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
8dc37c2
·
verified ·
1 Parent(s): e630d1b

Upload folder using huggingface_hub

Browse files
Files changed (27) hide show
  1. .gitattributes +2 -0
  2. checkpoint-238000/config.json +79 -0
  3. checkpoint-238000/generation_config.json +6 -0
  4. checkpoint-238000/model-00001-of-00002.safetensors +3 -0
  5. checkpoint-238000/model-00002-of-00002.safetensors +3 -0
  6. checkpoint-238000/model.safetensors.index.json +0 -0
  7. checkpoint-238000/optimizer.pt +3 -0
  8. checkpoint-238000/rng_state_0.pth +3 -0
  9. checkpoint-238000/rng_state_1.pth +3 -0
  10. checkpoint-238000/rng_state_2.pth +3 -0
  11. checkpoint-238000/rng_state_3.pth +3 -0
  12. checkpoint-238000/scheduler.pt +3 -0
  13. checkpoint-238000/trainer_state.json +3 -0
  14. checkpoint-238000/training_args.bin +3 -0
  15. checkpoint-240000/config.json +79 -0
  16. checkpoint-240000/generation_config.json +6 -0
  17. checkpoint-240000/model-00001-of-00002.safetensors +3 -0
  18. checkpoint-240000/model-00002-of-00002.safetensors +3 -0
  19. checkpoint-240000/model.safetensors.index.json +0 -0
  20. checkpoint-240000/optimizer.pt +3 -0
  21. checkpoint-240000/rng_state_0.pth +3 -0
  22. checkpoint-240000/rng_state_1.pth +3 -0
  23. checkpoint-240000/rng_state_2.pth +3 -0
  24. checkpoint-240000/rng_state_3.pth +3 -0
  25. checkpoint-240000/scheduler.pt +3 -0
  26. checkpoint-240000/trainer_state.json +3 -0
  27. checkpoint-240000/training_args.bin +3 -0
.gitattributes CHANGED
@@ -107,3 +107,5 @@ checkpoint-218000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
107
  checkpoint-220000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
108
  checkpoint-228000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
109
  checkpoint-230000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
 
 
 
107
  checkpoint-220000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
108
  checkpoint-228000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
109
  checkpoint-230000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
110
+ checkpoint-238000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
111
+ checkpoint-240000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
checkpoint-238000/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-238000/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-238000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb6e48bba1ad5a4efdce18a99abb211e7f5945b1564a392f95fcf37aaab89e2c
3
+ size 4974360904
checkpoint-238000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84febc4bc6b0912fae004f9311493a417b1bd9e1b1a8d1675239dcec79c29c05
3
+ size 3234250886
checkpoint-238000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-238000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12cfacf32cb8ef430e1d2dad11eb3727845606a42abf7ac4e9275520c82c82c0
3
+ size 7455061885
checkpoint-238000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a57d824b13404d32f137e9e831b04cbf1b98c3dd2514f2d7ea6e4207f2b58aad
3
+ size 15024
checkpoint-238000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:489c8b58f7db8ca8c163542aa279c2d79f02489f9886f348fbe48f26205cc3ee
3
+ size 15024
checkpoint-238000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d3b9626a04add14c6130ce8b2bd3bb53b2eb329753e2beed995dda620fb9e337
3
+ size 15024
checkpoint-238000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:43a76e00c2a4f6345d27edca55bbb265a2b8c6f862593a7cda975252dd1d16dc
3
+ size 15024
checkpoint-238000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60deae1edcda95ec635e6d15dd38264c057eed1ce479ba75e78bb137127c2f2c
3
+ size 1064
checkpoint-238000/trainer_state.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ca9719e7e15e15d73118d8d012c0ad4b5c2671f17c760092a300f70f4fd24cc
3
+ size 36679374
checkpoint-238000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c41d4e9f81a554b58bd6adb1e20c912dedf3bcb417ba4a712c692cad66ba3d5
3
+ size 6072
checkpoint-240000/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-240000/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-240000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b09f2d521ee9c4bdfd1c87f57da7a718c907b55e34e2cf5a6cdd56d24436463
3
+ size 4974360904
checkpoint-240000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93232b16537f4b1df93d50218a4dd0605cb6ab38b95fc0e977f481733095edb3
3
+ size 3234250886
checkpoint-240000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-240000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7e8c533706c01a16c5fd21a2a2ace40022fc9453b834797bfd978bffb234097
3
+ size 7455061885
checkpoint-240000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:39e487fafb970e6a57ba64ca239f09fdefeab8980b9b4d9162f411716edeaa8e
3
+ size 15024
checkpoint-240000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f68613f2c1b3ff348426efd3c478ed8947ae1195910ecf9786dd983b669957b
3
+ size 15024
checkpoint-240000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2d663956bdc57e1a6d20ded74ed2128cb7c0b1e0fa9ccf0b7c3f1ade7ca3437
3
+ size 15024
checkpoint-240000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d76d3862435084920e1633a57a0748ff6810aac00d6779ab3921a02fbf507e62
3
+ size 15024
checkpoint-240000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c7975591ecb1d9c1b39cc165ac932a43587177e97a2b83918da9ce9d6c4e02d
3
+ size 1064
checkpoint-240000/trainer_state.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:86e36ca7288ec62596d8202a6e37030794b490d69ae341cdd5bda7abaf230d23
3
+ size 36987716
checkpoint-240000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c41d4e9f81a554b58bd6adb1e20c912dedf3bcb417ba4a712c692cad66ba3d5
3
+ size 6072