Hugging Face's logo

tencent
/
HunyuanImage-3.0

Text-to-Image
Transformers
Safetensors
hunyuan_image_3_moe
text-generation
custom_code
Model card Files
xet
Community
11
root commited on
Commit
0e071a5
·
1 Parent(s): 02f7bed

upload

Browse files
Files changed (37) hide show
  1. config.json +270 -2
  2. generation_config.json +20 -0
  3. model-0001-of-0032.safetensors +3 -0
  4. model-0002-of-0032.safetensors +3 -0
  5. model-0003-of-0032.safetensors +3 -0
  6. model-0004-of-0032.safetensors +3 -0
  7. model-0005-of-0032.safetensors +3 -0
  8. model-0006-of-0032.safetensors +3 -0
  9. model-0007-of-0032.safetensors +3 -0
  10. model-0008-of-0032.safetensors +3 -0
  11. model-0009-of-0032.safetensors +3 -0
  12. model-0010-of-0032.safetensors +3 -0
  13. model-0011-of-0032.safetensors +3 -0
  14. model-0012-of-0032.safetensors +3 -0
  15. model-0013-of-0032.safetensors +3 -0
  16. model-0014-of-0032.safetensors +3 -0
  17. model-0015-of-0032.safetensors +3 -0
  18. model-0016-of-0032.safetensors +3 -0
  19. model-0017-of-0032.safetensors +3 -0
  20. model-0018-of-0032.safetensors +3 -0
  21. model-0019-of-0032.safetensors +3 -0
  22. model-0020-of-0032.safetensors +3 -0
  23. model-0021-of-0032.safetensors +3 -0
  24. model-0022-of-0032.safetensors +3 -0
  25. model-0023-of-0032.safetensors +3 -0
  26. model-0024-of-0032.safetensors +3 -0
  27. model-0025-of-0032.safetensors +3 -0
  28. model-0026-of-0032.safetensors +3 -0
  29. model-0027-of-0032.safetensors +3 -0
  30. model-0028-of-0032.safetensors +3 -0
  31. model-0029-of-0032.safetensors +3 -0
  32. model-0030-of-0032.safetensors +3 -0
  33. model-0031-of-0032.safetensors +3 -0
  34. model-0032-of-0032.safetensors +3 -0
  35. model.safetensors.index.json +0 -0
  36. tokenizer.json +3 -0
  37. tokenizer_config.json +8 -0
config.json CHANGED
@@ -1,5 +1,273 @@
1
  {
2
- "Name": [
3
- "HunyuanImage-3.0"
 
 
 
 
4
  ],
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5
  }
 
1
  {
2
+ "add_classification_head": false,
3
+ "anyres_pooling_size": 2,
4
+ "anyres_vit_max_image_size": null,
5
+ "anyres_vit_two_views": false,
6
+ "architectures": [
7
+ "HunyuanImage3ForCausalMM"
8
  ],
9
+ "attention_bias": false,
10
+ "attention_dropout": 0.0,
11
+ "attention_head_dim": 128,
12
+ "auto_map": {
13
+ "AutoConfig": "configuration_hunyuan.HunyuanImage3Config",
14
+ "AutoModel": "hunyuan.HunyuanImage3Model",
15
+ "AutoModelForCausalLM": "hunyuan.HunyuanImage3ForCausalMM"
16
+ },
17
+ "bos_token_id": 127958,
18
+ "cla_share_factor": 2,
19
+ "class_num": 0,
20
+ "dense_list": [
21
+ 4096,
22
+ 0
23
+ ],
24
+ "eod_token_id": 3,
25
+ "eos_token_id": 127957,
26
+ "group_limited_greedy": false,
27
+ "hidden_act": "silu",
28
+ "hidden_size": 4096,
29
+ "im_end_id": 128001,
30
+ "im_newline_id": 11,
31
+ "im_start_id": 128000,
32
+ "image_token_id": 128006,
33
+ "initializer_range": 0.02,
34
+ "intermediate_size": 3072,
35
+ "kv_lora_rank": null,
36
+ "mask_init_id": 12,
37
+ "max_position_embeddings": 12800,
38
+ "mlp_bias": false,
39
+ "model_type": "hunyuan_image_3_moe",
40
+ "moe_drop_tokens": false,
41
+ "moe_intermediate_size": [
42
+ 3072,
43
+ 3072,
44
+ 3072,
45
+ 3072,
46
+ 3072,
47
+ 3072,
48
+ 3072,
49
+ 3072,
50
+ 3072,
51
+ 3072,
52
+ 3072,
53
+ 3072,
54
+ 3072,
55
+ 3072,
56
+ 3072,
57
+ 3072,
58
+ 3072,
59
+ 3072,
60
+ 3072,
61
+ 3072,
62
+ 3072,
63
+ 3072,
64
+ 3072,
65
+ 3072,
66
+ 3072,
67
+ 3072,
68
+ 3072,
69
+ 3072,
70
+ 3072,
71
+ 3072,
72
+ 3072,
73
+ 3072
74
+ ],
75
+ "moe_layer_num_skipped": 0,
76
+ "moe_random_routing_dropped_token": false,
77
+ "moe_topk": [
78
+ 8,
79
+ 8,
80
+ 8,
81
+ 8,
82
+ 8,
83
+ 8,
84
+ 8,
85
+ 8,
86
+ 8,
87
+ 8,
88
+ 8,
89
+ 8,
90
+ 8,
91
+ 8,
92
+ 8,
93
+ 8,
94
+ 8,
95
+ 8,
96
+ 8,
97
+ 8,
98
+ 8,
99
+ 8,
100
+ 8,
101
+ 8,
102
+ 8,
103
+ 8,
104
+ 8,
105
+ 8,
106
+ 8,
107
+ 8,
108
+ 8,
109
+ 8
110
+ ],
111
+ "n_group": false,
112
+ "norm_topk_prob": true,
113
+ "norm_type": "rms",
114
+ "num_attention_heads": 32,
115
+ "num_experts": 64,
116
+ "num_hidden_layers": 32,
117
+ "num_key_value_heads": 8,
118
+ "num_media_embeds": 257,
119
+ "num_shared_expert": [
120
+ 1,
121
+ 1,
122
+ 1,
123
+ 1,
124
+ 1,
125
+ 1,
126
+ 1,
127
+ 1,
128
+ 1,
129
+ 1,
130
+ 1,
131
+ 1,
132
+ 1,
133
+ 1,
134
+ 1,
135
+ 1,
136
+ 1,
137
+ 1,
138
+ 1,
139
+ 1,
140
+ 1,
141
+ 1,
142
+ 1,
143
+ 1,
144
+ 1,
145
+ 1,
146
+ 1,
147
+ 1,
148
+ 1,
149
+ 1,
150
+ 1,
151
+ 1
152
+ ],
153
+ "pad_id": 128009,
154
+ "pad_token_id": 128009,
155
+ "pool_type": "last",
156
+ "position_embedding_xdrope": false,
157
+ "pretraining_tp": 1,
158
+ "q_lora_rank": null,
159
+ "qk_nope_head_dim": null,
160
+ "qk_rope_head_dim": null,
161
+ "rms_norm_eps": 1e-05,
162
+ "rope_scaling": {
163
+ "alpha": 1.0,
164
+ "beta_fast": 32,
165
+ "beta_slow": 1,
166
+ "factor": 1.0,
167
+ "mscale": 1.0,
168
+ "mscale_all_dim": 1.0,
169
+ "type": "custom"
170
+ },
171
+ "rope_theta": 10000.0,
172
+ "routed_scaling_factor": false,
173
+ "skip_cls_token": false,
174
+ "text_end_id": 7,
175
+ "text_start_id": 6,
176
+ "tie_word_embeddings": false,
177
+ "topk_group": false,
178
+ "torch_dtype": "bfloat16",
179
+ "transformers_version": "4.50.0",
180
+ "use_cache": true,
181
+ "use_cla": false,
182
+ "use_mixed_mlp_moe": true,
183
+ "use_mla": false,
184
+ "use_qk_norm": true,
185
+ "use_rotary_pos_emb": true,
186
+ "v_head_dim": null,
187
+ "video_end_id": 10,
188
+ "video_start_id": 9,
189
+ "vit_add_patchemb_bias": false,
190
+ "vit_input_resolution": 224,
191
+ "vit_mapping_type": "resampler",
192
+ "vit_norm_type": "fused",
193
+ "vit_patch": 1,
194
+ "vit_path": null,
195
+ "vit_remove_prenorm": false,
196
+ "vit_token": 64,
197
+ "vit_type": null,
198
+ "vit_used_rms_norm": false,
199
+ "vocab_size": 133120,
200
+ "xdrope_section": null,
201
+ "head_dim": 128,
202
+ "vae_downsample_factor": [
203
+ 16,
204
+ 16
205
+ ],
206
+ "vae": {
207
+ "_class_name": "AutoencoderKLConv3D",
208
+ "block_out_channels": [
209
+ 128,
210
+ 256,
211
+ 512,
212
+ 1024,
213
+ 1024
214
+ ],
215
+ "in_channels": 3,
216
+ "out_channels": 3,
217
+ "latent_channels": 32,
218
+ "layers_per_block": 2,
219
+ "ffactor_spatial": 16,
220
+ "ffactor_temporal": 4,
221
+ "sample_size": 384,
222
+ "sample_tsize": 96,
223
+ "downsample_match_channel": true,
224
+ "upsample_match_channel": true,
225
+ "scaling_factor": 0.562679178327931
226
+ },
227
+ "vit": {
228
+ "_attn_implementation": "sdpa",
229
+ "attention_dropout": 0.0,
230
+ "hidden_act": "gelu_pytorch_tanh",
231
+ "hidden_size": 1152,
232
+ "intermediate_size": 4304,
233
+ "layer_norm_eps": 1e-06,
234
+ "num_attention_heads": 16,
235
+ "num_channels": 3,
236
+ "num_hidden_layers": 27,
237
+ "num_patches": 256,
238
+ "patch_size": 16,
239
+ "torch_dtype": "float32",
240
+ "output_attentions": false,
241
+ "output_hidden_states": false,
242
+ "use_return_dict": true
243
+ },
244
+ "vit_processor": {
245
+ "do_convert_rgb": null,
246
+ "do_normalize": true,
247
+ "do_rescale": true,
248
+ "do_resize": true,
249
+ "image_mean": [
250
+ 0.5,
251
+ 0.5,
252
+ 0.5
253
+ ],
254
+ "image_processor_type": "Siglip2ImageProcessorFast",
255
+ "image_std": [
256
+ 0.5,
257
+ 0.5,
258
+ 0.5
259
+ ],
260
+ "max_num_patches": 1024,
261
+ "patch_size": 16,
262
+ "processor_class": "Siglip2Processor",
263
+ "resample": 2,
264
+ "rescale_factor": 0.00392156862745098
265
+ },
266
+ "vit_aligner": {
267
+ "projector_type": "mlp_gelu",
268
+ "input_dim": 1152,
269
+ "n_embed": 4096,
270
+ "depth": 2,
271
+ "torch_dtype": "float32"
272
+ }
273
  }
generation_config.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "disable_compile": true,
3
+ "eos_token_id": [
4
+ 127957
5
+ ],
6
+ "pad_token_id": 128009,
7
+ "do_sample": true,
8
+ "top_k": 1024,
9
+ "top_p": 0.95,
10
+ "temperature": 0.6,
11
+ "max_length": 12800,
12
+ "sequence_template": "pretrain",
13
+ "diff_infer_steps": 50,
14
+ "diff_guidance_scale": 5.0,
15
+ "flow_shift": 3.0,
16
+ "use_system_prompt": "None",
17
+ "drop_think": false,
18
+ "bot_task": "image",
19
+ "transformers_version": "4.50.0"
20
+ }
model-0001-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dad22fa5e99dcda532c242aa4d4875f9ea6fd8b2ed59e39776dec4ea55baf4e5
3
+ size 5363066616
model-0002-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9987e8220f81b70d07b62f06ac6c92bb0faf38ccb0ddd3f30b65ed895ad4a2fb
3
+ size 5318937248
model-0003-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:79f8d4d1b23562299da3360ac7e2437a4dd24be30b86bc8db580521b5f9b2616
3
+ size 5344627472
model-0004-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4faf1357831b25b9f9637594312e9024ee0fa1e87c734e20afdde2845fdaa516
3
+ size 5327343192
model-0005-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:46189f8777c117c431e46cc57ec2328fe72050452119ac7bb676bdaca3f76575
3
+ size 5344103080
model-0006-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f9d5f386b7c2d0b171bd8a25f3f08e3150936fde2dfd92e9aa1f6e27dbf2e0d
3
+ size 5318937248
model-0007-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d30616044acead06484eacace50a4cab66267feb13555f235bac63d2540cf471
3
+ size 5344103088
model-0008-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:740ccbff8fa1dbb2847fe8c342654f7d24fa81f058065e82dfbccb89ce2743c1
3
+ size 5318937256
model-0009-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5fc3df50de8591735d29f7acfece39b64b3735cccef176eb4a137f4ede68430
3
+ size 5344103088
model-0010-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f6058eb7527741d18c17131cb7810f11d8bd4c69cce10962e093e684413cd2a
3
+ size 5318937304
model-0011-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c38d5fd2f18191d849b444e873ff91d3f048d8c4bcd71b3035ff0f7973ac273
3
+ size 5344103232
model-0012-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:688a6a818f6d164d345e3bb37c4f3fcee40cc7d458027d2a37f7486463843ec3
3
+ size 5318937400
model-0013-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f77757aa32fa67f75f8f8ec5bc831d358093483c2a8692bff7477378aea00f28
3
+ size 5344103232
model-0014-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3308c079c20008e1ac8852cfb986764064077278754492f2fd9ec893857b6489
3
+ size 5318937400
model-0015-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e32b467eb49473c7f42696db0916ca3275c01984c48a10433d78be4d351b7ff8
3
+ size 5344103232
model-0016-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b97d98195a45518bae971bc43c224225b60e1fbb8b2eb93115024d2bdf328dca
3
+ size 5318937400
model-0017-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f00339bad7371e59f2d3642fd0575abafa92fc4509803f8fe5a64492185d2ab
3
+ size 5344103224
model-0018-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b48a59d090d396aa9801765485381f8255d442c2da2d9e98f1c21a68c6b83b1
3
+ size 5327859080
model-0019-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd4e5a082f3db3b61774ce86675cfb171f33319fd3dd8f942cd952633834d334
3
+ size 5344111888
model-0020-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f27fc2c0eedfc6b99ebe07e244c9689e89fa06dc65216d9c07aa6067783f86b5
3
+ size 5318937392
model-0021-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:128289d2f1b5787e19c7ff3defa7f8cd591304502d0db66c0cd57a214e5ecf13
3
+ size 5344103232
model-0022-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0535004052f37e3756ba186bff27d00accc22dad93a16edd40734579f030d92f
3
+ size 5318937400
model-0023-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a6d0b876aa4b9048a1f5b9cd6a72e386d1abd8c9b24d331c2fab8c6ff5fe8080
3
+ size 5344103232
model-0024-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c5c89a856e5d4b1e2d867e4b61cfb37311fa69848906be1b03e088ccc00a598
3
+ size 5318937400
model-0025-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2310c229c814a7ddc076e5bbf0b03e487b016ded864c2844b1562b081445f5f
3
+ size 5344103232
model-0026-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c6aa2a2dc78191899d1d9019f124058caf3537777950d3d3c3761ca616a85f79
3
+ size 5318937400
model-0027-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:784b14449c7af7674dc9f1a4e7e6b04e79afcf10250dc831a8f8ae87b4fb8827
3
+ size 5344103232
model-0028-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:74eaa13efc2eee80630067f9e54f18db6b3b3c1374b8cb0b719cab4d54996ee8
3
+ size 5318937400
model-0029-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f409aca2ec8d888bb2a7cb83c7f4b8d26ec97408f628a7268108ac028a45c31
3
+ size 5344103232
model-0030-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef654a129711bb1f98957f632de83d9dc145902a95826be1ca4e187841ee5781
3
+ size 5318937400
model-0031-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b468761f2b64c388c728a529692067bab5780938c542644f295b2b8f453c5a70
3
+ size 5354949964
model-0032-of-0032.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cafd49f6c37deea3d71f159c706ebc9fc5d2525b1f6b06491c3969c6aa3f4dec
3
+ size 3203019864
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f804f2be631f905a30f98f89c141abe1bbd3f916e1cda1e1563bb5dccd347b52
3
+ size 25022931
tokenizer_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|startoftext|>",
3
+ "clean_up_tokenization_spaces": true,
4
+ "eos_token": "<|endoftext|>",
5
+ "model_max_length": 1048576,
6
+ "pad_token": "<pad>",
7
+ "tokenizer_class": "PreTrainedTokenizerFast"
8
+ }