add GLM-Image pipeline support

- Add GLM-Image (zai-org/GLM-Image) model detection and loading
- Custom pipeline loader with proper component handling:
  - ByT5 text encoder (cannot use shared T5 due to different hidden size)
  - Vision-language encoder (9B AR model)
  - DiT transformer (7B)
- Fix EOS token early stopping in AR generation
- Add AR token generation progress tracking with terminal progress bar
- Fix uninitialized audio variable in processing
- Add TAESD support for GLM-Image (using f1 variant)

modules/modeldata.py

@@ -78,6 +78,8 @@ def get_model_type(pipe):
         model_type = 'prx'
     elif 'LongCat' in name:
         model_type = 'longcat'
+    elif 'GlmImage' in name:
+        model_type = 'glm_image'
     elif 'Ovis-Image' in name:
         model_type = 'ovis'
     # video models

modules/processing.py

@@ -400,6 +400,7 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
     infotexts = []
     output_images = []
     output_binary = None
+    audio = None
 
     process_init(p)
     if p.scripts is not None and isinstance(p.scripts, scripts_manager.ScriptRunner):

modules/processing_args.py

@@ -156,6 +156,8 @@ def task_specific_kwargs(p, model):
         task_args['reference_images'] = p.init_images
     if ('GoogleNanoBananaPipeline' in model_cls) and (p.init_images is not None) and (len(p.init_images) > 0):
         task_args['image'] = p.init_images[0]
+    if ('GlmImagePipeline' in model_cls) and (p.init_images is not None) and (len(p.init_images) > 0):
+        task_args['image'] = p.init_images
     if 'BlipDiffusionPipeline' in model_cls:
         if len(p.init_images) == 0:
             shared.log.error('BLiP diffusion requires init image')

modules/sd_detect.py

@@ -143,6 +143,8 @@ def guess_by_name(fn, current_guess):
         new_guess = 'LongCat'
     elif 'ovis-image' in fn.lower():
         new_guess = 'Ovis-Image'
+    elif 'glm-image' in fn.lower():
+        new_guess = 'GLM-Image'
     if debug_load:
         shared.log.trace(f'Autodetect: method=name file="{fn}" previous="{current_guess}" current="{new_guess}"')
     return new_guess or current_guess

modules/sd_models.py

@@ -483,6 +483,10 @@ def load_diffuser_force(detected_model_type, checkpoint_info, diffusers_load_con
             from pipelines.model_ovis import load_ovis
             sd_model = load_ovis(checkpoint_info, diffusers_load_config)
             allow_post_quant = False
+        elif model_type in ['GLM-Image']:
+            from pipelines.model_glm import load_glm_image
+            sd_model = load_glm_image(checkpoint_info, diffusers_load_config)
+            allow_post_quant = False
     except Exception as e:
         shared.log.error(f'Load {op}: path="{checkpoint_info.path}" {e}')
         if debug_load:

modules/sd_vae_taesd.py

@@ -38,7 +38,7 @@ prev_cls = ''
 prev_type = ''
 prev_model = ''
 lock = threading.Lock()
-supported = ['sd', 'sdxl', 'sd3', 'f1', 'h1', 'z_image', 'lumina2', 'hunyuanvideo', 'wanai', 'chrono', 'mochivideo', 'pixartsigma', 'pixartalpha', 'hunyuandit', 'omnigen', 'qwen', 'longcat']
+supported = ['sd', 'sdxl', 'sd3', 'f1', 'h1', 'z_image', 'lumina2', 'hunyuanvideo', 'wanai', 'chrono', 'mochivideo', 'pixartsigma', 'pixartalpha', 'hunyuandit', 'omnigen', 'qwen', 'longcat', 'glm_image']
 
 
 def warn_once(msg, variant=None):
@@ -59,7 +59,7 @@ def get_model(model_type = 'decoder', variant = None):
         model_cls = 'sd'
     elif model_cls in {'pixartsigma', 'hunyuandit', 'omnigen', 'auraflow'}:
         model_cls = 'sdxl'
-    elif model_cls in {'h1', 'z_image', 'lumina2', 'chroma', 'longcat'}:
+    elif model_cls in {'h1', 'z_image', 'lumina2', 'chroma', 'longcat', 'glm_image'}:
         model_cls = 'f1'
     elif model_cls in {'wanai', 'qwen', 'chrono'}:
         variant = variant or 'TAE WanVideo'

modules/shared_items.py

@@ -49,6 +49,7 @@ pipelines = {
     'HunyuanImage': getattr(diffusers, 'HunyuanImagePipeline', None),
     'Z-Image': getattr(diffusers, 'ZImagePipeline', None),
     'LongCat': getattr(diffusers, 'LongCatImagePipeline', None),
+    'GLM-Image': getattr(diffusers, 'GlmImagePipeline', None),
     # dynamically imported and redefined later
     'Meissonic': getattr(diffusers, 'DiffusionPipeline', None),
     'Monetico': getattr(diffusers, 'DiffusionPipeline', None),