Delete README.md

README.md

@@ -1,184 +0,0 @@
----
-library_name: Diffusers
-pipeline_tag: text-to-image
-inference: true
-base_model:
-- Qwen/Qwen-Image
----
-
-This tiny model is for debugging. It is randomly initialized with the config adapted from [Qwen/Qwen-Image](https://huggingface.co/Qwen/Qwen-Image).
-
-File size:
-- ~10MB  text_encoder/model.safetensors
-- ~200KB transformer/diffusion_pytorch_model.safetensors
-- ~5MB   vae/diffusion_pytorch_model.safetensors
-
-### Example usage:
-
-```python
-import torch
-from diffusers import DiffusionPipeline
-
-model_id = "tiny-random/Qwen-Image"
-torch_dtype = torch.bfloat16
-device = "cuda"
-pipe = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch_dtype)
-pipe = pipe.to(device)
-
-positive_magic = {
-    "en": "Ultra HD, 4K, cinematic composition.",  # for english prompt,
-    "zh": "超清，4K，电影级构图"  # for chinese prompt,
-}
-prompt = '''A coffee shop entrance features a chalkboard sign reading "Qwen Coffee 😊 $2 per cup," with a neon light beside it displaying "通义千问". Next to it hangs a poster showing a beautiful Chinese woman, and beneath the poster is written "π≈3.1415926-53589793-23846264-33832795-02384197". Ultra HD, 4K, cinematic composition.'''
-prompt += 'Some dummy random texts to make prompt long enough ' * 10
-negative_prompt = " "
-
-# Generate with different aspect ratios
-aspect_ratios = {
-    "1:1": (1328, 1328),
-    "16:9": (1664, 928),
-    "9:16": (928, 1664),
-    "4:3": (1472, 1140),
-    "3:4": (1140, 1472)
-}
-
-for width, height in aspect_ratios.values():
-    image = pipe(
-        prompt=prompt + positive_magic["en"],
-        negative_prompt=negative_prompt,
-        width=width,
-        height=height,
-        num_inference_steps=4,
-        true_cfg_scale=4.0,
-        generator=torch.Generator(device="cuda").manual_seed(42)
-    ).images[0]
-    print(image)
-```
-
-### Codes to create this repo:
-
-```python
-import json
-
-import torch
-from diffusers import (
-    AutoencoderKLQwenImage,
-    DiffusionPipeline,
-    FlowMatchEulerDiscreteScheduler,
-    QwenImagePipeline,
-    QwenImageTransformer2DModel,
-)
-from huggingface_hub import hf_hub_download
-from transformers import AutoConfig, AutoTokenizer, Qwen2_5_VLForConditionalGeneration
-from transformers.generation import GenerationConfig
-
-source_model_id = "Qwen/Qwen-Image"
-save_folder = "/tmp/tiny-random/Qwen-Image"
-
-torch.set_default_dtype(torch.bfloat16)
-scheduler = FlowMatchEulerDiscreteScheduler.from_pretrained(source_model_id, subfolder='scheduler')
-tokenizer = AutoTokenizer.from_pretrained(source_model_id, subfolder='tokenizer')
-
-def save_json(path, obj):
-    import json
-    from pathlib import Path
-    Path(path).parent.mkdir(parents=True, exist_ok=True)
-    with open(path, 'w', encoding='utf-8') as f:
-        json.dump(obj, f, indent=2, ensure_ascii=False)
-
-def init_weights(model):
-    import torch
-    torch.manual_seed(42)
-    with torch.no_grad():
-        for name, p in sorted(model.named_parameters()):
-            torch.nn.init.normal_(p, 0, 0.1)
-            print(name, p.shape, p.dtype, p.device)
-
-with open(hf_hub_download(source_model_id, filename='text_encoder/config.json', repo_type='model'), 'r', encoding='utf - 8') as f:
-    config = json.load(f)
-    config.update({
-        'hidden_size': 32,
-        'intermediate_size': 64,
-        'max_window_layers': 1,
-        'num_attention_heads': 2,
-        'num_hidden_layers': 2,
-        'num_key_value_heads': 1,
-        'sliding_window': 64,
-        'tie_word_embeddings': True,
-        'use_sliding_window': True,
-    })
-    del config['torch_dtype']
-    config['rope_scaling']['mrope_section'] = [4, 2, 2]
-    config['text_config'].update({
-        'hidden_size': 32,
-        'intermediate_size': 64,
-        'num_attention_heads': 2,
-        'num_hidden_layers': 2,
-        'num_key_value_heads': 1,
-        'sliding_window': 64,
-        'tie_word_embeddings': True,
-        'max_window_layers': 1,
-        'use_sliding_window': True,
-        'layer_types': ['full_attention', 'sliding_attention']
-    })
-    del config['text_config']['torch_dtype']
-    config['text_config']['rope_scaling']['mrope_section'] = [4, 2, 2]
-    config['vision_config'].update(
-        {
-            'depth': 2,
-            'fullatt_block_indexes': [0],
-            'hidden_size': 32,
-            'intermediate_size': 64,
-            'num_heads': 2,
-            'out_hidden_size': 32,
-        }
-    )
-    del config['vision_config']['torch_dtype']
-    save_json(f'{save_folder}/text_encoder/config.json', config)
-    text_encoder_config = AutoConfig.from_pretrained(f'{save_folder}/text_encoder')
-    text_encoder = Qwen2_5_VLForConditionalGeneration(text_encoder_config).to(torch.bfloat16)
-    generation_config = GenerationConfig.from_pretrained(source_model_id, subfolder='text_encoder')
-    # text_encoder.config.generation_config = generation_config
-    text_encoder.generation_config = generation_config
-    init_weights(text_encoder)
-
-with open(hf_hub_download(source_model_id, filename='transformer/config.json', repo_type='model'), 'r', encoding='utf-8') as f:
-    config = json.load(f)
-    config.update({
-        'attention_head_dim': 32,
-        'axes_dims_rope': [8, 12, 12],
-        'joint_attention_dim': 32,
-        'num_attention_heads': 1,
-        'num_layers': 2,
-    })
-    del config['pooled_projection_dim']  # not used
-    save_json(f'{save_folder}/transformer/config.json', config)
-    transformer_config = QwenImageTransformer2DModel.load_config(f'{save_folder}/transformer')
-    transformer = QwenImageTransformer2DModel.from_config(transformer_config)
-    init_weights(transformer)
-
-with open(hf_hub_download(source_model_id, filename='vae/config.json', repo_type='model'), 'r', encoding='utf-8') as f:
-    config = json.load(f)
-    config.update({
-        'num_res_blocks': 1,
-        'base_dim': 16,
-        'dim_mult': [1, 2, 4, 4],
-    })
-    del config['latents_mean']  # not used
-    del config['latents_std']  # not used
-    save_json(f'{save_folder}/vae/config.json', config)
-    vae_config = AutoencoderKLQwenImage.load_config(f'{save_folder}/vae')
-    vae = AutoencoderKLQwenImage.from_config(vae_config)
-    init_weights(vae)
-
-pipeline = QwenImagePipeline(
-    scheduler=scheduler,
-    text_encoder=text_encoder,
-    tokenizer=tokenizer,
-    transformer=transformer,
-    vae=vae,
-)
-pipeline = pipeline.to(torch.bfloat16)
-pipeline.save_pretrained(save_folder, safe_serialization=True)
-print(pipeline)
-```