keras-hub-nightly 0.22.0.dev202508110431__py3-none-any.whl → 0.22.0.dev202508140419__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (16) hide show
  1. keras_hub/src/models/deit/__init__.py +5 -0
  2. keras_hub/src/models/deit/deit_presets.py +8 -8
  3. keras_hub/src/models/esm/__init__.py +5 -0
  4. keras_hub/src/models/esm/esm_presets.py +53 -0
  5. keras_hub/src/models/mobilenet/mobilenet_presets.py +4 -4
  6. keras_hub/src/models/pali_gemma/pali_gemma_presets.py +17 -17
  7. keras_hub/src/models/qwen3/qwen3_presets.py +6 -6
  8. keras_hub/src/models/qwen_moe/qwen_moe_presets.py +1 -1
  9. keras_hub/src/models/roformer_v2/roformer_v2_attention.py +0 -3
  10. keras_hub/src/models/stable_diffusion_3/stable_diffusion_3_backbone.py +8 -0
  11. keras_hub/src/utils/transformers/convert_llama3.py +6 -0
  12. keras_hub/src/version.py +1 -1
  13. {keras_hub_nightly-0.22.0.dev202508110431.dist-info → keras_hub_nightly-0.22.0.dev202508140419.dist-info}/METADATA +4 -4
  14. {keras_hub_nightly-0.22.0.dev202508110431.dist-info → keras_hub_nightly-0.22.0.dev202508140419.dist-info}/RECORD +16 -16
  15. {keras_hub_nightly-0.22.0.dev202508110431.dist-info → keras_hub_nightly-0.22.0.dev202508140419.dist-info}/WHEEL +0 -0
  16. {keras_hub_nightly-0.22.0.dev202508110431.dist-info → keras_hub_nightly-0.22.0.dev202508140419.dist-info}/top_level.txt +0 -0
keras_hub/src/models/deit/__init__.py CHANGED
@@ -0,0 +1,5 @@
1
+ from keras_hub.src.models.deit.deit_backbone import DeiTBackbone
2
+ from keras_hub.src.models.deit.deit_presets import backbone_presets
3
+ from keras_hub.src.utils.preset_utils import register_presets
4
+
5
+ register_presets(backbone_presets, DeiTBackbone)
keras_hub/src/models/deit/deit_presets.py CHANGED
@@ -2,46 +2,46 @@
2
2
 
3
3
  # Metadata for loading pretrained model weights.
4
4
  backbone_presets = {
5
- "deit-base-distilled-patch16-384_imagenet": {
5
+ "deit_base_distilled_patch16_384_imagenet": {
6
6
  "metadata": {
7
7
  "description": (
8
8
  "DeiT-B16 model pre-trained on the ImageNet 1k dataset with "
9
9
  "image resolution of 384x384 "
10
10
  ),
11
- "params": 86092032,
11
+ "params": 86_092_032,
12
12
  "path": "deit",
13
13
  },
14
14
  "kaggle_handle": "kaggle://keras/deit/keras/deit_base_distilled_patch16_384_imagenet/1",
15
15
  },
16
- "deit-base-distilled-patch16-224_imagenet": {
16
+ "deit_base_distilled_patch16_224_imagenet": {
17
17
  "metadata": {
18
18
  "description": (
19
19
  "DeiT-B16 model pre-trained on the ImageNet 1k dataset with "
20
20
  "image resolution of 224x224 "
21
21
  ),
22
- "params": 85800192,
22
+ "params": 85_800_192,
23
23
  "path": "deit",
24
24
  },
25
25
  "kaggle_handle": "kaggle://keras/deit/keras/deit_base_distilled_patch16_224_imagenet/1",
26
26
  },
27
- "deit-tiny-distilled-patch16-224_imagenet": {
27
+ "deit_tiny_distilled_patch16_224_imagenet": {
28
28
  "metadata": {
29
29
  "description": (
30
30
  "DeiT-T16 model pre-trained on the ImageNet 1k dataset with "
31
31
  "image resolution of 224x224 "
32
32
  ),
33
- "params": 5524800,
33
+ "params": 5_524_800,
34
34
  "path": "deit",
35
35
  },
36
36
  "kaggle_handle": "kaggle://keras/deit/keras/deit_tiny_distilled_patch16_224_imagenet/1",
37
37
  },
38
- "deit-small-distilled-patch16-224_imagenet": {
38
+ "deit_small_distilled_patch16_224_imagenet": {
39
39
  "metadata": {
40
40
  "description": (
41
41
  "DeiT-S16 model pre-trained on the ImageNet 1k dataset with "
42
42
  "image resolution of 224x224 "
43
43
  ),
44
- "params": 21666432,
44
+ "params": 21_666_432,
45
45
  "path": "deit",
46
46
  },
47
47
  "kaggle_handle": "kaggle://keras/deit/keras/deit_small_distilled_patch16_224_imagenet/1",
keras_hub/src/models/esm/__init__.py CHANGED
@@ -0,0 +1,5 @@
1
+ from keras_hub.src.models.esm.esm_backbone import ESMBackbone
2
+ from keras_hub.src.models.esm.esm_presets import backbone_presets
3
+ from keras_hub.src.utils.preset_utils import register_presets
4
+
5
+ register_presets(backbone_presets, ESMBackbone)
keras_hub/src/models/esm/esm_presets.py CHANGED
@@ -0,0 +1,53 @@
1
+ """ESM model preset configurations."""
2
+
3
+ # Metadata for loading pretrained model weights.
4
+ backbone_presets = {
5
+ "esm2_t6_8M": {
6
+ "metadata": {
7
+ "description": (
8
+ "6 transformer layers version of the ESM-2 protein language "
9
+ "model, trained on the UniRef50 clustered protein sequence "
10
+ "dataset."
11
+ ),
12
+ "params": 7_408_960,
13
+ "path": "esm",
14
+ },
15
+ "kaggle_handle": "kaggle://keras/esm-2/keras/esm2_t6_8M/1",
16
+ },
17
+ "esm2_t12_35M": {
18
+ "metadata": {
19
+ "description": (
20
+ "12 transformer layers version of the ESM-2 protein language "
21
+ "model, trained on the UniRef50 clustered protein sequence "
22
+ "dataset."
23
+ ),
24
+ "params": 33_269_280,
25
+ "path": "esm",
26
+ },
27
+ "kaggle_handle": "kaggle://keras/esm-2/keras/esm2_t12_35M/1",
28
+ },
29
+ "esm2_t30_150M": {
30
+ "metadata": {
31
+ "description": (
32
+ "30 transformer layers version of the ESM-2 protein language "
33
+ "model, trained on the UniRef50 clustered protein sequence "
34
+ "dataset."
35
+ ),
36
+ "params": 147_728_000,
37
+ "path": "esm",
38
+ },
39
+ "kaggle_handle": "kaggle://keras/esm-2/keras/esm2_t30_150M/1",
40
+ },
41
+ "esm2_t33_650M": {
42
+ "metadata": {
43
+ "description": (
44
+ "33 transformer layers version of the ESM-2 protein language "
45
+ "model, trained on the UniRef50 clustered protein sequence "
46
+ "dataset."
47
+ ),
48
+ "params": 649_400_320,
49
+ "path": "esm",
50
+ },
51
+ "kaggle_handle": "kaggle://keras/esm-2/keras/esm2_t33_650M/1",
52
+ },
53
+ }
keras_hub/src/models/mobilenet/mobilenet_presets.py CHANGED
@@ -8,7 +8,7 @@ backbone_presets = {
8
8
  "dataset at a 224x224 resolution. Has half channel multiplier."
9
9
  ),
10
10
  "params": 278784,
11
- "path": "mobilenetv3",
11
+ "path": "mobilenet",
12
12
  },
13
13
  "kaggle_handle": "kaggle://keras/mobilenetv3/keras/mobilenet_v3_small_050_imagenet/1",
14
14
  },
@@ -20,7 +20,7 @@ backbone_presets = {
20
20
  "multiplier."
21
21
  ),
22
22
  "params": 939120,
23
- "path": "mobilenetv3",
23
+ "path": "mobilenet",
24
24
  },
25
25
  "kaggle_handle": "kaggle://keras/mobilenetv3/keras/mobilenet_v3_small_100_imagenet/1",
26
26
  },
@@ -32,7 +32,7 @@ backbone_presets = {
32
32
  "multiplier."
33
33
  ),
34
34
  "params": 2996352,
35
- "path": "mobilenetv3",
35
+ "path": "mobilenet",
36
36
  },
37
37
  "kaggle_handle": "kaggle://keras/mobilenetv3/keras/mobilenet_v3_large_100_imagenet/1",
38
38
  },
@@ -44,7 +44,7 @@ backbone_presets = {
44
44
  "multiplier."
45
45
  ),
46
46
  "params": 2996352,
47
- "path": "mobilenetv3",
47
+ "path": "mobilenet",
48
48
  },
49
49
  "kaggle_handle": "kaggle://keras/mobilenetv3/keras/mobilenet_v3_large_100_imagenet_21k/1",
50
50
  },
keras_hub/src/models/pali_gemma/pali_gemma_presets.py CHANGED
@@ -63,7 +63,7 @@ backbone_presets = {
63
63
  ),
64
64
  "params": 3032979696,
65
65
  "official_name": "PaliGemma2",
66
- "path": "pali_gemma2",
66
+ "path": "pali_gemma",
67
67
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
68
68
  },
69
69
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma_2_ft_docci_3b_448/2",
@@ -78,7 +78,7 @@ backbone_presets = {
78
78
  ),
79
79
  "params": 9663294192,
80
80
  "official_name": "PaliGemma2",
81
- "path": "pali_gemma2",
81
+ "path": "pali_gemma",
82
82
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
83
83
  },
84
84
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_ft_docci_10b_448/3",
@@ -93,7 +93,7 @@ backbone_presets = {
93
93
  ),
94
94
  "params": 3032094960,
95
95
  "official_name": "PaliGemma2",
96
- "path": "pali_gemma2",
96
+ "path": "pali_gemma",
97
97
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
98
98
  },
99
99
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_mix_3b_224/2",
@@ -108,7 +108,7 @@ backbone_presets = {
108
108
  ),
109
109
  "params": 3032979696,
110
110
  "official_name": "PaliGemma2",
111
- "path": "pali_gemma2",
111
+ "path": "pali_gemma",
112
112
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
113
113
  },
114
114
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_mix_3b_448/2",
@@ -123,7 +123,7 @@ backbone_presets = {
123
123
  ),
124
124
  "params": 9662409456,
125
125
  "official_name": "PaliGemma2",
126
- "path": "pali_gemma2",
126
+ "path": "pali_gemma",
127
127
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
128
128
  },
129
129
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_mix_10b_224/3",
@@ -138,7 +138,7 @@ backbone_presets = {
138
138
  ),
139
139
  "params": 9663294192,
140
140
  "official_name": "PaliGemma2",
141
- "path": "pali_gemma2",
141
+ "path": "pali_gemma",
142
142
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
143
143
  },
144
144
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_mix_10b_448/3",
@@ -153,7 +153,7 @@ backbone_presets = {
153
153
  ),
154
154
  "params": 27650192112,
155
155
  "official_name": "PaliGemma2",
156
- "path": "pali_gemma2",
156
+ "path": "pali_gemma",
157
157
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
158
158
  },
159
159
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_28b_mix_224/3",
@@ -168,7 +168,7 @@ backbone_presets = {
168
168
  ),
169
169
  "params": 27650192112,
170
170
  "official_name": "PaliGemma2",
171
- "path": "pali_gemma2",
171
+ "path": "pali_gemma",
172
172
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
173
173
  },
174
174
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_28b_mix_448/3",
@@ -183,7 +183,7 @@ backbone_presets = {
183
183
  ),
184
184
  "params": 3032094960,
185
185
  "official_name": "PaliGemma2",
186
- "path": "pali_gemma2",
186
+ "path": "pali_gemma",
187
187
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
188
188
  },
189
189
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_pt_3b_224/2",
@@ -198,7 +198,7 @@ backbone_presets = {
198
198
  ),
199
199
  "params": 3032979696,
200
200
  "official_name": "PaliGemma2",
201
- "path": "pali_gemma2",
201
+ "path": "pali_gemma",
202
202
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
203
203
  },
204
204
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_pt_3b_448/2",
@@ -213,7 +213,7 @@ backbone_presets = {
213
213
  ),
214
214
  "params": 3036518640,
215
215
  "official_name": "PaliGemma2",
216
- "path": "pali_gemma2",
216
+ "path": "pali_gemma",
217
217
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
218
218
  },
219
219
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_pt_3b_896/2",
@@ -228,7 +228,7 @@ backbone_presets = {
228
228
  ),
229
229
  "params": 9662409456,
230
230
  "official_name": "PaliGemma2",
231
- "path": "pali_gemma2",
231
+ "path": "pali_gemma",
232
232
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
233
233
  },
234
234
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_pt_10b_224/3",
@@ -243,7 +243,7 @@ backbone_presets = {
243
243
  ),
244
244
  "params": 9663294192,
245
245
  "official_name": "PaliGemma2",
246
- "path": "pali_gemma2",
246
+ "path": "pali_gemma",
247
247
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
248
248
  },
249
249
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_pt_10b_448/3",
@@ -258,7 +258,7 @@ backbone_presets = {
258
258
  ),
259
259
  "params": 9666833136,
260
260
  "official_name": "PaliGemma2",
261
- "path": "pali_gemma2",
261
+ "path": "pali_gemma",
262
262
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
263
263
  },
264
264
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_pt_10b_896/3",
@@ -273,7 +273,7 @@ backbone_presets = {
273
273
  ),
274
274
  "params": 27650192112,
275
275
  "official_name": "PaliGemma2",
276
- "path": "pali_gemma2",
276
+ "path": "pali_gemma",
277
277
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
278
278
  },
279
279
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_pt_28b_224/4",
@@ -288,7 +288,7 @@ backbone_presets = {
288
288
  ),
289
289
  "params": 27650192112,
290
290
  "official_name": "PaliGemma2",
291
- "path": "pali_gemma2",
291
+ "path": "pali_gemma",
292
292
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
293
293
  },
294
294
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_pt_28b_448/3",
@@ -303,7 +303,7 @@ backbone_presets = {
303
303
  ),
304
304
  "params": 27650192112,
305
305
  "official_name": "PaliGemma2",
306
- "path": "pali_gemma2",
306
+ "path": "pali_gemma",
307
307
  "model_card": "https://www.kaggle.com/models/google/paligemma-2",
308
308
  },
309
309
  "kaggle_handle": "kaggle://keras/paligemma2/keras/pali_gemma2_pt_28b_896/3",
keras_hub/src/models/qwen3/qwen3_presets.py CHANGED
@@ -8,7 +8,7 @@ backbone_presets = {
8
8
  "efficiency and fast inference on resource-constrained devices."
9
9
  ),
10
10
  "params": 596049920,
11
- "path": "qwen-3",
11
+ "path": "qwen3",
12
12
  },
13
13
  "kaggle_handle": "kaggle://keras/qwen-3/keras/qwen3_0.6b_en/1",
14
14
  },
@@ -19,7 +19,7 @@ backbone_presets = {
19
19
  "a good balance between performance and resource usage."
20
20
  ),
21
21
  "params": 1720574976,
22
- "path": "qwen-3",
22
+ "path": "qwen3",
23
23
  },
24
24
  "kaggle_handle": "kaggle://keras/qwen-3/keras/qwen3_1.7b_en/1",
25
25
  },
@@ -31,7 +31,7 @@ backbone_presets = {
31
31
  "variants."
32
32
  ),
33
33
  "params": 4022468096,
34
- "path": "qwen-3",
34
+ "path": "qwen3",
35
35
  },
36
36
  "kaggle_handle": "kaggle://keras/qwen-3/keras/qwen3_4b_en/1",
37
37
  },
@@ -43,7 +43,7 @@ backbone_presets = {
43
43
  "capabilities."
44
44
  ),
45
45
  "params": 8190735360,
46
- "path": "qwen-3",
46
+ "path": "qwen3",
47
47
  },
48
48
  "kaggle_handle": "kaggle://keras/qwen-3/keras/qwen3_8b_en/1",
49
49
  },
@@ -54,7 +54,7 @@ backbone_presets = {
54
54
  "advanced reasoning, coding, and multilingual capabilities."
55
55
  ),
56
56
  "params": 14768307200,
57
- "path": "qwen-3",
57
+ "path": "qwen3",
58
58
  },
59
59
  "kaggle_handle": "kaggle://keras/qwen-3/keras/qwen3_14b_en/1",
60
60
  },
@@ -66,7 +66,7 @@ backbone_presets = {
66
66
  "general language tasks."
67
67
  ),
68
68
  "params": 32762123264,
69
- "path": "qwen-3",
69
+ "path": "qwen3",
70
70
  },
71
71
  "kaggle_handle": "kaggle://keras/qwen-3/keras/qwen3_32b_en/1",
72
72
  },
keras_hub/src/models/qwen_moe/qwen_moe_presets.py CHANGED
@@ -8,7 +8,7 @@ backbone_presets = {
8
8
  "and 8 experts per MoE layer."
9
9
  ),
10
10
  "params": 14315784192,
11
- "path": "qwen-1.5-moe",
11
+ "path": "qwen_moe",
12
12
  },
13
13
  "kaggle_handle": "kaggle://keras/qwen-1.5-moe/Keras/qwen1.5_moe_2.7b_en/4",
14
14
  },
keras_hub/src/models/roformer_v2/roformer_v2_attention.py CHANGED
@@ -1,7 +1,6 @@
1
1
  import keras
2
2
  from keras import initializers
3
3
  from keras import ops
4
- from packaging import version
5
4
 
6
5
 
7
6
  class RoformerNorm(keras.layers.Layer):
@@ -180,8 +179,6 @@ class RoformerAttention(keras.layers.Layer):
180
179
  vw = ops.reshape(vw, (b, s, self.heads, self.head_size))
181
180
 
182
181
  qw, kw = self.rotary_embedding_layer([qw, kw])
183
- if version.parse(keras.__version__) < version.parse("3.6"):
184
- raise ("Please make sure your Keras version is >=3.6.")
185
182
  flash_attention = keras.config.is_flash_attention_enabled()
186
183
  attention_mask = ops.reshape(attention_mask, [b, 1, s, 1])
187
184
  if keras.config.backend() == "torch":
keras_hub/src/models/stable_diffusion_3/stable_diffusion_3_backbone.py CHANGED
@@ -1,4 +1,6 @@
1
1
  import keras
2
+ from keras import backend
3
+ from keras import distribution
2
4
  from keras import layers
3
5
  from keras import ops
4
6
 
@@ -621,11 +623,17 @@ class StableDiffusion3Backbone(Backbone):
621
623
  config["vae"]["config"]["dtype"] = dtype_config
622
624
 
623
625
  # Text encoders default to float16 dtype if not specified.
626
+ # TODO: JAX CPU doesn't support float16 in `nn.dot_product_attention`.
627
+ is_jax_cpu = (
628
+ backend.backend() == "jax"
629
+ and "cpu" in distribution.list_devices()[0].lower()
630
+ )
624
631
  for text_encoder in ("clip_l", "clip_g", "t5"):
625
632
  if (
626
633
  text_encoder in config
627
634
  and config[text_encoder] is not None
628
635
  and "dtype" not in config[text_encoder]["config"]
636
+ and not is_jax_cpu
629
637
  ):
630
638
  config[text_encoder]["config"]["dtype"] = "float16"
631
639
 
keras_hub/src/utils/transformers/convert_llama3.py CHANGED
@@ -127,6 +127,12 @@ def convert_tokenizer(cls, preset, **kwargs):
127
127
  vocab = tokenizer_config["model"]["vocab"]
128
128
  merges = tokenizer_config["model"]["merges"]
129
129
 
130
+ # Handle different merge formats
131
+ if merges and isinstance(merges[0], list) and len(merges[0]) == 2:
132
+ # Convert list of lists format [["Ġ", "a"], ["Ġ", "b"]]
133
+ # to space-separated strings
134
+ merges = [" ".join(merge) for merge in merges]
135
+
130
136
  # Load all special tokens with the exception of "reserved" ones.
131
137
  special_tokens = set()
132
138
  for token in tokenizer_config["added_tokens"]:
keras_hub/src/version.py CHANGED
@@ -1,7 +1,7 @@
1
1
  from keras_hub.src.api_export import keras_hub_export
2
2
 
3
3
  # Unique source of truth for the version number.
4
- __version__ = "0.22.0.dev202508110431"
4
+ __version__ = "0.22.0.dev202508140419"
5
5
 
6
6
 
7
7
  @keras_hub_export("keras_hub.version")
{keras_hub_nightly-0.22.0.dev202508110431.dist-info → keras_hub_nightly-0.22.0.dev202508140419.dist-info}/METADATA RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: keras-hub-nightly
3
- Version: 0.22.0.dev202508110431
3
+ Version: 0.22.0.dev202508140419
4
4
  Summary: Pretrained models for Keras.
5
5
  Author-email: Keras team <keras-users@googlegroups.com>
6
6
  License-Expression: Apache-2.0
@@ -18,9 +18,9 @@ Classifier: Operating System :: MacOS
18
18
  Classifier: Intended Audience :: Science/Research
19
19
  Classifier: Topic :: Scientific/Engineering
20
20
  Classifier: Topic :: Software Development
21
- Requires-Python: >=3.9
21
+ Requires-Python: >=3.10
22
22
  Description-Content-Type: text/markdown
23
- Requires-Dist: keras>=3.5
23
+ Requires-Dist: keras>=3.8
24
24
  Requires-Dist: absl-py
25
25
  Requires-Dist: numpy
26
26
  Requires-Dist: packaging
@@ -31,7 +31,7 @@ Requires-Dist: tensorflow-text; platform_system != "Windows"
31
31
 
32
32
  # KerasHub: Multi-framework Pretrained Models
33
33
  [![](https://github.com/keras-team/keras-hub/workflows/Tests/badge.svg?branch=master)](https://github.com/keras-team/keras-hub/actions?query=workflow%3ATests+branch%3Amaster)
34
- ![Python](https://img.shields.io/badge/python-v3.9.0+-success.svg)
34
+ ![Python](https://img.shields.io/badge/python-v3.10.0+-success.svg)
35
35
  [![contributions welcome](https://img.shields.io/badge/contributions-welcome-brightgreen.svg?style=flat)](https://github.com/keras-team/keras-hub/issues)
36
36
 
37
37
  > [!IMPORTANT]
{keras_hub_nightly-0.22.0.dev202508110431.dist-info → keras_hub_nightly-0.22.0.dev202508140419.dist-info}/RECORD RENAMED
@@ -5,7 +5,7 @@ keras_hub/models/__init__.py,sha256=UXMwKVZ7bg-AOrq2xsl8M0idUAS89pkdCvQKhzL-D3I,
5
5
  keras_hub/samplers/__init__.py,sha256=aFQIkiqbZpi8vjrPp2MVII4QUfE-eQjra5fMeHsoy7k,886
6
6
  keras_hub/src/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
7
7
  keras_hub/src/api_export.py,sha256=9pQZK27JObxWZ96QPLBp1OBsjWigh1iuV6RglPGMRk0,1499
8
- keras_hub/src/version.py,sha256=Kfnwa0XS7R0JFLM6j6sQxVyK9QBEZ0IxDVQU7992vGM,222
8
+ keras_hub/src/version.py,sha256=cvWGyJkZFd-yHG--qkMN58JFqhHaSPHDxkqsRUvNs0w,222
9
9
  keras_hub/src/layers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
10
10
  keras_hub/src/layers/modeling/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
11
11
  keras_hub/src/layers/modeling/alibi_bias.py,sha256=1XBTHI52L_iJDhN_w5ydu_iMhCuTgQAxEPwcLA6BPuk,4411
@@ -134,13 +134,13 @@ keras_hub/src/models/deeplab_v3/deeplab_v3_image_segmeter_preprocessor.py,sha256
134
134
  keras_hub/src/models/deeplab_v3/deeplab_v3_layers.py,sha256=mz9nG55gdXSTDE96AXgeTCwUFB95DIpTuqrvWIt5Lco,7840
135
135
  keras_hub/src/models/deeplab_v3/deeplab_v3_presets.py,sha256=ZKYY8A7mV2QvwXwjDUd9xAbVHo58-Hgj_IqNUbuyCIU,625
136
136
  keras_hub/src/models/deeplab_v3/deeplab_v3_segmenter.py,sha256=pubi30sPJKLOpz9fRQff2FZt_53KBvwf2uyaJ5YL7J8,3726
137
- keras_hub/src/models/deit/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
137
+ keras_hub/src/models/deit/__init__.py,sha256=5XUICYa-poqErbmMLArBKCrSxC7wsIiQwUpuCnvGt_E,245
138
138
  keras_hub/src/models/deit/deit_backbone.py,sha256=R5pBOqe8vcvD8VaRnsy_zIRIz6BLnUbkTeKUOoGNHPA,5942
139
139
  keras_hub/src/models/deit/deit_image_classifier.py,sha256=pUS2638yBAxEBxcJoHyLABsgjCWv_Y0Mj_8u0YgDPdI,5758
140
140
  keras_hub/src/models/deit/deit_image_classifier_preprocessor.py,sha256=s5pTcsUjlt1oIXFWIu-9gf2-sBesAyrjJIYmFOB96Xs,514
141
141
  keras_hub/src/models/deit/deit_image_converter.py,sha256=wEGCLHS_i4wF9WA4m7uUXcHNbwf6TYgvPoM6C_t0rpM,330
142
142
  keras_hub/src/models/deit/deit_layers.py,sha256=A80-UTHEUV8g5rEG-fr8OQpGe3HeoYlYwpoDCtq71ZU,17278
143
- keras_hub/src/models/deit/deit_presets.py,sha256=0c2jm2DIznOr6ciQoLM6QYopQTLiMx4jONGLaXvtt6g,1778
143
+ keras_hub/src/models/deit/deit_presets.py,sha256=5VwMAEg16RLWOjcdZ-BCYVlUlEzBfHz-6wCSOIhWGVQ,1786
144
144
  keras_hub/src/models/densenet/__init__.py,sha256=r7StyamnWeeZxOk9r4ZYNbS_YVhu9YGPyXhNxljvdPg,269
145
145
  keras_hub/src/models/densenet/densenet_backbone.py,sha256=f2nfsXyXQert2aYHq-L-JZtp8inq1fs1K47rzZQ9nTI,6744
146
146
  keras_hub/src/models/densenet/densenet_image_classifier.py,sha256=ye-Ix3oU42pfsDoh-h1PG4di1kzldO0ZO7Nj304p_X4,544
@@ -173,7 +173,7 @@ keras_hub/src/models/electra/__init__.py,sha256=vaXl_uQx_oLeKZWxmc1NRgCJfHpYJ35J
173
173
  keras_hub/src/models/electra/electra_backbone.py,sha256=h-QuFxACBvbMktkyGV2pIgn6dQ-kudJB1i14ekwEaL4,9004
174
174
  keras_hub/src/models/electra/electra_presets.py,sha256=6f0WAYtDx5To4gvi6btN8I8y7yfc9ANchTHRKgCyIkg,2697
175
175
  keras_hub/src/models/electra/electra_tokenizer.py,sha256=Ll_EW-14i-OZr6appQEt5ceMUCeEadF4yPJHMwaRfVs,2729
176
- keras_hub/src/models/esm/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
176
+ keras_hub/src/models/esm/__init__.py,sha256=_IlazeBwHkpetmLIZz3fFzC8CFcSnBRVQvw9nes4TN8,239
177
177
  keras_hub/src/models/esm/esm_attention.py,sha256=T21MVs9QDUe_8a53mcW3dJvJfaNZOg5lkMdxGhQdmFQ,3327
178
178
  keras_hub/src/models/esm/esm_backbone.py,sha256=ADIpeiYz16fw1PNvx2tX-51HsZ_AjR2wGLkXZHErWBg,8696
179
179
  keras_hub/src/models/esm/esm_classifier.py,sha256=35-_3U725JhzspQAO_4ZkTJ0Tuy0XKMVkSrpmFz2CaE,6049
@@ -181,7 +181,7 @@ keras_hub/src/models/esm/esm_classifier_preprocessor.py,sha256=TXjGH8ttElEsfBLOM
181
181
  keras_hub/src/models/esm/esm_encoder.py,sha256=FxqfM_amKnmzNJoTq-LKouKaf_huklbjLiQ37ip85Tc,4499
182
182
  keras_hub/src/models/esm/esm_masked_plm.py,sha256=FTNHrr0nRiuuO0Yqf5NSM48PehXWKMZvUVLBGET8X-8,3874
183
183
  keras_hub/src/models/esm/esm_masked_plm_preprocessor.py,sha256=jfpehbd1KN_s48KCPSUpzQf1YYeriuR6a81wmXSG8bE,6272
184
- keras_hub/src/models/esm/esm_presets.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
184
+ keras_hub/src/models/esm/esm_presets.py,sha256=f3O0qhHoHGx-xXS4DuSw8fqxVEKmDaj45jput7OMz9M,1792
185
185
  keras_hub/src/models/esm/esm_tokenizer.py,sha256=6hKDWanN4Hfl6eSNXHiHJUcwSMDRL4gHEWxenaMI3Os,3079
186
186
  keras_hub/src/models/f_net/__init__.py,sha256=a3OAwgEVy3Rv88ZlBE9RYLrPCNteImhGkW-lSAq5hyI,249
187
187
  keras_hub/src/models/f_net/f_net_backbone.py,sha256=6vZEq2UgoJxU2-aEesdXZnyRbACxpMZQ1akyVbGH8wg,8290
@@ -295,7 +295,7 @@ keras_hub/src/models/mobilenet/mobilenet_backbone.py,sha256=aZBSFeLUObYYoi3od9DI
295
295
  keras_hub/src/models/mobilenet/mobilenet_image_classifier.py,sha256=rgPVJeSRqyp3-Fgf5ERbg_97c4cSawRmAtoJpdBN8WA,2437
296
296
  keras_hub/src/models/mobilenet/mobilenet_image_classifier_preprocessor.py,sha256=yqM4wQ3ae7wXTBO0aMuvJx6XqllA7Psqzjvpm2NABXM,573
297
297
  keras_hub/src/models/mobilenet/mobilenet_image_converter.py,sha256=a3Ka0UYYK5wHSOjf2oMHSgofRazTAeUfttklVefq14w,360
298
- keras_hub/src/models/mobilenet/mobilenet_presets.py,sha256=--nhaM6LmaiCtQlZPDwoQTHW7ciU0igzS4f9ssdD9Lo,1903
298
+ keras_hub/src/models/mobilenet/mobilenet_presets.py,sha256=hR_3xxI_PigE8UprXW4lAuKRa3LFGdidBaN8LklxwRQ,1895
299
299
  keras_hub/src/models/mobilenet/util.py,sha256=S7j4UacmVIJ3fU8cymyAoK49eHcpWIKTOyUQiEjcbzQ,721
300
300
  keras_hub/src/models/moonshine/__init__.py,sha256=WK_9Cy1dp5KplNAaTsaJbd-2DGLsiHQsIL5ZnXuCbDQ,275
301
301
  keras_hub/src/models/moonshine/moonshine_audio_converter.py,sha256=FnvR7SP44uVOsA3g9azUhQjsVg809eJ5nqoJZQ-DAq0,11854
@@ -320,7 +320,7 @@ keras_hub/src/models/pali_gemma/pali_gemma_causal_lm.py,sha256=aT075qRyFmuo5Jwph
320
320
  keras_hub/src/models/pali_gemma/pali_gemma_causal_lm_preprocessor.py,sha256=F57y0fZ0wYYxfGIjfrJc1W9uQpViYFx5bvFjj5CqUbI,4814
321
321
  keras_hub/src/models/pali_gemma/pali_gemma_decoder_block.py,sha256=24ABQ1vGlppV-KfWh0YqJjzM_Lu2GIwvyJ4X2XXie_A,5616
322
322
  keras_hub/src/models/pali_gemma/pali_gemma_image_converter.py,sha256=5yM_jUtrFsWIieiwfFBoP7mtPmQAwywkeLKbd7fhmzk,371
323
- keras_hub/src/models/pali_gemma/pali_gemma_presets.py,sha256=DAaSzquR4_AnSjToDjgXj2zbrT5skUpXmzKoyATwwHk,13006
323
+ keras_hub/src/models/pali_gemma/pali_gemma_presets.py,sha256=Q_zfHEjGTtXEiCwjoJc2g6HjmoNoLgSDRNfRvIsf0dA,12989
324
324
  keras_hub/src/models/pali_gemma/pali_gemma_tokenizer.py,sha256=ljTiADHo0Ok88q-jVzwJIle2C8xcxnudLTsBLzIySaM,2415
325
325
  keras_hub/src/models/pali_gemma/pali_gemma_vit.py,sha256=SbWanwCoONSwgiWQsc6lFdvhqKZ-zDW42XzQt8CNMtU,18311
326
326
  keras_hub/src/models/phi3/__init__.py,sha256=zIbf1MU-ks91mEkjTRJAsk51N3BBnXDF2JM1vO-13PQ,245
@@ -349,7 +349,7 @@ keras_hub/src/models/qwen3/qwen3_causal_lm.py,sha256=cn_4WFVxhlOArtIGAaqkNzIz9Rx
349
349
  keras_hub/src/models/qwen3/qwen3_causal_lm_preprocessor.py,sha256=H4g-bgvuhAUnDwjJovydK16Kes38ZFZWPvflrgHqZis,458
350
350
  keras_hub/src/models/qwen3/qwen3_decoder.py,sha256=68s9jQj53zFmXE4-SGXKYHu546fXOyi9LUbnKk-HGYY,11595
351
351
  keras_hub/src/models/qwen3/qwen3_layernorm.py,sha256=EJxjf7Pr6ufPQnNeuYQxkExzPjPk4PQxqMsoBeSEkDo,1073
352
- keras_hub/src/models/qwen3/qwen3_presets.py,sha256=ZcsmPFj3Z4TBoa7ZkJK4JN1D6iHHZ6kCrqXxQE8IH_k,2524
352
+ keras_hub/src/models/qwen3/qwen3_presets.py,sha256=eAqRbjLyRTSXcN-jnGHqoCHejKm2gmt8_zL4EPoE-JA,2518
353
353
  keras_hub/src/models/qwen3/qwen3_tokenizer.py,sha256=LmPtg0vprMchDvYfTj8m5PraXI2QS3-YgdIIpIm5iAs,1448
354
354
  keras_hub/src/models/qwen_moe/__init__.py,sha256=5D8GUmVDsJs0J4sVZHcXOLkZf12U96l-WtwyVee4lu8,267
355
355
  keras_hub/src/models/qwen_moe/qwen_moe_attention.py,sha256=o0mcVTDMtElMYq3NSYRCfuYVdF-W8YDSU5ogensrVJg,13277
@@ -358,7 +358,7 @@ keras_hub/src/models/qwen_moe/qwen_moe_causal_lm.py,sha256=MeP60v7GcN_SmH5_ULRpq
358
358
  keras_hub/src/models/qwen_moe/qwen_moe_causal_lm_preprocessor.py,sha256=9P6TT7W_fqf4HsXcmlHF-DW_anR-XoDrRN2ZFGA7Ai4,3168
359
359
  keras_hub/src/models/qwen_moe/qwen_moe_decoder.py,sha256=kmUjLpYTbJQ3J_31qWhLOd0Dg2_9cl_JX_zM8ZMH1Qo,23130
360
360
  keras_hub/src/models/qwen_moe/qwen_moe_layernorm.py,sha256=DbkWJo7U0-cwdZwHPeAnFznYwtao6o0fjpoDJ9UWnpc,927
361
- keras_hub/src/models/qwen_moe/qwen_moe_presets.py,sha256=ZL5LHXtvcxgY38ZWZPLAPzW742sxNYER_sHn2At9nLk,455
361
+ keras_hub/src/models/qwen_moe/qwen_moe_presets.py,sha256=nGQ0azaOJAjBorR_6_Qtb1yCSXPdFJdRp0_ULYT4_04,451
362
362
  keras_hub/src/models/qwen_moe/qwen_moe_tokenizer.py,sha256=2c3X8jNGO0q0UL5NtUqSgHWLqhyJGi2ohNcTeOGhd84,1407
363
363
  keras_hub/src/models/resnet/__init__.py,sha256=C5UqlQ6apm8WSp1bnrxB6Bi3BGaknxRQs-r3b2wpaGA,257
364
364
  keras_hub/src/models/resnet/resnet_backbone.py,sha256=Q7nlqcTXZzjqd0e-DsjHC4ok58yOX7qxseotym3uZpM,31276
@@ -384,7 +384,7 @@ keras_hub/src/models/roberta/roberta_text_classifier.py,sha256=x36hU84P-ROReZniU
384
384
  keras_hub/src/models/roberta/roberta_text_classifier_preprocessor.py,sha256=gAJa8JdPUmT1N7nxBqtaIbnfXV-xlNjTtkEevQhfjNU,5993
385
385
  keras_hub/src/models/roberta/roberta_tokenizer.py,sha256=VKPrgXVT9aMKP7et2DIWKlTN8g4tIzjya0MHqNz9BwQ,2712
386
386
  keras_hub/src/models/roformer_v2/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
387
- keras_hub/src/models/roformer_v2/roformer_v2_attention.py,sha256=3n3XYbL3_Ro1JWBAD7Mh4WoOltQOCV2ebcx2tH6-TVY,7226
387
+ keras_hub/src/models/roformer_v2/roformer_v2_attention.py,sha256=C8wwCw0FpmOWQq8H1IeTn25-6_EzDRD-8UXAN77-5gk,7060
388
388
  keras_hub/src/models/roformer_v2/roformer_v2_backbone.py,sha256=a5gG47Gvo-dFoToMe6Q3oOYJz8HypPZWIhY-cGwS9_c,7187
389
389
  keras_hub/src/models/roformer_v2/roformer_v2_encoder.py,sha256=o_M3dDtebBtXRAxwhiRmdWA59tu1_MNKLINf4GQYfeA,4218
390
390
  keras_hub/src/models/roformer_v2/roformer_v2_masked_lm.py,sha256=4uQ6DKFDdBOu0bHaL45bqtpL-CMZw59inXirD9zWFlI,5950
@@ -422,7 +422,7 @@ keras_hub/src/models/siglip/siglip_vision_encoder.py,sha256=CaNaFq5thBC3TUXXOf2q
422
422
  keras_hub/src/models/stable_diffusion_3/__init__.py,sha256=ZKYQuaRObyhKq8GVAHmoRvlXp6FpU8ChvutVCHyXKuc,343
423
423
  keras_hub/src/models/stable_diffusion_3/flow_match_euler_discrete_scheduler.py,sha256=1K_B3d3fNn50eY84OgxVHyIHHZhmlJY03b71pMSmE9s,3246
424
424
  keras_hub/src/models/stable_diffusion_3/mmdit.py,sha256=emyDmtpJiFU_9crSDBC5CaXoZnM1Eti8uAQtwv2v8B0,42794
425
- keras_hub/src/models/stable_diffusion_3/stable_diffusion_3_backbone.py,sha256=odg1nA02dmINSjAfwKnzPt_HC6b7nQfP000swHxIfaI,24055
425
+ keras_hub/src/models/stable_diffusion_3/stable_diffusion_3_backbone.py,sha256=BEtMwYaxrJxHpNT_E1wK-SPCBCp4hgbnX-UjgqGrQ7g,24362
426
426
  keras_hub/src/models/stable_diffusion_3/stable_diffusion_3_image_to_image.py,sha256=uNsNSQ4EFceGfIMzgjYWFMuL0XdfM58rubTcrCVPrts,5532
427
427
  keras_hub/src/models/stable_diffusion_3/stable_diffusion_3_inpaint.py,sha256=2UIRz11DRbHJ7IVbkjpBjtbkZGC3-eYhMtVUWTmWMH8,6437
428
428
  keras_hub/src/models/stable_diffusion_3/stable_diffusion_3_presets.py,sha256=x7Ez4L955MJE4ABtBy-63YpU9XpR0Ro8QWPzYYJs1yE,2167
@@ -532,7 +532,7 @@ keras_hub/src/utils/transformers/convert_distilbert.py,sha256=SlfIRhSRk5c1ir2HGi
532
532
  keras_hub/src/utils/transformers/convert_esm.py,sha256=rOgGnNY37ZbYnoVC3L-Y-yGGAxTRmYtQV0nJoandH2Y,6214
533
533
  keras_hub/src/utils/transformers/convert_gemma.py,sha256=ElCgwBpSN5Q7rV5PJawTsoytPzs5ZjuwoY60YAe8y_A,6533
534
534
  keras_hub/src/utils/transformers/convert_gpt2.py,sha256=HCeHN_-GiQJRxLCM9OCJJ1watPVpIBF8ujS8pGbBOWc,5703
535
- keras_hub/src/utils/transformers/convert_llama3.py,sha256=c5phNl-QayQ_BS0s-lenbu6oHxqfwDShKJoh9DluxUU,6146
535
+ keras_hub/src/utils/transformers/convert_llama3.py,sha256=DjVUyQbl4AV-h8VqSIzmxiCd7cYOKIJTYoLM__NtyY0,6413
536
536
  keras_hub/src/utils/transformers/convert_mistral.py,sha256=kVhN9h1ZFVhwkNW8p3wnS7eANJUXIsNy1RxWXy20Gqw,4760
537
537
  keras_hub/src/utils/transformers/convert_mixtral.py,sha256=PxeCY8Xe7U_caICugwOCEjuSZ51ZUtmef6rUxh-Wt54,5508
538
538
  keras_hub/src/utils/transformers/convert_pali_gemma.py,sha256=B1leeDw96Yvu81hYumf66hIid07k5NLqoeWAJgPnaLs,10649
@@ -546,7 +546,7 @@ keras_hub/src/utils/transformers/export/gemma.py,sha256=NpTSgRUSWp3WXQil1CjYUVFV
546
546
  keras_hub/src/utils/transformers/export/hf_exporter.py,sha256=oTdRS8SalPCbi_cZPC55aZUBc-1_pdviUIp0XysA4cI,3234
547
547
  keras_hub/tokenizers/__init__.py,sha256=gQIESc4erRLuwxHyxtYy_Z0ePQXw_uhXAa4GVHMffYk,4244
548
548
  keras_hub/utils/__init__.py,sha256=jXPqVGBpJr_PpYmqD8aDG-fRMlxH-ulqCR2SZMn288Y,646
549
- keras_hub_nightly-0.22.0.dev202508110431.dist-info/METADATA,sha256=SZuAT_7BaOIA770Dvi8XRNjsZpz-fWCy3SEEIylMKNs,7393
550
- keras_hub_nightly-0.22.0.dev202508110431.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
551
- keras_hub_nightly-0.22.0.dev202508110431.dist-info/top_level.txt,sha256=N4J6piIWBKa38A4uV-CnIopnOEf8mHAbkNXafXm_CuA,10
552
- keras_hub_nightly-0.22.0.dev202508110431.dist-info/RECORD,,
549
+ keras_hub_nightly-0.22.0.dev202508140419.dist-info/METADATA,sha256=fWp2bszOsVXkWuBituiIzhaEi47PdPePBMvy0ufhYP8,7395
550
+ keras_hub_nightly-0.22.0.dev202508140419.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
551
+ keras_hub_nightly-0.22.0.dev202508140419.dist-info/top_level.txt,sha256=N4J6piIWBKa38A4uV-CnIopnOEf8mHAbkNXafXm_CuA,10
552
+ keras_hub_nightly-0.22.0.dev202508140419.dist-info/RECORD,,