管好【SD-WEBUI】中大量的模型：名称+预览图+备注+分组管理

大家好，欢迎来到IT知识分享网。

（零）前言

（一）模型管理

我们前面知道了通过生成按钮下面的桃红色显示附加网络按钮功能，可以用缩略图/卡牌的形式直观的管理这些模型。
卡牌默认很大但可以设置小一些。如下：

（1.1）模型名称（文件名）

.extra-network-cards .card .actions .name{ 
    font-size: 10px; /*font-weight: bold;*/ line-break: anywhere; } 

（1.2）模型缩略图

这段应该是大家都知道的内容：

• 点击一个基础模型的卡牌，相当于在基础模型下拉列表中选择它。
• 点击一个LoRA模型，相当于在当前输入框（提示词）中输入和启用这个模型，用于生成图片。

但如果你先把鼠标指向模型的名字，然后再指向替换预览图片（这里翻译可能不同，意思一样就行了）。
就可以把刚刚生成的图片，设置成为这个模型的预览图。

（1.3）模型备注文字

但是大家可能不知道，将一个模型同名的txt文件放入模型所在目录（如上图）。
那么这个文本文件中的内容，就将展示为模型的备注。

（1.4）模型分组（子目录）

（1.5）模型详细信息

好像没啥用，其实可以作为自己训练模型的参考。

{ 
    "ss_sd_model_name": "chilloutmix_NiPrunedFp32Fix.safetensors", "ss_resolution": "(1024, 576)", "ss_clip_skip": "2", "ss_num_train_images": "400", "ss_tag_frequency": { 
    "5_adaupper": { 
    "ada_re4": 46, "1girl": 46, "upper body": 46, "(solo:0.964)": 46, "black belt": 46, "(black background:1.3)": 46, "(simple background)": 46, "black gloves": 46, "black hair": 46, "medium breasts": 46, "handgun": 46, "handgun in hand": 46, "short hair": 46, "one arm lift": 46, "asian": 46, "red sweater": 46, "red leotard": 46 }, "5_adafull": { 
    "ada_re4": 34, "1 girl": 34, "solo": 34, "full body": 34, "(black background:1.3)": 34, "(simple background)": 34, "short hair": 34, "black hair": 34, "medium breasts": 34, "asian": 34, "red sweater": 34, "red leotard": 34, "long black leather boots": 34, "black belts": 34, "handgun": 34, "hand gun in hand": 34, "holster": 34, "black gloves": 34, "one arm lift": 34 } }, "ss_batch_size_per_device": "1", "ss_bucket_info": { 
    "buckets": { 
    "0": { 
    "resolution": [ 1024, 576 ], "count": 400 } }, "mean_img_ar_error": 0.0 }, "ss_bucket_no_upscale": "False", "ss_cache_latents": "True", "ss_caption_dropout_every_n_epochs": "0", "ss_caption_dropout_rate": "0.0", "ss_caption_tag_dropout_rate": "0.0", "ss_color_aug": "False", "ss_dataset_dirs": { 
    "5_adaupper": { 
    "n_repeats": 5, "img_count": 46 }, "5_adafull": { 
    "n_repeats": 5, "img_count": 34 } }, "ss_enable_bucket": "True", "ss_epoch": "12", "ss_face_crop_aug_range": "None", "ss_flip_aug": "False", "ss_full_fp16": "False", "ss_gradient_accumulation_steps": "1", "ss_gradient_checkpointing": "False", "ss_keep_tokens": "0", "ss_learning_rate": "0.0001", "ss_lowram": "False", "ss_lr_scheduler": "cosine_with_restarts", "ss_lr_warmup_steps": "0", "ss_max_bucket_reso": "1024", "ss_max_grad_norm": "1.0", "ss_max_token_length": "225", "ss_max_train_steps": "8000", "ss_min_bucket_reso": "256", "ss_mixed_precision": "fp16", "ss_network_alpha": "128.0", "ss_network_dim": "128", "ss_network_module": "networks.lora", "ss_new_sd_model_hash": "fca54c5e80b89ab03e0ab4b98d051ab187f92860f3cd664dc9d08b271", "ss_noise_offset": "0.0", "ss_num_batches_per_epoch": "400", "ss_num_epochs": "20", "ss_num_reg_images": "0", "ss_optimizer": "bitsandbytes.optim.adamw.AdamW8bit", "ss_output_name": "adawong", "ss_prior_loss_weight": "1.0", "ss_random_crop": "False", "ss_reg_dataset_dirs": { 
   }, "ss_sd_model_hash": "95f8d0a7", "ss_sd_scripts_commit_hash": "45945f698a7c57b83564e7ad3845c7ba5b", "ss_seed": "1337", "ss_session_id": "", "ss_shuffle_caption": "True", "ss_text_encoder_lr": "1e-05", "ss_total_batch_size": "1", "ss_training_comment": "None", "ss_training_finished_at": ".", "ss_training_started_at": ".", "ss_unet_lr": "0.0001", "ss_v2": "False", "sshs_legacy_hash": "6f4fe396", "sshs_model_hash": "482dc88150e305f3f19941d696d55e7be9f4db9361ca84c482a76a63fd6749ce" } 

（二）第二部分

（三）第三部分

🤪 to be continued…