Hugging Face's logo

zhiyang1
/
blip3o_f32c32-sana-1.0-768_patch-1_group-7_fusion_residual_attn

Safetensors
Model card Files
xet
Community
zhiyang1 commited on
Commit
7e33131
·
verified ·
1 Parent(s): 0c5a3c5

Upload folder using huggingface_hub

Browse files
Files changed (27) hide show
  1. .gitattributes +2 -0
  2. checkpoint-208000/config.json +79 -0
  3. checkpoint-208000/generation_config.json +6 -0
  4. checkpoint-208000/model-00001-of-00002.safetensors +3 -0
  5. checkpoint-208000/model-00002-of-00002.safetensors +3 -0
  6. checkpoint-208000/model.safetensors.index.json +0 -0
  7. checkpoint-208000/optimizer.pt +3 -0
  8. checkpoint-208000/rng_state_0.pth +3 -0
  9. checkpoint-208000/rng_state_1.pth +3 -0
  10. checkpoint-208000/rng_state_2.pth +3 -0
  11. checkpoint-208000/rng_state_3.pth +3 -0
  12. checkpoint-208000/scheduler.pt +3 -0
  13. checkpoint-208000/trainer_state.json +3 -0
  14. checkpoint-208000/training_args.bin +3 -0
  15. checkpoint-210000/config.json +79 -0
  16. checkpoint-210000/generation_config.json +6 -0
  17. checkpoint-210000/model-00001-of-00002.safetensors +3 -0
  18. checkpoint-210000/model-00002-of-00002.safetensors +3 -0
  19. checkpoint-210000/model.safetensors.index.json +0 -0
  20. checkpoint-210000/optimizer.pt +3 -0
  21. checkpoint-210000/rng_state_0.pth +3 -0
  22. checkpoint-210000/rng_state_1.pth +3 -0
  23. checkpoint-210000/rng_state_2.pth +3 -0
  24. checkpoint-210000/rng_state_3.pth +3 -0
  25. checkpoint-210000/scheduler.pt +3 -0
  26. checkpoint-210000/trainer_state.json +3 -0
  27. checkpoint-210000/training_args.bin +3 -0
.gitattributes CHANGED
@@ -101,3 +101,5 @@ checkpoint-198000/coco_steps-40_cfg-5_0/coco_8_LOOKS[[:space:]]LIKE[[:space:]]A[
101
  checkpoint-198000/coco_steps-40_cfg-5_0/coco_9_Three[[:space:]]people[[:space:]]are[[:space:]]walking[[:space:]]down[[:space:]]the[[:space:]]road[[:space:]]with[[:space:]]three[[:space:]]horses._1.png filter=lfs diff=lfs merge=lfs -text
102
  checkpoint-198000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
103
  checkpoint-200000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
 
 
 
101
  checkpoint-198000/coco_steps-40_cfg-5_0/coco_9_Three[[:space:]]people[[:space:]]are[[:space:]]walking[[:space:]]down[[:space:]]the[[:space:]]road[[:space:]]with[[:space:]]three[[:space:]]horses._1.png filter=lfs diff=lfs merge=lfs -text
102
  checkpoint-198000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
103
  checkpoint-200000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
104
+ checkpoint-208000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
105
+ checkpoint-210000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
checkpoint-208000/config.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ar_steps": 1,
3
+ "architectures": [
4
+ "DiffVLMDiffusion"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "condition_layer": -1,
9
+ "eos_token_id": 151645,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 1536,
12
+ "image_token_id": 151655,
13
+ "img_cross_attention_dim": 2048,
14
+ "img_diffuser_depth": 1,
15
+ "img_ffn_dim_multiplier": null,
16
+ "img_hidden_size": 1536,
17
+ "img_multiple_of": 256,
18
+ "img_norm_eps": 1e-05,
19
+ "img_num_attention_heads": 12,
20
+ "img_num_kv_heads": 12,
21
+ "img_qk_norm": true,
22
+ "in_channels": 32,
23
+ "initializer_range": 0.02,
24
+ "inject_img_diffuser": false,
25
+ "input_size": 32,
26
+ "intermediate_size": 8960,
27
+ "layer_group_size": 7,
28
+ "layerwise_start_idx": 0,
29
+ "lora_alpha": 16,
30
+ "lora_bias": "none",
31
+ "lora_dropout": 0.05,
32
+ "lora_enable": false,
33
+ "lora_r": 64,
34
+ "max_position_embeddings": 32768,
35
+ "max_window_layers": 28,
36
+ "model_type": "qwen2_vl",
37
+ "non_linearity": 1,
38
+ "norm_elementwise_affine": true,
39
+ "num_attention_heads": 12,
40
+ "num_hidden_layers": 28,
41
+ "num_key_value_heads": 2,
42
+ "patch_size": 1,
43
+ "repa_coeff": 0.5,
44
+ "repa_layers": null,
45
+ "repa_shared": false,
46
+ "rms_norm_eps": 1e-06,
47
+ "rope_scaling": {
48
+ "mrope_section": [
49
+ 16,
50
+ 24,
51
+ 24
52
+ ],
53
+ "rope_type": "default",
54
+ "type": "default"
55
+ },
56
+ "rope_theta": 1000000.0,
57
+ "sample_size": 128,
58
+ "sampling_steps": 28,
59
+ "sliding_window": null,
60
+ "tie_word_embeddings": true,
61
+ "torch_dtype": "bfloat16",
62
+ "transformers_version": "4.47.0",
63
+ "use_cache": true,
64
+ "use_repa": false,
65
+ "use_residual_attn": true,
66
+ "use_sliding_window": false,
67
+ "vae_path": "mit-han-lab/dc-ae-f32c32-sana-1.1-diffusers",
68
+ "video_token_id": 151656,
69
+ "vision_config": {
70
+ "hidden_size": 1536,
71
+ "in_chans": 3,
72
+ "model_type": "qwen2_vl",
73
+ "spatial_patch_size": 14
74
+ },
75
+ "vision_end_token_id": 151653,
76
+ "vision_start_token_id": 151652,
77
+ "vision_token_id": 151654,
78
+ "vocab_size": 151936
79
+ }
checkpoint-208000/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 151643,
4
+ "eos_token_id": 151645,
5
+ "transformers_version": "4.47.0"
6
+ }
checkpoint-208000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:731f244136f8b35b62cd6a351883478a4634a89591d9ea037c3713eec8df08f9
3
+ size 4974360904
checkpoint-208000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d3a6e45be235ca7066e6dd49ea2e0ca938761c9ed464d084aeb90f304207cc56
3
+ size 3234250886
checkpoint-208000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-208000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:baad445f9533638c550f0a06ba9315e5a532390eb26ac717939108fac937e22d
3
+ size 7455061885
checkpoint-208000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:92ba6022ab9fc39ca07a91ff6a6ec70f7b8537795e034ed5e70f328c6c6a8b9e
3
+ size 15024
checkpoint-208000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4e2d53d52a48d26c6bfea0ee562555cfbe5dcecb6b823553d9612c827a07dd3
3
+ size 15024
checkpoint-208000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8515ca2d215e60906df5326f484931e399cb8a4f18f59d8c0da8b83837dfb0f
3
+ size 15024
checkpoint-208000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:66f48ecadca5b174c46e93473865948f7840065d2376c833d46abddfb5112b13
3
+ size 15024
checkpoint-208000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63df0663fc26b123b4ad06332880999a0edd752b2d6e3a329e27417b70fd3740
3
+ size 1064
checkpoint-208000/trainer_state.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7045a3bd7a313418b736eda6b48c251f7b015a259401c038c9d04181b4057be0
3
+ size 32049833
checkpoint-208000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8cbac9a56577b729748dc6ff37fc2da66bfa74a51be7a8423dc40fe0ae473d78
3
+ size 6008
checkpoint-210000/config.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ar_steps": 1,
3
+ "architectures": [
4
+ "DiffVLMDiffusion"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "condition_layer": -1,
9
+ "eos_token_id": 151645,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 1536,
12
+ "image_token_id": 151655,
13
+ "img_cross_attention_dim": 2048,
14
+ "img_diffuser_depth": 1,
15
+ "img_ffn_dim_multiplier": null,
16
+ "img_hidden_size": 1536,
17
+ "img_multiple_of": 256,
18
+ "img_norm_eps": 1e-05,
19
+ "img_num_attention_heads": 12,
20
+ "img_num_kv_heads": 12,
21
+ "img_qk_norm": true,
22
+ "in_channels": 32,
23
+ "initializer_range": 0.02,
24
+ "inject_img_diffuser": false,
25
+ "input_size": 32,
26
+ "intermediate_size": 8960,
27
+ "layer_group_size": 7,
28
+ "layerwise_start_idx": 0,
29
+ "lora_alpha": 16,
30
+ "lora_bias": "none",
31
+ "lora_dropout": 0.05,
32
+ "lora_enable": false,
33
+ "lora_r": 64,
34
+ "max_position_embeddings": 32768,
35
+ "max_window_layers": 28,
36
+ "model_type": "qwen2_vl",
37
+ "non_linearity": 1,
38
+ "norm_elementwise_affine": true,
39
+ "num_attention_heads": 12,
40
+ "num_hidden_layers": 28,
41
+ "num_key_value_heads": 2,
42
+ "patch_size": 1,
43
+ "repa_coeff": 0.5,
44
+ "repa_layers": null,
45
+ "repa_shared": false,
46
+ "rms_norm_eps": 1e-06,
47
+ "rope_scaling": {
48
+ "mrope_section": [
49
+ 16,
50
+ 24,
51
+ 24
52
+ ],
53
+ "rope_type": "default",
54
+ "type": "default"
55
+ },
56
+ "rope_theta": 1000000.0,
57
+ "sample_size": 128,
58
+ "sampling_steps": 28,
59
+ "sliding_window": null,
60
+ "tie_word_embeddings": true,
61
+ "torch_dtype": "bfloat16",
62
+ "transformers_version": "4.47.0",
63
+ "use_cache": true,
64
+ "use_repa": false,
65
+ "use_residual_attn": true,
66
+ "use_sliding_window": false,
67
+ "vae_path": "mit-han-lab/dc-ae-f32c32-sana-1.1-diffusers",
68
+ "video_token_id": 151656,
69
+ "vision_config": {
70
+ "hidden_size": 1536,
71
+ "in_chans": 3,
72
+ "model_type": "qwen2_vl",
73
+ "spatial_patch_size": 14
74
+ },
75
+ "vision_end_token_id": 151653,
76
+ "vision_start_token_id": 151652,
77
+ "vision_token_id": 151654,
78
+ "vocab_size": 151936
79
+ }
checkpoint-210000/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 151643,
4
+ "eos_token_id": 151645,
5
+ "transformers_version": "4.47.0"
6
+ }
checkpoint-210000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f80fb5276bef91679c6779dd5c914cabca8aec49d57d615e4d46bcaef929f85c
3
+ size 4974360904
checkpoint-210000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a924e08394e573c84ea636167f8b03b069e4620f71809517ab694cdaed8790d
3
+ size 3234250886
checkpoint-210000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-210000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9844a736fe0e76e71995120476b07720a87e87e3e756eaeb2eadc01a88f890f
3
+ size 7455061885
checkpoint-210000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b41bcbe5e03af6bcdc8702985f43b609aabbbd0a7453297ed321a4bd48ac5e3
3
+ size 15024
checkpoint-210000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8959b18340fdf09a6a487f40d8ac0810bd6c5d8036714b0d6040e50732af854a
3
+ size 15024
checkpoint-210000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09531dcbf4a07ef182b52ef73880908e6f5555a9a6c7be553699abe3a86da754
3
+ size 15024
checkpoint-210000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35ed51cae6665a1eb1ac66f07765d30a6f78a32651a99cedf50994446fb8d427
3
+ size 15024
checkpoint-210000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5ead708e8a92b2004d0f2a90155606e70dd9afe6eefe9cd5a9e82c083885698
3
+ size 1064
checkpoint-210000/trainer_state.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0fc452df935ec9f7bd89677295309e495e06075a10625e99a5f34efbb17ea78
3
+ size 32358507
checkpoint-210000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8cbac9a56577b729748dc6ff37fc2da66bfa74a51be7a8423dc40fe0ae473d78
3
+ size 6008