tf-models-nightly 2.19.0.dev20241105__py2.py3-none-any.whl → 2.19.0.dev20241107__py2.py3-none-any.whl

official/modeling/multitask/configs.py

@@ -25,7 +25,7 @@ from official.modeling.privacy import configs as dp_configs
 class TaskRoutine(hyperparams.Config):
   # TODO(hongkuny): deprecate the task_name once we migrated client code.
   task_name: str = ""
-  task_config: cfg.TaskConfig = None
+  task_config: cfg.TaskConfig | None = None
   eval_steps: Optional[int] = None
   task_weight: Optional[float] = 1.0
 
@@ -33,7 +33,7 @@ class TaskRoutine(hyperparams.Config):
 @dataclasses.dataclass
 class MultiTaskConfig(hyperparams.Config):
   init_checkpoint: str = ""
-  model: hyperparams.Config = None
+  model: hyperparams.Config | None = None
   task_routines: Tuple[TaskRoutine, ...] = ()
   # Configs for differential privacy
   # These configs are only effective if you use create_optimizer in

official/nlp/modeling/layers/rezero_transformer.py

@@ -82,6 +82,10 @@ class ReZeroTransformer(tf_keras.layers.Layer):
                num_kv_heads=None,
                src_block_size=None,
                tgt_block_size=None,
+               linformer_dim=None,
+               linformer_shared_kv_projection=True,
+               use_sigmoid_attn=False,
+               sigmoid_attn_bias=None,
                **kwargs):
     # attention_dropout will override attention_dropout_rate.
     # This is to unify the input params with TransformerEncoderBlock.
@@ -115,6 +119,15 @@ class ReZeroTransformer(tf_keras.layers.Layer):
     self._num_kv_heads = num_kv_heads
     self._src_block_size = src_block_size
     self._tgt_block_size = tgt_block_size
+    self._linformer_dim = linformer_dim
+    self._linformer_shared_kv_projection = linformer_shared_kv_projection
+    self._use_sigmoid_attn = use_sigmoid_attn
+    self._sigmoid_attn_bias = sigmoid_attn_bias
+    if self._linformer_dim is not None or self._use_sigmoid_attn:
+      raise ValueError(
+          "Linformer and Sigmoid attention are not supported in ReZero"
+          " Transformer."
+      )
     if self._num_kv_heads is not None and self._src_block_size is not None:
       raise ValueError(
           "Block sparse attention does not support Multi-query attention."
@@ -284,6 +297,12 @@ class ReZeroTransformer(tf_keras.layers.Layer):
             tf_keras.constraints.serialize(self._kernel_constraint),
         "bias_constraint":
             tf_keras.constraints.serialize(self._bias_constraint),
+        "linformer_dim": self._linformer_dim,
+        "linformer_shared_kv_projection": (
+            self._linformer_shared_kv_projection
+        ),
+        "use_sigmoid_attn": self._use_sigmoid_attn,
+        "sigmoid_attn_bias": self._sigmoid_attn_bias,
     }
     base_config = super().get_config()
     return dict(list(base_config.items()) + list(config.items()))

{tf_models_nightly-2.19.0.dev20241105.dist-info → tf_models_nightly-2.19.0.dev20241107.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: tf-models-nightly
-Version: 2.19.0.dev20241105
+Version: 2.19.0.dev20241107
 Summary: TensorFlow Official Models
 Home-page: https://github.com/tensorflow/models
 Author: Google Inc.

{tf_models_nightly-2.19.0.dev20241105.dist-info → tf_models_nightly-2.19.0.dev20241107.dist-info}/RECORD

@@ -215,7 +215,7 @@ official/modeling/multitask/__init__.py,sha256=7oiypy0N82PDw9aSdcJBLVoGTd_oRSUOd
 official/modeling/multitask/base_model.py,sha256=QI8qb8ipj75IUj6bKNjcAFHPjeqmNjqHr7nUbPd6a-o,1946
 official/modeling/multitask/base_trainer.py,sha256=83cLDajiyS2lJPMhllTdIsKXqiVTFLDaGZaherTPCa8,5858
 official/modeling/multitask/base_trainer_test.py,sha256=qJ7z4kid2XAX6hOIvUHa7dwqxouemMekS9ZXhPjWW9w,3663
-official/modeling/multitask/configs.py,sha256=ZO2waQrMn9CAgyFpsmeQvplCF5VeXz7tCPmIuy5jvlc,3164
+official/modeling/multitask/configs.py,sha256=LfgGxw9lHHbtaQjnYRTeBDApY-29w5g4CfJZgm5zezU,3178
 official/modeling/multitask/evaluator.py,sha256=spDm2X8EX62qsxI2ehVjrkIKoo-omQQOYcAVKZNgxHc,6078
 official/modeling/multitask/evaluator_test.py,sha256=vU-q-gM7GqiMqE5zbBnOT8mPFhQmHjniMyNnwganhso,4643
 official/modeling/multitask/interleaving_trainer.py,sha256=f111ZhknyS34hpP0FfdWjX3_iiLViHfBd0VSuC715s0,4635
@@ -347,7 +347,7 @@ official/nlp/modeling/layers/reuse_attention.py,sha256=qvAC-Dr2uPbpQWOvaf0RVN7t6
 official/nlp/modeling/layers/reuse_attention_test.py,sha256=rKr-dl05DqQesYdvYaCYYahIp0ObP4Xgi4Lno4jsl3Y,14329
 official/nlp/modeling/layers/reuse_transformer.py,sha256=S0IxI8LzjAnZ5L3MDy32oanI3oGQhxQjdkC3ff-zlmc,15697
 official/nlp/modeling/layers/reuse_transformer_test.py,sha256=GXuJWfNrqsOwxAi0xSyuziD3kreVWGPCr0LHmxxe0Mk,17201
-official/nlp/modeling/layers/rezero_transformer.py,sha256=SE6iDIlguTxEBdKR79XGOZaqHJ79UY5VZuolgQXQz8g,14249
+official/nlp/modeling/layers/rezero_transformer.py,sha256=hS6LHmIJCZxdRpdeJ__c0jT3mqO9u4Kl5kvj2pQvtqg,15074
 official/nlp/modeling/layers/rezero_transformer_test.py,sha256=-Ib-PpZvWgP5aiH3EhyY69OxXLldxCMsnT8oAhojXzI,8620
 official/nlp/modeling/layers/routing.py,sha256=hV6RHVBU0lEgwx180Q78mDncuXTcyWPyaVBqatWCtQw,4469
 official/nlp/modeling/layers/routing_test.py,sha256=ViRCnFWPdwM4Kam0k8aDZbyoJqeqyIiQKEmlhNE7LgI,2226
@@ -1222,9 +1222,9 @@ tensorflow_models/tensorflow_models_test.py,sha256=nc6A9K53OGqF25xN5St8EiWvdVbda
 tensorflow_models/nlp/__init__.py,sha256=4tA5Pf4qaFwT-fIFOpX7x7FHJpnyJT-5UgOeFYTyMlc,807
 tensorflow_models/uplift/__init__.py,sha256=mqfa55gweOdpKoaQyid4A_4u7xw__FcQeSIF0k_pYmI,999
 tensorflow_models/vision/__init__.py,sha256=zBorY_v5xva1uI-qxhZO3Qh-Dii-Suq6wEYh6hKHDfc,833
-tf_models_nightly-2.19.0.dev20241105.dist-info/AUTHORS,sha256=1dG3fXVu9jlo7bul8xuix5F5vOnczMk7_yWn4y70uw0,337
-tf_models_nightly-2.19.0.dev20241105.dist-info/LICENSE,sha256=WxeBS_DejPZQabxtfMOM_xn8qoZNJDQjrT7z2wG1I4U,11512
-tf_models_nightly-2.19.0.dev20241105.dist-info/METADATA,sha256=eYlauAwsK97PC2gPtfD9IGgV1qA8O2KXUsotBaQUiAI,1432
-tf_models_nightly-2.19.0.dev20241105.dist-info/WHEEL,sha256=kGT74LWyRUZrL4VgLh6_g12IeVl_9u9ZVhadrgXZUEY,110
-tf_models_nightly-2.19.0.dev20241105.dist-info/top_level.txt,sha256=gum2FfO5R4cvjl2-QtP-S1aNmsvIZaFFT6VFzU0f4-g,33
-tf_models_nightly-2.19.0.dev20241105.dist-info/RECORD,,
+tf_models_nightly-2.19.0.dev20241107.dist-info/AUTHORS,sha256=1dG3fXVu9jlo7bul8xuix5F5vOnczMk7_yWn4y70uw0,337
+tf_models_nightly-2.19.0.dev20241107.dist-info/LICENSE,sha256=WxeBS_DejPZQabxtfMOM_xn8qoZNJDQjrT7z2wG1I4U,11512
+tf_models_nightly-2.19.0.dev20241107.dist-info/METADATA,sha256=nmdMccnfdKHt9Pdvqu5_yFkPIhz2duSMK6a3CMgQs9A,1432
+tf_models_nightly-2.19.0.dev20241107.dist-info/WHEEL,sha256=kGT74LWyRUZrL4VgLh6_g12IeVl_9u9ZVhadrgXZUEY,110
+tf_models_nightly-2.19.0.dev20241107.dist-info/top_level.txt,sha256=gum2FfO5R4cvjl2-QtP-S1aNmsvIZaFFT6VFzU0f4-g,33
+tf_models_nightly-2.19.0.dev20241107.dist-info/RECORD,,