Merge pull request #3328 from vladmandic/dev

merge

CHANGELOG.md

@@ -67,6 +67,7 @@ This release is primary service release with cumulative fixes and several improv
 - fix control input type video  
 - fix reset pipeline at the end of each iteration  
 - fix faceswap when no faces detected  
+- fix civitai search
 - multiple ModernUI fixes
 
 ## Update for 2024-06-23

modules/model_kolors.py

@@ -1,26 +1,25 @@
 import torch
-import transformers
 import diffusers
 
 
 repo_id = 'Kwai-Kolors/Kolors'
-encoder_id = 'THUDM/chatglm3-6b'
 
 
 def load_kolors(_checkpoint_info, diffusers_load_config={}):
     from modules import shared, devices, modelloader
     modelloader.hf_login()
     diffusers_load_config['variant'] = "fp16"
     if 'torch_dtype' not in diffusers_load_config:
-        diffusers_load_config['torch_dtype'] = 'torch.float16'
+        diffusers_load_config['torch_dtype'] = torch.float16
 
-    text_encoder = transformers.AutoModel.from_pretrained(encoder_id, torch_dtype=torch.float16, trust_remote_code=True, cache_dir=shared.opts.diffusers_dir)
+    # import torch
+    # import transformers
+    # encoder_id = 'THUDM/chatglm3-6b'
+    # text_encoder = transformers.AutoModel.from_pretrained(encoder_id, torch_dtype=torch.float16, trust_remote_code=True, cache_dir=shared.opts.diffusers_dir)
     # text_encoder = transformers.AutoModel.from_pretrained("THUDM/chatglm3-6b", torch_dtype=torch.float16, trust_remote_code=True).quantize(4).cuda()
-    tokenizer = transformers.AutoTokenizer.from_pretrained(encoder_id, trust_remote_code=True, cache_dir=shared.opts.diffusers_dir)
-    pipe = diffusers.StableDiffusionXLPipeline.from_pretrained(
+    # tokenizer = transformers.AutoTokenizer.from_pretrained(encoder_id, trust_remote_code=True, cache_dir=shared.opts.diffusers_dir)
+    pipe = diffusers.KolorsPipeline.from_pretrained(
         repo_id,
-        tokenizer=tokenizer,
-        text_encoder=text_encoder,
         cache_dir = shared.opts.diffusers_dir,
         **diffusers_load_config,
     )

modules/modelloader.py

@@ -101,6 +101,9 @@ def download_civit_model_thread(model_name, model_url, model_path, model_type, t
     elif model_type == 'Embedding':
         model_file = os.path.join(shared.opts.embeddings_dir, model_path, model_name)
         temp_file = os.path.join(shared.opts.embeddings_dir, model_path, temp_file)
+    elif model_type == 'VAE':
+        model_file = os.path.join(shared.opts.vae_dir, model_path, model_name)
+        temp_file = os.path.join(shared.opts.vae_dir, model_path, temp_file)
     else:
         model_file = os.path.join(shared.opts.ckpt_dir, model_path, model_name)
         temp_file = os.path.join(shared.opts.ckpt_dir, model_path, temp_file)

modules/textual_inversion/textual_inversion.py

@@ -121,7 +121,8 @@ def insert_tokens(embeddings: list, tokenizers: list):
     """
     tokens = []
     for embedding in embeddings:
-        tokens += embedding.tokens
+        if embedding is not None:
+            tokens += embedding.tokens
     for tokenizer in tokenizers:
         tokenizer.add_tokens(tokens)
 
@@ -295,7 +296,8 @@ def load_diffusers_embedding(self, filename: Union[str, List[str]] = None, data:
                         (len(embedding.vector_sizes) < len(hiddensizes) and len(embedding.vector_sizes) != 2)):  # SD3 no T5
                     embedding.tokens = []
                     self.skipped_embeddings[embedding.name] = embedding
-            except Exception:
+            except Exception as e:
+                shared.log.error(f'Embedding invalid: name="{embedding.name}" fn="{filename}" {e}')
                 self.skipped_embeddings[embedding.name] = embedding
         if overwrite:
             shared.log.info(f"Loading Bundled embeddings: {list(data.keys())}")

modules/ui_models.py

@@ -418,41 +418,41 @@ def hf_download_model(hub_id: str, token, variant, revision, mirror, custom_pipe
                 def civit_search_model(name, tag, model_type):
                     # types = 'LORA' if model_type == 'LoRA' else 'Checkpoint'
                     url = 'https://civitai.com/api/v1/models?limit=25&Sort=Newest'
-                    if model_type == 'SD 1.5' or model_type == 'SD XL':
+                    if model_type == 'Model':
                         url += '&types=Checkpoint'
                     elif model_type == 'LoRA':
-                        url += '&types=LORA'
+                        url += '&types=LORA&types=DoRA&types=LoCon'
                     elif model_type == 'Embedding':
                         url += '&types=TextualInversion'
+                    elif model_type == 'VAE':
+                        url += '&types=VAE'
                     if name is not None and len(name) > 0:
                         url += f'&query={name}'
                     if tag is not None and len(tag) > 0:
                         url += f'&tag={tag}'
                     r = req(url)
                     log.debug(f'CivitAI search: name="{name}" tag={tag or "none"} url="{url}" status={r.status_code}')
                     if r.status_code != 200:
+                        log.warning(f'CivitAI search: name="{name}" tag={tag} status={r.status_code}')
+                        return [], gr.update(visible=False, value=[]), gr.update(visible=False, value=None), gr.update(visible=False, value=None)
+                    try:
+                        body = r.json()
+                    except Exception as e:
+                        log.error(f'CivitAI search: name="{name}" tag={tag} {e}')
                         return [], gr.update(visible=False, value=[]), gr.update(visible=False, value=None), gr.update(visible=False, value=None)
-                    body = r.json()
                     nonlocal data
                     data = body.get('items', [])
                     data1 = []
                     for model in data:
                         found = 0
-                        if model_type == 'LoRA' and model['type'] in ['LORA', 'LoCon']:
+                        if model_type == 'LoRA' and model['type'].lower() in ['lora', 'locon', 'dora', 'lycoris']:
                             found += 1
-                        elif model_type == 'Embedding' and model['type'] == 'TextualInversion':
+                        elif model_type == 'Embedding' and model['type'].lower() in ['textualinversion', 'embedding']:
+                            found += 1
+                        elif model_type == 'Model' and model['type'].lower() in ['checkpoint']:
+                            found += 1
+                        elif model_type == 'VAE' and model['type'].lower() in ['vae']:
                             found += 1
-                        elif model_type.startswith('SD') and model['type'] == 'Checkpoint':
-                            for variant in model['modelVersions']:
-                                if model_type == 'SD 1.5':
-                                    if 'SD 1.' in variant['baseModel']:
-                                        found += 1
-                                if model_type == 'SD XL':
-                                    if 'SDXL' in variant['baseModel']:
-                                        found += 1
-                                else:
-                                    if 'SD 1.' not in variant['baseModel'] and 'SDXL' not in variant['baseModel']:
-                                        found += 1
                         elif model_type == 'Other':
                             found += 1
                         if found > 0:
@@ -464,8 +464,7 @@ def civit_search_model(name, tag, model_type):
                                 model['stats']['rating']
                             ])
                     res = f'Search result: name={name} tag={tag or "none"} type={model_type} models={len(data1)}'
-                    return res, gr.update(visible=len(data1) > 0, value=data1 if len(data1) > 0 else []), gr.update(
-                        visible=False, value=None), gr.update(visible=False, value=None)
+                    return res, gr.update(visible=len(data1) > 0, value=data1 if len(data1) > 0 else []), gr.update(visible=False, value=None), gr.update(visible=False, value=None)
 
                 def civit_select1(evt: gr.SelectData, in_data):
                     model_id = in_data[evt.index[0]][0]
@@ -596,7 +595,7 @@ def civit_search_metadata(rehash, title):
                     gr.HTML('<h2>Search for models</h2>')
                 with gr.Row():
                     with gr.Column(scale=1):
-                        civit_model_type = gr.Dropdown(label='Model type', choices=['SD 1.5', 'SD XL', 'LoRA', 'Embedding', 'Other'], value='LoRA')
+                        civit_model_type = gr.Dropdown(label='Model type', choices=['Model', 'LoRA', 'Embedding', 'VAE', 'Other'], value='Model')
                     with gr.Column(scale=15):
                         with gr.Row():
                             civit_search_text = gr.Textbox('', label='Search models', placeholder='keyword')