Update inference_manager.py

inference_manager.py

@@ -26,7 +26,7 @@ import cv2
 import re
 import gradio as gr
 from PIL import Image
-MAX_SEED = np.iinfo(np.int32).max
+MAX_SEED = 12211231#np.iinfo(np.int32).max
 #from onediffx import compile_pipe, save_pipe, load_pipe
 
 HF_TOKEN = os.getenv('HF_TOKEN')
@@ -509,7 +509,7 @@ class ModelManager:
         p = remove_child_related_content(p)
         prompt_str = cfg.get("prompt", "{prompt}").replace("{prompt}", p)
         generator = torch.Generator(model.base_model_pipeline.device).manual_seed(seed)
-        print(f"generate: p={p}, np={np}, steps={steps}, guidance_scale={guidance_scale}, size={width},{height}, seed={seed}")
+        print(f"generate: p={p}, np={negative_prompt}, steps={steps}, guidance_scale={guidance_scale}, size={width},{height}, seed={seed}")
         images = ip_model.generate(
             prompt=prompt_str,
             negative_prompt=negative_prompt,
@@ -558,7 +558,7 @@ class ModelManager:
         lora_list = inference_params.get("loras", [])
         seed = inference_params.get("seed", 0)
 
-        pipe = model.build_pipeline_with_lora(lora_list, sampler, lora_list)
+        pipe = model.build_pipeline_with_lora(lora_list, sampler)
 
         start = time.time()
         pipe.to("cuda")
@@ -571,7 +571,7 @@ class ModelManager:
         p = remove_child_related_content(p)
         prompt_str = cfg.get("prompt", "{prompt}").replace("{prompt}", p)
         generator = torch.Generator(pipe.device).manual_seed(seed)
-        print(f"generate: p={p}, np={np}, steps={steps}, guidance_scale={guidance_scale}, size={width},{height}, seed={seed}")
+        print(f"generate: p={p}, np={negative_prompt}, steps={steps}, guidance_scale={guidance_scale}, size={width},{height}, seed={seed}")
         images = pipe(
             prompt=prompt_str,
             negative_prompt=negative_prompt,