JLaw

args.json

@@ -0,0 +1,60 @@
+{
+  "pretrained_model_name_or_path": "/opt/ml/stable-diffusion-v1-5",
+  "pretrained_vae_name_or_path": "stabilityai/sd-vae-ft-mse",
+  "revision": "fp16",
+  "tokenizer_name": null,
+  "instance_data_dir": null,
+  "class_data_dir": null,
+  "instance_prompt": "photo of xyz person",
+  "class_prompt": "photo of a person",
+  "save_sample_prompt": null,
+  "save_sample_negative_prompt": null,
+  "n_save_sample": 4,
+  "save_guidance_scale": 7.5,
+  "save_infer_steps": 50,
+  "pad_tokens": false,
+  "with_prior_preservation": true,
+  "prior_loss_weight": 1.0,
+  "num_class_images": 50,
+  "output_dir": "/opt/ml/model",
+  "image_output_dir": "image-samples",
+  "seed": 15213,
+  "resolution": 512,
+  "center_crop": false,
+  "train_text_encoder": true,
+  "train_batch_size": 2,
+  "sample_batch_size": 4,
+  "num_train_epochs": 60,
+  "max_train_steps": 1500,
+  "gradient_accumulation_steps": 1,
+  "gradient_checkpointing": true,
+  "learning_rate": 1e-06,
+  "scale_lr": false,
+  "lr_scheduler": "constant",
+  "lr_warmup_steps": 0,
+  "use_8bit_adam": true,
+  "adam_beta1": 0.9,
+  "adam_beta2": 0.999,
+  "adam_weight_decay": 0.01,
+  "adam_epsilon": 1e-08,
+  "max_grad_norm": 1.0,
+  "push_to_hub": false,
+  "hub_token": null,
+  "hub_model_id": null,
+  "logging_dir": "logs",
+  "log_interval": 10,
+  "save_interval": 10000,
+  "save_min_steps": 0,
+  "mixed_precision": "fp16",
+  "not_cache_latents": false,
+  "hflip": false,
+  "local_rank": -1,
+  "concepts_list": [
+    {
+      "instance_prompt": "photo of zwx model",
+      "class_prompt": "photo of a model",
+      "instance_data_dir": "/opt/ml/input/data/zwx_jl",
+      "class_data_dir": "/opt/ml/input/data/female_jl"
+    }
+  ]
+}

code/inference.py

@@ -0,0 +1,53 @@
+"""
+Additional inference script for stable diffusion
+Edited by Xiaoxiao
+"""
+
+from diffusers import StableDiffusionPipeline, DDIMScheduler
+import torch
+import base64
+import numpy as np
+
+def process_data(data: dict) -> dict:
+    g_cuda = None
+    g_cuda = torch.Generator(device='cuda')
+    
+    return {
+        "prompt": data.pop("prompt", data),
+        "negative_prompt": data.pop("negative_prompt", ""),
+        "num_images_per_prompt": min(data.pop("num_samples", 2),5),
+        "guidance_scale": data.pop("guidance_scale", 7.5),
+        "num_inference_steps": min(data.pop("num_inference_steps", 50), 50),
+        "height": 512,
+        "width": 512,
+        "generator":g_cuda.manual_seed(data.pop("seed",15213))
+    }
+
+
+def model_fn(model_dir: str):
+    scheduler = DDIMScheduler(beta_start=0.00085, beta_end=0.012, beta_schedule="scaled_linear", clip_sample=False, set_alpha_to_one=False)
+    t2i_pipe = StableDiffusionPipeline.from_pretrained(
+        model_dir,
+        scheduler=scheduler, 
+        safety_checker=None, 
+        torch_dtype=torch.float16
+    )
+    if torch.cuda.is_available():
+        t2i_pipe = t2i_pipe.to("cuda")
+
+    t2i_pipe.enable_attention_slicing()
+    return t2i_pipe
+
+
+def predict_fn(data: dict, hgf_pipe) -> dict:
+
+    with torch.autocast("cuda"):
+        images = hgf_pipe(**process_data(data))["images"]
+
+    # return dictionary, which will be json serializable
+    return {
+        "images": [
+            base64.b64encode(np.array(image).astype(np.uint8)).decode("utf-8")
+            for image in images
+        ]
+    }

code/requirements.txt

@@ -0,0 +1,7 @@
+numpy==1.23.4
+torch==1.12.1
+diffusers==0.9.0
+transformers==4.23.1
+spacy==3.4.2
+ftfy==6.1.1
+huggingface-hub==0.11.1

model_index.json

@@ -0,0 +1,25 @@
+{
+  "_class_name": "StableDiffusionPipeline",
+  "_diffusers_version": "0.9.0",
+  "requires_safety_checker": false,
+  "scheduler": [
+    "diffusers",
+    "DDIMScheduler"
+  ],
+  "text_encoder": [
+    "transformers",
+    "CLIPTextModel"
+  ],
+  "tokenizer": [
+    "transformers",
+    "CLIPTokenizer"
+  ],
+  "unet": [
+    "diffusers",
+    "UNet2DConditionModel"
+  ],
+  "vae": [
+    "diffusers",
+    "AutoencoderKL"
+  ]
+}

scheduler/scheduler_config.json

@@ -0,0 +1,13 @@
+{
+  "_class_name": "DDIMScheduler",
+  "_diffusers_version": "0.9.0",
+  "beta_end": 0.012,
+  "beta_schedule": "scaled_linear",
+  "beta_start": 0.00085,
+  "clip_sample": false,
+  "num_train_timesteps": 1000,
+  "prediction_type": "epsilon",
+  "set_alpha_to_one": false,
+  "steps_offset": 1,
+  "trained_betas": null
+}

text_encoder/config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "/opt/ml/stable-diffusion-v1-5",
+  "architectures": [
+    "CLIPTextModel"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "dropout": 0.0,
+  "eos_token_id": 2,
+  "hidden_act": "quick_gelu",
+  "hidden_size": 768,
+  "initializer_factor": 1.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 77,
+  "model_type": "clip_text_model",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 1,
+  "projection_dim": 768,
+  "torch_dtype": "float32",
+  "transformers_version": "4.23.1",
+  "vocab_size": 49408
+}

text_encoder/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a6edb965c48d852c06dffe144c820a54ebc69fbc6950d69265654648a22a42b
+size 492308087

tokenizer/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<|startoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|endoftext|>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer/tokenizer_config.json

@@ -0,0 +1,34 @@
+{
+  "add_prefix_space": false,
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "<|startoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "do_lower_case": true,
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "errors": "replace",
+  "model_max_length": 77,
+  "name_or_path": "/opt/ml/stable-diffusion-v1-5/tokenizer",
+  "pad_token": "<|endoftext|>",
+  "special_tokens_map_file": "./special_tokens_map.json",
+  "tokenizer_class": "CLIPTokenizer",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

unet/diffusion_pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edce6d1f2df31d7a12d0242225853097e7e0f6f69ce97b8cb26a2e27f3e4318c
+size 1013039104