End of training

README.md

@@ -8,6 +8,11 @@ tags:
 - text-to-image
 - diffusers
 - diffusers-training
+- stable-diffusion-xl
+- stable-diffusion-xl-diffusers
+- text-to-image
+- diffusers-training
+- diffusers
 inference: true
 ---
 

checkpoint-500/optimizer.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f9e6ab86922cb7b72c19bd045adf022fe9fe8111bf99d064dcefa8b961fbb070
-size 6876750164
+oid sha256:7c31f59f4b78a785597572a6dcb431e57ab65ae580da3d306026b69caf29301d
+size 20541173898

checkpoint-500/random_states_0.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:840ca222f2219dc9e25c50fa0a0887565341eef7af58dbffbbd38e7bc19cbfe6
+oid sha256:fba2490b444e3f7bc495867f068da5aefac52c301f12a4a277353371d86233b9
 size 14344

checkpoint-500/unet/config.json

@@ -1,17 +1,20 @@
 {
   "_class_name": "UNet2DConditionModel",
-  "_diffusers_version": "0.32.0.dev0",
-  "_name_or_path": "CompVis/stable-diffusion-v1-4",
+  "_diffusers_version": "0.31.0.dev0",
+  "_name_or_path": "stabilityai/stable-diffusion-xl-base-1.0",
   "act_fn": "silu",
-  "addition_embed_type": null,
+  "addition_embed_type": "text_time",
   "addition_embed_type_num_heads": 64,
-  "addition_time_embed_dim": null,
-  "attention_head_dim": 8,
+  "addition_time_embed_dim": 256,
+  "attention_head_dim": [
+    5,
+    10,
+    20
+  ],
   "attention_type": "default",
   "block_out_channels": [
     320,
     640,
-    1280,
     1280
   ],
   "center_input_sample": false,
@@ -19,13 +22,12 @@
   "class_embeddings_concat": false,
   "conv_in_kernel": 3,
   "conv_out_kernel": 3,
-  "cross_attention_dim": 768,
+  "cross_attention_dim": 2048,
   "cross_attention_norm": null,
   "down_block_types": [
+    "DownBlock2D",
     "CrossAttnDownBlock2D",
-    "CrossAttnDownBlock2D",
-    "CrossAttnDownBlock2D",
-    "DownBlock2D"
+    "CrossAttnDownBlock2D"
   ],
   "downsample_padding": 1,
   "dropout": 0.0,
@@ -45,24 +47,27 @@
   "num_class_embeds": null,
   "only_cross_attention": false,
   "out_channels": 4,
-  "projection_class_embeddings_input_dim": null,
+  "projection_class_embeddings_input_dim": 2816,
   "resnet_out_scale_factor": 1.0,
   "resnet_skip_time_act": false,
   "resnet_time_scale_shift": "default",
   "reverse_transformer_layers_per_block": null,
-  "sample_size": 64,
+  "sample_size": 128,
   "time_cond_proj_dim": null,
   "time_embedding_act_fn": null,
   "time_embedding_dim": null,
   "time_embedding_type": "positional",
   "timestep_post_act": null,
-  "transformer_layers_per_block": 1,
+  "transformer_layers_per_block": [
+    1,
+    2,
+    10
+  ],
   "up_block_types": [
-    "UpBlock2D",
     "CrossAttnUpBlock2D",
     "CrossAttnUpBlock2D",
-    "CrossAttnUpBlock2D"
+    "UpBlock2D"
   ],
-  "upcast_attention": false,
-  "use_linear_projection": false
+  "upcast_attention": null,
+  "use_linear_projection": true
 }

checkpoint-500/unet/diffusion_pytorch_model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43f198254173d0c698b86c3a66632a5b48f92342d9247d32323fe7e2eca92281
+size 9988139160

checkpoint-500/unet/diffusion_pytorch_model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da38e9c423bc2de37a0994b625352213f88ea99e62040805e7f99d55f26a3f78
+size 281938304

checkpoint-500/unet_ema/config.json

@@ -1,17 +1,20 @@
 {
   "_class_name": "UNet2DConditionModel",
-  "_diffusers_version": "0.32.0.dev0",
-  "_name_or_path": "CompVis/stable-diffusion-v1-4",
+  "_diffusers_version": "0.31.0.dev0",
+  "_name_or_path": "stabilityai/stable-diffusion-xl-base-1.0",
   "act_fn": "silu",
-  "addition_embed_type": null,
+  "addition_embed_type": "text_time",
   "addition_embed_type_num_heads": 64,
-  "addition_time_embed_dim": null,
-  "attention_head_dim": 8,
+  "addition_time_embed_dim": 256,
+  "attention_head_dim": [
+    5,
+    10,
+    20
+  ],
   "attention_type": "default",
   "block_out_channels": [
     320,
     640,
-    1280,
     1280
   ],
   "center_input_sample": false,
@@ -19,14 +22,13 @@
   "class_embeddings_concat": false,
   "conv_in_kernel": 3,
   "conv_out_kernel": 3,
-  "cross_attention_dim": 768,
+  "cross_attention_dim": 2048,
   "cross_attention_norm": null,
   "decay": 0.9999,
   "down_block_types": [
+    "DownBlock2D",
     "CrossAttnDownBlock2D",
-    "CrossAttnDownBlock2D",
-    "CrossAttnDownBlock2D",
-    "DownBlock2D"
+    "CrossAttnDownBlock2D"
   ],
   "downsample_padding": 1,
   "dropout": 0.0,
@@ -50,26 +52,29 @@
   "optimization_step": 500,
   "out_channels": 4,
   "power": 0.6666666666666666,
-  "projection_class_embeddings_input_dim": null,
+  "projection_class_embeddings_input_dim": 2816,
   "resnet_out_scale_factor": 1.0,
   "resnet_skip_time_act": false,
   "resnet_time_scale_shift": "default",
   "reverse_transformer_layers_per_block": null,
-  "sample_size": 64,
+  "sample_size": 128,
   "time_cond_proj_dim": null,
   "time_embedding_act_fn": null,
   "time_embedding_dim": null,
   "time_embedding_type": "positional",
   "timestep_post_act": null,
-  "transformer_layers_per_block": 1,
+  "transformer_layers_per_block": [
+    1,
+    2,
+    10
+  ],
   "up_block_types": [
-    "UpBlock2D",
     "CrossAttnUpBlock2D",
     "CrossAttnUpBlock2D",
-    "CrossAttnUpBlock2D"
+    "UpBlock2D"
   ],
-  "upcast_attention": false,
+  "upcast_attention": null,
   "update_after_step": 0,
   "use_ema_warmup": false,
-  "use_linear_projection": false
+  "use_linear_projection": true
 }

checkpoint-500/unet_ema/diffusion_pytorch_model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b331ad90ad28f1a76431966846cf0b4f2720c537486e1a9b9218cf7f2de18e61
+size 9988139160

checkpoint-500/unet_ema/diffusion_pytorch_model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87ecff22104df4101a137f89236d5c515ff4c5e2485b4d0974416aca7cfc8b69
+size 281938304

logs/text2image-fine-tune-sdxl/1732684849.244847/events.out.tfevents.1732684849.datasec2.engr.uconn.edu.3908411.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dca698a951687f906ae26bd108185d7498f6b3cceea8cb104196606826169ef1
+size 2833

logs/text2image-fine-tune-sdxl/1732684849.2467508/hparams.yml

@@ -0,0 +1,58 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+adam_epsilon: 1.0e-08
+adam_weight_decay: 0.01
+allow_tf32: false
+cache_dir: null
+caption_column: text
+center_crop: true
+checkpointing_steps: 500
+checkpoints_total_limit: null
+dataloader_num_workers: 0
+dataset_config_name: null
+dataset_name: haorandai/Nov_Clean_Bicycle_UF_1samples_with1constraints
+enable_npu_flash_attention: false
+enable_xformers_memory_efficient_attention: false
+gradient_accumulation_steps: 4
+gradient_checkpointing: true
+hub_model_id: null
+hub_token: null
+image_column: image
+learning_rate: 1.0e-05
+local_rank: -1
+logging_dir: logs
+lr_scheduler: constant
+lr_warmup_steps: 500
+max_grad_norm: 1.0
+max_train_samples: null
+max_train_steps: 500
+mixed_precision: fp16
+noise_offset: 0
+num_train_epochs: 500
+num_validation_images: 4
+output_dir: /scr/home/biw23007/haoran/nov25_finetune/clean_target/SDXL/png_finetune_xl_model_dir_1e-05_500_withconstraint/haorandai/Nov_Clean_Bicycle_UF_1samples_with1constraints/
+prediction_type: null
+pretrained_model_name_or_path: stabilityai/stable-diffusion-xl-base-1.0
+pretrained_vae_model_name_or_path: null
+proportion_empty_prompts: 0
+push_to_hub: true
+random_flip: true
+report_to: tensorboard
+resolution: 224
+resume_from_checkpoint: null
+revision: null
+scale_lr: false
+seed: null
+snr_gamma: null
+timestep_bias_begin: 0
+timestep_bias_end: 1000
+timestep_bias_multiplier: 1.0
+timestep_bias_portion: 0.25
+timestep_bias_strategy: none
+train_batch_size: 1
+train_data_dir: null
+use_8bit_adam: false
+use_ema: true
+validation_epochs: 1
+validation_prompt: null
+variant: null

logs/text2image-fine-tune-sdxl/events.out.tfevents.1732684849.datasec2.engr.uconn.edu.3908411.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53c43f8d5a05931aa6ff1f9698988aadffc083f0b4c9d1027cc312320abccdd8
+size 24461

model_index.json

@@ -1,32 +1,36 @@
 {
-  "_class_name": "StableDiffusionPipeline",
-  "_diffusers_version": "0.32.0.dev0",
-  "_name_or_path": "CompVis/stable-diffusion-v1-4",
+  "_class_name": "StableDiffusionXLPipeline",
+  "_diffusers_version": "0.31.0.dev0",
+  "_name_or_path": "stabilityai/stable-diffusion-xl-base-1.0",
   "feature_extractor": [
-    "transformers",
-    "CLIPImageProcessor"
+    null,
+    null
   ],
+  "force_zeros_for_empty_prompt": true,
   "image_encoder": [
     null,
     null
   ],
-  "requires_safety_checker": true,
-  "safety_checker": [
-    "stable_diffusion",
-    "StableDiffusionSafetyChecker"
-  ],
   "scheduler": [
     "diffusers",
-    "PNDMScheduler"
+    "EulerDiscreteScheduler"
   ],
   "text_encoder": [
     "transformers",
     "CLIPTextModel"
   ],
+  "text_encoder_2": [
+    "transformers",
+    "CLIPTextModelWithProjection"
+  ],
   "tokenizer": [
     "transformers",
     "CLIPTokenizer"
   ],
+  "tokenizer_2": [
+    "transformers",
+    "CLIPTokenizer"
+  ],
   "unet": [
     "diffusers",
     "UNet2DConditionModel"

scheduler/scheduler_config.json

@@ -1,15 +1,23 @@
 {
-  "_class_name": "PNDMScheduler",
-  "_diffusers_version": "0.32.0.dev0",
+  "_class_name": "EulerDiscreteScheduler",
+  "_diffusers_version": "0.31.0.dev0",
   "beta_end": 0.012,
   "beta_schedule": "scaled_linear",
   "beta_start": 0.00085,
   "clip_sample": false,
+  "final_sigmas_type": "zero",
+  "interpolation_type": "linear",
   "num_train_timesteps": 1000,
   "prediction_type": "epsilon",
+  "rescale_betas_zero_snr": false,
+  "sample_max_value": 1.0,
   "set_alpha_to_one": false,
+  "sigma_max": null,
+  "sigma_min": null,
   "skip_prk_steps": true,
   "steps_offset": 1,
   "timestep_spacing": "leading",
-  "trained_betas": null
+  "timestep_type": "discrete",
+  "trained_betas": null,
+  "use_karras_sigmas": false
 }

text_encoder/config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "CompVis/stable-diffusion-v1-4",
+  "_name_or_path": "/home/biw23007/.cache/huggingface/hub/models--stabilityai--stable-diffusion-xl-base-1.0/snapshots/462165984030d82259a11f4367a4eed129e94a7b/text_encoder",
   "architectures": [
     "CLIPTextModel"
   ],
@@ -18,7 +18,7 @@
   "num_attention_heads": 12,
   "num_hidden_layers": 12,
   "pad_token_id": 1,
-  "projection_dim": 512,
+  "projection_dim": 768,
   "torch_dtype": "float16",
   "transformers_version": "4.44.2",
   "vocab_size": 49408

text_encoder_2/config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "/home/biw23007/.cache/huggingface/hub/models--stabilityai--stable-diffusion-xl-base-1.0/snapshots/462165984030d82259a11f4367a4eed129e94a7b/text_encoder_2",
+  "architectures": [
+    "CLIPTextModelWithProjection"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "dropout": 0.0,
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_size": 1280,
+  "initializer_factor": 1.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 5120,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 77,
+  "model_type": "clip_text_model",
+  "num_attention_heads": 20,
+  "num_hidden_layers": 32,
+  "pad_token_id": 1,
+  "projection_dim": 1280,
+  "torch_dtype": "float16",
+  "transformers_version": "4.44.2",
+  "vocab_size": 49408
+}

text_encoder_2/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec310df2af79c318e24d20511b601a591ca8cd4f1fce1d8dff822a356bcdb1f4
+size 1389382176

tokenizer_2/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": "!",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_2/tokenizer_config.json

@@ -0,0 +1,38 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "!",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "49406": {
+      "content": "<|startoftext|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "49407": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|startoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "do_lower_case": true,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 77,
+  "pad_token": "!",
+  "tokenizer_class": "CLIPTokenizer",
+  "unk_token": "<|endoftext|>"
+}

unet/config.json

@@ -1,17 +1,20 @@
 {
   "_class_name": "UNet2DConditionModel",
-  "_diffusers_version": "0.32.0.dev0",
-  "_name_or_path": "CompVis/stable-diffusion-v1-4",
+  "_diffusers_version": "0.31.0.dev0",
+  "_name_or_path": "stabilityai/stable-diffusion-xl-base-1.0",
   "act_fn": "silu",
-  "addition_embed_type": null,
+  "addition_embed_type": "text_time",
   "addition_embed_type_num_heads": 64,
-  "addition_time_embed_dim": null,
-  "attention_head_dim": 8,
+  "addition_time_embed_dim": 256,
+  "attention_head_dim": [
+    5,
+    10,
+    20
+  ],
   "attention_type": "default",
   "block_out_channels": [
     320,
     640,
-    1280,
     1280
   ],
   "center_input_sample": false,
@@ -19,13 +22,12 @@
   "class_embeddings_concat": false,
   "conv_in_kernel": 3,
   "conv_out_kernel": 3,
-  "cross_attention_dim": 768,
+  "cross_attention_dim": 2048,
   "cross_attention_norm": null,
   "down_block_types": [
+    "DownBlock2D",
     "CrossAttnDownBlock2D",
-    "CrossAttnDownBlock2D",
-    "CrossAttnDownBlock2D",
-    "DownBlock2D"
+    "CrossAttnDownBlock2D"
   ],
   "downsample_padding": 1,
   "dropout": 0.0,
@@ -45,24 +47,27 @@
   "num_class_embeds": null,
   "only_cross_attention": false,
   "out_channels": 4,
-  "projection_class_embeddings_input_dim": null,
+  "projection_class_embeddings_input_dim": 2816,
   "resnet_out_scale_factor": 1.0,
   "resnet_skip_time_act": false,
   "resnet_time_scale_shift": "default",
   "reverse_transformer_layers_per_block": null,
-  "sample_size": 64,
+  "sample_size": 128,
   "time_cond_proj_dim": null,
   "time_embedding_act_fn": null,
   "time_embedding_dim": null,
   "time_embedding_type": "positional",
   "timestep_post_act": null,
-  "transformer_layers_per_block": 1,
+  "transformer_layers_per_block": [
+    1,
+    2,
+    10
+  ],
   "up_block_types": [
-    "UpBlock2D",
     "CrossAttnUpBlock2D",
     "CrossAttnUpBlock2D",
-    "CrossAttnUpBlock2D"
+    "UpBlock2D"
   ],
-  "upcast_attention": false,
-  "use_linear_projection": false
+  "upcast_attention": null,
+  "use_linear_projection": true
 }

unet/diffusion_pytorch_model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b331ad90ad28f1a76431966846cf0b4f2720c537486e1a9b9218cf7f2de18e61
+size 9988139160

unet/diffusion_pytorch_model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87ecff22104df4101a137f89236d5c515ff4c5e2485b4d0974416aca7cfc8b69
+size 281938304

vae/config.json

@@ -1,7 +1,7 @@
 {
   "_class_name": "AutoencoderKL",
-  "_diffusers_version": "0.32.0.dev0",
-  "_name_or_path": "CompVis/stable-diffusion-v1-4",
+  "_diffusers_version": "0.31.0.dev0",
+  "_name_or_path": "stabilityai/stable-diffusion-xl-base-1.0",
   "act_fn": "silu",
   "block_out_channels": [
     128,
@@ -24,8 +24,8 @@
   "mid_block_add_attention": true,
   "norm_num_groups": 32,
   "out_channels": 3,
-  "sample_size": 512,
-  "scaling_factor": 0.18215,
+  "sample_size": 1024,
+  "scaling_factor": 0.13025,
   "shift_factor": null,
   "up_block_types": [
     "UpDecoderBlock2D",

vae/diffusion_pytorch_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4fbcf0ebe55a0984f5a5e00d8c4521d52359af7229bb4d81890039d2aa16dd7c
+oid sha256:bcb60880a46b63dea58e9bc591abe15f8350bde47b405f9c38f4be70c6161e68
 size 167335342