Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
Community
650
dacorvo HF Staff commited on
Commit
c1fe719
·
verified ·
1 Parent(s): 3292f30

Synchronizing local compiler cache.

Browse files
Files changed (8) hide show
  1. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/stable-diffusion/stabilityai/sdxl-turbo/7c1121df63c546f81f0f.json +170 -0
  2. neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/text_encoder/model.neuron +1 -1
  3. neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/text_encoder_2/model.neuron +1 -1
  4. neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/unet/model.neuron +1 -1
  5. neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/vae_decoder/config.json +4 -4
  6. neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/vae_decoder/model.neuron +1 -1
  7. neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/vae_encoder/config.json +4 -4
  8. neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/vae_encoder/model.neuron +1 -1
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/stable-diffusion/stabilityai/sdxl-turbo/7c1121df63c546f81f0f.json ADDED
@@ -0,0 +1,170 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "MultiModelCacheEntry",
3
+ "_model_id": "stabilityai/sdxl-turbo",
4
+ "_task": null,
5
+ "text_encoder": {
6
+ "architectures": [
7
+ "CLIPTextModel"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "dropout": 0.0,
11
+ "export_model_type": "clip-text-model",
12
+ "hidden_act": "quick_gelu",
13
+ "hidden_size": 768,
14
+ "initializer_factor": 1.0,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 3072,
17
+ "layer_norm_eps": 1e-05,
18
+ "max_position_embeddings": 77,
19
+ "model_type": "clip_text_model",
20
+ "neuron": {
21
+ "auto_cast": "matmul",
22
+ "auto_cast_type": "bf16",
23
+ "compiler_type": "neuronx-cc",
24
+ "compiler_version": "2.19.8089.0+8ab9f450",
25
+ "dynamic_batch_size": false,
26
+ "float_dtype": "fp32",
27
+ "inline_weights_to_neff": false,
28
+ "int_dtype": "int64",
29
+ "optlevel": "2",
30
+ "output_attentions": false,
31
+ "output_hidden_states": false,
32
+ "static_batch_size": 1,
33
+ "static_sequence_length": 77,
34
+ "task": "feature-extraction",
35
+ "tensor_parallel_size": 1
36
+ },
37
+ "num_attention_heads": 12,
38
+ "num_hidden_layers": 12,
39
+ "output_hidden_states": true,
40
+ "vocab_size": 49408
41
+ },
42
+ "text_encoder_2": {
43
+ "architectures": [
44
+ "CLIPTextModelWithProjection"
45
+ ],
46
+ "attention_dropout": 0.0,
47
+ "dropout": 0.0,
48
+ "export_model_type": "clip-text-with-projection",
49
+ "hidden_act": "gelu",
50
+ "hidden_size": 1280,
51
+ "initializer_factor": 1.0,
52
+ "initializer_range": 0.02,
53
+ "intermediate_size": 5120,
54
+ "layer_norm_eps": 1e-05,
55
+ "max_position_embeddings": 77,
56
+ "model_type": "clip_text_model",
57
+ "neuron": {
58
+ "auto_cast": "matmul",
59
+ "auto_cast_type": "bf16",
60
+ "compiler_type": "neuronx-cc",
61
+ "compiler_version": "2.19.8089.0+8ab9f450",
62
+ "dynamic_batch_size": false,
63
+ "float_dtype": "fp32",
64
+ "inline_weights_to_neff": false,
65
+ "int_dtype": "int64",
66
+ "optlevel": "2",
67
+ "output_attentions": false,
68
+ "output_hidden_states": false,
69
+ "static_batch_size": 1,
70
+ "static_sequence_length": 77,
71
+ "task": "feature-extraction",
72
+ "tensor_parallel_size": 1
73
+ },
74
+ "num_attention_heads": 20,
75
+ "num_hidden_layers": 32,
76
+ "output_hidden_states": true,
77
+ "vocab_size": 49408
78
+ },
79
+ "unet": {
80
+ "_class_name": "UNet2DConditionModel",
81
+ "act_fn": "silu",
82
+ "addition_embed_type": "text_time",
83
+ "addition_embed_type_num_heads": 64,
84
+ "addition_time_embed_dim": 256,
85
+ "attention_head_dim": [
86
+ 5,
87
+ 10,
88
+ 20
89
+ ],
90
+ "attention_type": "default",
91
+ "block_out_channels": [
92
+ 320,
93
+ 640,
94
+ 1280
95
+ ],
96
+ "center_input_sample": false,
97
+ "class_embed_type": null,
98
+ "class_embeddings_concat": false,
99
+ "conv_in_kernel": 3,
100
+ "conv_out_kernel": 3,
101
+ "cross_attention_dim": 2048,
102
+ "cross_attention_norm": null,
103
+ "down_block_types": [
104
+ "DownBlock2D",
105
+ "CrossAttnDownBlock2D",
106
+ "CrossAttnDownBlock2D"
107
+ ],
108
+ "downsample_padding": 1,
109
+ "dropout": 0.0,
110
+ "dual_cross_attention": false,
111
+ "encoder_hid_dim": null,
112
+ "encoder_hid_dim_type": null,
113
+ "flip_sin_to_cos": true,
114
+ "freq_shift": 0,
115
+ "in_channels": 4,
116
+ "layers_per_block": 2,
117
+ "mid_block_only_cross_attention": null,
118
+ "mid_block_scale_factor": 1,
119
+ "mid_block_type": "UNetMidBlock2DCrossAttn",
120
+ "neuron": {
121
+ "auto_cast": "matmul",
122
+ "auto_cast_type": "bf16",
123
+ "compiler_type": "neuronx-cc",
124
+ "compiler_version": "2.19.8089.0+8ab9f450",
125
+ "dynamic_batch_size": false,
126
+ "float_dtype": "fp32",
127
+ "inline_weights_to_neff": true,
128
+ "int_dtype": "int64",
129
+ "optlevel": "2",
130
+ "output_attentions": false,
131
+ "output_hidden_states": false,
132
+ "static_batch_size": 1,
133
+ "static_height": 96,
134
+ "static_num_channels": 4,
135
+ "static_sequence_length": 77,
136
+ "static_vae_scale_factor": 8,
137
+ "static_width": 96,
138
+ "task": "semantic-segmentation",
139
+ "tensor_parallel_size": 1
140
+ },
141
+ "norm_eps": 1e-05,
142
+ "norm_num_groups": 32,
143
+ "num_attention_heads": null,
144
+ "num_class_embeds": null,
145
+ "only_cross_attention": false,
146
+ "out_channels": 4,
147
+ "projection_class_embeddings_input_dim": 2816,
148
+ "resnet_out_scale_factor": 1.0,
149
+ "resnet_skip_time_act": false,
150
+ "resnet_time_scale_shift": "default",
151
+ "reverse_transformer_layers_per_block": null,
152
+ "time_cond_proj_dim": null,
153
+ "time_embedding_act_fn": null,
154
+ "time_embedding_dim": null,
155
+ "time_embedding_type": "positional",
156
+ "timestep_post_act": null,
157
+ "transformer_layers_per_block": [
158
+ 1,
159
+ 2,
160
+ 10
161
+ ],
162
+ "up_block_types": [
163
+ "CrossAttnUpBlock2D",
164
+ "CrossAttnUpBlock2D",
165
+ "UpBlock2D"
166
+ ],
167
+ "upcast_attention": null,
168
+ "use_linear_projection": true
169
+ }
170
+ }
neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/text_encoder/model.neuron CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:62fa1d90f218e31a6779db8fd5158cf5e96adecda2fc03efd0ea1d2ef1d60640
3
  size 496092729
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db4debf1af68b61b5e7c6b5b85107c8f8a736bfee0ee4011daca1d3be90169f5
3
  size 496092729
neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/text_encoder_2/model.neuron CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:858b9a5b53a5437d6f1741c687ed59ee88a68666efe75d97299f194bed60d283
3
  size 2801582571
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94445aa6f8743f3d05b2f1b00d2376a43607f5832f525f7aaeb3f0a7f9ca9511
3
  size 2801582571
neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/unet/model.neuron CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b7356baceb6018beb09d703e2b8032d4a53c1786268b752ec2ed8492bc9801a6
3
  size 4157656637
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7d54e2b38b792251505236cc5842f4e6d4d1f9812328c90c4811f6d95940d30
3
  size 4157656637
neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/vae_decoder/config.json CHANGED
@@ -3,12 +3,12 @@
3
  "_commit_hash": null,
4
  "_diffusers_version": "0.34.0",
5
  "_use_default_values": [
6
- "latents_std",
7
  "shift_factor",
8
- "mid_block_add_attention",
9
  "use_quant_conv",
10
- "latents_mean",
11
- "use_post_quant_conv"
12
  ],
13
  "act_fn": "silu",
14
  "block_out_channels": [
 
3
  "_commit_hash": null,
4
  "_diffusers_version": "0.34.0",
5
  "_use_default_values": [
6
+ "latents_mean",
7
  "shift_factor",
8
+ "use_post_quant_conv",
9
  "use_quant_conv",
10
+ "latents_std",
11
+ "mid_block_add_attention"
12
  ],
13
  "act_fn": "silu",
14
  "block_out_channels": [
neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/vae_decoder/model.neuron CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9110cdfdbddf629c0b638ce6d3f27860587924e43cd9c582b743e4f78d2ecd9f
3
  size 349015667
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dcf94f31e95ce250169a921c84a9606b2a8257a22664ded8f300f81efa60e085
3
  size 349015667
neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/vae_encoder/config.json CHANGED
@@ -3,12 +3,12 @@
3
  "_commit_hash": null,
4
  "_diffusers_version": "0.34.0",
5
  "_use_default_values": [
6
- "latents_std",
7
  "shift_factor",
8
- "mid_block_add_attention",
9
  "use_quant_conv",
10
- "latents_mean",
11
- "use_post_quant_conv"
12
  ],
13
  "act_fn": "silu",
14
  "block_out_channels": [
 
3
  "_commit_hash": null,
4
  "_diffusers_version": "0.34.0",
5
  "_use_default_values": [
6
+ "latents_mean",
7
  "shift_factor",
8
+ "use_post_quant_conv",
9
  "use_quant_conv",
10
+ "latents_std",
11
+ "mid_block_add_attention"
12
  ],
13
  "act_fn": "silu",
14
  "block_out_channels": [
neuronxcc-2.19.8089.0+8ab9f450/MODULE_7c1121df63c546f81f0f/vae_encoder/model.neuron CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7a4d91824d33622eb70eb3b39c58f59ca96fd7786ce028ee359662721b35d7dd
3
  size 201275507
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99136300b4e42d08ac7c41a9692b2a020736de2f1b2f4d27695bee3a3c758544
3
  size 201275507