler", False, "", "", 0.35, True, True, False, 4, 4, 32,
         False, "", "", 0.35, True, True, False, 4, 4, 32,
-        True, None, None, "plus_face", "original", 0.7, None, None, "base", "style", 0.7, 0.0, progress
+        True, None, None, "plus_face", "original", 0.7, None, None, "base", "style", 0.7, 0.0, load_lora_cpu, verbose_info, gpu_duration, progress
     )
 
     progress(1, desc="Inference completed.")
@@ -820,7 +880,7 @@ def get_t2i_model_info(repo_id: str):
         if " " in repo_id or not api.repo_exists(repo_id): return ""
         model = api.model_info(repo_id=repo_id)
     except Exception as e:
-        print(f"Error: Failed to get {repo_id}'s info. ")
+        print(f"Error: Failed to get {repo_id}'s info. {e}")
         return ""
     if model.private or model.gated: return ""
     tags = model.tags
@@ -1013,13 +1073,13 @@ def download_my_lora(dl_urls: str, lora1: str, lora2: str, lora3: str, lora4: st
         gr.update(value=lora4, choices=choices), gr.update(value=lora5, choices=choices)
 
 
-def set_prompt_loras(prompt, lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora4, lora4_wt, lora5, lora5_wt):
+def set_prompt_loras(prompt, model_name, lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora4, lora4_wt, lora5, lora5_wt):
     import re
-    lora1 = get_valid_lora_name(lora1)
-    lora2 = get_valid_lora_name(lora2)
-    lora3 = get_valid_lora_name(lora3)
-    lora4 = get_valid_lora_name(lora4)
-    lora5 = get_valid_lora_name(lora5)
+    lora1 = get_valid_lora_name(lora1, model_name)
+    lora2 = get_valid_lora_name(lora2, model_name)
+    lora3 = get_valid_lora_name(lora3, model_name)
+    lora4 = get_valid_lora_name(lora4, model_name)
+    lora5 = get_valid_lora_name(lora5, model_name)
     if not "<lora" in prompt: return lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora4, lora4_wt, lora5, lora5_wt
     lora1_wt = get_valid_lora_wt(prompt, lora1, lora1_wt)
     lora2_wt = get_valid_lora_wt(prompt, lora2, lora2_wt)
@@ -1129,9 +1189,9 @@ def update_loras(prompt, lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora
      gr.update(value=tag5, label=label5, visible=on5), gr.update(visible=on5), gr.update(value=md5, visible=on5)
 
 
-def search_civitai_lora(query, base_model):
+def search_civitai_lora(query, base_model, sort="Highest Rated", period="AllTime", tag=""):
     global civitai_lora_last_results
-    items = search_lora_on_civitai(query, base_model)
+    items = search_lora_on_civitai(query, base_model, 100, sort, period, tag)
     if not items: return gr.update(choices=[("", "")], value="", visible=False),\
           gr.update(value="", visible=False), gr.update(visible=True), gr.update(visible=True)
     civitai_lora_last_results = {}
@@ -1324,7 +1384,6 @@ def process_style_prompt(prompt: str, neg_prompt: str, styles_key: str = "None",
     return gr.update(value=prompt), gr.update(value=neg_prompt)
 
 
-from PIL import Image
 def save_images(images: list[Image.Image], metadatas: list[str]):
     from PIL import PngImagePlugin
     try:

env.py

@@ -102,6 +102,11 @@ load_diffusers_format_model = [
     "Raelina/Raemu-Flux",
 ]
 
+DIFFUSERS_FORMAT_LORAS = [
+    "nerijs/animation2k-flux",
+    "XLabs-AI/flux-RealismLora",
+]
+
 # List all Models for specified user
 HF_MODEL_USER_LIKES = ["votepurchase"] # sorted by number of likes
 HF_MODEL_USER_EX = ["John6666"] # sorted by a special rule

llmdolphin.py

@@ -7,6 +7,7 @@ from llama_cpp_agent.chat_history import BasicChatHistory
 from llama_cpp_agent.chat_history.messages import Roles
 from ja_to_danbooru.ja_to_danbooru import jatags_to_danbooru_tags
 import wrapt_timeout_decorator
+from pathlib import Path
 
 
 llm_models_dir = "./llm_models"
@@ -19,6 +20,8 @@ llm_models = {
     #"": ["", MessagesFormatterType.PHI_3],
     "mn-12b-lyra-v2a1-q5_k_m.gguf": ["HalleyStarbun/MN-12B-Lyra-v2a1-Q5_K_M-GGUF", MessagesFormatterType.CHATML],
     "L3-8B-Tamamo-v1.i1-Q5_K_M.gguf": ["mradermacher/L3-8B-Tamamo-v1-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "MN-12B-Mag-Mell-Q4_K_M.gguf": ["inflatebot/MN-12B-Mag-Mell-R1-GGUF", MessagesFormatterType.MISTRAL],
+    "Qwen2.5-14B-Instruct-Q4_K_M.gguf": ["bartowski/Qwen2.5-14B-Instruct-GGUF", MessagesFormatterType.OPEN_CHAT],
     "Instant-RP-Noodles-12B-v1.3.Q4_K_M.gguf": ["mradermacher/Instant-RP-Noodles-12B-v1.3-GGUF", MessagesFormatterType.MISTRAL],
     "MN-12B-Lyra-v4-Q4_K_M.gguf": ["bartowski/MN-12B-Lyra-v4-GGUF", MessagesFormatterType.CHATML],
     "Lyra4-Gutenberg-12B.Q4_K_M.gguf": ["mradermacher/Lyra4-Gutenberg-12B-GGUF", MessagesFormatterType.CHATML],
@@ -50,9 +53,108 @@ llm_models = {
     "StarDust-12b-v2.i1-Q5_K_M.gguf": ["mradermacher/StarDust-12b-v2-i1-GGUF", MessagesFormatterType.CHATML],
     "Rocinante-12B-v2c-Q4_K_M.gguf": ["TheDrummer/UnslopNemo-v1-GGUF", MessagesFormatterType.MISTRAL],
     "mn-maghin-12b-q6_k.gguf": ["rityak/MN-Maghin-12B-Q6_K-GGUF", MessagesFormatterType.MISTRAL],
+    "qwen2.5-lumen-14b-q4_k_m.gguf": ["Lambent/Qwen2.5-Lumen-14B-Q4_K_M-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "Qwen2.5-14B_Uncensored_Instruct.Q4_K_M.gguf": ["mradermacher/Qwen2.5-14B_Uncensored_Instruct-GGUF", MessagesFormatterType.OPEN_CHAT],
     "Trinas_Nectar-8B-model_stock.i1-Q4_K_M.gguf": ["mradermacher/Trinas_Nectar-8B-model_stock-i1-GGUF", MessagesFormatterType.MISTRAL],
     "ChatWaifu_v1.4.Q5_K_M.gguf": ["mradermacher/ChatWaifu_v1.4-GGUF", MessagesFormatterType.MISTRAL],
     "ChatWaifu_v1.3.1.Q4_K_M.gguf": ["mradermacher/ChatWaifu_v1.3.1-GGUF", MessagesFormatterType.MISTRAL],
+    "Collaiborator-MEDLLM-Llama-3-8B-v1.i1-Q5_K_M.gguf": ["mradermacher/Collaiborator-MEDLLM-Llama-3-8B-v1-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "Chili_Dog_8B.i1-Q4_K_M.gguf": ["mradermacher/Chili_Dog_8B-i1-GGUF", MessagesFormatterType.CHATML],
+    "astra-v1-12b-q5_k_m.gguf": ["P0x0/Astra-v1-12B-GGUF", MessagesFormatterType.MISTRAL],
+    "Llama-3.1-8B-TitanFusion-Mix.Q4_K_S.gguf": ["mradermacher/Llama-3.1-8B-TitanFusion-Mix-GGUF", MessagesFormatterType.LLAMA_3],
+    "Qwen2.5-7B-Instruct-kowiki-qa-Q5_K_M.gguf": ["teddylee777/Qwen2.5-7B-Instruct-kowiki-qa-gguf", MessagesFormatterType.OPEN_CHAT],
+    "Qwen2.5-7B-Instruct-abliterated-v2.Q5_K_M.gguf": ["mradermacher/Qwen2.5-7B-Instruct-abliterated-v2-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "Mistral-Small-Instruct-2409-abliterated.Q4_K_S.gguf": ["mradermacher/Mistral-Small-Instruct-2409-abliterated-GGUF", MessagesFormatterType.MISTRAL],
+    "Qwen2.5-14b-web.Q4_K_S.gguf": ["mradermacher/Qwen2.5-14b-web-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "EVA-Yi-1.5-9B-32K-V1.i1-Q4_K_M.gguf": ["mradermacher/EVA-Yi-1.5-9B-32K-V1-i1-GGUF", MessagesFormatterType.MISTRAL],
+    "Eunoia_Vespera-8B-LINEAR.i1-Q5_K_M.gguf": ["mradermacher/Eunoia_Vespera-8B-LINEAR-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "fatgirlmagicv2.i1-Q4_K_M.gguf": ["mradermacher/fatgirlmagicv2-i1-GGUF", MessagesFormatterType.CHATML],
+    "Pullulation-2-9B.Q4_K_S.gguf": ["mradermacher/Pullulation-2-9B-GGUF", MessagesFormatterType.ALPACA],
+    "Neithabet-9G.Q4_K_M.gguf": ["mradermacher/Neithabet-9G-GGUF", MessagesFormatterType.ALPACA],
+    "magicfatgirlv2.Q4_K_M.gguf": ["mradermacher/magicfatgirlv2-GGUF", MessagesFormatterType.CHATML],
+    "Llama-3.1-8B-TitanFusion-v3.Q5_K_M.gguf": ["mradermacher/Llama-3.1-8B-TitanFusion-v3-GGUF", MessagesFormatterType.LLAMA_3],
+    "Fatgirl_v2_8B.Q5_K_M.gguf": ["mradermacher/Fatgirl_v2_8B-GGUF", MessagesFormatterType.CHATML],
+    "wip-test_pending_4-q4_k_m.gguf": ["DreadPoor/WIP-TEST_PENDING_4-Q4_K_M-GGUF", MessagesFormatterType.LLAMA_3],
+    "himeyuri12b-01-Q4_K_M.gguf": ["Elizezen/Himeyuri-v0.1-12B-GGUF", MessagesFormatterType.MISTRAL],
+    "Qwen2.5-14B-Gutenberg-1e-Delta.Q4_K_M.gguf": ["QuantFactory/Qwen2.5-14B-Gutenberg-1e-Delta-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "mn-12b-stellar-veil-q6_k.gguf": ["v000000/MN-12B-Stellar-Veil-Q6_K-GGUF", MessagesFormatterType.MISTRAL],
+    "stock.g2mma-q4_k_m.gguf": ["ClaudioItaly/Stock.G2mma-Q4_K_M-GGUF", MessagesFormatterType.ALPACA],
+    "Athena-gemma-2-9b-it-Philos-KTO.i1-Q4_K_M.gguf": ["mradermacher/Athena-gemma-2-9b-it-Philos-KTO-i1-GGUF", MessagesFormatterType.ALPACA],
+    "Qwen2.5-14B-Gutenberg-Instruct-Slerpeno.Q4_K_M.gguf": ["mradermacher/Qwen2.5-14B-Gutenberg-Instruct-Slerpeno-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "typressai-9b-q4_k_m.gguf": ["ClaudioItaly/TypressAI-9B-Q4_K_M-GGUF", MessagesFormatterType.ALPACA],
+    "josiefied-qwen2.5-7b-instruct-abliterated-v2.Q5_K_M.gguf": ["Isaak-Carter/Josiefied-Qwen2.5-7B-Instruct-abliterated-v2-gguf", MessagesFormatterType.OPEN_CHAT],
+    "josiefied-qwen2.5-7b-instruct-abliterated.Q5_K_M.gguf": ["Isaak-Carter/Josiefied-Qwen2.5-7B-Instruct-abliterated-gguf", MessagesFormatterType.OPEN_CHAT],
+    "Fireball-Llama-3.1-8B-Philos-Reflection-v0.2.i1-Q5_K_M.gguf": ["mradermacher/Fireball-Llama-3.1-8B-Philos-Reflection-v0.2-i1-GGUF", MessagesFormatterType.MISTRAL],
+    "ArliAI-RPMax-Formax-v1.Q5_K_M.gguf": ["mradermacher/ArliAI-RPMax-Formax-v1-GGUF", MessagesFormatterType.LLAMA_3],
+    "Agente-Llama-3.1-Asistant-16bit-v2.Q5_K_M.gguf": ["mradermacher/Agente-Llama-3.1-Asistant-16bit-v2-GGUF", MessagesFormatterType.LLAMA_3],
+    "Magnum-Blackout-Ataraxy-4-9b.Q4_K_M.gguf": ["mradermacher/Magnum-Blackout-Ataraxy-4-9b-GGUF", MessagesFormatterType.ALPACA],
+    "Magnum-Blackout-Ataraxy-5-9b.Q4_K_M.gguf": ["mradermacher/Magnum-Blackout-Ataraxy-5-9b-GGUF", MessagesFormatterType.ALPACA],
+    "Vapor_7B.Q5_K_M.gguf": ["mradermacher/Vapor_7B-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "fusion-guide-12b-0.1.Q4_K_M.gguf": ["mradermacher/fusion-guide-12b-0.1-GGUF", MessagesFormatterType.MISTRAL],
+    "Meta-Llama-3.1-8B-Instruct-HalfAbliterated-TIES.i1-Q4_K_M.gguf": ["mradermacher/Meta-Llama-3.1-8B-Instruct-HalfAbliterated-TIES-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "Qwen2.5-14B_Uncencored-Q4_K_M.gguf": ["bartowski/Qwen2.5-14B_Uncencored-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "L3.1-Niitorm-8B-DPO-t0.0001.i1-Q5_K_M.gguf": ["mradermacher/L3.1-Niitorm-8B-DPO-t0.0001-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "Hermes-ClimateStorm-Sauerkraut-abliterated.i1-Q5_K_M.gguf": ["mradermacher/Hermes-ClimateStorm-Sauerkraut-abliterated-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "qwen2.5-boosted-q5_k_m.gguf": ["ClaudioItaly/Qwen2.5-Boosted-Q5_K_M-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "qwen2.5-14b-q4_k_m.gguf": ["Triangle104/Qwen2.5-14B-Q4_K_M-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "Qwen2.5-7B-Instruct-Q4_K_M.gguf": ["bartowski/Qwen2.5-7B-Instruct-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "Hermes-Storm-lorablated.i1-Q5_K_M.gguf": ["mradermacher/Hermes-Storm-lorablated-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "Gemma-Ataraxy-Dare-NoBase-9b.i1-Q4_K_M.gguf": ["mradermacher/Gemma-Ataraxy-Dare-NoBase-9b-i1-GGUF", MessagesFormatterType.ALPACA],
+    "Axolotl-Llama-3.1-8B-instruct-finetuned-V3-merged.Q4_K_M.gguf": ["mradermacher/Axolotl-Llama-3.1-8B-instruct-finetuned-V3-merged-GGUF", MessagesFormatterType.LLAMA_3],
+    "Llama-3.1-8B-TitanFusion-v2.i1-Q5_K_M.gguf": ["mradermacher/Llama-3.1-8B-TitanFusion-v2-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "Magnum-Blackout-Ataraxy-2-9b.Q4_K_M.gguf": ["mradermacher/Magnum-Blackout-Ataraxy-2-9b-GGUF", MessagesFormatterType.ALPACA],
+    "Aspire-8B-model_stock.i1-Q5_K_M.gguf": ["mradermacher/Aspire-8B-model_stock-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "L3.1-Celestial-Stone-2x8B-DPO.Q4_K_M.gguf": ["mradermacher/L3.1-Celestial-Stone-2x8B-DPO-GGUF", MessagesFormatterType.LLAMA_3],
+    "L3-Dark-Planet-8B-V2-Eight-Orbs-Of-Power.i1-Q4_K_M.gguf": ["mradermacher/L3-Dark-Planet-8B-V2-Eight-Orbs-Of-Power-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "l3-luna-8b-q5_k_m-imat.gguf": ["Casual-Autopsy/L3-Luna-8B-Q5_K_M-GGUF", MessagesFormatterType.LLAMA_3],
+    "Insanity.i1-Q4_K_M.gguf": ["mradermacher/Insanity-i1-GGUF", MessagesFormatterType.MISTRAL],
+    "Rocinante-12B-v2d-Q4_K_M.gguf": ["TheDrummer/UnslopNemo-v2-GGUF", MessagesFormatterType.MISTRAL],
+    "Fatgirl_8B.Q4_K_M.gguf": ["QuantFactory/Fatgirl_8B-GGUF", MessagesFormatterType.MISTRAL],
+    "stormclimate-q4_k_m.gguf": ["MotherEarth/stormclimate-Q4_K_M-GGUF", MessagesFormatterType.LLAMA_3],
+    "InsanityB.Q4_K_M.gguf": ["mradermacher/InsanityB-GGUF", MessagesFormatterType.MISTRAL],
+    "mistral-nemo-gutades-12B.Q4_K_M.gguf": ["mradermacher/mistral-nemo-gutades-12B-GGUF", MessagesFormatterType.MISTRAL],
+    "FourFictionGemma-9.Q5_K_M.gguf": ["mradermacher/FourFictionGemma-9-GGUF", MessagesFormatterType.ALPACA],
+    "blackmagnumataraxy-9b-q6_k.gguf": ["BarBarickoza/BlackMagnumAtaraxy-9B-Q6_K-GGUF", MessagesFormatterType.ALPACA],
+    "fourfastgemma-9-q4_k_m-imat.gguf": ["ClaudioItaly/FourFastGemma-9-Q4_K_M-GGUF", MessagesFormatterType.ALPACA],
+    "ThinkingMistral-gen.i1-Q4_K_M.gguf": ["mradermacher/ThinkingMistral-gen-i1-GGUF", MessagesFormatterType.MISTRAL],
+    "L3.1-Dark-Planet-10.7B-ExxxxxxxxTended-D_AU-Q4_k_m.gguf": ["DavidAU/L3.1-Dark-Planet-10.7B-ExxxxxxxxTended-GGUF", MessagesFormatterType.LLAMA_3],
+    "Gemma-The-Writer-9B-D_AU-Q4_k_m.gguf": ["DavidAU/Gemma-The-Writer-9B-GGUF", MessagesFormatterType.ALPACA],
+    "CleverBoi-Nemo-12B-v2.i1-Q4_K_M.gguf": ["mradermacher/CleverBoi-Nemo-12B-v2-i1-GGUF", MessagesFormatterType.MISTRAL],
+    "ORPO-EdgeRunner-Tactical-7B-GSM8K.Q5_K_M.gguf": ["mradermacher/ORPO-EdgeRunner-Tactical-7B-GSM8K-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "wip-test_pending_c-q4_k_m.gguf": ["DreadPoor/WIP-TEST_PENDING_C-Q4_K_M-GGUF", MessagesFormatterType.LLAMA_3],
+    "L3.1-8B-komorebi-q8_0.gguf": ["crestf411/L3.1-8B-komorebi-gguf", MessagesFormatterType.LLAMA_3],
+    "Canada_Reimbursement_Model.Q5_K_M.gguf": ["mradermacher/Canada_Reimbursement_Model-GGUF", MessagesFormatterType.LLAMA_3],
+    "BuddyGlassNeverSleeps.Q5_K_M.gguf": ["mradermacher/BuddyGlassNeverSleeps-GGUF", MessagesFormatterType.LLAMA_3],
+    "gemma-writer-stock-no-ifable-9b-q6_k.gguf": ["BarBarickoza/Gemma-writer-stock-no-Ifable-9b-Q6_K-GGUF", MessagesFormatterType.ALPACA],
+    "Gemma-Ataraxy-Dare-9b.i1-Q4_K_M.gguf": ["mradermacher/Gemma-Ataraxy-Dare-9b-i1-GGUF", MessagesFormatterType.ALPACA],
+    "MotherEarth-Hermes-8B.i1-Q5_K_M.gguf": ["mradermacher/MotherEarth-Hermes-8B-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "albacus-q5_k_m-imat.gguf": ["ClaudioItaly/Albacus-V2-Imatrix", MessagesFormatterType.MISTRAL],
+    "mn-12b-siskin-test2-q8_0.gguf": ["Nohobby/MN-12B-Siskin-TEST2-Q8_0-GGUF", MessagesFormatterType.MISTRAL],
+    "mn-12b-siskin-test3-q8_0.gguf": ["Nohobby/MN-12B-Siskin-TEST3-Q8_0-GGUF", MessagesFormatterType.MISTRAL],
+    "Violet_Twilight-v0.2.Q4_K_M.gguf": ["Epiculous/Violet_Twilight-v0.2-GGUF", MessagesFormatterType.CHATML],
+    "albacus-q4_k_m-imat.gguf": ["ClaudioItaly/Albacus-Imatrix-Q4_K_M-GGUF", MessagesFormatterType.MISTRAL],
+    "mn-chinofun-q4_k_m.gguf": ["djuna/MN-Chinofun-Q4_K_M-GGUF", MessagesFormatterType.MISTRAL],
+    "NarraThinker12B.i1-Q4_K_M.gguf": ["mradermacher/NarraThinker12B-i1-GGUF", MessagesFormatterType.MISTRAL],
+    "IceDrinkNameNotFoundRP-7b-Model_Stock.Q4_K_S.gguf": ["mradermacher/IceDrinkNameNotFoundRP-7b-Model_Stock-GGUF", MessagesFormatterType.MISTRAL],
+    "Hatheno_Max_1.1-ALT-8B-model_stock.i1-Q5_K_M.gguf": ["mradermacher/Hatheno_Max_1.1-ALT-8B-model_stock-i1-GGUF", MessagesFormatterType.MISTRAL],
+    "Gluon-8B.i1-Q5_K_M.gguf": ["mradermacher/Gluon-8B-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "mergekit-model_stock-yhrnwcb-q4_k_m.gguf": ["DreadPoor/WIP-TEST_PENDING_A-Q4_K_M-GGUF", MessagesFormatterType.LLAMA_3],
+    "wip-test_pending_b-q4_k_m.gguf": ["DreadPoor/WIP-TEST_PENDING_B-Q4_K_M-GGUF", MessagesFormatterType.LLAMA_3],
+    "Hatheno_Max_1.1-8B-model_stock.i1-Q5_K_M.gguf": ["mradermacher/Hatheno_Max_1.1-8B-model_stock-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "MN-12B-Siskin-v0.2.i1-Q4_K_M.gguf": ["mradermacher/MN-12B-Siskin-v0.2-i1-GGUF", MessagesFormatterType.MISTRAL],
+    "Hatheno_Max-8B-model_stock.i1-Q4_K_M.gguf": ["mradermacher/Hatheno_Max-8B-model_stock-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "L3.1-SuperNovabliterated-8B-model_stock.i1-Q4_K_M.gguf": ["mradermacher/L3.1-SuperNovabliterated-8B-model_stock-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "Hatheno_Max-ALT-8B-model_stock.i1-Q5_K_M.gguf": ["mradermacher/Hatheno_Max-ALT-8B-model_stock-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "experiment_x-wip-q4_k_m.gguf": ["DreadPoor/EXPERIMENT_X-WIP-Q4_K_M-GGUF", MessagesFormatterType.MISTRAL],
+    "narrathinker12b-q4_k_m.gguf": ["ClaudioItaly/NarraThinker12B-Q4_K_M-GGUF", MessagesFormatterType.MISTRAL],
+    "llama-3.1-8b-matrix-q5_k_m.gguf": ["bunnycore/LLama-3.1-8B-Matrix-Q5_K_M-GGUF", MessagesFormatterType.LLAMA_3],
+    "Barcenas-8b-Cartas.Q5_K_M.gguf": ["mradermacher/Barcenas-8b-Cartas-GGUF", MessagesFormatterType.LLAMA_3],
+    "HannaOpenHermes-2.5-Mistral-7B.Q5_K_M.gguf": ["mradermacher/HannaOpenHermes-2.5-Mistral-7B-GGUF", MessagesFormatterType.MISTRAL],
+    "IceDrinkNameGoesHereRP-7b-Model_Stock.i1-Q4_K_M.gguf": ["mradermacher/IceDrinkNameGoesHereRP-7b-Model_Stock-i1-GGUF", MessagesFormatterType.ALPACA],
+    "Llama-3.1-Literotica-8B.Q4_K_S.gguf": ["mradermacher/Llama-3.1-Literotica-8B-GGUF", MessagesFormatterType.LLAMA_3],
+    "project-12-q4_k_m.gguf": ["ClaudioItaly/Project-12-Q4_K_M-GGUF", MessagesFormatterType.MISTRAL],
+    "L3.1-Celestial-Stone-2x8B.i1-Q4_K_M.gguf": ["mradermacher/L3.1-Celestial-Stone-2x8B-i1-GGUF", MessagesFormatterType.LLAMA_3],
+    "experiment_y-wip-q4_k_m.gguf": ["DreadPoor/EXPERIMENT_Y-WIP-Q4_K_M-GGUF", MessagesFormatterType.LLAMA_3],
     "L3.1-Vulca-Epith-Bluegrade-v0.2-8B.q8_0.gguf": ["kromquant/L3.1-Vulca-Epith-Bluegrade-v0.2-8B-GGUFs", MessagesFormatterType.LLAMA_3],
     "llama-3.1-8b-omnimatrix-iq4_nl-imat.gguf": ["bunnycore/Llama-3.1-8B-OmniMatrix-IQ4_NL-GGUF", MessagesFormatterType.LLAMA_3],
     "L3.1-Artemis-d-8B.i1-Q5_K_M.gguf": ["mradermacher/L3.1-Artemis-d-8B-i1-GGUF", MessagesFormatterType.LLAMA_3],
@@ -481,7 +583,9 @@ llm_models = {
     "Meta-Llama-3.1-8B-Claude-iMat-Q5_K_M.gguf": ["InferenceIllusionist/Meta-Llama-3.1-8B-Claude-iMat-GGUF", MessagesFormatterType.LLAMA_3],
     "Phi-3.1-mini-128k-instruct-Q6_K_L.gguf": ["bartowski/Phi-3.1-mini-128k-instruct-GGUF", MessagesFormatterType.PHI_3],
     "tifa-7b-qwen2-v0.1.q4_k_m.gguf": ["Tifa-RP/Tifa-7B-Qwen2-v0.1-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "Holland-Magnum-Merge-R2.i1-Q5_K_M.gguf": ["mradermacher/Holland-Magnum-Merge-R2-i1-GGUF", MessagesFormatterType.LLAMA_3],
     "Oumuamua-7b-RP_Q5_K_M.gguf": ["Aratako/Oumuamua-7b-RP-GGUF", MessagesFormatterType.MISTRAL],
+    "Berghof-NSFW-7B.Q5_K_M.gguf": ["QuantFactory/Berghof-NSFW-7B-GGUF", MessagesFormatterType.MISTRAL],
     "Japanese-TextGen-Kage-v0.1.2-2x7B-NSFW_iMat_Ch200_IQ4_XS.gguf": ["dddump/Japanese-TextGen-Kage-v0.1.2-2x7B-NSFW-gguf", MessagesFormatterType.VICUNA],
     "ChatWaifu_v1.2.1.Q5_K_M.gguf": ["mradermacher/ChatWaifu_v1.2.1-GGUF", MessagesFormatterType.MISTRAL],
     "ChatWaifu_v1.1.Q5_K_M.gguf": ["mradermacher/ChatWaifu_v1.1-GGUF", MessagesFormatterType.MISTRAL],
@@ -1087,3 +1191,75 @@ def dolphin_parse_simple(
     else:
         prompts = list_uniq(to_list(raw_prompt) + ["nsfw", "explicit", "rating_explicit"])
     return ", ".join(prompts)
+
+
+# https://huggingface.co/spaces/CaioXapelaum/GGUF-Playground
+import cv2
+cv2.setNumThreads(1)
+
+@spaces.GPU()
+def respond_playground(
+    message,
+    history: list[tuple[str, str]],
+    model,
+    system_message,
+    max_tokens,
+    temperature,
+    top_p,
+    top_k,
+    repeat_penalty,
+):
+    if override_llm_format:
+        chat_template = override_llm_format
+    else:
+        chat_template = llm_models[model][1]
+
+    llm = Llama(
+        model_path=str(Path(f"{llm_models_dir}/{model}")),
+        flash_attn=True,
+        n_gpu_layers=81, # 81
+        n_batch=1024,
+        n_ctx=8192, #8192
+    )
+    provider = LlamaCppPythonProvider(llm)
+
+    agent = LlamaCppAgent(
+        provider,
+        system_prompt=f"{system_message}",
+        predefined_messages_formatter_type=chat_template,
+        debug_output=False
+    )
+    
+    settings = provider.get_provider_default_settings()
+    settings.temperature = temperature
+    settings.top_k = top_k
+    settings.top_p = top_p
+    settings.max_tokens = max_tokens
+    settings.repeat_penalty = repeat_penalty
+    settings.stream = True
+
+    messages = BasicChatHistory()
+
+    # Add user and assistant messages to the history
+    for msn in history:
+        user = {'role': Roles.user, 'content': msn[0]}
+        assistant = {'role': Roles.assistant, 'content': msn[1]}
+        messages.add_message(user)
+        messages.add_message(assistant)
+
+    # Stream the response
+    try:
+        stream = agent.get_chat_response(
+            message,
+            llm_sampling_settings=settings,
+            chat_history=messages,
+            returns_streaming_generator=True,
+            print_output=False
+        )
+
+        outputs = ""
+        for output in stream:
+            outputs += output
+            yield outputs
+    except Exception as e:
+        yield f"Error during response generation: {str(e)}"

modutils.py

@@ -4,13 +4,21 @@ import gradio as gr
 from huggingface_hub import HfApi
 import os
 from pathlib import Path
+from PIL import Image
 
 
 from env import (HF_LORA_PRIVATE_REPOS1, HF_LORA_PRIVATE_REPOS2,
-    HF_MODEL_USER_EX, HF_MODEL_USER_LIKES,
+    HF_MODEL_USER_EX, HF_MODEL_USER_LIKES, DIFFUSERS_FORMAT_LORAS,
     directory_loras, hf_read_token, HF_TOKEN, CIVITAI_API_KEY)
 
 
+MODEL_TYPE_DICT = {
+    "diffusers:StableDiffusionPipeline": "SD 1.5",
+    "diffusers:StableDiffusionXLPipeline": "SDXL",
+    "diffusers:FluxPipeline": "FLUX",
+}
+
+
 def get_user_agent():
     return 'Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:127.0) Gecko/20100101 Firefox/127.0'
 
@@ -27,6 +35,11 @@ def list_sub(a, b):
     return [e for e in a if e not in b]
 
 
+def is_repo_name(s):
+    import re
+    return re.fullmatch(r'^[^/]+?/[^/]+?$', s)
+
+
 from translatepy import Translator
 translator = Translator()
 def translate_to_en(input: str):
@@ -64,7 +77,7 @@ def download_things(directory, url, hf_token="", civitai_api_key=""):
         if hf_token:
             os.system(f"aria2c --console-log-level=error --summary-interval=10 --header={user_header} -c -x 16 -k 1M -s 16 {url} -d {directory}  -o {url.split('/')[-1]}")
         else:
-            os.system (f"aria2c --optimize-concurrent-downloads --console-log-level=error --summary-interval=10 -c -x 16 -k 1M -s 16 {url} -d {directory}  -o {url.split('/')[-1]}")
+            os.system(f"aria2c --optimize-concurrent-downloads --console-log-level=error --summary-interval=10 -c -x 16 -k 1M -s 16 {url} -d {directory}  -o {url.split('/')[-1]}")
     elif "civitai.com" in url:
         if "?" in url:
             url = url.split("?")[0]
@@ -100,7 +113,6 @@ def safe_float(input):
     return output
 
 
-from PIL import Image
 def save_images(images: list[Image.Image], metadatas: list[str]):
     from PIL import PngImagePlugin
     import uuid
@@ -245,10 +257,10 @@ model_id_list = get_model_id_list()
 
 
 def get_t2i_model_info(repo_id: str):
-    api = HfApi()
+    api = HfApi(token=HF_TOKEN)
     try:
-        if " " in repo_id or not api.repo_exists(repo_id): return ""
-        model = api.model_info(repo_id=repo_id)
+        if not is_repo_name(repo_id): return ""
+        model = api.model_info(repo_id=repo_id, timeout=5.0)
     except Exception as e:
         print(f"Error: Failed to get {repo_id}'s info.")
         print(e)
@@ -258,9 +270,8 @@ def get_t2i_model_info(repo_id: str):
     info = []
     url = f"https://huggingface.co/{repo_id}/"
     if not 'diffusers' in tags: return ""
-    if 'diffusers:FluxPipeline' in tags: info.append("FLUX.1")
-    elif 'diffusers:StableDiffusionXLPipeline' in tags: info.append("SDXL")
-    elif 'diffusers:StableDiffusionPipeline' in tags: info.append("SD1.5")
+    for k, v in MODEL_TYPE_DICT.items():
+        if k in tags: info.append(v)
     if model.card_data and model.card_data.tags:
         info.extend(list_sub(model.card_data.tags, ['text-to-image', 'stable-diffusion', 'stable-diffusion-api', 'safetensors', 'stable-diffusion-xl']))
     info.append(f"DLs: {model.downloads}")
@@ -285,12 +296,8 @@ def get_tupled_model_list(model_list):
         tags = model.tags
         info = []
         if not 'diffusers' in tags: continue
-        if 'diffusers:FluxPipeline' in tags:
-            info.append("FLUX.1")
-        if 'diffusers:StableDiffusionXLPipeline' in tags:
-            info.append("SDXL")
-        elif 'diffusers:StableDiffusionPipeline' in tags:
-            info.append("SD1.5")
+        for k, v in MODEL_TYPE_DICT.items():
+            if k in tags: info.append(v)
         if model.card_data and model.card_data.tags:
             info.extend(list_sub(model.card_data.tags, ['text-to-image', 'stable-diffusion', 'stable-diffusion-api', 'safetensors', 'stable-diffusion-xl']))
         if "pony" in info:
@@ -374,7 +381,7 @@ def get_civitai_info(path):
 
 
 def get_lora_model_list():
-    loras = list_uniq(get_private_lora_model_lists() + get_local_model_list(directory_loras))
+    loras = list_uniq(get_private_lora_model_lists() + get_local_model_list(directory_loras) + DIFFUSERS_FORMAT_LORAS)
     loras.insert(0, "None")
     loras.insert(0, "")
     return loras
@@ -483,7 +490,7 @@ def download_my_lora(dl_urls: str, lora1: str, lora2: str, lora3: str, lora4: st
         gr.update(value=lora4, choices=choices), gr.update(value=lora5, choices=choices)
 
 
-def get_valid_lora_name(query: str):
+def get_valid_lora_name(query: str, model_name: str):
     path = "None"
     if not query or query == "None": return "None"
     if to_lora_key(query) in loras_dict.keys(): return query
@@ -497,7 +504,7 @@ def get_valid_lora_name(query: str):
         dl_file = download_lora(query)
         if dl_file and Path(dl_file).exists(): return dl_file
     else:
-        dl_file = find_similar_lora(query)
+        dl_file = find_similar_lora(query, model_name)
         if dl_file and Path(dl_file).exists(): return dl_file
     return "None"
 
@@ -521,14 +528,14 @@ def get_valid_lora_wt(prompt: str, lora_path: str, lora_wt: float):
     return wt
 
 
-def set_prompt_loras(prompt, prompt_syntax, lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora4, lora4_wt, lora5, lora5_wt):
+def set_prompt_loras(prompt, prompt_syntax, model_name, lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora4, lora4_wt, lora5, lora5_wt):
     import re
     if not "Classic" in str(prompt_syntax):  return lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora4, lora4_wt, lora5, lora5_wt
-    lora1 = get_valid_lora_name(lora1)
-    lora2 = get_valid_lora_name(lora2)
-    lora3 = get_valid_lora_name(lora3)
-    lora4 = get_valid_lora_name(lora4)
-    lora5 = get_valid_lora_name(lora5)
+    lora1 = get_valid_lora_name(lora1, model_name)
+    lora2 = get_valid_lora_name(lora2, model_name)
+    lora3 = get_valid_lora_name(lora3, model_name)
+    lora4 = get_valid_lora_name(lora4, model_name)
+    lora5 = get_valid_lora_name(lora5, model_name)
     if not "<lora" in prompt: return lora1, lora1_wt, lora2, lora2_wt, lora3, lora3_wt, lora4, lora4_wt, lora5, lora5_wt
     lora1_wt = get_valid_lora_wt(prompt, lora1, lora1_wt)
     lora2_wt = get_valid_lora_wt(prompt, lora2, lora2_wt)
@@ -790,16 +797,17 @@ def get_civitai_info(path):
         return items
 
 
-def search_lora_on_civitai(query: str, allow_model: list[str] = ["Pony", "SDXL 1.0"], limit: int = 100):
+def search_lora_on_civitai(query: str, allow_model: list[str] = ["Pony", "SDXL 1.0"], limit: int = 100,
+                           sort: str = "Highest Rated", period: str = "AllTime", tag: str = ""):
     import requests
     from requests.adapters import HTTPAdapter
     from urllib3.util import Retry
-    if not query: return None
     user_agent = get_user_agent()
     headers = {'User-Agent': user_agent, 'content-type': 'application/json'}
     base_url = 'https://civitai.com/api/v1/models'
-    params = {'query': query, 'types': ['LORA'], 'sort': 'Highest Rated', 'period': 'AllTime',
-              'nsfw': 'true', 'supportsGeneration ': 'true', 'limit': limit}
+    params = {'types': ['LORA'], 'sort': sort, 'period': period, 'limit': limit, 'nsfw': 'true'}
+    if query: params["query"] = query
+    if tag: params["tag"] = tag
     session = requests.Session()
     retries = Retry(total=5, backoff_factor=1, status_forcelist=[500, 502, 503, 504])
     session.mount("https://", HTTPAdapter(max_retries=retries))
@@ -828,9 +836,9 @@ def search_lora_on_civitai(query: str, allow_model: list[str] = ["Pony", "SDXL 1
         return items
 
 
-def search_civitai_lora(query, base_model):
+def search_civitai_lora(query, base_model, sort="Highest Rated", period="AllTime", tag=""):
     global civitai_lora_last_results
-    items = search_lora_on_civitai(query, base_model)
+    items = search_lora_on_civitai(query, base_model, 100, sort, period, tag)
     if not items: return gr.update(choices=[("", "")], value="", visible=False),\
           gr.update(value="", visible=False), gr.update(visible=True), gr.update(visible=True)
     civitai_lora_last_results = {}
@@ -856,7 +864,27 @@ def select_civitai_lora(search_result):
     return gr.update(value=search_result), gr.update(value=md, visible=True)
 
 
-def find_similar_lora(q: str):
+LORA_BASE_MODEL_DICT = {
+    "diffusers:StableDiffusionPipeline": ["SD 1.5"],
+    "diffusers:StableDiffusionXLPipeline": ["Pony", "SDXL 1.0"],
+    "diffusers:FluxPipeline": ["Flux.1 D", "Flux.1 S"],
+}
+
+
+def get_lora_base_model(model_name: str):
+    api = HfApi(token=HF_TOKEN)
+    default = ["Pony", "SDXL 1.0"]
+    try:
+        model = api.model_info(repo_id=model_name, timeout=5.0)
+        tags = model.tags
+        for tag in tags:
+            if tag in LORA_BASE_MODEL_DICT.keys(): return LORA_BASE_MODEL_DICT.get(tag, default)
+    except Exception:
+        return default
+    return default
+
+
+def find_similar_lora(q: str, model_name: str):
     from rapidfuzz.process import extractOne
     from rapidfuzz.utils import default_process
     query = to_lora_key(q)
@@ -879,7 +907,7 @@ def find_similar_lora(q: str):
     print(f"Finding <lora:{query}:...> on Civitai...")
     civitai_query = Path(query).stem if Path(query).is_file() else query
     civitai_query = civitai_query.replace("_", " ").replace("-", " ")
-    base_model = ["Pony", "SDXL 1.0"]
+    base_model = get_lora_base_model(model_name)
     items = search_lora_on_civitai(civitai_query, base_model, 1)
     if items:
         item = items[0]
@@ -1241,11 +1269,11 @@ def set_textual_inversion_prompt(textual_inversion_gui, prompt_gui, neg_prompt_g
 
 def get_model_pipeline(repo_id: str):
     from huggingface_hub import HfApi
-    api = HfApi()
+    api = HfApi(token=HF_TOKEN)
     default = "StableDiffusionPipeline"
     try:
-        if " " in repo_id or not api.repo_exists(repo_id): return default
-        model = api.model_info(repo_id=repo_id)
+        if not is_repo_name(repo_id): return default
+        model = api.model_info(repo_id=repo_id, timeout=5.0)
     except Exception:
         return default
     if model.private or model.gated: return default