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
2ad6701
·
verified ·
1 Parent(s): 8dc37c2

Upload folder using huggingface_hub

Browse files
Files changed (27) hide show
  1. .gitattributes +2 -0
  2. checkpoint-248000/config.json +79 -0
  3. checkpoint-248000/generation_config.json +6 -0
  4. checkpoint-248000/model-00001-of-00002.safetensors +3 -0
  5. checkpoint-248000/model-00002-of-00002.safetensors +3 -0
  6. checkpoint-248000/model.safetensors.index.json +0 -0
  7. checkpoint-248000/optimizer.pt +3 -0
  8. checkpoint-248000/rng_state_0.pth +3 -0
  9. checkpoint-248000/rng_state_1.pth +3 -0
  10. checkpoint-248000/rng_state_2.pth +3 -0
  11. checkpoint-248000/rng_state_3.pth +3 -0
  12. checkpoint-248000/scheduler.pt +3 -0
  13. checkpoint-248000/trainer_state.json +3 -0
  14. checkpoint-248000/training_args.bin +3 -0
  15. checkpoint-250000/config.json +79 -0
  16. checkpoint-250000/generation_config.json +6 -0
  17. checkpoint-250000/model-00001-of-00002.safetensors +3 -0
  18. checkpoint-250000/model-00002-of-00002.safetensors +3 -0
  19. checkpoint-250000/model.safetensors.index.json +0 -0
  20. checkpoint-250000/optimizer.pt +3 -0
  21. checkpoint-250000/rng_state_0.pth +3 -0
  22. checkpoint-250000/rng_state_1.pth +3 -0
  23. checkpoint-250000/rng_state_2.pth +3 -0
  24. checkpoint-250000/rng_state_3.pth +3 -0
  25. checkpoint-250000/scheduler.pt +3 -0
  26. checkpoint-250000/trainer_state.json +3 -0
  27. checkpoint-250000/training_args.bin +3 -0
.gitattributes CHANGED
@@ -109,3 +109,5 @@ 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
 
 
 
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
112
+ checkpoint-248000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
113
+ checkpoint-250000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
checkpoint-248000/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-248000/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-248000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6d56b5aa9d2860a402afdd40cb2ab1b7d3e228acc2e44ffa6a4d5f8d7b5e6844
3
+ size 4974360904
checkpoint-248000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:222e807d8f50cf887abe792808c49b64a9a7f5cdcd264e7352c08a6c68920551
3
+ size 3234250886
checkpoint-248000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-248000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f014e702efd371a1a5b4657073d4307eaefdf4d66111d4b9378d0c7cf9c9b1f
3
+ size 7455061885
checkpoint-248000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4decc7efc7ce579f84cd35c269b930c611f2992e0f0a9d79f7c5b2850097806e
3
+ size 15024
checkpoint-248000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37083ac56e0a45a595a3dd4b2109a2614300576ff961399d3fce51d1b4333bef
3
+ size 15024
checkpoint-248000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c2275efb9ee2a1c07e4a9a7b6c17b9375d9881ef4132c8cba099b3a85683387
3
+ size 15024
checkpoint-248000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3280016304b3004d86f932d2a8dc9c3ad6abbc38b6e68eb525814d5e89c9f1f2
3
+ size 15024
checkpoint-248000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7645e71f0dc89e82131ae909b0c6a1df5de6c694f33da11b62657d3d69756d13
3
+ size 1064
checkpoint-248000/trainer_state.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a1633a5e53cc6290de5feeb9711db5b9b618de9a10578ff8e092d4be6ea80f86
3
+ size 38220378
checkpoint-248000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8154a02c05578fecad2fc05e54e14b6fa4d2886901c30f8c487e39f2aa7da8b9
3
+ size 6072
checkpoint-250000/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-250000/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-250000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f3fdfd03eb0640b1c4d768832cee6b2337b886254d3f6938507c4ecfc29aa486
3
+ size 4974360904
checkpoint-250000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:57b4b0426e206a3f3c30702ee7a05baf116f07d513404662b559c639371e9d56
3
+ size 3234250886
checkpoint-250000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-250000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea8a162ddffeeab8a35fab27025300d6bebac8fe78bce5d34ceb488f6bc0ed7d
3
+ size 7455061885
checkpoint-250000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de3cf3a1225ac8e3d557adc61fbdbdf8c82ac88ba9ee154d5fb2b6f3547e746d
3
+ size 15024
checkpoint-250000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf9f29c7874355a04860c572a6ebba6448029d7ead7b65840fdb64cde77027a9
3
+ size 15024
checkpoint-250000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f7e980197c627a9ff00fd466be9ebb11d16968b5600f4d32e55e2d11e0defcd5
3
+ size 15024
checkpoint-250000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d6a88240148079c2bab333323e58d2c891602de9334fb541cf090d63395d6444
3
+ size 15024
checkpoint-250000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15b81eb4eef0bf07eb5fdab4f1345862cc210205041ce235fbccd23ce9dfc6ff
3
+ size 1064
checkpoint-250000/trainer_state.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2354259a7f018bb028c58e82f5826670449a7ed1f7fc4304484a3594a05a2f82
3
+ size 38528784
checkpoint-250000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8154a02c05578fecad2fc05e54e14b6fa4d2886901c30f8c487e39f2aa7da8b9
3
+ size 6072