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
e630d1b
·
verified ·
1 Parent(s): ab6ca7a

Upload folder using huggingface_hub

Browse files
Files changed (27) hide show
  1. .gitattributes +2 -0
  2. checkpoint-228000/config.json +79 -0
  3. checkpoint-228000/generation_config.json +6 -0
  4. checkpoint-228000/model-00001-of-00002.safetensors +3 -0
  5. checkpoint-228000/model-00002-of-00002.safetensors +3 -0
  6. checkpoint-228000/model.safetensors.index.json +0 -0
  7. checkpoint-228000/optimizer.pt +3 -0
  8. checkpoint-228000/rng_state_0.pth +3 -0
  9. checkpoint-228000/rng_state_1.pth +3 -0
  10. checkpoint-228000/rng_state_2.pth +3 -0
  11. checkpoint-228000/rng_state_3.pth +3 -0
  12. checkpoint-228000/scheduler.pt +3 -0
  13. checkpoint-228000/trainer_state.json +3 -0
  14. checkpoint-228000/training_args.bin +3 -0
  15. checkpoint-230000/config.json +79 -0
  16. checkpoint-230000/generation_config.json +6 -0
  17. checkpoint-230000/model-00001-of-00002.safetensors +3 -0
  18. checkpoint-230000/model-00002-of-00002.safetensors +3 -0
  19. checkpoint-230000/model.safetensors.index.json +0 -0
  20. checkpoint-230000/optimizer.pt +3 -0
  21. checkpoint-230000/rng_state_0.pth +3 -0
  22. checkpoint-230000/rng_state_1.pth +3 -0
  23. checkpoint-230000/rng_state_2.pth +3 -0
  24. checkpoint-230000/rng_state_3.pth +3 -0
  25. checkpoint-230000/scheduler.pt +3 -0
  26. checkpoint-230000/trainer_state.json +3 -0
  27. checkpoint-230000/training_args.bin +3 -0
.gitattributes CHANGED
@@ -105,3 +105,5 @@ 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
106
  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
 
 
 
105
  checkpoint-210000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
106
  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
checkpoint-228000/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-228000/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-228000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c53eb400e1ae4717aa81410e5bb897ee7a65accf58da13b18a5999581823e95
3
+ size 4974360904
checkpoint-228000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80fec52d66518a696b5c4fbaba87b9913fcbbdf2804f83bd5b721e498f6f25a4
3
+ size 3234250886
checkpoint-228000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-228000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b57241c14019f648d3bdda6927f11a9ba5a24ada899eccfe9722b5364c7af94d
3
+ size 7455061885
checkpoint-228000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b9aa478d4e1b4dd081cee3c774cb8621dd58dc4367d7b44eeffa8f901b3b13fc
3
+ size 15024
checkpoint-228000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:793c9f247256807cfcdf36520565f284121334dc1f467a6c65f8a42b2fb00eaa
3
+ size 15024
checkpoint-228000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ddc9951302eb155f78ca533ef9a3a7c60445c27c0586877722c11bc33880e440
3
+ size 15024
checkpoint-228000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be2495bccca729d51f07f1089029c5bbf752d0558dac57e0ec4800c1eb9ed8cb
3
+ size 15024
checkpoint-228000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4334f1b9d036935e9c86a7ec818dc9edd6d9afc4ce0fddc8e80e36b8724fddc7
3
+ size 1064
checkpoint-228000/trainer_state.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:147001334149ac4444167842f9b4fcd3ffb90d7a567711c60eb7b300f1473967
3
+ size 35137830
checkpoint-228000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25f1ac425f4b484983830a206749895e0ec401c30eddfe94d53723058d79e7b6
3
+ size 6072
checkpoint-230000/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-230000/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-230000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d767587214fe6ec24a29281733c8fe07fc200b6fcca6022fc7500d89c2c43a84
3
+ size 4974360904
checkpoint-230000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9cafabca5cfa9cef974088309bb2501a0b088e3b36abf2d1083d50b0cf0a445e
3
+ size 3234250886
checkpoint-230000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-230000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5773fd8f8988d13367c9e2cd0c1ecaa260c7d31a7da9dd5f38f152d8e1857b70
3
+ size 7455061885
checkpoint-230000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15a3d18b018a29b15d0652366ee9eddd45e5be49674daefe001db834425c8b66
3
+ size 15024
checkpoint-230000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ed110d51b5c25890d23a4d59be66ceddd227a441b56da116b02e324ff08c505
3
+ size 15024
checkpoint-230000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8072063cc42e2ae93f6ac9d70ea6c3ec8d8a9316555d6286652fcbfa6719b6d9
3
+ size 15024
checkpoint-230000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:baa28696b379ce49e21e6f44b1ef9f2e54df329705f11edb4feaeebe7f14e34a
3
+ size 15024
checkpoint-230000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b345e2ccd01353d7702ec0c974915715bc022b497fde302ac150d2a88a849611
3
+ size 1064
checkpoint-230000/trainer_state.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a02063fdf89a48543e58f2fb3c50435f6e8ec9cf06e678cdf53e18f7cec020c8
3
+ size 35446275
checkpoint-230000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25f1ac425f4b484983830a206749895e0ec401c30eddfe94d53723058d79e7b6
3
+ size 6072