Update files

app.py

@@ -1,18 +1,110 @@
+# import gradio as gr
+# from transformers import pipeline
+
+# pipeline = pipeline(task="image-classification", model="julien-c/hotdog-not-hotdog")
+
+# def predict(input_img):
+#     predictions = pipeline(input_img)
+#     return input_img, {p["label"]: p["score"] for p in predictions} 
+
+# gradio_app = gr.Interface(
+#     predict,
+#     inputs=gr.Image(label="Select hot dog candidate", sources=['upload', 'webcam'], type="pil"),
+#     outputs=[gr.Image(label="Processed Image"), gr.Label(label="Result", num_top_classes=2)],
+#     title="Hot Dog? Or Not?",
+# )
+
+import random
 import gradio as gr
-from transformers import pipeline
+from PIL import Image
+import torch
+import uuid
+import numpy as np
+from diffusers import AutoPipelineForText2Image, AutoPipelineForInpainting, StableDiffusionXLInpaintPipeline, StableDiffusionXLPipeline
+
+
+model_id = "stabilityai/stable-diffusion-xl-base-1.0"
+
+# we have give lora weights here
+adapter_id = "qwe cat_20240307-085154.json"
+
+pipe = AutoPipelineForText2Image.from_pretrained(model_id, 
+                                                        torch_dtype=torch.float16, 
+                                                        variant="fp16")
+# pipe.enable_freeu(s1=0.9, s2=0.2, b1=1.2, b2=1.4)
+
+pipe.to("cuda")
+
 
-pipeline = pipeline(task="image-classification", model="julien-c/hotdog-not-hotdog")
 
-def predict(input_img):
-    predictions = pipeline(input_img)
-    return input_img, {p["label"]: p["score"] for p in predictions} 
+pipe.load_lora_weights('lora_weights',
+                       weight_name=adapter_id, 
+                       adapter_name="qwe")
+pipe.fuse_lora() #lora_scale=0.7
 
-gradio_app = gr.Interface(
-    predict,
-    inputs=gr.Image(label="Select hot dog candidate", sources=['upload', 'webcam'], type="pil"),
-    outputs=[gr.Image(label="Processed Image"), gr.Label(label="Result", num_top_classes=2)],
-    title="Hot Dog? Or Not?",
-)
+def set_lora_weight(lora_scale):
+    pipe.unfuse_lora(True)
+    pipe.load_lora_weights('lora_weights',
+                       weight_name=adapter_id, 
+                       adapter_name="qwe")
+    pipe.fuse_lora(lora_scale=lora_scale) #lora_scale=0.7
+    print('DONE')
+
+def generate(text, guidance_scale, num_images_per_prompt, height, width, generator_seed):
+    
+    generator = torch.Generator("cuda").manual_seed(generator_seed)
+
+    
+    prompt = text
+    image = pipe(prompt=prompt, 
+                 negative_prompt='worst quality, normal quality, low quality, low res, blurry, text, watermark, logo, banner, extra digits, cropped, jpeg artifacts, signature, username, error, sketch ,duplicate, ugly, monochrome, horror, geometry, mutation, disgusting',
+                 guidance_scale=guidance_scale,
+                 num_images_per_prompt=num_images_per_prompt,
+                 height=height,
+                 width=width,
+                 num_inference_steps=20,
+                 generator=generator).images
+    return image
+
+
+
+
+with gr.Blocks() as demo:
+    with gr.Row():
+        with gr.Column():
+            
+            gallery = gr.Gallery(
+                label="Generate",
+             object_fit="contain", height="512")
+            
+            text = gr.Textbox(
+                    label="Enter Prompt...")
+            btn = gr.Button("Generate", scale=0)
+            guidance_scale = gr.Slider(minimum=0, maximum=15, value=7.5, label='guidance scale')
+            num_images_per_prompt = gr.Slider(minimum=1, maximum=4, value=2, step=1, label = 'number of images per prompt')
+            height = gr.Slider(minimum=512, maximum=2048, value=1024, label = 'Image height')
+            width = gr.Slider(minimum=512, maximum=2048, value=1024,step=8,label = 'Image width')
+            lora_scale = gr.Slider(minimum=0.1, maximum=1, value=1,step=0.01,label = 'Lora scale')
+            generator_seed = gr.Slider(minimum=-1, maximum=100, value=1,step=1,label = 'generator_seed')
+            
+            
+        
+        # with gr.Column():
+        #     gallery = gr.Gallery(
+        #         label="Generate",
+        #      object_fit="contain", height="2048")
+
+
+    btn.click(generate, 
+              inputs=[text,guidance_scale,num_images_per_prompt, height, width, generator_seed], 
+              outputs=gallery)
+    
+    lora_scale.change(set_lora_weight,
+                      inputs=lora_scale)
+    
 
 if __name__ == "__main__":
-    gradio_app.launch()
+    demo.launch(share=True)
+
+# if __name__ == "__main__":
+#     gradio_app.launch()

qwe cat_20240307-085154.json

@@ -0,0 +1,115 @@
+{
+  "LoRA_type": "Standard",
+  "LyCORIS_preset": "full",
+  "adaptive_noise_scale": 0,
+  "additional_parameters": "",
+  "block_alphas": "",
+  "block_dims": "",
+  "block_lr_zero_threshold": "",
+  "bucket_no_upscale": false,
+  "bucket_reso_steps": 32,
+  "cache_latents": true,
+  "cache_latents_to_disk": true,
+  "caption_dropout_every_n_epochs": 0.0,
+  "caption_dropout_rate": 0,
+  "caption_extension": ".txt",
+  "clip_skip": "1",
+  "color_aug": false,
+  "conv_alpha": 32,
+  "conv_block_alphas": "",
+  "conv_block_dims": "",
+  "conv_dim": 32,
+  "debiased_estimation_loss": false,
+  "decompose_both": false,
+  "dim_from_weights": false,
+  "down_lr_weight": "",
+  "enable_bucket": true,
+  "epoch": 3,
+  "factor": -1,
+  "flip_aug": false,
+  "full_bf16": false,
+  "full_fp16": false,
+  "gradient_accumulation_steps": 1,
+  "gradient_checkpointing": true,
+  "keep_tokens": "0",
+  "learning_rate": 0.0001,
+  "logging_dir": "/home/rohan/Projects/Lora/doll/log",
+  "lora_network_weights": "",
+  "lr_scheduler": "constant",
+  "lr_scheduler_args": "",
+  "lr_scheduler_num_cycles": "1",
+  "lr_scheduler_power": "",
+  "lr_warmup": 0,
+  "max_bucket_reso": 2048,
+  "max_data_loader_n_workers": "0",
+  "max_resolution": "1024,1024",
+  "max_timestep": 1000,
+  "max_token_length": "75",
+  "max_train_epochs": "1",
+  "max_train_steps": "",
+  "mem_eff_attn": false,
+  "mid_lr_weight": "",
+  "min_bucket_reso": 256,
+  "min_snr_gamma": 5,
+  "min_timestep": 0,
+  "mixed_precision": "bf16",
+  "model_list": "custom",
+  "module_dropout": 0,
+  "multires_noise_discount": 0,
+  "multires_noise_iterations": 0,
+  "network_alpha": 13,
+  "network_dim": 128,
+  "network_dropout": 0,
+  "no_token_padding": false,
+  "noise_offset": 0,
+  "noise_offset_type": "Original",
+  "num_cpu_threads_per_process": 2,
+  "optimizer": "AdamW8bit",
+  "optimizer_args": "",
+  "output_dir": "/home/rohan/Projects/Lora/qwe_toy/model/",
+  "output_name": "qwe cat",
+  "persistent_data_loader_workers": false,
+  "pretrained_model_name_or_path": "stabilityai/stable-diffusion-xl-base-1.0",
+  "prior_loss_weight": 1.0,
+  "random_crop": false,
+  "rank_dropout": 0,
+  "reg_data_dir": "/home/rohan/Projects/Lora/qwe_toy/log/",
+  "resume": "",
+  "sample_every_n_epochs": 0,
+  "sample_every_n_steps": 0,
+  "sample_prompts": "",
+  "sample_sampler": "euler_a",
+  "save_every_n_epochs": 2,
+  "save_every_n_steps": 0,
+  "save_last_n_steps": 0,
+  "save_last_n_steps_state": 0,
+  "save_model_as": "safetensors",
+  "save_precision": "bf16",
+  "save_state": false,
+  "scale_v_pred_loss_like_noise_pred": false,
+  "scale_weight_norms": 2.5,
+  "sdxl": true,
+  "sdxl_cache_text_encoder_outputs": false,
+  "sdxl_no_half_vae": true,
+  "seed": "",
+  "shuffle_caption": false,
+  "stop_text_encoder_training_pct": 0,
+  "text_encoder_lr": 0.0001,
+  "train_batch_size": 1,
+  "train_data_dir": "/home/rohan/Projects/Lora/qwe_toy/img/",
+  "train_on_input": true,
+  "training_comment": "trigger: playboy centerfold",
+  "unet_lr": 0.0001,
+  "unit": 1,
+  "up_lr_weight": "",
+  "use_cp": false,
+  "use_wandb": false,
+  "v2": false,
+  "v_parameterization": false,
+  "v_pred_like_loss": 0,
+  "vae": "",
+  "vae_batch_size": 0,
+  "wandb_api_key": "",
+  "weighted_captions": false,
+  "xformers": "xformers"
+}

requirements.txt

@@ -1,2 +1,4 @@
-transformers
-torch
+torch
+torchvision 
+torchaudio
+diffusers