Hugging Face's logo

kmpartner
/
bkv2tpcmlr-test

PEFT
TensorBoard
Diffusers
Safetensors
Model card Files
xet
 Metrics Community
kmpartner commited on
Commit
3c3724d
·
verified ·
1 Parent(s): 8dae345

Upload folder using huggingface_hub

Browse files
Files changed (14) hide show
  1. checkpoint-500/optimizer.bin +3 -0
  2. checkpoint-500/random_states_0.pkl +3 -0
  3. checkpoint-500/scaler.pt +3 -0
  4. checkpoint-500/scheduler.bin +3 -0
  5. checkpoint-500/unet/config.json +69 -0
  6. checkpoint-500/unet/diffusion_pytorch_model.safetensors +3 -0
  7. checkpoint-500/unet_target/config.json +69 -0
  8. checkpoint-500/unet_target/diffusion_pytorch_model.safetensors +3 -0
  9. logs/text2image-fine-tune/1757230976.8246334/events.out.tfevents.1757230976.aa48bcb8bd56.3032.1 +3 -0
  10. logs/text2image-fine-tune/1757230976.8271024/hparams.yml +51 -0
  11. logs/text2image-fine-tune/1757231048.9554732/events.out.tfevents.1757231048.aa48bcb8bd56.3460.1 +3 -0
  12. logs/text2image-fine-tune/1757231048.9570954/hparams.yml +51 -0
  13. logs/text2image-fine-tune/events.out.tfevents.1757230976.aa48bcb8bd56.3032.0 +3 -0
  14. logs/text2image-fine-tune/events.out.tfevents.1757231048.aa48bcb8bd56.3460.0 +3 -0
checkpoint-500/optimizer.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d655547e98cab8223205a8f521af5feb51708923dc955e7897ee9226db8b1a60
3
+ size 656198874
checkpoint-500/random_states_0.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01c1a38ecc8d1d80ff588111e035e28959c5fdb7ab7b0665af1191f95f26d925
3
+ size 14408
checkpoint-500/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:18b984273ea2d45b7ffb1d047bb359d93111e41fcad70d16a1b453fd38f72636
3
+ size 988
checkpoint-500/scheduler.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65115106fbd9463fa6c7133b543dece71c31fcba363ff15b8c73f8c7eb0739c5
3
+ size 1000
checkpoint-500/unet/config.json ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "UNet2DConditionModel",
3
+ "_diffusers_version": "0.26.3",
4
+ "_name_or_path": "nota-ai/bk-sdm-v2-tiny",
5
+ "act_fn": "silu",
6
+ "addition_embed_type": null,
7
+ "addition_embed_type_num_heads": 64,
8
+ "addition_time_embed_dim": null,
9
+ "attention_head_dim": [
10
+ 5,
11
+ 10,
12
+ 20
13
+ ],
14
+ "attention_type": "default",
15
+ "block_out_channels": [
16
+ 320,
17
+ 640,
18
+ 1280
19
+ ],
20
+ "center_input_sample": false,
21
+ "class_embed_type": null,
22
+ "class_embeddings_concat": false,
23
+ "conv_in_kernel": 3,
24
+ "conv_out_kernel": 3,
25
+ "cross_attention_dim": 1024,
26
+ "cross_attention_norm": null,
27
+ "down_block_types": [
28
+ "CrossAttnDownBlock2D",
29
+ "CrossAttnDownBlock2D",
30
+ "CrossAttnDownBlock2D"
31
+ ],
32
+ "downsample_padding": 1,
33
+ "dropout": 0.0,
34
+ "dual_cross_attention": false,
35
+ "encoder_hid_dim": null,
36
+ "encoder_hid_dim_type": null,
37
+ "flip_sin_to_cos": true,
38
+ "freq_shift": 0,
39
+ "in_channels": 4,
40
+ "layers_per_block": 1,
41
+ "mid_block_only_cross_attention": null,
42
+ "mid_block_scale_factor": 1,
43
+ "mid_block_type": null,
44
+ "norm_eps": 1e-05,
45
+ "norm_num_groups": 32,
46
+ "num_attention_heads": null,
47
+ "num_class_embeds": null,
48
+ "only_cross_attention": false,
49
+ "out_channels": 4,
50
+ "projection_class_embeddings_input_dim": null,
51
+ "resnet_out_scale_factor": 1.0,
52
+ "resnet_skip_time_act": false,
53
+ "resnet_time_scale_shift": "default",
54
+ "reverse_transformer_layers_per_block": null,
55
+ "sample_size": 64,
56
+ "time_cond_proj_dim": null,
57
+ "time_embedding_act_fn": null,
58
+ "time_embedding_dim": null,
59
+ "time_embedding_type": "positional",
60
+ "timestep_post_act": null,
61
+ "transformer_layers_per_block": 1,
62
+ "up_block_types": [
63
+ "CrossAttnUpBlock2D",
64
+ "CrossAttnUpBlock2D",
65
+ "CrossAttnUpBlock2D"
66
+ ],
67
+ "upcast_attention": false,
68
+ "use_linear_projection": true
69
+ }
checkpoint-500/unet/diffusion_pytorch_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ccf0d8b2b6d9efb893569047de775c97d36de67682a19242c12e414fc1b33bba
3
+ size 1307346120
checkpoint-500/unet_target/config.json ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "UNet2DConditionModel",
3
+ "_diffusers_version": "0.26.3",
4
+ "_name_or_path": "nota-ai/bk-sdm-v2-tiny",
5
+ "act_fn": "silu",
6
+ "addition_embed_type": null,
7
+ "addition_embed_type_num_heads": 64,
8
+ "addition_time_embed_dim": null,
9
+ "attention_head_dim": [
10
+ 5,
11
+ 10,
12
+ 20
13
+ ],
14
+ "attention_type": "default",
15
+ "block_out_channels": [
16
+ 320,
17
+ 640,
18
+ 1280
19
+ ],
20
+ "center_input_sample": false,
21
+ "class_embed_type": null,
22
+ "class_embeddings_concat": false,
23
+ "conv_in_kernel": 3,
24
+ "conv_out_kernel": 3,
25
+ "cross_attention_dim": 1024,
26
+ "cross_attention_norm": null,
27
+ "down_block_types": [
28
+ "CrossAttnDownBlock2D",
29
+ "CrossAttnDownBlock2D",
30
+ "CrossAttnDownBlock2D"
31
+ ],
32
+ "downsample_padding": 1,
33
+ "dropout": 0.0,
34
+ "dual_cross_attention": false,
35
+ "encoder_hid_dim": null,
36
+ "encoder_hid_dim_type": null,
37
+ "flip_sin_to_cos": true,
38
+ "freq_shift": 0,
39
+ "in_channels": 4,
40
+ "layers_per_block": 1,
41
+ "mid_block_only_cross_attention": null,
42
+ "mid_block_scale_factor": 1,
43
+ "mid_block_type": null,
44
+ "norm_eps": 1e-05,
45
+ "norm_num_groups": 32,
46
+ "num_attention_heads": null,
47
+ "num_class_embeds": null,
48
+ "only_cross_attention": false,
49
+ "out_channels": 4,
50
+ "projection_class_embeddings_input_dim": null,
51
+ "resnet_out_scale_factor": 1.0,
52
+ "resnet_skip_time_act": false,
53
+ "resnet_time_scale_shift": "default",
54
+ "reverse_transformer_layers_per_block": null,
55
+ "sample_size": 64,
56
+ "time_cond_proj_dim": null,
57
+ "time_embedding_act_fn": null,
58
+ "time_embedding_dim": null,
59
+ "time_embedding_type": "positional",
60
+ "timestep_post_act": null,
61
+ "transformer_layers_per_block": 1,
62
+ "up_block_types": [
63
+ "CrossAttnUpBlock2D",
64
+ "CrossAttnUpBlock2D",
65
+ "CrossAttnUpBlock2D"
66
+ ],
67
+ "upcast_attention": false,
68
+ "use_linear_projection": true
69
+ }
checkpoint-500/unet_target/diffusion_pytorch_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a952089fc056f156973e957af76e7ce19f49f3e8b80e1f70e33bcff9ece634cd
3
+ size 1307346120
logs/text2image-fine-tune/1757230976.8246334/events.out.tfevents.1757230976.aa48bcb8bd56.3032.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:28fb480648d568d7955d4ed575425d73a11dfbe9b808c5599e0a353e7385e87d
3
+ size 2524
logs/text2image-fine-tune/1757230976.8271024/hparams.yml ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ adam_beta1: 0.9
2
+ adam_beta2: 0.999
3
+ adam_epsilon: 1.0e-08
4
+ adam_weight_decay: 0.001
5
+ allow_tf32: false
6
+ cache_dir: null
7
+ cast_teacher_unet: false
8
+ center_crop: false
9
+ checkpointing_steps: 500
10
+ checkpoints_total_limit: 10
11
+ dataloader_num_workers: 1
12
+ enable_xformers_memory_efficient_attention: true
13
+ gradient_accumulation_steps: 4
14
+ gradient_checkpointing: true
15
+ hub_model_id: null
16
+ hub_token: null
17
+ huber_c: 0.001
18
+ learning_rate: 5.0e-06
19
+ local_rank: -1
20
+ logging_dir: logs
21
+ lora_rank: 64
22
+ loss_type: huber
23
+ lr_scheduler: constant
24
+ lr_warmup_steps: 500
25
+ max_grad_norm: 1.0
26
+ max_train_samples: 4000000
27
+ max_train_steps: 500
28
+ mixed_precision: fp16
29
+ multiphase: 1
30
+ not_apply_cfg_solver: false
31
+ num_ddim_timesteps: 50
32
+ num_train_epochs: 1
33
+ output_dir: bkv2tpcmlr-test
34
+ pretrained_teacher_model: nota-ai/bk-sdm-v2-tiny
35
+ pretrained_vae_model_name_or_path: null
36
+ proportion_empty_prompts: 0
37
+ push_to_hub: false
38
+ random_flip: false
39
+ report_to: tensorboard
40
+ resolution: 512
41
+ resume_from_checkpoint: null
42
+ revision: null
43
+ scale_lr: false
44
+ seed: null
45
+ teacher_revision: null
46
+ tracker_project_name: text2image-fine-tune
47
+ train_batch_size: 4
48
+ use_8bit_adam: true
49
+ validation_steps: 5000000
50
+ w_max: 5.0
51
+ w_min: 4.0
logs/text2image-fine-tune/1757231048.9554732/events.out.tfevents.1757231048.aa48bcb8bd56.3460.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8037e49e07c4f32034bfac6c021553d76ee0b4752819d7ea12c6f4b6dab2a86
3
+ size 2524
logs/text2image-fine-tune/1757231048.9570954/hparams.yml ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ adam_beta1: 0.9
2
+ adam_beta2: 0.999
3
+ adam_epsilon: 1.0e-08
4
+ adam_weight_decay: 0.001
5
+ allow_tf32: false
6
+ cache_dir: null
7
+ cast_teacher_unet: false
8
+ center_crop: false
9
+ checkpointing_steps: 500
10
+ checkpoints_total_limit: 10
11
+ dataloader_num_workers: 1
12
+ enable_xformers_memory_efficient_attention: true
13
+ gradient_accumulation_steps: 1
14
+ gradient_checkpointing: true
15
+ hub_model_id: null
16
+ hub_token: null
17
+ huber_c: 0.001
18
+ learning_rate: 5.0e-06
19
+ local_rank: -1
20
+ logging_dir: logs
21
+ lora_rank: 64
22
+ loss_type: huber
23
+ lr_scheduler: constant
24
+ lr_warmup_steps: 500
25
+ max_grad_norm: 1.0
26
+ max_train_samples: 4000000
27
+ max_train_steps: 500
28
+ mixed_precision: fp16
29
+ multiphase: 1
30
+ not_apply_cfg_solver: false
31
+ num_ddim_timesteps: 50
32
+ num_train_epochs: 1
33
+ output_dir: bkv2tpcmlr-test
34
+ pretrained_teacher_model: nota-ai/bk-sdm-v2-tiny
35
+ pretrained_vae_model_name_or_path: null
36
+ proportion_empty_prompts: 0
37
+ push_to_hub: false
38
+ random_flip: false
39
+ report_to: tensorboard
40
+ resolution: 512
41
+ resume_from_checkpoint: null
42
+ revision: null
43
+ scale_lr: false
44
+ seed: null
45
+ teacher_revision: null
46
+ tracker_project_name: text2image-fine-tune
47
+ train_batch_size: 16
48
+ use_8bit_adam: true
49
+ validation_steps: 5000000
50
+ w_max: 5.0
51
+ w_min: 4.0
logs/text2image-fine-tune/events.out.tfevents.1757230976.aa48bcb8bd56.3032.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4e1b25c160fbdff01aed4000de270d356797972a5191fc5abac1a3487887ce5c
3
+ size 1634
logs/text2image-fine-tune/events.out.tfevents.1757231048.aa48bcb8bd56.3460.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12c1b3afa06eb85cd0ce55207a16fa0cec7d0ed059409e18a0d7c6e307867717
3
+ size 41834