Upload folder using huggingface_hub

checkpoint-12000/heads.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1eeadf82255b8e4c157f03d7a92b3398b2a03fb33753f607b82c7aaaba55964c
+size 30420614

checkpoint-12000/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dcd3186c5146819f452753183373ed914ebd92c856d8956c8e4950f9d838f50
+size 1496291698

checkpoint-12000/optimizer_1.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:287f06e3484097249b37cfde128263f7b2b583eb67442c97e9fbe5b44bc868f0
+size 15431330

checkpoint-12000/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:494cf9b7998d4d9f6bc1837ff6675b1a702ae9c2379ecbf1b3168b308cf53f89
+size 14408

checkpoint-12000/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45e321b0253d10a375f8e12d6f89a2c3428616cdedd4f712dbf5245439f903bb
+size 988

checkpoint-12000/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6bb6146ab99be3f9944095fc78ca3055d6ae4580925c82421ef7024e377e4b3
+size 1000

checkpoint-12000/unet/config.json

@@ -0,0 +1,73 @@
+{
+  "_class_name": "UNet2DConditionModel",
+  "_diffusers_version": "0.26.3",
+  "_name_or_path": "k5pcmlra4-test",
+  "act_fn": "silu",
+  "addition_embed_type": "text_time",
+  "addition_embed_type_num_heads": 64,
+  "addition_time_embed_dim": 256,
+  "attention_head_dim": [
+    5,
+    10,
+    20
+  ],
+  "attention_type": "default",
+  "block_out_channels": [
+    320,
+    640,
+    1280
+  ],
+  "center_input_sample": false,
+  "class_embed_type": null,
+  "class_embeddings_concat": false,
+  "conv_in_kernel": 3,
+  "conv_out_kernel": 3,
+  "cross_attention_dim": 2048,
+  "cross_attention_norm": null,
+  "down_block_types": [
+    "DownBlock2D",
+    "CrossAttnDownBlock2D",
+    "CrossAttnDownBlock2D"
+  ],
+  "downsample_padding": 1,
+  "dropout": 0.0,
+  "dual_cross_attention": false,
+  "encoder_hid_dim": null,
+  "encoder_hid_dim_type": null,
+  "flip_sin_to_cos": true,
+  "freq_shift": 0,
+  "in_channels": 4,
+  "layers_per_block": 2,
+  "mid_block_only_cross_attention": null,
+  "mid_block_scale_factor": 1,
+  "mid_block_type": "UNetMidBlock2D",
+  "norm_eps": 1e-05,
+  "norm_num_groups": 32,
+  "num_attention_heads": null,
+  "num_class_embeds": null,
+  "only_cross_attention": false,
+  "out_channels": 4,
+  "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": 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,
+    1,
+    2
+  ],
+  "up_block_types": [
+    "CrossAttnUpBlock2D",
+    "CrossAttnUpBlock2D",
+    "UpBlock2D"
+  ],
+  "upcast_attention": null,
+  "use_linear_projection": true
+}

checkpoint-12000/unet/diffusion_pytorch_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97a2f97b6370a83a7154af4bb0ecc29d099da2476e21145ddacdec59f0b46d50
+size 2980972328

logs/text2image-fine-tune/1761896259.308622/events.out.tfevents.1761896259.8cc40be14ed0.3620.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2655ece0a80473e2986e84d1345543a607ea4101cbafb8c4a13826d84d20a721
+size 2929

logs/text2image-fine-tune/1761896259.3188834/hparams.yml

@@ -0,0 +1,58 @@
+adam_beta1: 0.9
+adam_beta2: 0.999
+adam_epsilon: 1.0e-08
+adam_weight_decay: 0.0
+adv_lr: 1.0e-05
+adv_weight: 0.025
+allow_tf32: false
+cache_dir: null
+cast_teacher_unet: false
+center_crop: false
+checkpointing_steps: 400
+checkpoints_total_limit: null
+dataloader_num_workers: 1
+ema_decay: 0.95
+enable_xformers_memory_efficient_attention: true
+gradient_accumulation_steps: 5
+gradient_checkpointing: true
+hub_model_id: null
+hub_token: null
+huber_c: 0.001
+learning_rate: 2.0e-06
+local_rank: -1
+logging_dir: logs
+lora_rank: 64
+loss_type: huber
+lr_scheduler: constant
+lr_warmup_steps: 500
+max_grad_norm: 1.0
+max_train_samples: 50000000
+max_train_steps: 12000
+mixed_precision: fp16
+num_ddim_timesteps: 250
+num_inference_steps_max: 6
+num_inference_steps_min: 1
+num_train_epochs: 30
+output_dir: k5pcmlra4-test
+pretrained_teacher_model: segmind/Segmind-Vega
+pretrained_vae_model_name_or_path: madebyollin/sdxl-vae-fp16-fix
+proportion_empty_prompts: 0.2
+push_to_hub: false
+random_flip: false
+report_to: tensorboard
+resolution: 512
+resume_from_checkpoint: checkpoint-11600
+revision: null
+s_ratio: 0.3
+scale_lr: false
+seed: null
+teacher_revision: null
+tracker_project_name: text2image-fine-tune
+train_batch_size: 4
+train_shards_path_or_url: null
+use_8bit_adam: true
+use_fix_crop_and_size: true
+val_infer_step: 4
+validation_steps: 500000
+w_max: 9.0
+w_min: 2.0

logs/text2image-fine-tune/events.out.tfevents.1761896259.8cc40be14ed0.3620.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:100248475125b9297e677ce01901c0dfe30a7829dbe979dd27f86de532c7da7f
+size 43688