code
stringlengths 81
54k
| code_codestyle
int64 0
721
| style_context
stringlengths 91
41.9k
| style_context_codestyle
int64 0
699
| label
int64 0
1
|
|---|---|---|---|---|
"""simple docstring"""
import gc
import unittest
import numpy as np
import torch
from diffusers import AutoencoderKL, DDIMScheduler, DiTPipeline, DPMSolverMultistepScheduler, TransformeraDModel
from diffusers.utils import is_xformers_available, load_numpy, slow, torch_device
from diffusers.utils.testing_utils import enable_full_determinism, require_torch_gpu
from ..pipeline_params import (
CLASS_CONDITIONED_IMAGE_GENERATION_BATCH_PARAMS,
CLASS_CONDITIONED_IMAGE_GENERATION_PARAMS,
)
from ..test_pipelines_common import PipelineTesterMixin
enable_full_determinism()
class A__ ( _snake_case , unittest.TestCase):
"""simple docstring"""
snake_case__ : Tuple =DiTPipeline
snake_case__ : Optional[int] =CLASS_CONDITIONED_IMAGE_GENERATION_PARAMS
snake_case__ : Optional[Any] =PipelineTesterMixin.required_optional_params - {
'''latents''',
'''num_images_per_prompt''',
'''callback''',
'''callback_steps''',
}
snake_case__ : Tuple =CLASS_CONDITIONED_IMAGE_GENERATION_BATCH_PARAMS
snake_case__ : Optional[int] =False
def a__ ( self: List[Any] )-> Any:
torch.manual_seed(0 )
lowerCamelCase : int = TransformeraDModel(
sample_size=16 , num_layers=2 , patch_size=4 , attention_head_dim=8 , num_attention_heads=2 , in_channels=4 , out_channels=8 , attention_bias=lowerCAmelCase__ , activation_fn="""gelu-approximate""" , num_embeds_ada_norm=1_000 , norm_type="""ada_norm_zero""" , norm_elementwise_affine=lowerCAmelCase__ , )
lowerCamelCase : List[Any] = AutoencoderKL()
lowerCamelCase : Optional[Any] = DDIMScheduler()
lowerCamelCase : Optional[Any] = {"""transformer""": transformer.eval(), """vae""": vae.eval(), """scheduler""": scheduler}
return components
def a__ ( self: int , __a: Optional[int] , __a: List[Any]=0 )-> Tuple:
if str(lowerCAmelCase__ ).startswith("""mps""" ):
lowerCamelCase : int = torch.manual_seed(lowerCAmelCase__ )
else:
lowerCamelCase : Union[str, Any] = torch.Generator(device=lowerCAmelCase__ ).manual_seed(lowerCAmelCase__ )
lowerCamelCase : Optional[int] = {
"""class_labels""": [1],
"""generator""": generator,
"""num_inference_steps""": 2,
"""output_type""": """numpy""",
}
return inputs
def a__ ( self: Union[str, Any] )-> Any:
lowerCamelCase : str = """cpu"""
lowerCamelCase : Optional[int] = self.get_dummy_components()
lowerCamelCase : str = self.pipeline_class(**lowerCAmelCase__ )
pipe.to(lowerCAmelCase__ )
pipe.set_progress_bar_config(disable=lowerCAmelCase__ )
lowerCamelCase : Dict = self.get_dummy_inputs(lowerCAmelCase__ )
lowerCamelCase : Dict = pipe(**lowerCAmelCase__ ).images
lowerCamelCase : Union[str, Any] = image[0, -3:, -3:, -1]
self.assertEqual(image.shape , (1, 16, 16, 3) )
lowerCamelCase : Optional[Any] = np.array([0.29_46, 0.66_01, 0.43_29, 0.32_96, 0.41_44, 0.53_19, 0.72_73, 0.50_13, 0.44_57] )
lowerCamelCase : Optional[Any] = np.abs(image_slice.flatten() - expected_slice ).max()
self.assertLessEqual(lowerCAmelCase__ , 1e-3 )
def a__ ( self: Optional[int] )-> Union[str, Any]:
self._test_inference_batch_single_identical(relax_max_difference=lowerCAmelCase__ , expected_max_diff=1e-3 )
@unittest.skipIf(
torch_device != """cuda""" or not is_xformers_available() , reason="""XFormers attention is only available with CUDA and `xformers` installed""" , )
def a__ ( self: Any )-> str:
self._test_xformers_attention_forwardGenerator_pass(expected_max_diff=1e-3 )
@require_torch_gpu
@slow
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: str )-> Dict:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: List[str] )-> Optional[Any]:
lowerCamelCase : List[str] = torch.manual_seed(0 )
lowerCamelCase : Dict = DiTPipeline.from_pretrained("""facebook/DiT-XL-2-256""" )
pipe.to("""cuda""" )
lowerCamelCase : Any = ["""vase""", """umbrella""", """white shark""", """white wolf"""]
lowerCamelCase : List[str] = pipe.get_label_ids(lowerCAmelCase__ )
lowerCamelCase : List[Any] = pipe(lowerCAmelCase__ , generator=lowerCAmelCase__ , num_inference_steps=40 , output_type="""np""" ).images
for word, image in zip(lowerCAmelCase__ , lowerCAmelCase__ ):
lowerCamelCase : Optional[Any] = load_numpy(
f'https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/dit/{word}.npy' )
assert np.abs((expected_image - image).max() ) < 1e-2
def a__ ( self: Optional[int] )-> List[Any]:
lowerCamelCase : List[str] = DiTPipeline.from_pretrained("""facebook/DiT-XL-2-512""" )
lowerCamelCase : str = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config )
pipe.to("""cuda""" )
lowerCamelCase : Optional[Any] = ["""vase""", """umbrella"""]
lowerCamelCase : Any = pipe.get_label_ids(lowerCAmelCase__ )
lowerCamelCase : Dict = torch.manual_seed(0 )
lowerCamelCase : str = pipe(lowerCAmelCase__ , generator=lowerCAmelCase__ , num_inference_steps=25 , output_type="""np""" ).images
for word, image in zip(lowerCAmelCase__ , lowerCAmelCase__ ):
lowerCamelCase : Dict = load_numpy(
"""https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main"""
f'/dit/{word}_512.npy' )
assert np.abs((expected_image - image).max() ) < 1e-1
| 701
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_flax_available,
is_tf_available,
is_tokenizers_available,
is_torch_available,
is_vision_available,
)
__lowerCamelCase :List[str] = {
'configuration_owlvit': [
'OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP',
'OwlViTConfig',
'OwlViTOnnxConfig',
'OwlViTTextConfig',
'OwlViTVisionConfig',
],
'processing_owlvit': ['OwlViTProcessor'],
}
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[int] = ['OwlViTFeatureExtractor']
__lowerCamelCase :List[str] = ['OwlViTImageProcessor']
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[Any] = [
'OWLVIT_PRETRAINED_MODEL_ARCHIVE_LIST',
'OwlViTModel',
'OwlViTPreTrainedModel',
'OwlViTTextModel',
'OwlViTVisionModel',
'OwlViTForObjectDetection',
]
if TYPE_CHECKING:
from .configuration_owlvit import (
OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP,
OwlViTConfig,
OwlViTOnnxConfig,
OwlViTTextConfig,
OwlViTVisionConfig,
)
from .processing_owlvit import OwlViTProcessor
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .feature_extraction_owlvit import OwlViTFeatureExtractor
from .image_processing_owlvit import OwlViTImageProcessor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_owlvit import (
OWLVIT_PRETRAINED_MODEL_ARCHIVE_LIST,
OwlViTForObjectDetection,
OwlViTModel,
OwlViTPreTrainedModel,
OwlViTTextModel,
OwlViTVisionModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
import unittest
from transformers import PegasusConfig, PegasusTokenizer, is_flax_available
from transformers.testing_utils import require_flax, slow
from ...test_configuration_common import ConfigTester
from ...test_modeling_flax_common import FlaxModelTesterMixin, ids_tensor
if is_flax_available():
import os
# The slow tests are often failing with OOM error on GPU
# This makes JAX allocate exactly what is needed on demand, and deallocate memory that is no longer needed
# but will be slower as stated here https://jax.readthedocs.io/en/latest/gpu_memory_allocation.html
__lowerCamelCase :List[str] = 'platform'
import jax
import jax.numpy as jnp
import numpy as np
from transformers import FlaxPegasusForConditionalGeneration, FlaxPegasusModel
@require_flax
class A__ :
"""simple docstring"""
snake_case__ : List[str] =PegasusConfig
snake_case__ : Dict ={}
snake_case__ : Optional[Any] ="""gelu"""
def __init__( self: List[Any] , __a: Optional[int] , __a: str=13 , __a: Dict=7 , __a: Dict=True , __a: List[str]=False , __a: List[Any]=99 , __a: Union[str, Any]=32 , __a: List[str]=5 , __a: str=4 , __a: Any=37 , __a: int=0.1 , __a: Optional[Any]=0.1 , __a: Union[str, Any]=20 , __a: Dict=2 , __a: Optional[Any]=1 , __a: Tuple=0 , )-> Optional[int]:
lowerCamelCase : Union[str, Any] = parent
lowerCamelCase : Dict = batch_size
lowerCamelCase : Optional[int] = seq_length
lowerCamelCase : Optional[Any] = is_training
lowerCamelCase : Union[str, Any] = use_labels
lowerCamelCase : Optional[int] = vocab_size
lowerCamelCase : int = hidden_size
lowerCamelCase : str = num_hidden_layers
lowerCamelCase : Optional[Any] = num_attention_heads
lowerCamelCase : Union[str, Any] = intermediate_size
lowerCamelCase : Tuple = hidden_dropout_prob
lowerCamelCase : List[Any] = attention_probs_dropout_prob
lowerCamelCase : List[str] = max_position_embeddings
lowerCamelCase : int = eos_token_id
lowerCamelCase : Any = pad_token_id
lowerCamelCase : int = bos_token_id
def a__ ( self: int )-> Any:
lowerCamelCase : Tuple = ids_tensor([self.batch_size, self.seq_length - 1] , self.vocab_size ).clip(3 , self.vocab_size )
lowerCamelCase : int = np.expand_dims(np.array([self.eos_token_id] * self.batch_size ) , 1 )
lowerCamelCase : List[Any] = np.concatenate([input_ids, eos_tensor] , axis=1 )
lowerCamelCase : Tuple = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : Optional[int] = self.config_cls(
vocab_size=self.vocab_size , d_model=self.hidden_size , encoder_layers=self.num_hidden_layers , decoder_layers=self.num_hidden_layers , encoder_attention_heads=self.num_attention_heads , decoder_attention_heads=self.num_attention_heads , encoder_ffn_dim=self.intermediate_size , decoder_ffn_dim=self.intermediate_size , dropout=self.hidden_dropout_prob , attention_dropout=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , eos_token_ids=[2] , bos_token_id=self.bos_token_id , pad_token_id=self.pad_token_id , decoder_start_token_id=self.pad_token_id , **self.config_updates , )
lowerCamelCase : Tuple = prepare_pegasus_inputs_dict(UpperCAmelCase_ , UpperCAmelCase_ , UpperCAmelCase_ )
return config, inputs_dict
def a__ ( self: Optional[int] , __a: Any , __a: int , __a: str )-> str:
lowerCamelCase : Optional[int] = 20
lowerCamelCase : Union[str, Any] = model_class_name(UpperCAmelCase_ )
lowerCamelCase : str = model.encode(inputs_dict["""input_ids"""] )
lowerCamelCase , lowerCamelCase : Optional[Any] = (
inputs_dict["""decoder_input_ids"""],
inputs_dict["""decoder_attention_mask"""],
)
lowerCamelCase : int = model.init_cache(decoder_input_ids.shape[0] , UpperCAmelCase_ , UpperCAmelCase_ )
lowerCamelCase : int = jnp.ones((decoder_input_ids.shape[0], max_decoder_length) , dtype="""i4""" )
lowerCamelCase : Dict = jnp.broadcast_to(
jnp.arange(decoder_input_ids.shape[-1] - 1 )[None, :] , (decoder_input_ids.shape[0], decoder_input_ids.shape[-1] - 1) , )
lowerCamelCase : Optional[Any] = model.decode(
decoder_input_ids[:, :-1] , UpperCAmelCase_ , decoder_attention_mask=UpperCAmelCase_ , past_key_values=UpperCAmelCase_ , decoder_position_ids=UpperCAmelCase_ , )
lowerCamelCase : Dict = jnp.array(decoder_input_ids.shape[0] * [[decoder_input_ids.shape[-1] - 1]] , dtype="""i4""" )
lowerCamelCase : str = model.decode(
decoder_input_ids[:, -1:] , UpperCAmelCase_ , decoder_attention_mask=UpperCAmelCase_ , past_key_values=outputs_cache.past_key_values , decoder_position_ids=UpperCAmelCase_ , )
lowerCamelCase : Tuple = model.decode(UpperCAmelCase_ , UpperCAmelCase_ )
lowerCamelCase : Optional[Any] = np.max(np.abs((outputs_cache_next[0][:, -1, :5] - outputs[0][:, -1, :5]) ) )
self.parent.assertTrue(diff < 1e-3 , msg=f'Max diff is {diff}' )
def a__ ( self: List[str] , __a: int , __a: Tuple , __a: List[Any] )-> Dict:
lowerCamelCase : str = 20
lowerCamelCase : Union[str, Any] = model_class_name(UpperCAmelCase_ )
lowerCamelCase : Optional[int] = model.encode(inputs_dict["""input_ids"""] )
lowerCamelCase , lowerCamelCase : Any = (
inputs_dict["""decoder_input_ids"""],
inputs_dict["""decoder_attention_mask"""],
)
lowerCamelCase : Union[str, Any] = jnp.concatenate(
[
decoder_attention_mask,
jnp.zeros((decoder_attention_mask.shape[0], max_decoder_length - decoder_attention_mask.shape[1]) ),
] , axis=-1 , )
lowerCamelCase : Any = model.init_cache(decoder_input_ids.shape[0] , UpperCAmelCase_ , UpperCAmelCase_ )
lowerCamelCase : Dict = jnp.broadcast_to(
jnp.arange(decoder_input_ids.shape[-1] - 1 )[None, :] , (decoder_input_ids.shape[0], decoder_input_ids.shape[-1] - 1) , )
lowerCamelCase : Optional[Any] = model.decode(
decoder_input_ids[:, :-1] , UpperCAmelCase_ , decoder_attention_mask=UpperCAmelCase_ , past_key_values=UpperCAmelCase_ , decoder_position_ids=UpperCAmelCase_ , )
lowerCamelCase : List[Any] = jnp.array(decoder_input_ids.shape[0] * [[decoder_input_ids.shape[-1] - 1]] , dtype="""i4""" )
lowerCamelCase : Optional[Any] = model.decode(
decoder_input_ids[:, -1:] , UpperCAmelCase_ , past_key_values=outputs_cache.past_key_values , decoder_attention_mask=UpperCAmelCase_ , decoder_position_ids=UpperCAmelCase_ , )
lowerCamelCase : Union[str, Any] = model.decode(UpperCAmelCase_ , UpperCAmelCase_ , decoder_attention_mask=UpperCAmelCase_ )
lowerCamelCase : Tuple = np.max(np.abs((outputs_cache_next[0][:, -1, :5] - outputs[0][:, -1, :5]) ) )
self.parent.assertTrue(diff < 1e-3 , msg=f'Max diff is {diff}' )
def snake_case ( UpperCamelCase__ : Optional[int] , UpperCamelCase__ : List[str] , UpperCamelCase__ : List[Any] , UpperCamelCase__ : Any=None , UpperCamelCase__ : Optional[int]=None , ) -> Dict:
if attention_mask is None:
lowerCamelCase : int = np.not_equal(_SCREAMING_SNAKE_CASE , config.pad_token_id ).astype(np.inta )
if decoder_attention_mask is None:
lowerCamelCase : Optional[int] = np.concatenate(
[
np.ones(decoder_input_ids[:, :1].shape , dtype=np.inta ),
np.not_equal(decoder_input_ids[:, 1:] , config.pad_token_id ).astype(np.inta ),
] , axis=-1 , )
return {
"input_ids": input_ids,
"decoder_input_ids": decoder_input_ids,
"attention_mask": attention_mask,
"decoder_attention_mask": decoder_attention_mask,
}
@require_flax
class A__ ( snake_case__ , unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[int] =(
(
FlaxPegasusForConditionalGeneration,
FlaxPegasusModel,
)
if is_flax_available()
else ()
)
snake_case__ : Union[str, Any] =(FlaxPegasusForConditionalGeneration,) if is_flax_available() else ()
snake_case__ : List[str] =True
snake_case__ : str =False
snake_case__ : Optional[int] =False
snake_case__ : List[str] =False
def a__ ( self: Tuple )-> int:
lowerCamelCase : List[Any] = FlaxPegasusModelTester(self )
lowerCamelCase : str = ConfigTester(self , config_class=UpperCAmelCase_ )
def a__ ( self: Union[str, Any] )-> Optional[Any]:
self.config_tester.run_common_tests()
def a__ ( self: Tuple )-> Dict:
lowerCamelCase , lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
self.model_tester.check_use_cache_forward(UpperCAmelCase_ , UpperCAmelCase_ , UpperCAmelCase_ )
def a__ ( self: Optional[int] )-> List[Any]:
lowerCamelCase , lowerCamelCase : int = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
self.model_tester.check_use_cache_forward_with_attn_mask(UpperCAmelCase_ , UpperCAmelCase_ , UpperCAmelCase_ )
def a__ ( self: Any )-> List[Any]:
lowerCamelCase , lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
with self.subTest(model_class.__name__ ):
lowerCamelCase : Dict = self._prepare_for_class(UpperCAmelCase_ , UpperCAmelCase_ )
lowerCamelCase : Union[str, Any] = model_class(UpperCAmelCase_ )
@jax.jit
def encode_jitted(__a: Optional[int] , __a: Optional[Any]=None , **__a: Optional[Any] ):
return model.encode(input_ids=UpperCAmelCase_ , attention_mask=UpperCAmelCase_ )
with self.subTest("""JIT Enabled""" ):
lowerCamelCase : Dict = encode_jitted(**UpperCAmelCase_ ).to_tuple()
with self.subTest("""JIT Disabled""" ):
with jax.disable_jit():
lowerCamelCase : Tuple = encode_jitted(**UpperCAmelCase_ ).to_tuple()
self.assertEqual(len(UpperCAmelCase_ ) , len(UpperCAmelCase_ ) )
for jitted_output, output in zip(UpperCAmelCase_ , UpperCAmelCase_ ):
self.assertEqual(jitted_output.shape , output.shape )
def a__ ( self: List[Any] )-> Any:
lowerCamelCase , lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
with self.subTest(model_class.__name__ ):
lowerCamelCase : Union[str, Any] = model_class(UpperCAmelCase_ )
lowerCamelCase : Dict = model.encode(inputs_dict["""input_ids"""] , inputs_dict["""attention_mask"""] )
lowerCamelCase : List[Any] = {
"""decoder_input_ids""": inputs_dict["""decoder_input_ids"""],
"""decoder_attention_mask""": inputs_dict["""decoder_attention_mask"""],
"""encoder_outputs""": encoder_outputs,
}
@jax.jit
def decode_jitted(__a: Optional[int] , __a: Dict , __a: List[Any] ):
return model.decode(
decoder_input_ids=UpperCAmelCase_ , decoder_attention_mask=UpperCAmelCase_ , encoder_outputs=UpperCAmelCase_ , )
with self.subTest("""JIT Enabled""" ):
lowerCamelCase : Optional[int] = decode_jitted(**UpperCAmelCase_ ).to_tuple()
with self.subTest("""JIT Disabled""" ):
with jax.disable_jit():
lowerCamelCase : Optional[Any] = decode_jitted(**UpperCAmelCase_ ).to_tuple()
self.assertEqual(len(UpperCAmelCase_ ) , len(UpperCAmelCase_ ) )
for jitted_output, output in zip(UpperCAmelCase_ , UpperCAmelCase_ ):
self.assertEqual(jitted_output.shape , output.shape )
@slow
def a__ ( self: Any )-> List[str]:
for model_class_name in self.all_model_classes:
lowerCamelCase : int = model_class_name.from_pretrained("""google/pegasus-large""" , from_pt=UpperCAmelCase_ )
lowerCamelCase : List[Any] = np.ones((1, 1) )
lowerCamelCase : Any = model(UpperCAmelCase_ )
self.assertIsNotNone(UpperCAmelCase_ )
@slow
def a__ ( self: Dict )-> Tuple:
lowerCamelCase : Optional[Any] = FlaxPegasusForConditionalGeneration.from_pretrained("""google/pegasus-xsum""" )
lowerCamelCase : Optional[int] = PegasusTokenizer.from_pretrained("""google/pegasus-xsum""" )
lowerCamelCase : str = [
""" PG&E stated it scheduled the blackouts in response to forecasts for high winds amid dry conditions. The aim is to reduce the risk of wildfires. Nearly 800 thousand customers were scheduled to be affected by the shutoffs which were expected to last through at least midday tomorrow.""",
""" The London trio are up for best UK act and best album, as well as getting two nominations in the best song category.\"We got told like this morning 'Oh I think you're nominated'\", said Dappy.\"And I was like 'Oh yeah, which one?' And now we've got nominated for four awards. I mean, wow!\"Bandmate Fazer added: \"We thought it's best of us to come down and mingle with everyone and say hello to the cameras. And now we find we've got four nominations.\"The band have two shots at the best song prize, getting the nod for their Tynchy Stryder collaboration Number One, and single Strong Again.Their album Uncle B will also go up against records by the likes of Beyonce and Kanye West.N-Dubz picked up the best newcomer Mobo in 2007, but female member Tulisa said they wouldn't be too disappointed if they didn't win this time around.\"At the end of the day we're grateful to be where we are in our careers.\"If it don't happen then it don't happen - live to fight another day and keep on making albums and hits for the fans.\"Dappy also revealed they could be performing live several times on the night.The group will be doing Number One and also a possible rendition of the War Child single, I Got Soul.The charity song is a re-working of The Killers' All These Things That I've Done and is set to feature artists like Chipmunk, Ironik and Pixie Lott.This year's Mobos will be held outside of London for the first time, in Glasgow on 30 September.N-Dubz said they were looking forward to performing for their Scottish fans and boasted about their recent shows north of the border.\"We just done Edinburgh the other day,\" said Dappy.\"We smashed up an N-Dubz show over there. We done Aberdeen about three or four months ago - we smashed up that show over there! Everywhere we go we smash it up!\" """,
]
lowerCamelCase : Optional[int] = [
"""California's largest electricity provider has turned off power to hundreds of thousands of customers.""",
"""Pop group N-Dubz have revealed they were surprised to get four nominations for this year's Mobo Awards.""",
]
lowerCamelCase : str = tokenizer(UpperCAmelCase_ , return_tensors="""np""" , truncation=UpperCAmelCase_ , max_length=512 , padding=UpperCAmelCase_ )
lowerCamelCase : Union[str, Any] = model.generate(**UpperCAmelCase_ , num_beams=2 ).sequences
lowerCamelCase : str = tokenizer.batch_decode(UpperCAmelCase_ , skip_special_tokens=UpperCAmelCase_ )
assert tgt_text == decoded
| 702
|
"""simple docstring"""
import collections
import inspect
import unittest
from transformers import FocalNetConfig
from transformers.testing_utils import require_torch, require_vision, slow, torch_device
from transformers.utils import cached_property, is_torch_available, is_vision_available
from ...test_backbone_common import BackboneTesterMixin
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, _config_zero_init, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from torch import nn
from transformers import (
FocalNetBackbone,
FocalNetForImageClassification,
FocalNetForMaskedImageModeling,
FocalNetModel,
)
from transformers.models.focalnet.modeling_focalnet import FOCALNET_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import AutoImageProcessor
class A__ :
"""simple docstring"""
def __init__( self: List[Any] , __a: List[str] , __a: Optional[int]=13 , __a: List[str]=32 , __a: int=2 , __a: List[str]=3 , __a: Union[str, Any]=16 , __a: int=[32, 64, 128] , __a: Optional[Any]=[1, 2, 1] , __a: Optional[int]=[2, 2, 4] , __a: Tuple=2 , __a: Dict=2.0 , __a: List[str]=True , __a: Optional[Any]=0.0 , __a: Any=0.0 , __a: List[Any]=0.1 , __a: List[str]="gelu" , __a: Tuple=False , __a: Union[str, Any]=True , __a: Optional[int]=0.02 , __a: Tuple=1e-5 , __a: int=True , __a: List[Any]=None , __a: Optional[int]=True , __a: Dict=10 , __a: List[str]=8 , __a: Any=["stage1", "stage2"] , __a: Union[str, Any]=[1, 2] , )-> Dict:
lowerCamelCase : Dict = parent
lowerCamelCase : Optional[Any] = batch_size
lowerCamelCase : Union[str, Any] = image_size
lowerCamelCase : Optional[int] = patch_size
lowerCamelCase : Any = num_channels
lowerCamelCase : Any = embed_dim
lowerCamelCase : Dict = hidden_sizes
lowerCamelCase : List[Any] = depths
lowerCamelCase : Tuple = num_heads
lowerCamelCase : List[Any] = window_size
lowerCamelCase : str = mlp_ratio
lowerCamelCase : str = qkv_bias
lowerCamelCase : str = hidden_dropout_prob
lowerCamelCase : Dict = attention_probs_dropout_prob
lowerCamelCase : Tuple = drop_path_rate
lowerCamelCase : Dict = hidden_act
lowerCamelCase : Tuple = use_absolute_embeddings
lowerCamelCase : List[str] = patch_norm
lowerCamelCase : List[str] = layer_norm_eps
lowerCamelCase : str = initializer_range
lowerCamelCase : Tuple = is_training
lowerCamelCase : int = scope
lowerCamelCase : Union[str, Any] = use_labels
lowerCamelCase : List[str] = type_sequence_label_size
lowerCamelCase : str = encoder_stride
lowerCamelCase : List[str] = out_features
lowerCamelCase : Optional[int] = out_indices
def a__ ( self: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : str = None
if self.use_labels:
lowerCamelCase : List[str] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : str = self.get_config()
return config, pixel_values, labels
def a__ ( self: List[Any] )-> Optional[int]:
return FocalNetConfig(
image_size=self.image_size , patch_size=self.patch_size , num_channels=self.num_channels , embed_dim=self.embed_dim , hidden_sizes=self.hidden_sizes , depths=self.depths , num_heads=self.num_heads , window_size=self.window_size , mlp_ratio=self.mlp_ratio , qkv_bias=self.qkv_bias , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , drop_path_rate=self.drop_path_rate , hidden_act=self.hidden_act , use_absolute_embeddings=self.use_absolute_embeddings , path_norm=self.patch_norm , layer_norm_eps=self.layer_norm_eps , initializer_range=self.initializer_range , encoder_stride=self.encoder_stride , out_features=self.out_features , out_indices=self.out_indices , )
def a__ ( self: Tuple , __a: Optional[int] , __a: Optional[int] , __a: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = FocalNetModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Tuple = model(__a )
lowerCamelCase : Any = ((config.image_size // config.patch_size) ** 2) // (4 ** (len(config.depths ) - 1))
lowerCamelCase : List[Any] = int(config.embed_dim * 2 ** (len(config.depths ) - 1) )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, expected_seq_len, expected_dim) )
def a__ ( self: Optional[int] , __a: Dict , __a: Tuple , __a: List[Any] )-> int:
lowerCamelCase : List[Any] = FocalNetBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Optional[Any] = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , len(config.out_features ) )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.image_size, 8, 8] )
# verify channels
self.parent.assertEqual(len(model.channels ) , len(config.out_features ) )
self.parent.assertListEqual(model.channels , config.hidden_sizes[:-1] )
# verify backbone works with out_features=None
lowerCamelCase : Dict = None
lowerCamelCase : Dict = FocalNetBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , 1 )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.image_size * 2, 4, 4] )
# verify channels
self.parent.assertEqual(len(model.channels ) , 1 )
self.parent.assertListEqual(model.channels , [config.hidden_sizes[-1]] )
def a__ ( self: Optional[int] , __a: Optional[int] , __a: Optional[int] , __a: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = FocalNetForMaskedImageModeling(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[str] = model(__a )
self.parent.assertEqual(
result.reconstruction.shape , (self.batch_size, self.num_channels, self.image_size, self.image_size) )
# test greyscale images
lowerCamelCase : List[str] = 1
lowerCamelCase : Any = FocalNetForMaskedImageModeling(__a )
model.to(__a )
model.eval()
lowerCamelCase : str = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Tuple = model(__a )
self.parent.assertEqual(result.reconstruction.shape , (self.batch_size, 1, self.image_size, self.image_size) )
def a__ ( self: str , __a: Optional[Any] , __a: Optional[Any] , __a: Tuple )-> str:
lowerCamelCase : Optional[Any] = self.type_sequence_label_size
lowerCamelCase : Optional[Any] = FocalNetForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : List[str] = model(__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
# test greyscale images
lowerCamelCase : int = 1
lowerCamelCase : List[Any] = FocalNetForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Optional[Any] = model(__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
def a__ ( self: int )-> Optional[int]:
lowerCamelCase : str = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = config_and_inputs
lowerCamelCase : List[Any] = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =(
(
FocalNetModel,
FocalNetForImageClassification,
FocalNetForMaskedImageModeling,
FocalNetBackbone,
)
if is_torch_available()
else ()
)
snake_case__ : Optional[int] =(
{'''feature-extraction''': FocalNetModel, '''image-classification''': FocalNetForImageClassification}
if is_torch_available()
else {}
)
snake_case__ : Tuple =False
snake_case__ : Dict =False
snake_case__ : Dict =False
snake_case__ : Tuple =False
snake_case__ : Optional[int] =False
def a__ ( self: Union[str, Any] )-> Optional[int]:
lowerCamelCase : List[str] = FocalNetModelTester(self )
lowerCamelCase : Optional[Any] = ConfigTester(self , config_class=__a , embed_dim=37 , has_text_modality=__a )
def a__ ( self: List[str] )-> List[str]:
self.create_and_test_config_common_properties()
self.config_tester.create_and_test_config_to_json_string()
self.config_tester.create_and_test_config_to_json_file()
self.config_tester.create_and_test_config_from_and_save_pretrained()
self.config_tester.create_and_test_config_with_num_labels()
self.config_tester.check_config_can_be_init_without_params()
self.config_tester.check_config_arguments_init()
def a__ ( self: List[str] )-> Union[str, Any]:
return
def a__ ( self: Tuple )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: List[Any] )-> Dict:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_backbone(*__a )
def a__ ( self: List[Any] )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_masked_image_modeling(*__a )
def a__ ( self: List[str] )-> Dict:
lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_classification(*__a )
@unittest.skip(reason="""FocalNet does not use inputs_embeds""" )
def a__ ( self: Optional[Any] )-> str:
pass
@unittest.skip(reason="""FocalNet does not use feedforward chunking""" )
def a__ ( self: Optional[Any] )-> Dict:
pass
def a__ ( self: Optional[Any] )-> Dict:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : Any = model_class(__a )
self.assertIsInstance(model.get_input_embeddings() , (nn.Module) )
lowerCamelCase : Dict = model.get_output_embeddings()
self.assertTrue(x is None or isinstance(__a , nn.Linear ) )
def a__ ( self: Tuple )-> Optional[int]:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : int = model_class(__a )
lowerCamelCase : int = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Any = [*signature.parameters.keys()]
lowerCamelCase : List[Any] = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: str , __a: Union[str, Any] , __a: int , __a: Tuple , __a: List[str] )-> Union[str, Any]:
lowerCamelCase : List[Any] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : List[str] = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : List[str] = outputs.hidden_states
lowerCamelCase : Tuple = getattr(
self.model_tester , """expected_num_hidden_layers""" , len(self.model_tester.depths ) + 1 )
self.assertEqual(len(__a ) , __a )
# FocalNet has a different seq_length
lowerCamelCase : Tuple = (
config.patch_size
if isinstance(config.patch_size , collections.abc.Iterable )
else (config.patch_size, config.patch_size)
)
lowerCamelCase : Dict = (image_size[1] // patch_size[1]) * (image_size[0] // patch_size[0])
self.assertListEqual(
list(hidden_states[0].shape[-2:] ) , [num_patches, self.model_tester.embed_dim] , )
lowerCamelCase : Optional[Any] = outputs.reshaped_hidden_states
self.assertEqual(len(__a ) , __a )
lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = reshaped_hidden_states[0].shape
lowerCamelCase : Tuple = (
reshaped_hidden_states[0].view(__a , __a , height * width ).permute(0 , 2 , 1 )
)
self.assertListEqual(
list(reshaped_hidden_states.shape[-2:] ) , [num_patches, self.model_tester.embed_dim] , )
def a__ ( self: Any )-> Any:
lowerCamelCase , lowerCamelCase : str = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Union[str, Any] = (
self.model_tester.image_size
if isinstance(self.model_tester.image_size , collections.abc.Iterable )
else (self.model_tester.image_size, self.model_tester.image_size)
)
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : List[str] = True
self.check_hidden_states_output(__a , __a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : List[Any] = True
self.check_hidden_states_output(__a , __a , __a , __a )
def a__ ( self: str )-> Union[str, Any]:
lowerCamelCase , lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : List[str] = 3
lowerCamelCase : Any = (
self.model_tester.image_size
if isinstance(self.model_tester.image_size , collections.abc.Iterable )
else (self.model_tester.image_size, self.model_tester.image_size)
)
lowerCamelCase : Optional[int] = (
config.patch_size
if isinstance(config.patch_size , collections.abc.Iterable )
else (config.patch_size, config.patch_size)
)
lowerCamelCase : Optional[Any] = image_size[0] + patch_size[0] - (image_size[0] % patch_size[0])
lowerCamelCase : List[str] = image_size[1] + patch_size[1] - (image_size[1] % patch_size[1])
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : str = True
self.check_hidden_states_output(__a , __a , __a , (padded_height, padded_width) )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : Union[str, Any] = True
self.check_hidden_states_output(__a , __a , __a , (padded_height, padded_width) )
@slow
def a__ ( self: Optional[int] )-> List[Any]:
for model_name in FOCALNET_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : List[str] = FocalNetModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def a__ ( self: str )-> Any:
lowerCamelCase , lowerCamelCase : Optional[Any] = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : int = _config_zero_init(__a )
for model_class in self.all_model_classes:
lowerCamelCase : int = model_class(config=__a )
for name, param in model.named_parameters():
if "embeddings" not in name and param.requires_grad:
self.assertIn(
((param.data.mean() * 1e9).round() / 1e9).item() , [0.0, 1.0] , msg=f'Parameter {name} of model {model_class} seems not properly initialized' , )
@require_vision
@require_torch
class A__ ( unittest.TestCase):
"""simple docstring"""
@cached_property
def a__ ( self: Optional[int] )-> Optional[Any]:
# TODO update organization
return AutoImageProcessor.from_pretrained("""microsoft/focalnet-tiny""" ) if is_vision_available() else None
@slow
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : Tuple = FocalNetForImageClassification.from_pretrained("""microsoft/focalnet-tiny""" ).to(__a )
lowerCamelCase : Any = self.default_image_processor
lowerCamelCase : Union[str, Any] = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
lowerCamelCase : int = image_processor(images=__a , return_tensors="""pt""" ).to(__a )
# forward pass
with torch.no_grad():
lowerCamelCase : Any = model(**__a )
# verify the logits
lowerCamelCase : Tuple = torch.Size((1, 1_000) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : List[str] = torch.tensor([0.21_66, -0.43_68, 0.21_91] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3] , __a , atol=1e-4 ) )
self.assertTrue(outputs.logits.argmax(dim=-1 ).item() , 281 )
@require_torch
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : str =(FocalNetBackbone,) if is_torch_available() else ()
snake_case__ : Optional[int] =FocalNetConfig
snake_case__ : str =False
def a__ ( self: Union[str, Any] )-> Tuple:
lowerCamelCase : str = FocalNetModelTester(self )
| 42
| 0
|
"""simple docstring"""
import argparse
import os
from pathlib import Path
from typing import Dict
import tensorflow as tf
import torch
from tqdm import tqdm
from transformers import PegasusConfig, PegasusForConditionalGeneration, PegasusTokenizer
from transformers.models.pegasus.configuration_pegasus import DEFAULTS, task_specific_params
__lowerCamelCase :List[str] = [
# replace left string with right string to get the relevant state_dict key (identical state dict to bart)
["memory_attention", "encoder_attn"],
["attention", "attn"],
["/", "."],
[".LayerNorm.gamma", "_layer_norm.weight"],
[".LayerNorm.beta", "_layer_norm.bias"],
["r.layer_", "r.layers."],
["output_proj", "out_proj"],
["ffn.dense_1.", "fc2."],
["ffn.dense.", "fc1."],
["ffn_layer_norm", "final_layer_norm"],
["kernel", "weight"],
["encoder_layer_norm.", "encoder.layer_norm."],
["decoder_layer_norm.", "decoder.layer_norm."],
["embeddings.weights", "shared.weight"],
]
def snake_case ( UpperCamelCase__ : int ) -> List[str]:
'''simple docstring'''
for pegasus_name, hf_name in PATTERNS:
lowerCamelCase : List[Any] = k.replace(UpperCamelCase__ , UpperCamelCase__ )
return k
def snake_case ( UpperCamelCase__ : Tuple , UpperCamelCase__ : List[Any] ) -> PegasusForConditionalGeneration:
'''simple docstring'''
lowerCamelCase : Union[str, Any] = DEFAULTS.copy()
cfg_kwargs.update(UpperCamelCase__ )
lowerCamelCase : List[Any] = PegasusConfig(**UpperCamelCase__ )
lowerCamelCase : Optional[Any] = PegasusForConditionalGeneration(UpperCamelCase__ )
lowerCamelCase : str = torch_model.model.state_dict()
lowerCamelCase : Any = {}
for k, v in tf_weights.items():
lowerCamelCase : Optional[Any] = rename_state_dict_key(UpperCamelCase__ )
if new_k not in sd:
raise ValueError(F'could not find new key {new_k} in state dict. (converted from {k})' )
if "dense" in k or "proj" in new_k:
lowerCamelCase : Dict = v.T
lowerCamelCase : Optional[Any] = torch.tensor(UpperCamelCase__ , dtype=sd[new_k].dtype )
assert v.shape == sd[new_k].shape, F'{new_k}, {k}, {v.shape}, {sd[new_k].shape}'
# make sure embedding.padding_idx is respected
lowerCamelCase : int = torch.zeros_like(mapping["""shared.weight"""][cfg.pad_token_id + 1] )
lowerCamelCase : Tuple = mapping["""shared.weight"""]
lowerCamelCase : List[str] = mapping["""shared.weight"""]
lowerCamelCase : str = {k: torch.zeros_like(UpperCamelCase__ ) for k, v in sd.items() if k.endswith("""bias""" ) and k not in mapping}
mapping.update(**UpperCamelCase__ )
lowerCamelCase , lowerCamelCase : Tuple = torch_model.model.load_state_dict(UpperCamelCase__ , strict=UpperCamelCase__ )
lowerCamelCase : List[str] = [
k for k in missing if k not in ["""encoder.embed_positions.weight""", """decoder.embed_positions.weight"""]
]
assert unexpected_missing == [], F'no matches found for the following torch keys {unexpected_missing}'
assert extra == [], F'no matches found for the following tf keys {extra}'
return torch_model
def snake_case ( UpperCamelCase__ : Optional[int]="./ckpt/aeslc/model.ckpt-32000" ) -> Dict:
'''simple docstring'''
lowerCamelCase : int = tf.train.list_variables(UpperCamelCase__ )
lowerCamelCase : Union[str, Any] = {}
lowerCamelCase : Dict = ["""Adafactor""", """global_step"""]
for name, shape in tqdm(UpperCamelCase__ , desc="""converting tf checkpoint to dict""" ):
lowerCamelCase : Optional[int] = any(pat in name for pat in ignore_name )
if skip_key:
continue
lowerCamelCase : Any = tf.train.load_variable(UpperCamelCase__ , UpperCamelCase__ )
lowerCamelCase : Union[str, Any] = array
return tf_weights
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : str ) -> Union[str, Any]:
'''simple docstring'''
lowerCamelCase : Any = Path(UpperCamelCase__ ).parent.name
lowerCamelCase : List[Any] = task_specific_params[F'summarization_{dataset}']["""max_position_embeddings"""]
lowerCamelCase : List[Any] = PegasusTokenizer.from_pretrained("""sshleifer/pegasus""" , model_max_length=UpperCamelCase__ )
assert tok.model_max_length == desired_max_model_length
tok.save_pretrained(UpperCamelCase__ )
# convert model
lowerCamelCase : Tuple = get_tf_weights_as_numpy(UpperCamelCase__ )
lowerCamelCase : Dict = task_specific_params[F'summarization_{dataset}']
if dataset == "large":
lowerCamelCase : Any = task_specific_params
lowerCamelCase : Optional[int] = convert_pegasus(UpperCamelCase__ , UpperCamelCase__ )
torch_model.save_pretrained(UpperCamelCase__ )
lowerCamelCase : int = torch_model.state_dict()
sd.pop("""model.decoder.embed_positions.weight""" )
sd.pop("""model.encoder.embed_positions.weight""" )
torch.save(UpperCamelCase__ , Path(UpperCamelCase__ ) / """pytorch_model.bin""" )
if __name__ == "__main__":
__lowerCamelCase :str = argparse.ArgumentParser()
# Required parameters
parser.add_argument('tf_ckpt_path', type=str, help='passed to tf.train.list_variables')
parser.add_argument('save_dir', default=None, type=str, help='Path to the output PyTorch model.')
__lowerCamelCase :Tuple = parser.parse_args()
if args.save_dir is None:
__lowerCamelCase :Optional[int] = Path(args.tf_ckpt_path).parent.name
__lowerCamelCase :Union[str, Any] = os.path.join('pegasus', dataset)
convert_pegasus_ckpt_to_pytorch(args.tf_ckpt_path, args.save_dir)
| 703
|
"""simple docstring"""
import os
def snake_case ( ) -> Optional[Any]:
with open(os.path.dirname(UpperCamelCase__ ) + """/grid.txt""" ) as f:
lowerCamelCase : int = [] # noqa: E741
for _ in range(20 ):
l.append([int(UpperCamelCase__ ) for x in f.readline().split()] )
lowerCamelCase : Union[str, Any] = 0
# right
for i in range(20 ):
for j in range(17 ):
lowerCamelCase : Dict = l[i][j] * l[i][j + 1] * l[i][j + 2] * l[i][j + 3]
if temp > maximum:
lowerCamelCase : Tuple = temp
# down
for i in range(17 ):
for j in range(20 ):
lowerCamelCase : Any = l[i][j] * l[i + 1][j] * l[i + 2][j] * l[i + 3][j]
if temp > maximum:
lowerCamelCase : Optional[Any] = temp
# diagonal 1
for i in range(17 ):
for j in range(17 ):
lowerCamelCase : List[Any] = l[i][j] * l[i + 1][j + 1] * l[i + 2][j + 2] * l[i + 3][j + 3]
if temp > maximum:
lowerCamelCase : List[str] = temp
# diagonal 2
for i in range(17 ):
for j in range(3 , 20 ):
lowerCamelCase : List[str] = l[i][j] * l[i + 1][j - 1] * l[i + 2][j - 2] * l[i + 3][j - 3]
if temp > maximum:
lowerCamelCase : List[Any] = temp
return maximum
if __name__ == "__main__":
print(solution())
| 42
| 0
|
"""simple docstring"""
from __future__ import annotations
import unittest
from transformers import BlenderbotSmallConfig, BlenderbotSmallTokenizer, is_tf_available
from transformers.testing_utils import require_tf, require_tokenizers, slow
from transformers.utils import cached_property
from ...test_configuration_common import ConfigTester
from ...test_modeling_tf_common import TFModelTesterMixin, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_tf_available():
import tensorflow as tf
from transformers import TFAutoModelForSeqaSeqLM, TFBlenderbotSmallForConditionalGeneration, TFBlenderbotSmallModel
@require_tf
class A__ :
"""simple docstring"""
snake_case__ : Tuple =BlenderbotSmallConfig
snake_case__ : List[Any] ={}
snake_case__ : Union[str, Any] ='''gelu'''
def __init__( self: Optional[Any] , __a: int , __a: Optional[Any]=13 , __a: Union[str, Any]=7 , __a: Any=True , __a: Optional[int]=False , __a: str=99 , __a: Optional[int]=32 , __a: Optional[Any]=2 , __a: str=4 , __a: Union[str, Any]=37 , __a: List[Any]=0.1 , __a: Any=0.1 , __a: Dict=20 , __a: Optional[Any]=2 , __a: Any=1 , __a: Optional[int]=0 , )-> str:
lowerCamelCase : Tuple = parent
lowerCamelCase : Optional[Any] = batch_size
lowerCamelCase : Union[str, Any] = seq_length
lowerCamelCase : Dict = is_training
lowerCamelCase : List[str] = use_labels
lowerCamelCase : int = vocab_size
lowerCamelCase : Any = hidden_size
lowerCamelCase : Union[str, Any] = num_hidden_layers
lowerCamelCase : Optional[int] = num_attention_heads
lowerCamelCase : Any = intermediate_size
lowerCamelCase : str = hidden_dropout_prob
lowerCamelCase : Tuple = attention_probs_dropout_prob
lowerCamelCase : Optional[Any] = max_position_embeddings
lowerCamelCase : str = eos_token_id
lowerCamelCase : List[str] = pad_token_id
lowerCamelCase : str = bos_token_id
def a__ ( self: int )-> List[Any]:
lowerCamelCase : Optional[int] = ids_tensor([self.batch_size, self.seq_length - 1] , self.vocab_size )
lowerCamelCase : List[Any] = tf.expand_dims(tf.constant([self.eos_token_id] * self.batch_size ) , 1 )
lowerCamelCase : Dict = tf.concat([input_ids, eos_tensor] , axis=1 )
lowerCamelCase : str = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : Any = self.config_cls(
vocab_size=self.vocab_size , d_model=self.hidden_size , encoder_layers=self.num_hidden_layers , decoder_layers=self.num_hidden_layers , encoder_attention_heads=self.num_attention_heads , decoder_attention_heads=self.num_attention_heads , encoder_ffn_dim=self.intermediate_size , decoder_ffn_dim=self.intermediate_size , dropout=self.hidden_dropout_prob , attention_dropout=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , eos_token_ids=[2] , bos_token_id=self.bos_token_id , pad_token_id=self.pad_token_id , decoder_start_token_id=self.pad_token_id , **self.config_updates , )
lowerCamelCase : int = prepare_blenderbot_small_inputs_dict(UpperCamelCase_ , UpperCamelCase_ , UpperCamelCase_ )
return config, inputs_dict
def a__ ( self: Tuple , __a: Optional[int] , __a: Optional[Any] )-> List[str]:
lowerCamelCase : List[Any] = TFBlenderbotSmallModel(config=UpperCamelCase_ ).get_decoder()
lowerCamelCase : str = inputs_dict['input_ids']
lowerCamelCase : Union[str, Any] = input_ids[:1, :]
lowerCamelCase : List[str] = inputs_dict['attention_mask'][:1, :]
lowerCamelCase : Optional[Any] = inputs_dict['head_mask']
lowerCamelCase : Optional[int] = 1
# first forward pass
lowerCamelCase : Any = model(UpperCamelCase_ , attention_mask=UpperCamelCase_ , head_mask=UpperCamelCase_ , use_cache=UpperCamelCase_ )
lowerCamelCase : Tuple = outputs.to_tuple()
# create hypothetical next token and extent to next_input_ids
lowerCamelCase : Tuple = ids_tensor((self.batch_size, 3) , config.vocab_size )
lowerCamelCase : List[Any] = tf.cast(ids_tensor((self.batch_size, 3) , 2 ) , tf.inta )
# append to next input_ids and
lowerCamelCase : str = tf.concat([input_ids, next_tokens] , axis=-1 )
lowerCamelCase : Dict = tf.concat([attention_mask, next_attn_mask] , axis=-1 )
lowerCamelCase : Optional[int] = model(UpperCamelCase_ , attention_mask=UpperCamelCase_ )[0]
lowerCamelCase : Optional[int] = model(UpperCamelCase_ , attention_mask=UpperCamelCase_ , past_key_values=UpperCamelCase_ )[0]
self.parent.assertEqual(next_tokens.shape[1] , output_from_past.shape[1] )
# select random slice
lowerCamelCase : Union[str, Any] = int(ids_tensor((1,) , output_from_past.shape[-1] ) )
lowerCamelCase : Union[str, Any] = output_from_no_past[:, -3:, random_slice_idx]
lowerCamelCase : Any = output_from_past[:, :, random_slice_idx]
# test that outputs are equal for slice
tf.debugging.assert_near(UpperCamelCase_ , UpperCamelCase_ , rtol=1e-3 )
def snake_case ( UpperCamelCase__ : List[Any] , UpperCamelCase__ : Any , UpperCamelCase__ : Tuple , UpperCamelCase__ : Dict=None , UpperCamelCase__ : List[Any]=None , UpperCamelCase__ : Tuple=None , UpperCamelCase__ : List[Any]=None , UpperCamelCase__ : str=None , ) -> Tuple:
if attention_mask is None:
lowerCamelCase : Any = tf.cast(tf.math.not_equal(_lowercase , config.pad_token_id ) , tf.inta )
if decoder_attention_mask is None:
lowerCamelCase : Optional[Any] = tf.concat(
[
tf.ones(decoder_input_ids[:, :1].shape , dtype=tf.inta ),
tf.cast(tf.math.not_equal(decoder_input_ids[:, 1:] , config.pad_token_id ) , tf.inta ),
] , axis=-1 , )
if head_mask is None:
lowerCamelCase : str = tf.ones((config.encoder_layers, config.encoder_attention_heads) )
if decoder_head_mask is None:
lowerCamelCase : Optional[int] = tf.ones((config.decoder_layers, config.decoder_attention_heads) )
if cross_attn_head_mask is None:
lowerCamelCase : Dict = tf.ones((config.decoder_layers, config.decoder_attention_heads) )
return {
"input_ids": input_ids,
"decoder_input_ids": decoder_input_ids,
"attention_mask": attention_mask,
"decoder_attention_mask": decoder_attention_mask,
"head_mask": head_mask,
"decoder_head_mask": decoder_head_mask,
"cross_attn_head_mask": cross_attn_head_mask,
}
@require_tf
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[int] =(
(TFBlenderbotSmallForConditionalGeneration, TFBlenderbotSmallModel) if is_tf_available() else ()
)
snake_case__ : List[str] =(TFBlenderbotSmallForConditionalGeneration,) if is_tf_available() else ()
snake_case__ : Dict =(
{
'''conversational''': TFBlenderbotSmallForConditionalGeneration,
'''feature-extraction''': TFBlenderbotSmallModel,
'''summarization''': TFBlenderbotSmallForConditionalGeneration,
'''text2text-generation''': TFBlenderbotSmallForConditionalGeneration,
'''translation''': TFBlenderbotSmallForConditionalGeneration,
}
if is_tf_available()
else {}
)
snake_case__ : int =True
snake_case__ : Optional[Any] =False
snake_case__ : Optional[Any] =False
def a__ ( self: int )-> List[str]:
lowerCamelCase : Union[str, Any] = TFBlenderbotSmallModelTester(self )
lowerCamelCase : Tuple = ConfigTester(self , config_class=UpperCamelCase_ )
def a__ ( self: Optional[int] )-> Dict:
self.config_tester.run_common_tests()
def a__ ( self: Optional[int] )-> Union[str, Any]:
lowerCamelCase : str = self.model_tester.prepare_config_and_inputs_for_common()
self.model_tester.check_decoder_model_past_large_inputs(*UpperCamelCase_ )
@require_tokenizers
@require_tf
class A__ ( unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[int] =[
'''Social anxiety\nWow, I am never shy. Do you have anxiety?\nYes. I end up sweating and blushing and feel like '''
''' i\'m going to throw up.\nand why is that?'''
]
snake_case__ : str ='''facebook/blenderbot_small-90M'''
@cached_property
def a__ ( self: Optional[int] )-> Any:
return BlenderbotSmallTokenizer.from_pretrained("""facebook/blenderbot-90M""" )
@cached_property
def a__ ( self: Union[str, Any] )-> List[str]:
lowerCamelCase : Tuple = TFAutoModelForSeqaSeqLM.from_pretrained(self.model_name )
return model
@slow
def a__ ( self: Union[str, Any] )-> List[Any]:
lowerCamelCase : Optional[Any] = self.tokenizer(self.src_text , return_tensors="""tf""" )
lowerCamelCase : List[Any] = self.model.generate(
model_inputs.input_ids , attention_mask=model_inputs.attention_mask , num_beams=2 , use_cache=UpperCamelCase_ , )
lowerCamelCase : Optional[Any] = self.tokenizer.batch_decode(generated_ids.numpy() , skip_special_tokens=UpperCamelCase_ )[0]
assert generated_words in (
"i don't know. i just feel like i'm going to throw up. it's not fun.",
"i'm not sure. i just feel like i've been feeling like i have to be in a certain place",
"i'm not sure. i just feel like i've been in a bad situation.",
)
| 704
|
"""simple docstring"""
import gc
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DDIMScheduler,
StableDiffusionAttendAndExcitePipeline,
UNetaDConditionModel,
)
from diffusers.utils import load_numpy, skip_mps, slow
from diffusers.utils.testing_utils import require_torch_gpu
from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
from ..test_pipelines_common import PipelineKarrasSchedulerTesterMixin, PipelineLatentTesterMixin, PipelineTesterMixin
__lowerCamelCase :Any = False
@skip_mps
class A__ ( __lowercase , __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[Any] =StableDiffusionAttendAndExcitePipeline
snake_case__ : Any =False
snake_case__ : Dict =TEXT_TO_IMAGE_PARAMS
snake_case__ : Any =TEXT_TO_IMAGE_BATCH_PARAMS.union({'''token_indices'''})
snake_case__ : Dict =TEXT_TO_IMAGE_IMAGE_PARAMS
snake_case__ : str =TEXT_TO_IMAGE_IMAGE_PARAMS
@classmethod
def a__ ( cls: Dict )-> Tuple:
super().setUpClass()
torch.use_deterministic_algorithms(__a )
@classmethod
def a__ ( cls: Union[str, Any] )-> Any:
super().tearDownClass()
torch.use_deterministic_algorithms(__a )
def a__ ( self: Tuple )-> Union[str, Any]:
torch.manual_seed(0 )
lowerCamelCase : str = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=1 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , cross_attention_dim=32 , attention_head_dim=(2, 4) , use_linear_projection=__a , )
lowerCamelCase : Union[str, Any] = DDIMScheduler(
beta_start=0.0_00_85 , beta_end=0.0_12 , beta_schedule="""scaled_linear""" , clip_sample=__a , set_alpha_to_one=__a , )
torch.manual_seed(0 )
lowerCamelCase : Union[str, Any] = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
lowerCamelCase : str = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-0_5 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1_000 , hidden_act="""gelu""" , projection_dim=512 , )
lowerCamelCase : Optional[int] = CLIPTextModel(__a )
lowerCamelCase : str = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" )
lowerCamelCase : List[str] = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""safety_checker""": None,
"""feature_extractor""": None,
}
return components
def a__ ( self: Tuple , __a: int , __a: Union[str, Any]=0 )-> Optional[Any]:
if str(__a ).startswith("""mps""" ):
lowerCamelCase : Tuple = torch.manual_seed(__a )
else:
lowerCamelCase : str = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : Dict = {
"""prompt""": """a cat and a frog""",
"""token_indices""": [2, 5],
"""generator""": generator,
"""num_inference_steps""": 1,
"""guidance_scale""": 6.0,
"""output_type""": """numpy""",
"""max_iter_to_alter""": 2,
"""thresholds""": {0: 0.7},
}
return inputs
def a__ ( self: Dict )-> str:
lowerCamelCase : Tuple = """cpu"""
lowerCamelCase : List[str] = self.get_dummy_components()
lowerCamelCase : List[Any] = self.pipeline_class(**__a )
pipe.to(__a )
pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Any = self.get_dummy_inputs(__a )
lowerCamelCase : Union[str, Any] = pipe(**__a ).images
lowerCamelCase : Tuple = image[0, -3:, -3:, -1]
self.assertEqual(image.shape , (1, 64, 64, 3) )
lowerCamelCase : Optional[Any] = np.array(
[0.63_90_53_64, 0.62_89_73_07, 0.48_59_90_17, 0.5_13_36_24, 0.5_55_00_48, 0.45_76_95_16, 0.50_32_69_73, 0.5_02_31_39, 0.45_38_44_96] )
lowerCamelCase : Optional[Any] = np.abs(image_slice.flatten() - expected_slice ).max()
self.assertLessEqual(__a , 1e-3 )
def a__ ( self: int )-> Optional[Any]:
super().test_cpu_offload_forward_pass(expected_max_diff=5e-4 )
def a__ ( self: Union[str, Any] )-> Optional[int]:
# NOTE: Larger batch sizes cause this test to timeout, only test on smaller batches
self._test_inference_batch_consistent(batch_sizes=[1, 2] )
def a__ ( self: Tuple )-> int:
self._test_inference_batch_single_identical(batch_size=2 , expected_max_diff=7e-4 )
def a__ ( self: Dict )-> List[Any]:
super().test_dict_tuple_outputs_equivalent(expected_max_difference=3e-3 )
def a__ ( self: Optional[int] )-> Dict:
super().test_pt_np_pil_outputs_equivalent(expected_max_diff=5e-4 )
def a__ ( self: Any )-> Tuple:
super().test_save_load_local(expected_max_difference=5e-4 )
def a__ ( self: str )-> str:
super().test_save_load_optional_components(expected_max_difference=4e-4 )
@require_torch_gpu
@slow
class A__ ( unittest.TestCase):
"""simple docstring"""
@classmethod
def a__ ( cls: Any )-> Tuple:
super().setUpClass()
torch.use_deterministic_algorithms(__a )
@classmethod
def a__ ( cls: Dict )-> Optional[int]:
super().tearDownClass()
torch.use_deterministic_algorithms(__a )
def a__ ( self: int )-> Optional[int]:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : List[Any] = torch.manual_seed(51 )
lowerCamelCase : List[str] = StableDiffusionAttendAndExcitePipeline.from_pretrained(
"""CompVis/stable-diffusion-v1-4""" , safety_checker=__a , torch_dtype=torch.floataa )
pipe.to("""cuda""" )
lowerCamelCase : Dict = """a painting of an elephant with glasses"""
lowerCamelCase : Any = [5, 7]
lowerCamelCase : Tuple = pipe(
prompt=__a , token_indices=__a , guidance_scale=7.5 , generator=__a , num_inference_steps=5 , max_iter_to_alter=5 , output_type="""numpy""" , ).images[0]
lowerCamelCase : Union[str, Any] = load_numpy(
"""https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/attend-and-excite/elephant_glasses.npy""" )
assert np.abs((expected_image - image).max() ) < 5e-1
| 42
| 0
|
"""simple docstring"""
import warnings
from ...utils import logging
from .image_processing_videomae import VideoMAEImageProcessor
__lowerCamelCase :str = logging.get_logger(__name__)
class A__ ( lowercase__):
"""simple docstring"""
def __init__( self: int , *__a: Union[str, Any] , **__a: List[Any] )-> List[Any]:
warnings.warn(
"""The class VideoMAEFeatureExtractor is deprecated and will be removed in version 5 of Transformers."""
""" Please use VideoMAEImageProcessor instead.""" , __lowerCamelCase , )
super().__init__(*__lowerCamelCase , **__lowerCamelCase )
| 705
|
"""simple docstring"""
import unittest
from transformers import EsmConfig, is_torch_available
from transformers.testing_utils import TestCasePlus, require_torch, slow, torch_device
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, ids_tensor, random_attention_mask
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import EsmForMaskedLM, EsmForSequenceClassification, EsmForTokenClassification, EsmModel
from transformers.models.esm.modeling_esm import (
ESM_PRETRAINED_MODEL_ARCHIVE_LIST,
EsmEmbeddings,
create_position_ids_from_input_ids,
)
class A__ :
"""simple docstring"""
def __init__( self: List[str] , __a: List[str] , __a: Dict=13 , __a: Tuple=7 , __a: Dict=False , __a: str=True , __a: List[Any]=False , __a: Dict=True , __a: Any=33 , __a: Optional[Any]=32 , __a: List[Any]=5 , __a: Any=4 , __a: Dict=37 , __a: str="gelu" , __a: str=0.1 , __a: int=0.1 , __a: Optional[int]=512 , __a: List[Any]=16 , __a: int=2 , __a: int=0.02 , __a: Optional[int]=3 , __a: str=4 , __a: Tuple=None , )-> Tuple:
lowerCamelCase : Union[str, Any] = parent
lowerCamelCase : Tuple = batch_size
lowerCamelCase : Any = seq_length
lowerCamelCase : Any = is_training
lowerCamelCase : Tuple = use_input_mask
lowerCamelCase : int = use_token_type_ids
lowerCamelCase : List[str] = use_labels
lowerCamelCase : Optional[int] = vocab_size
lowerCamelCase : Tuple = hidden_size
lowerCamelCase : List[str] = num_hidden_layers
lowerCamelCase : Optional[int] = num_attention_heads
lowerCamelCase : Optional[Any] = intermediate_size
lowerCamelCase : Optional[Any] = hidden_act
lowerCamelCase : Union[str, Any] = hidden_dropout_prob
lowerCamelCase : Optional[Any] = attention_probs_dropout_prob
lowerCamelCase : Any = max_position_embeddings
lowerCamelCase : str = type_vocab_size
lowerCamelCase : List[Any] = type_sequence_label_size
lowerCamelCase : Optional[Any] = initializer_range
lowerCamelCase : Union[str, Any] = num_labels
lowerCamelCase : Optional[Any] = num_choices
lowerCamelCase : Any = scope
def a__ ( self: Optional[int] )-> List[Any]:
lowerCamelCase : List[Any] = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : Dict = None
if self.use_input_mask:
lowerCamelCase : Tuple = random_attention_mask([self.batch_size, self.seq_length] )
lowerCamelCase : Any = None
lowerCamelCase : int = None
lowerCamelCase : Union[str, Any] = None
if self.use_labels:
lowerCamelCase : Optional[Any] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : Optional[int] = ids_tensor([self.batch_size, self.seq_length] , self.num_labels )
lowerCamelCase : Union[str, Any] = ids_tensor([self.batch_size] , self.num_choices )
lowerCamelCase : List[str] = self.get_config()
return config, input_ids, input_mask, sequence_labels, token_labels, choice_labels
def a__ ( self: Tuple )-> Union[str, Any]:
return EsmConfig(
vocab_size=self.vocab_size , hidden_size=self.hidden_size , pad_token_id=1 , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , initializer_range=self.initializer_range , )
def a__ ( self: List[Any] , __a: List[str] , __a: str , __a: Tuple , __a: List[str] , __a: List[str] , __a: str )-> int:
lowerCamelCase : Optional[int] = EsmModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a , attention_mask=__a )
lowerCamelCase : str = model(__a )
lowerCamelCase : Optional[Any] = model(__a )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) )
self.parent.assertEqual(result.pooler_output.shape , (self.batch_size, self.hidden_size) )
def a__ ( self: int , __a: Union[str, Any] , __a: Optional[int] , __a: List[str] , __a: str , __a: List[str] , __a: Tuple )-> int:
lowerCamelCase : str = EsmForMaskedLM(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[Any] = model(__a , attention_mask=__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) )
def a__ ( self: List[str] , __a: List[Any] , __a: List[str] , __a: int , __a: Union[str, Any] , __a: List[Any] , __a: Tuple )-> List[str]:
lowerCamelCase : Tuple = self.num_labels
lowerCamelCase : Dict = EsmForTokenClassification(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a , attention_mask=__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) )
def a__ ( self: Optional[int] )-> Optional[int]:
lowerCamelCase : Any = self.prepare_config_and_inputs()
(
(
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) ,
) : Tuple = config_and_inputs
lowerCamelCase : List[Any] = {"""input_ids""": input_ids, """attention_mask""": input_mask}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Any =False
snake_case__ : Dict =(
(
EsmForMaskedLM,
EsmModel,
EsmForSequenceClassification,
EsmForTokenClassification,
)
if is_torch_available()
else ()
)
snake_case__ : Dict =()
snake_case__ : Optional[int] =(
{
'''feature-extraction''': EsmModel,
'''fill-mask''': EsmForMaskedLM,
'''text-classification''': EsmForSequenceClassification,
'''token-classification''': EsmForTokenClassification,
'''zero-shot''': EsmForSequenceClassification,
}
if is_torch_available()
else {}
)
snake_case__ : Any =True
def a__ ( self: Optional[int] )-> Optional[int]:
lowerCamelCase : Optional[Any] = EsmModelTester(self )
lowerCamelCase : Any = ConfigTester(self , config_class=__a , hidden_size=37 )
def a__ ( self: List[Any] )-> Optional[Any]:
self.config_tester.run_common_tests()
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: Tuple )-> Any:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
for type in ["absolute", "relative_key", "relative_key_query"]:
lowerCamelCase : Tuple = type
self.model_tester.create_and_check_model(*__a )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Optional[int] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_masked_lm(*__a )
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_token_classification(*__a )
@slow
def a__ ( self: Any )-> List[Any]:
for model_name in ESM_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : int = EsmModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def a__ ( self: str )-> List[str]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()[0]
lowerCamelCase : Union[str, Any] = EsmEmbeddings(config=__a )
lowerCamelCase : List[str] = torch.as_tensor([[12, 31, 13, model.padding_idx]] )
lowerCamelCase : Union[str, Any] = torch.as_tensor(
[
[
0 + model.padding_idx + 1,
1 + model.padding_idx + 1,
2 + model.padding_idx + 1,
model.padding_idx,
]
] )
lowerCamelCase : Optional[Any] = create_position_ids_from_input_ids(__a , model.padding_idx )
self.assertEqual(position_ids.shape , expected_positions.shape )
self.assertTrue(torch.all(torch.eq(__a , __a ) ) )
def a__ ( self: Optional[int] )-> int:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()[0]
lowerCamelCase : Any = EsmEmbeddings(config=__a )
lowerCamelCase : Dict = torch.empty(2 , 4 , 30 )
lowerCamelCase : List[Any] = [
0 + embeddings.padding_idx + 1,
1 + embeddings.padding_idx + 1,
2 + embeddings.padding_idx + 1,
3 + embeddings.padding_idx + 1,
]
lowerCamelCase : Any = torch.as_tensor([expected_single_positions, expected_single_positions] )
lowerCamelCase : List[str] = embeddings.create_position_ids_from_inputs_embeds(__a )
self.assertEqual(position_ids.shape , expected_positions.shape )
self.assertTrue(torch.all(torch.eq(__a , __a ) ) )
@unittest.skip("""Esm does not support embedding resizing""" )
def a__ ( self: Any )-> Optional[Any]:
pass
@unittest.skip("""Esm does not support embedding resizing""" )
def a__ ( self: Dict )-> Dict:
pass
@unittest.skip("""Will be fixed soon by reducing the size of the model used for common tests.""" )
def a__ ( self: List[str] )-> Dict:
pass
@require_torch
class A__ ( __lowercase):
"""simple docstring"""
@slow
def a__ ( self: Any )-> Union[str, Any]:
with torch.no_grad():
lowerCamelCase : Union[str, Any] = EsmForMaskedLM.from_pretrained("""facebook/esm2_t6_8M_UR50D""" )
model.eval()
lowerCamelCase : List[str] = torch.tensor([[0, 1, 2, 3, 4, 5]] )
lowerCamelCase : Tuple = model(__a )[0]
lowerCamelCase : Dict = 33
lowerCamelCase : List[str] = torch.Size((1, 6, vocab_size) )
self.assertEqual(output.shape , __a )
lowerCamelCase : Tuple = torch.tensor(
[[[8.92_15, -10.58_98, -6.46_71], [-6.39_67, -13.91_14, -1.12_12], [-7.78_12, -13.95_16, -3.74_06]]] )
self.assertTrue(torch.allclose(output[:, :3, :3] , __a , atol=1e-4 ) )
@slow
def a__ ( self: Dict )-> str:
with torch.no_grad():
lowerCamelCase : Any = EsmModel.from_pretrained("""facebook/esm2_t6_8M_UR50D""" )
model.eval()
lowerCamelCase : Optional[Any] = torch.tensor([[0, 6, 4, 13, 5, 4, 16, 12, 11, 7, 2]] )
lowerCamelCase : Any = model(__a )[0]
# compare the actual values for a slice.
lowerCamelCase : Tuple = torch.tensor(
[[[0.14_44, 0.54_13, 0.32_48], [0.30_34, 0.00_53, 0.31_08], [0.32_28, -0.24_99, 0.34_15]]] )
self.assertTrue(torch.allclose(output[:, :3, :3] , __a , atol=1e-4 ) )
| 42
| 0
|
"""simple docstring"""
import argparse
import re
import numpy as np
import requests
import torch
from huggingface_hub import hf_hub_download
from PIL import Image
from transformers import (
SamConfig,
SamImageProcessor,
SamModel,
SamProcessor,
SamVisionConfig,
)
__lowerCamelCase :Tuple = {
'iou_prediction_head.layers.0': 'iou_prediction_head.proj_in',
'iou_prediction_head.layers.1': 'iou_prediction_head.layers.0',
'iou_prediction_head.layers.2': 'iou_prediction_head.proj_out',
'mask_decoder.output_upscaling.0': 'mask_decoder.upscale_conv1',
'mask_decoder.output_upscaling.1': 'mask_decoder.upscale_layer_norm',
'mask_decoder.output_upscaling.3': 'mask_decoder.upscale_conv2',
'mask_downscaling.0': 'mask_embed.conv1',
'mask_downscaling.1': 'mask_embed.layer_norm1',
'mask_downscaling.3': 'mask_embed.conv2',
'mask_downscaling.4': 'mask_embed.layer_norm2',
'mask_downscaling.6': 'mask_embed.conv3',
'point_embeddings': 'point_embed',
'pe_layer.positional_encoding_gaussian_matrix': 'shared_embedding.positional_embedding',
'image_encoder': 'vision_encoder',
'neck.0': 'neck.conv1',
'neck.1': 'neck.layer_norm1',
'neck.2': 'neck.conv2',
'neck.3': 'neck.layer_norm2',
'patch_embed.proj': 'patch_embed.projection',
'.norm': '.layer_norm',
'blocks': 'layers',
}
def snake_case ( UpperCamelCase__ : List[Any] ) -> int:
lowerCamelCase : Union[str, Any] = {}
state_dict.pop("""pixel_mean""" , lowerCamelCase_ )
state_dict.pop("""pixel_std""" , lowerCamelCase_ )
lowerCamelCase : str = R'.*.output_hypernetworks_mlps.(\d+).layers.(\d+).*'
for key, value in state_dict.items():
for key_to_modify, new_key in KEYS_TO_MODIFY_MAPPING.items():
if key_to_modify in key:
lowerCamelCase : int = key.replace(lowerCamelCase_ , lowerCamelCase_ )
if re.match(lowerCamelCase_ , lowerCamelCase_ ):
lowerCamelCase : int = int(re.match(lowerCamelCase_ , lowerCamelCase_ ).group(2 ) )
if layer_nb == 0:
lowerCamelCase : Optional[Any] = key.replace("""layers.0""" , """proj_in""" )
elif layer_nb == 1:
lowerCamelCase : List[str] = key.replace("""layers.1""" , """layers.0""" )
elif layer_nb == 2:
lowerCamelCase : List[Any] = key.replace("""layers.2""" , """proj_out""" )
lowerCamelCase : Any = value
lowerCamelCase : int = model_state_dict[
'prompt_encoder.shared_embedding.positional_embedding'
]
return model_state_dict
def snake_case ( UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : int , UpperCamelCase__ : Dict , UpperCamelCase__ : int="ybelkada/segment-anything" ) -> Any:
lowerCamelCase : Dict = hf_hub_download(lowerCamelCase_ , F'checkpoints/{model_name}.pth' )
if "sam_vit_b" in model_name:
lowerCamelCase : Tuple = SamConfig()
elif "sam_vit_l" in model_name:
lowerCamelCase : Union[str, Any] = SamVisionConfig(
hidden_size=1024 , num_hidden_layers=24 , num_attention_heads=16 , global_attn_indexes=[5, 11, 17, 23] , )
lowerCamelCase : Tuple = SamConfig(
vision_config=lowerCamelCase_ , )
elif "sam_vit_h" in model_name:
lowerCamelCase : Optional[Any] = SamVisionConfig(
hidden_size=1280 , num_hidden_layers=32 , num_attention_heads=16 , global_attn_indexes=[7, 15, 23, 31] , )
lowerCamelCase : str = SamConfig(
vision_config=lowerCamelCase_ , )
lowerCamelCase : Optional[int] = torch.load(lowerCamelCase_ , map_location="""cpu""" )
lowerCamelCase : Union[str, Any] = replace_keys(lowerCamelCase_ )
lowerCamelCase : int = SamImageProcessor()
lowerCamelCase : Dict = SamProcessor(image_processor=lowerCamelCase_ )
lowerCamelCase : Union[str, Any] = SamModel(lowerCamelCase_ )
hf_model.load_state_dict(lowerCamelCase_ )
lowerCamelCase : Dict = hf_model.to("""cuda""" )
lowerCamelCase : Dict = 'https://huggingface.co/ybelkada/segment-anything/resolve/main/assets/car.png'
lowerCamelCase : Any = Image.open(requests.get(lowerCamelCase_ , stream=lowerCamelCase_ ).raw ).convert("""RGB""" )
lowerCamelCase : Dict = [[[400, 650]]]
lowerCamelCase : List[str] = [[1]]
lowerCamelCase : Any = processor(images=np.array(lowerCamelCase_ ) , return_tensors="""pt""" ).to("""cuda""" )
with torch.no_grad():
lowerCamelCase : Tuple = hf_model(**lowerCamelCase_ )
lowerCamelCase : str = output.iou_scores.squeeze()
if model_name == "sam_vit_h_4b8939":
assert scores[-1].item() == 0.5_7_9_8_9_0_2_5_1_1_5_9_6_6_8
lowerCamelCase : Union[str, Any] = processor(
images=np.array(lowerCamelCase_ ) , input_points=lowerCamelCase_ , input_labels=lowerCamelCase_ , return_tensors="""pt""" ).to("""cuda""" )
with torch.no_grad():
lowerCamelCase : List[str] = hf_model(**lowerCamelCase_ )
lowerCamelCase : List[Any] = output.iou_scores.squeeze()
assert scores[-1].item() == 0.9_7_1_2_6_0_3_0_9_2_1_9_3_6_0_4
lowerCamelCase : Tuple = ((75, 275, 1725, 850),)
lowerCamelCase : str = processor(images=np.array(lowerCamelCase_ ) , input_boxes=lowerCamelCase_ , return_tensors="""pt""" ).to("""cuda""" )
with torch.no_grad():
lowerCamelCase : Tuple = hf_model(**lowerCamelCase_ )
lowerCamelCase : Union[str, Any] = output.iou_scores.squeeze()
assert scores[-1].item() == 0.8_6_8_6_0_1_5_6_0_5_9_2_6_5_1_4
# Test with 2 points and 1 image.
lowerCamelCase : str = [[[400, 650], [800, 650]]]
lowerCamelCase : Optional[int] = [[1, 1]]
lowerCamelCase : Any = processor(
images=np.array(lowerCamelCase_ ) , input_points=lowerCamelCase_ , input_labels=lowerCamelCase_ , return_tensors="""pt""" ).to("""cuda""" )
with torch.no_grad():
lowerCamelCase : List[str] = hf_model(**lowerCamelCase_ )
lowerCamelCase : Optional[int] = output.iou_scores.squeeze()
assert scores[-1].item() == 0.9_9_3_6_0_4_7_7_9_2_4_3_4_6_9_2
if __name__ == "__main__":
__lowerCamelCase :Union[str, Any] = argparse.ArgumentParser()
__lowerCamelCase :Dict = ['sam_vit_b_01ec64', 'sam_vit_h_4b8939', 'sam_vit_l_0b3195']
parser.add_argument(
'--model_name',
default='sam_vit_h_4b8939',
choices=choices,
type=str,
help='Path to hf config.json of model to convert',
)
parser.add_argument('--pytorch_dump_folder_path', default=None, type=str, help='Path to the output PyTorch model.')
parser.add_argument(
'--push_to_hub',
action='store_true',
help='Whether to push the model and processor to the hub after converting',
)
parser.add_argument(
'--model_hub_id',
default='ybelkada/segment-anything',
choices=choices,
type=str,
help='Path to hf config.json of model to convert',
)
__lowerCamelCase :Union[str, Any] = parser.parse_args()
convert_sam_checkpoint(args.model_name, args.pytorch_dump_folder_path, args.push_to_hub, args.model_hub_id)
| 706
|
"""simple docstring"""
import unittest
from transformers import AlbertTokenizer, AlbertTokenizerFast
from transformers.testing_utils import get_tests_dir, require_sentencepiece, require_tokenizers, slow
from ...test_tokenization_common import TokenizerTesterMixin
__lowerCamelCase :str = get_tests_dir('fixtures/spiece.model')
@require_sentencepiece
@require_tokenizers
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =AlbertTokenizer
snake_case__ : Optional[Any] =AlbertTokenizerFast
snake_case__ : Optional[int] =True
snake_case__ : Any =True
snake_case__ : Optional[int] =True
def a__ ( self: Dict )-> Optional[Any]:
super().setUp()
# We have a SentencePiece fixture for testing
lowerCamelCase : int = AlbertTokenizer(__a )
tokenizer.save_pretrained(self.tmpdirname )
def a__ ( self: Tuple , __a: Tuple )-> Union[str, Any]:
lowerCamelCase : List[str] = """this is a test"""
lowerCamelCase : int = """this is a test"""
return input_text, output_text
def a__ ( self: Any )-> List[Any]:
lowerCamelCase : int = """<pad>"""
lowerCamelCase : Optional[int] = 0
self.assertEqual(self.get_tokenizer()._convert_token_to_id(__a ) , __a )
self.assertEqual(self.get_tokenizer()._convert_id_to_token(__a ) , __a )
def a__ ( self: Tuple )-> str:
lowerCamelCase : Optional[int] = list(self.get_tokenizer().get_vocab().keys() )
self.assertEqual(vocab_keys[0] , """<pad>""" )
self.assertEqual(vocab_keys[1] , """<unk>""" )
self.assertEqual(vocab_keys[-1] , """▁eloquent""" )
self.assertEqual(len(__a ) , 30_000 )
def a__ ( self: List[str] )-> Any:
self.assertEqual(self.get_tokenizer().vocab_size , 30_000 )
def a__ ( self: Optional[Any] )-> Union[str, Any]:
if not self.test_rust_tokenizer:
return
lowerCamelCase : str = self.get_tokenizer()
lowerCamelCase : Tuple = self.get_rust_tokenizer()
lowerCamelCase : Union[str, Any] = """I was born in 92000, and this is falsé."""
lowerCamelCase : List[str] = tokenizer.tokenize(__a )
lowerCamelCase : Tuple = rust_tokenizer.tokenize(__a )
self.assertListEqual(__a , __a )
lowerCamelCase : Dict = tokenizer.encode(__a , add_special_tokens=__a )
lowerCamelCase : List[str] = rust_tokenizer.encode(__a , add_special_tokens=__a )
self.assertListEqual(__a , __a )
lowerCamelCase : Any = self.get_rust_tokenizer()
lowerCamelCase : List[str] = tokenizer.encode(__a )
lowerCamelCase : str = rust_tokenizer.encode(__a )
self.assertListEqual(__a , __a )
def a__ ( self: Tuple )-> List[Any]:
lowerCamelCase : List[str] = AlbertTokenizer(__a , keep_accents=__a )
lowerCamelCase : int = tokenizer.tokenize("""This is a test""" )
self.assertListEqual(__a , ["""▁this""", """▁is""", """▁a""", """▁test"""] )
self.assertListEqual(tokenizer.convert_tokens_to_ids(__a ) , [48, 25, 21, 1_289] )
lowerCamelCase : List[str] = tokenizer.tokenize("""I was born in 92000, and this is falsé.""" )
self.assertListEqual(
__a , ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """é""", """."""] )
lowerCamelCase : List[str] = tokenizer.convert_tokens_to_ids(__a )
self.assertListEqual(__a , [31, 23, 386, 19, 561, 3_050, 15, 17, 48, 25, 8_256, 18, 1, 9] )
lowerCamelCase : Tuple = tokenizer.convert_ids_to_tokens(__a )
self.assertListEqual(
__a , ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """<unk>""", """."""] , )
def a__ ( self: Tuple )-> str:
lowerCamelCase : str = AlbertTokenizer(__a )
lowerCamelCase : Union[str, Any] = tokenizer.encode("""sequence builders""" )
lowerCamelCase : List[Any] = tokenizer.encode("""multi-sequence build""" )
lowerCamelCase : Any = tokenizer.build_inputs_with_special_tokens(__a )
lowerCamelCase : str = tokenizer.build_inputs_with_special_tokens(__a , __a )
assert encoded_sentence == [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id]
assert encoded_pair == [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id] + text_a + [
tokenizer.sep_token_id
]
@slow
def a__ ( self: Any )-> Dict:
# fmt: off
lowerCamelCase : Optional[Any] = {"""attention_mask""": [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """input_ids""": [[2, 21_970, 13, 5, 6_092, 167, 28, 7_103, 2_153, 673, 8, 7_028, 12_051, 18, 17, 7_103, 2_153, 673, 8, 3_515, 18_684, 8, 4_461, 6, 1_927, 297, 8, 12_060, 2_607, 18, 13, 5, 4_461, 15, 10_538, 38, 8, 135, 15, 822, 58, 15, 993, 10_363, 15, 1_460, 8_005, 4_461, 15, 993, 255, 2_328, 9, 9, 9, 6, 26, 1_112, 816, 3_260, 13, 5, 103, 2_377, 6, 17, 1_112, 816, 2_782, 13, 5, 103, 10_641, 6, 29, 84, 2_512, 2_430, 782, 18_684, 2_761, 19, 808, 2_430, 2_556, 17, 855, 1_480, 9_477, 4_091, 128, 11_712, 15, 7_103, 2_153, 673, 17, 24_883, 9_990, 9, 3], [2, 11_502, 25, 1_006, 20, 782, 8, 11_809, 855, 1_732, 19_393, 18_667, 37, 367, 21_018, 69, 1_854, 34, 11_860, 19_124, 27, 156, 225, 17, 193, 4_141, 19, 65, 9_124, 9, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [2, 14, 2_231, 886, 2_385, 17_659, 84, 14, 16_792, 1_952, 9, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """token_type_ids""": [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]} # noqa: E501
# fmt: on
self.tokenizer_integration_test_util(
expected_encoding=__a , model_name="""albert-base-v2""" , revision="""6b6560eaf5ff2e250b00c50f380c5389a9c2d82e""" , )
| 42
| 0
|
import os
import numpy
import onnx
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Any ) -> List[str]:
lowerCamelCase : Optional[Any] = a.name
lowerCamelCase : List[str] = b.name
lowerCamelCase : int = ''
lowerCamelCase : Dict = ''
lowerCamelCase : Optional[Any] = a == b
lowerCamelCase : Optional[Any] = name_a
lowerCamelCase : Optional[Any] = name_b
return res
def snake_case ( UpperCamelCase__ : Any , UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Any ) -> Union[str, Any]:
for i, input_name in enumerate(node_proto.input ):
if input_name == name:
node_proto.input.insert(UpperCamelCase__ , UpperCamelCase__ )
node_proto.input.pop(i + 1 )
if node_proto.op_type == "If":
_graph_replace_input_with(node_proto.attribute[0].g , UpperCamelCase__ , UpperCamelCase__ )
_graph_replace_input_with(node_proto.attribute[1].g , UpperCamelCase__ , UpperCamelCase__ )
if node_proto.op_type == "Loop":
_graph_replace_input_with(node_proto.attribute[0].g , UpperCamelCase__ , UpperCamelCase__ )
def snake_case ( UpperCamelCase__ : Dict , UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : Any ) -> Tuple:
for n in graph_proto.node:
_node_replace_input_with(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
def snake_case ( UpperCamelCase__ : List[Any] , UpperCamelCase__ : Dict , UpperCamelCase__ : str ) -> List[Any]:
lowerCamelCase : Tuple = list(model.graph.initializer )
lowerCamelCase : List[str] = list(model_without_ext.graph.initializer )
for i, ref_i in ind_to_replace:
assert inits_with_data[i].name == inits[i].name
assert inits_with_data[ref_i].name == inits[ref_i].name
assert i > ref_i
lowerCamelCase : List[str] = inits[i].name
lowerCamelCase : Dict = inits[ref_i].name
model_without_ext.graph.initializer.remove(inits[i] )
# for n in model.graph.node:
_graph_replace_input_with(model_without_ext.graph , UpperCamelCase__ , UpperCamelCase__ )
def snake_case ( UpperCamelCase__ : List[Any] ) -> str:
lowerCamelCase : List[Any] = os.path.dirname(UpperCamelCase__ )
lowerCamelCase : List[str] = os.path.basename(UpperCamelCase__ )
lowerCamelCase : Dict = onnx.load(os.path.join(UpperCamelCase__ , UpperCamelCase__ ) )
lowerCamelCase : Optional[int] = list(model.graph.initializer )
lowerCamelCase : Tuple = set()
lowerCamelCase : int = {}
lowerCamelCase : Optional[Any] = []
lowerCamelCase : Optional[Any] = 0
for i in range(len(UpperCamelCase__ ) ):
if i in dup_set:
continue
for j in range(i + 1 , len(UpperCamelCase__ ) ):
if j in dup_set:
continue
if _is_equal_tensor_proto(inits[i] , inits[j] ):
dup_set.add(UpperCamelCase__ )
dup_set.add(UpperCamelCase__ )
lowerCamelCase : Any = inits[j].data_type
lowerCamelCase : Dict = numpy.prod(inits[j].dims )
if dtype == 1:
mem_size *= 4
elif dtype == 6:
mem_size *= 4
elif dtype == 7 or dtype == 11:
mem_size *= 8
else:
print("""unexpected data type: """ , UpperCamelCase__ )
total_reduced_size += mem_size
lowerCamelCase : List[str] = inits[i].name
lowerCamelCase : List[Any] = inits[j].name
if name_i in dup_map:
dup_map[name_i].append(UpperCamelCase__ )
else:
lowerCamelCase : str = [name_j]
ind_to_replace.append((j, i) )
print("""total reduced size: """ , total_reduced_size / 1024 / 1024 / 1024 , """GB""" )
lowerCamelCase : Optional[Any] = sorted(UpperCamelCase__ )
_remove_dup_initializers_from_model(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
lowerCamelCase : Dict = 'optimized_' + model_file_name
lowerCamelCase : str = os.path.join(UpperCamelCase__ , UpperCamelCase__ )
onnx.save(UpperCamelCase__ , UpperCamelCase__ )
return new_model
| 707
|
"""simple docstring"""
__lowerCamelCase :List[Any] = {0: [2, 3], 1: [0], 2: [1], 3: [4], 4: []}
__lowerCamelCase :Union[str, Any] = {0: [1, 2, 3], 1: [2], 2: [0], 3: [4], 4: [5], 5: [3]}
def snake_case ( UpperCamelCase__ : dict[int, list[int]] , UpperCamelCase__ : int , UpperCamelCase__ : list[bool] ) -> list[int]:
lowerCamelCase : Tuple = True
lowerCamelCase : Any = []
for neighbour in graph[vert]:
if not visited[neighbour]:
order += topology_sort(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
order.append(UpperCamelCase__ )
return order
def snake_case ( UpperCamelCase__ : dict[int, list[int]] , UpperCamelCase__ : int , UpperCamelCase__ : list[bool] ) -> list[int]:
lowerCamelCase : List[Any] = True
lowerCamelCase : int = [vert]
for neighbour in reversed_graph[vert]:
if not visited[neighbour]:
component += find_components(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return component
def snake_case ( UpperCamelCase__ : dict[int, list[int]] ) -> list[list[int]]:
lowerCamelCase : int = len(UpperCamelCase__ ) * [False]
lowerCamelCase : dict[int, list[int]] = {vert: [] for vert in range(len(UpperCamelCase__ ) )}
for vert, neighbours in graph.items():
for neighbour in neighbours:
reversed_graph[neighbour].append(UpperCamelCase__ )
lowerCamelCase : int = []
for i, was_visited in enumerate(UpperCamelCase__ ):
if not was_visited:
order += topology_sort(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
lowerCamelCase : Tuple = []
lowerCamelCase : str = len(UpperCamelCase__ ) * [False]
for i in range(len(UpperCamelCase__ ) ):
lowerCamelCase : Any = order[len(UpperCamelCase__ ) - i - 1]
if not visited[vert]:
lowerCamelCase : List[str] = find_components(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
components_list.append(UpperCamelCase__ )
return components_list
| 42
| 0
|
"""simple docstring"""
import datasets
from .evaluate import evaluate
__lowerCamelCase :int = '\\n@inproceedings{Rajpurkar2016SQuAD10,\n title={SQuAD: 100, 000+ Questions for Machine Comprehension of Text},\n author={Pranav Rajpurkar and Jian Zhang and Konstantin Lopyrev and Percy Liang},\n booktitle={EMNLP},\n year={2016}\n}\n'
__lowerCamelCase :Optional[Any] = '\nThis metric wrap the official scoring script for version 1 of the Stanford Question Answering Dataset (SQuAD).\n\nStanford Question Answering Dataset (SQuAD) is a reading comprehension dataset, consisting of questions posed by\ncrowdworkers on a set of Wikipedia articles, where the answer to every question is a segment of text, or span,\nfrom the corresponding reading passage, or the question might be unanswerable.\n'
__lowerCamelCase :str = '\nComputes SQuAD scores (F1 and EM).\nArgs:\n predictions: List of question-answers dictionaries with the following key-values:\n - \'id\': id of the question-answer pair as given in the references (see below)\n - \'prediction_text\': the text of the answer\n references: List of question-answers dictionaries with the following key-values:\n - \'id\': id of the question-answer pair (see above),\n - \'answers\': a Dict in the SQuAD dataset format\n {\n \'text\': list of possible texts for the answer, as a list of strings\n \'answer_start\': list of start positions for the answer, as a list of ints\n }\n Note that answer_start values are not taken into account to compute the metric.\nReturns:\n \'exact_match\': Exact match (the normalized answer exactly match the gold answer)\n \'f1\': The F-score of predicted tokens versus the gold answer\nExamples:\n\n >>> predictions = [{\'prediction_text\': \'1976\', \'id\': \'56e10a3be3433e1400422b22\'}]\n >>> references = [{\'answers\': {\'answer_start\': [97], \'text\': [\'1976\']}, \'id\': \'56e10a3be3433e1400422b22\'}]\n >>> squad_metric = datasets.load_metric("squad")\n >>> results = squad_metric.compute(predictions=predictions, references=references)\n >>> print(results)\n {\'exact_match\': 100.0, \'f1\': 100.0}\n'
@datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION , _KWARGS_DESCRIPTION)
class A__ ( datasets.Metric):
"""simple docstring"""
def a__ ( self: str )-> List[str]:
return datasets.MetricInfo(
description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features(
{
"""predictions""": {"""id""": datasets.Value("""string""" ), """prediction_text""": datasets.Value("""string""" )},
"""references""": {
"""id""": datasets.Value("""string""" ),
"""answers""": datasets.features.Sequence(
{
"""text""": datasets.Value("""string""" ),
"""answer_start""": datasets.Value("""int32""" ),
} ),
},
} ) , codebase_urls=["""https://rajpurkar.github.io/SQuAD-explorer/"""] , reference_urls=["""https://rajpurkar.github.io/SQuAD-explorer/"""] , )
def a__ ( self: Dict , __a: Optional[Any] , __a: List[str] )-> Tuple:
lowerCamelCase : str = {prediction['''id''']: prediction['''prediction_text'''] for prediction in predictions}
lowerCamelCase : int = [
{
'''paragraphs''': [
{
'''qas''': [
{
'''answers''': [{'''text''': answer_text} for answer_text in ref['''answers''']['''text''']],
'''id''': ref['''id'''],
}
for ref in references
]
}
]
}
]
lowerCamelCase : Union[str, Any] = evaluate(dataset=__a , predictions=__a )
return score
| 708
|
"""simple docstring"""
from typing import List, Optional, Union
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :str = logging.get_logger(__name__)
__lowerCamelCase :Any = {
'huggingface/time-series-transformer-tourism-monthly': (
'https://huggingface.co/huggingface/time-series-transformer-tourism-monthly/resolve/main/config.json'
),
# See all TimeSeriesTransformer models at https://huggingface.co/models?filter=time_series_transformer
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : List[Any] ='''time_series_transformer'''
snake_case__ : List[Any] ={
'''hidden_size''': '''d_model''',
'''num_attention_heads''': '''encoder_attention_heads''',
'''num_hidden_layers''': '''encoder_layers''',
}
def __init__( self: List[str] , __a: Optional[int] = None , __a: Optional[int] = None , __a: str = "student_t" , __a: str = "nll" , __a: int = 1 , __a: List[int] = [1, 2, 3, 4, 5, 6, 7] , __a: Optional[Union[str, bool]] = "mean" , __a: int = 0 , __a: int = 0 , __a: int = 0 , __a: int = 0 , __a: Optional[List[int]] = None , __a: Optional[List[int]] = None , __a: int = 32 , __a: int = 32 , __a: int = 2 , __a: int = 2 , __a: int = 2 , __a: int = 2 , __a: bool = True , __a: str = "gelu" , __a: int = 64 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: int = 100 , __a: float = 0.02 , __a: Tuple=True , **__a: str , )-> Any:
# time series specific configuration
lowerCamelCase : str = prediction_length
lowerCamelCase : Optional[Any] = context_length or prediction_length
lowerCamelCase : Tuple = distribution_output
lowerCamelCase : Any = loss
lowerCamelCase : List[Any] = input_size
lowerCamelCase : int = num_time_features
lowerCamelCase : Dict = lags_sequence
lowerCamelCase : Optional[int] = scaling
lowerCamelCase : int = num_dynamic_real_features
lowerCamelCase : Tuple = num_static_real_features
lowerCamelCase : Any = num_static_categorical_features
if cardinality and num_static_categorical_features > 0:
if len(__a ) != num_static_categorical_features:
raise ValueError(
"""The cardinality should be a list of the same length as `num_static_categorical_features`""" )
lowerCamelCase : int = cardinality
else:
lowerCamelCase : Dict = [0]
if embedding_dimension and num_static_categorical_features > 0:
if len(__a ) != num_static_categorical_features:
raise ValueError(
"""The embedding dimension should be a list of the same length as `num_static_categorical_features`""" )
lowerCamelCase : str = embedding_dimension
else:
lowerCamelCase : str = [min(50 , (cat + 1) // 2 ) for cat in self.cardinality]
lowerCamelCase : Any = num_parallel_samples
# Transformer architecture configuration
lowerCamelCase : Any = input_size * len(__a ) + self._number_of_features
lowerCamelCase : List[str] = d_model
lowerCamelCase : Tuple = encoder_attention_heads
lowerCamelCase : Optional[int] = decoder_attention_heads
lowerCamelCase : Union[str, Any] = encoder_ffn_dim
lowerCamelCase : str = decoder_ffn_dim
lowerCamelCase : str = encoder_layers
lowerCamelCase : Any = decoder_layers
lowerCamelCase : Optional[int] = dropout
lowerCamelCase : List[str] = attention_dropout
lowerCamelCase : Tuple = activation_dropout
lowerCamelCase : Optional[int] = encoder_layerdrop
lowerCamelCase : int = decoder_layerdrop
lowerCamelCase : Optional[int] = activation_function
lowerCamelCase : Optional[Any] = init_std
lowerCamelCase : Optional[Any] = use_cache
super().__init__(is_encoder_decoder=__a , **__a )
@property
def a__ ( self: int )-> int:
return (
sum(self.embedding_dimension )
+ self.num_dynamic_real_features
+ self.num_time_features
+ self.num_static_real_features
+ self.input_size * 2 # the log1p(abs(loc)) and log(scale) features
)
| 42
| 0
|
"""simple docstring"""
import cva
import numpy as np
class A__ :
"""simple docstring"""
def __init__( self: List[Any] , __a: List[Any] , __a: Any )-> Tuple:
if k in (0.04, 0.06):
lowerCamelCase : Dict = k
lowerCamelCase : str = window_size
else:
raise ValueError("""invalid k value""" )
def __str__( self: Optional[int] )-> Tuple:
return str(self.k )
def a__ ( self: List[Any] , __a: Dict )-> Union[str, Any]:
lowerCamelCase : Tuple = cva.imread(_lowerCAmelCase , 0 )
lowerCamelCase , lowerCamelCase : str = img.shape
lowerCamelCase : List[str] = []
lowerCamelCase : Any = img.copy()
lowerCamelCase : Tuple = cva.cvtColor(_lowerCAmelCase , cva.COLOR_GRAY2RGB )
lowerCamelCase , lowerCamelCase : int = np.gradient(_lowerCAmelCase )
lowerCamelCase : Optional[int] = dx**2
lowerCamelCase : Tuple = dy**2
lowerCamelCase : Dict = dx * dy
lowerCamelCase : Optional[int] = 0.04
lowerCamelCase : str = self.window_size // 2
for y in range(_lowerCAmelCase , h - offset ):
for x in range(_lowerCAmelCase , w - offset ):
lowerCamelCase : List[Any] = ixx[
y - offset : y + offset + 1, x - offset : x + offset + 1
].sum()
lowerCamelCase : Union[str, Any] = iyy[
y - offset : y + offset + 1, x - offset : x + offset + 1
].sum()
lowerCamelCase : Tuple = ixy[
y - offset : y + offset + 1, x - offset : x + offset + 1
].sum()
lowerCamelCase : List[Any] = (wxx * wyy) - (wxy**2)
lowerCamelCase : Tuple = wxx + wyy
lowerCamelCase : Dict = det - k * (trace**2)
# Can change the value
if r > 0.5:
corner_list.append([x, y, r] )
color_img.itemset((y, x, 0) , 0 )
color_img.itemset((y, x, 1) , 0 )
color_img.itemset((y, x, 2) , 255 )
return color_img, corner_list
if __name__ == "__main__":
__lowerCamelCase :Optional[Any] = HarrisCorner(0.04, 3)
__lowerCamelCase :Tuple = edge_detect.detect('path_to_image')
cva.imwrite('detect.png', color_img)
| 709
|
"""simple docstring"""
from __future__ import annotations
__lowerCamelCase :int = 10
def snake_case ( UpperCamelCase__ : list[int] ) -> list[int]:
lowerCamelCase : int = 1
lowerCamelCase : Union[str, Any] = max(UpperCamelCase__ )
while placement <= max_digit:
# declare and initialize empty buckets
lowerCamelCase : list[list] = [[] for _ in range(UpperCamelCase__ )]
# split list_of_ints between the buckets
for i in list_of_ints:
lowerCamelCase : Any = int((i / placement) % RADIX )
buckets[tmp].append(UpperCamelCase__ )
# put each buckets' contents into list_of_ints
lowerCamelCase : Dict = 0
for b in range(UpperCamelCase__ ):
for i in buckets[b]:
lowerCamelCase : List[str] = i
a += 1
# move to next
placement *= RADIX
return list_of_ints
if __name__ == "__main__":
import doctest
doctest.testmod()
| 42
| 0
|
"""simple docstring"""
def snake_case ( UpperCamelCase__ : Union[str, Any] = 10**12 ) -> int:
lowerCamelCase : Any = 1
lowerCamelCase : int = 0
lowerCamelCase : str = 1
lowerCamelCase : Union[str, Any] = 1
while numerator <= 2 * min_total - 1:
prev_numerator += 2 * numerator
numerator += 2 * prev_numerator
prev_denominator += 2 * denominator
denominator += 2 * prev_denominator
return (denominator + 1) // 2
if __name__ == "__main__":
print(f"""{solution() = }""")
| 710
|
"""simple docstring"""
import argparse
import pickle
import numpy as np
import torch
from torch import nn
from transformers import ReformerConfig, ReformerModelWithLMHead
from transformers.utils import logging
logging.set_verbosity_info()
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Tuple , UpperCamelCase__ : Optional[Any]=None ) -> Tuple:
# set parameter of one layer
assert torch_layer.weight.shape == weight.shape, F'{torch_layer} layer.weight does not match'
lowerCamelCase : Dict = nn.Parameter(UpperCamelCase__ )
if bias is not None:
assert torch_layer.bias.shape == bias.shape, F'{torch_layer} layer.bias does not match'
lowerCamelCase : Any = nn.Parameter(UpperCamelCase__ )
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : Any , UpperCamelCase__ : Tuple ) -> Union[str, Any]:
# set torch weights for 1-to-1 comparison
lowerCamelCase : Dict = np.asarray(weights[0] )
lowerCamelCase : List[Any] = np.asarray(weights[1] )
lowerCamelCase : List[str] = np.asarray(weights[2] )
set_param(
torch_layer.self_attention.query_key , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.value , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.output.dense , torch.tensor(UpperCamelCase__ ).view(-1 , UpperCamelCase__ ).contiguous().transpose(0 , 1 ) , )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Any ) -> List[Any]:
# set torch weights for 1-to-1 comparison
lowerCamelCase : Tuple = np.asarray(weights[0] )
lowerCamelCase : Any = np.asarray(weights[1] )
lowerCamelCase : List[Any] = np.asarray(weights[2] )
lowerCamelCase : List[str] = np.asarray(weights[3] )
set_param(
torch_layer.self_attention.query , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.key , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.value , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.output.dense , torch.tensor(UpperCamelCase__ ).view(-1 , UpperCamelCase__ ).contiguous().transpose(0 , 1 ) , )
def snake_case ( UpperCamelCase__ : Dict , UpperCamelCase__ : Optional[int] , UpperCamelCase__ : Any ) -> Optional[Any]:
# layernorm 1
lowerCamelCase : str = weights[0][0][0]
lowerCamelCase : Optional[int] = np.asarray(layer_norm_a[0] )
lowerCamelCase : Tuple = np.asarray(layer_norm_a[1] )
set_param(
torch_block.attention.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# lsh weights + output
lowerCamelCase : List[Any] = weights[0][1]
if len(UpperCamelCase__ ) < 4:
set_layer_weights_in_torch_lsh(UpperCamelCase__ , torch_block.attention , UpperCamelCase__ )
else:
set_layer_weights_in_torch_local(UpperCamelCase__ , torch_block.attention , UpperCamelCase__ )
# intermediate weighs
lowerCamelCase : int = weights[2][0][1][2]
# Chunked Feed Forward
if len(UpperCamelCase__ ) == 4:
lowerCamelCase : Dict = intermediate_weights[2]
# layernorm 2
lowerCamelCase : Optional[int] = np.asarray(intermediate_weights[0][0] )
lowerCamelCase : Tuple = np.asarray(intermediate_weights[0][1] )
set_param(
torch_block.feed_forward.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# intermediate dense
lowerCamelCase : Optional[Any] = np.asarray(intermediate_weights[1][0] )
lowerCamelCase : Union[str, Any] = np.asarray(intermediate_weights[1][1] )
set_param(
torch_block.feed_forward.dense.dense , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
# intermediate out
lowerCamelCase : Optional[int] = np.asarray(intermediate_weights[4][0] )
lowerCamelCase : List[Any] = np.asarray(intermediate_weights[4][1] )
set_param(
torch_block.feed_forward.output.dense , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : List[str] , UpperCamelCase__ : int ) -> List[Any]:
# reformer model
lowerCamelCase : List[Any] = torch_model.reformer
# word embeds
lowerCamelCase : Union[str, Any] = np.asarray(weights[1] )
set_param(
torch_model_reformer.embeddings.word_embeddings , torch.tensor(UpperCamelCase__ ) , )
if isinstance(weights[3] , UpperCamelCase__ ):
lowerCamelCase : Union[str, Any] = torch_model_reformer.embeddings.position_embeddings
for emb_idx in range(len(position_embeddings.weights ) ):
lowerCamelCase : str = np.asarray(weights[3][emb_idx][0] )
assert (
position_embeddings.weights[emb_idx].shape == emb_weights.shape
), F'{position_embeddings[emb_idx]} emb does not match'
lowerCamelCase : Dict = nn.Parameter(torch.tensor(UpperCamelCase__ ) )
lowerCamelCase : int = weights[5]
assert len(torch_model_reformer.encoder.layers ) * 4 == len(
UpperCamelCase__ ), "HF and trax model do not have the same number of layers"
for layer_idx, layer in enumerate(torch_model_reformer.encoder.layers ):
lowerCamelCase : Dict = trax_layer_weights[4 * layer_idx : 4 * (layer_idx + 1)]
set_block_weights_in_torch(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
# output layer norm
lowerCamelCase : Any = np.asarray(weights[7][0] )
lowerCamelCase : List[str] = np.asarray(weights[7][1] )
set_param(
torch_model_reformer.encoder.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# output embeddings
lowerCamelCase : List[Any] = np.asarray(weights[9][0] )
lowerCamelCase : Optional[int] = np.asarray(weights[9][1] )
set_param(
torch_model.lm_head.decoder , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
def snake_case ( UpperCamelCase__ : str , UpperCamelCase__ : int , UpperCamelCase__ : str ) -> Optional[int]:
# Initialise PyTorch model
lowerCamelCase : Union[str, Any] = ReformerConfig.from_json_file(UpperCamelCase__ )
print(F'Building PyTorch model from configuration: {config}' )
lowerCamelCase : str = ReformerModelWithLMHead(UpperCamelCase__ )
with open(UpperCamelCase__ , """rb""" ) as f:
lowerCamelCase : str = pickle.load(UpperCamelCase__ )["""weights"""]
set_model_weights_in_torch(UpperCamelCase__ , UpperCamelCase__ , config.hidden_size )
# Save pytorch-model
print(F'Save PyTorch model to {pytorch_dump_path}' )
torch.save(model.state_dict() , UpperCamelCase__ )
if __name__ == "__main__":
__lowerCamelCase :Dict = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--trax_model_pkl_path', default=None, type=str, required=True, help='Path to the TensorFlow checkpoint path.'
)
parser.add_argument(
'--config_file',
default=None,
type=str,
required=True,
help=(
'The config json file corresponding to the pre-trained Reformer model. \n'
'This specifies the model architecture.'
),
)
parser.add_argument(
'--pytorch_dump_path', default=None, type=str, required=True, help='Path to the output PyTorch model.'
)
__lowerCamelCase :Optional[int] = parser.parse_args()
convert_trax_checkpoint_to_pytorch(args.trax_model_pkl_path, args.config_file, args.pytorch_dump_path)
| 42
| 0
|
"""simple docstring"""
import copy
from collections import OrderedDict
from typing import Dict, Mapping
from packaging import version
from ...configuration_utils import PretrainedConfig
from ...onnx import OnnxConfig
from ...utils import logging
from ..auto import CONFIG_MAPPING
__lowerCamelCase :int = logging.get_logger(__name__)
__lowerCamelCase :Optional[Any] = {
'facebook/detr-resnet-50': 'https://huggingface.co/facebook/detr-resnet-50/resolve/main/config.json',
# See all DETR models at https://huggingface.co/models?filter=detr
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : str ='''detr'''
snake_case__ : List[str] =['''past_key_values''']
snake_case__ : str ={
'''hidden_size''': '''d_model''',
'''num_attention_heads''': '''encoder_attention_heads''',
}
def __init__( self: Optional[int] , __a: Tuple=True , __a: List[Any]=None , __a: List[str]=3 , __a: Any=100 , __a: Union[str, Any]=6 , __a: Union[str, Any]=2_048 , __a: str=8 , __a: Dict=6 , __a: Tuple=2_048 , __a: Optional[int]=8 , __a: int=0.0 , __a: Union[str, Any]=0.0 , __a: Optional[Any]=True , __a: Union[str, Any]="relu" , __a: str=256 , __a: int=0.1 , __a: List[str]=0.0 , __a: Union[str, Any]=0.0 , __a: int=0.02 , __a: Union[str, Any]=1.0 , __a: str=False , __a: int="sine" , __a: List[str]="resnet50" , __a: List[Any]=True , __a: Tuple=False , __a: Dict=1 , __a: Optional[Any]=5 , __a: Optional[int]=2 , __a: List[str]=1 , __a: int=1 , __a: str=5 , __a: List[Any]=2 , __a: Union[str, Any]=0.1 , **__a: Union[str, Any] , )-> int:
if backbone_config is not None and use_timm_backbone:
raise ValueError("""You can't specify both `backbone_config` and `use_timm_backbone`.""" )
if not use_timm_backbone:
if backbone_config is None:
logger.info("""`backbone_config` is `None`. Initializing the config with the default `ResNet` backbone.""" )
lowerCamelCase : int = CONFIG_MAPPING["resnet"](out_features=["""stage4"""] )
elif isinstance(__A , __A ):
lowerCamelCase : Union[str, Any] = backbone_config.get("""model_type""" )
lowerCamelCase : Optional[Any] = CONFIG_MAPPING[backbone_model_type]
lowerCamelCase : Optional[int] = config_class.from_dict(__A )
# set timm attributes to None
lowerCamelCase : List[Any] = None, None, None
lowerCamelCase : Tuple = use_timm_backbone
lowerCamelCase : Dict = backbone_config
lowerCamelCase : List[Any] = num_channels
lowerCamelCase : Union[str, Any] = num_queries
lowerCamelCase : Optional[Any] = d_model
lowerCamelCase : Any = encoder_ffn_dim
lowerCamelCase : Optional[Any] = encoder_layers
lowerCamelCase : int = encoder_attention_heads
lowerCamelCase : Union[str, Any] = decoder_ffn_dim
lowerCamelCase : Dict = decoder_layers
lowerCamelCase : Dict = decoder_attention_heads
lowerCamelCase : Optional[Any] = dropout
lowerCamelCase : Tuple = attention_dropout
lowerCamelCase : Union[str, Any] = activation_dropout
lowerCamelCase : List[Any] = activation_function
lowerCamelCase : str = init_std
lowerCamelCase : Dict = init_xavier_std
lowerCamelCase : Tuple = encoder_layerdrop
lowerCamelCase : List[str] = decoder_layerdrop
lowerCamelCase : Optional[int] = encoder_layers
lowerCamelCase : List[str] = auxiliary_loss
lowerCamelCase : Optional[int] = position_embedding_type
lowerCamelCase : Union[str, Any] = backbone
lowerCamelCase : str = use_pretrained_backbone
lowerCamelCase : Union[str, Any] = dilation
# Hungarian matcher
lowerCamelCase : Any = class_cost
lowerCamelCase : int = bbox_cost
lowerCamelCase : Tuple = giou_cost
# Loss coefficients
lowerCamelCase : Union[str, Any] = mask_loss_coefficient
lowerCamelCase : Tuple = dice_loss_coefficient
lowerCamelCase : str = bbox_loss_coefficient
lowerCamelCase : Dict = giou_loss_coefficient
lowerCamelCase : Any = eos_coefficient
super().__init__(is_encoder_decoder=__A , **__A )
@property
def a__ ( self: Tuple )-> int:
return self.encoder_attention_heads
@property
def a__ ( self: str )-> Union[str, Any]:
return self.d_model
@classmethod
def a__ ( cls: str , __a: PretrainedConfig , **__a: Optional[Any] )-> List[Any]:
return cls(backbone_config=__A , **__A )
def a__ ( self: Optional[Any] )-> str:
lowerCamelCase : Union[str, Any] = copy.deepcopy(self.__dict__ )
if output["backbone_config"] is not None:
lowerCamelCase : List[Any] = self.backbone_config.to_dict()
lowerCamelCase : List[str] = self.__class__.model_type
return output
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : str =version.parse('''1.11''')
@property
def a__ ( self: Union[str, Any] )-> str:
return OrderedDict(
[
("""pixel_values""", {0: """batch""", 1: """num_channels""", 2: """height""", 3: """width"""}),
("""pixel_mask""", {0: """batch"""}),
] )
@property
def a__ ( self: Union[str, Any] )-> Union[str, Any]:
return 1e-5
@property
def a__ ( self: Optional[int] )-> List[str]:
return 12
| 711
|
"""simple docstring"""
import inspect
import unittest
import torch
import torch.nn as nn
from accelerate.hooks import (
AlignDevicesHook,
ModelHook,
SequentialHook,
add_hook_to_module,
attach_align_device_hook,
remove_hook_from_module,
remove_hook_from_submodules,
)
from accelerate.test_utils import require_multi_gpu
class A__ ( nn.Module):
"""simple docstring"""
def __init__( self: Dict )-> Dict:
super().__init__()
lowerCamelCase : Tuple = nn.Linear(3 , 4 )
lowerCamelCase : Optional[Any] = nn.BatchNormad(4 )
lowerCamelCase : Optional[Any] = nn.Linear(4 , 5 )
def a__ ( self: List[str] , __a: List[Any] )-> Optional[Any]:
return self.lineara(self.batchnorm(self.lineara(__a ) ) )
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Tuple , __a: int , *__a: Any , **__a: Tuple )-> Tuple:
return (args[0] + 1,) + args[1:], kwargs
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Optional[int] , __a: List[str] , __a: List[Any] )-> List[str]:
return output + 1
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: int )-> str:
lowerCamelCase : List[str] = ModelForTest()
lowerCamelCase : Dict = ModelHook()
add_hook_to_module(__a , __a )
self.assertEqual(test_model._hf_hook , __a )
self.assertTrue(hasattr(__a , """_old_forward""" ) )
# Check adding the hook did not change the name or the signature
self.assertEqual(test_model.forward.__name__ , """forward""" )
self.assertListEqual(list(inspect.signature(test_model.forward ).parameters ) , ["""x"""] )
remove_hook_from_module(__a )
self.assertFalse(hasattr(__a , """_hf_hook""" ) )
self.assertFalse(hasattr(__a , """_old_forward""" ) )
def a__ ( self: int )-> str:
lowerCamelCase : List[str] = ModelForTest()
lowerCamelCase : Union[str, Any] = ModelHook()
add_hook_to_module(__a , __a )
add_hook_to_module(__a , __a , append=__a )
self.assertEqual(isinstance(test_model._hf_hook , __a ) , __a )
self.assertEqual(len(test_model._hf_hook.hooks ) , 2 )
self.assertTrue(hasattr(__a , """_old_forward""" ) )
# Check adding the hook did not change the name or the signature
self.assertEqual(test_model.forward.__name__ , """forward""" )
self.assertListEqual(list(inspect.signature(test_model.forward ).parameters ) , ["""x"""] )
remove_hook_from_module(__a )
self.assertFalse(hasattr(__a , """_hf_hook""" ) )
self.assertFalse(hasattr(__a , """_old_forward""" ) )
def a__ ( self: List[Any] )-> List[str]:
lowerCamelCase : str = ModelForTest()
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : Union[str, Any] = test_model(x + 1 )
lowerCamelCase : Optional[int] = test_model(x + 2 )
lowerCamelCase : List[Any] = PreForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[int] = test_model(__a )
self.assertTrue(torch.allclose(__a , __a , atol=1e-5 ) )
# Attaching a hook to a model when it already has one replaces, does not chain
lowerCamelCase : Dict = PreForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Tuple = test_model(__a )
self.assertTrue(torch.allclose(__a , __a , atol=1e-5 ) )
# You need to use the sequential hook to chain two or more hooks
lowerCamelCase : Any = SequentialHook(PreForwardHook() , PreForwardHook() )
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[Any] = test_model(__a )
assert torch.allclose(__a , __a , atol=1e-5 )
def a__ ( self: Any )-> Optional[int]:
lowerCamelCase : str = ModelForTest()
lowerCamelCase : List[str] = torch.randn(2 , 3 )
lowerCamelCase : int = test_model(__a )
lowerCamelCase : Dict = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Tuple = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 , atol=1e-5 ) )
# Attaching a hook to a model when it already has one replaces, does not chain
lowerCamelCase : str = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[Any] = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 , atol=1e-5 ) )
# You need to use the sequential hook to chain two or more hooks
lowerCamelCase : Union[str, Any] = SequentialHook(PostForwardHook() , PostForwardHook() )
add_hook_to_module(__a , __a )
lowerCamelCase : str = test_model(__a )
assert torch.allclose(__a , output + 2 , atol=1e-5 )
def a__ ( self: int )-> Dict:
lowerCamelCase : List[Any] = ModelForTest()
lowerCamelCase : Optional[int] = torch.randn(2 , 3 )
lowerCamelCase : List[str] = test_model(__a )
lowerCamelCase : Any = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : str = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 ) )
self.assertTrue(outputa.requires_grad )
lowerCamelCase : Optional[int] = True
lowerCamelCase : Optional[int] = test_model(__a )
self.assertFalse(outputa.requires_grad )
@require_multi_gpu
def a__ ( self: List[str] )-> Union[str, Any]:
lowerCamelCase : int = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
add_hook_to_module(model.lineara , AlignDevicesHook(execution_device=0 ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(execution_device=0 ) )
add_hook_to_module(model.lineara , AlignDevicesHook(execution_device=1 ) )
self.assertEqual(model.lineara.weight.device , torch.device(0 ) )
self.assertEqual(model.batchnorm.weight.device , torch.device(0 ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device(0 ) )
self.assertEqual(model.lineara.weight.device , torch.device(1 ) )
# We can still make a forward pass. The input does not need to be on any particular device
lowerCamelCase : str = torch.randn(2 , 3 )
lowerCamelCase : Dict = model(__a )
self.assertEqual(output.device , torch.device(1 ) )
# We can add a general hook to put back output on same device as input.
add_hook_to_module(__a , AlignDevicesHook(io_same_device=__a ) )
lowerCamelCase : Optional[int] = torch.randn(2 , 3 ).to(0 )
lowerCamelCase : str = model(__a )
self.assertEqual(output.device , torch.device(0 ) )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Union[str, Any] = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : Tuple = {"""execution_device""": 0 if torch.cuda.is_available() else """cpu""", """offload""": True}
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(**__a ) )
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(hook_kwargs["""execution_device"""] )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Optional[Any] = torch.randn(2 , 3 )
lowerCamelCase : Optional[Any] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_module(model.lineara )
remove_hook_from_module(model.batchnorm )
remove_hook_from_module(model.lineara )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
lowerCamelCase : Any = {
"""execution_device""": 0 if torch.cuda.is_available() else """cpu""",
"""offload""": True,
"""offload_buffers""": True,
}
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(**__a ) )
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : int = torch.randn(2 , 3 )
lowerCamelCase : Optional[int] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_module(model.lineara )
remove_hook_from_module(model.batchnorm )
remove_hook_from_module(model.lineara )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
def a__ ( self: Any )-> List[str]:
lowerCamelCase : int = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : int = 0 if torch.cuda.is_available() else """cpu"""
attach_align_device_hook(__a , execution_device=__a , offload=__a )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(__a )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : Optional[Any] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
attach_align_device_hook(__a , execution_device=__a , offload=__a , offload_buffers=__a )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : Optional[int] = torch.randn(2 , 3 )
lowerCamelCase : int = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
def a__ ( self: Optional[Any] )-> List[Any]:
lowerCamelCase : List[Any] = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : Any = 0 if torch.cuda.is_available() else """cpu"""
attach_align_device_hook(
__a , execution_device=__a , offload=__a , weights_map=model.state_dict() )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(__a )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : int = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
attach_align_device_hook(
__a , execution_device=__a , offload=__a , weights_map=model.state_dict() , offload_buffers=__a , )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : Tuple = torch.randn(2 , 3 )
lowerCamelCase : Any = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
| 42
| 0
|
"""simple docstring"""
from math import factorial
def snake_case ( UpperCamelCase__ : Tuple = 20 ) -> int:
lowerCamelCase : int = 2 * n # middle entry of odd rows starting at row 3 is the solution for n = 1,
# 2, 3,...
lowerCamelCase : Dict = n // 2
return int(factorial(__lowerCAmelCase ) / (factorial(__lowerCAmelCase ) * factorial(n - k )) )
if __name__ == "__main__":
import sys
if len(sys.argv) == 1:
print(solution(20))
else:
try:
__lowerCamelCase :List[str] = int(sys.argv[1])
print(solution(n))
except ValueError:
print('Invalid entry - please enter a number.')
| 712
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_torch_available,
)
__lowerCamelCase :Optional[Any] = {
'configuration_encodec': [
'ENCODEC_PRETRAINED_CONFIG_ARCHIVE_MAP',
'EncodecConfig',
],
'feature_extraction_encodec': ['EncodecFeatureExtractor'],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Union[str, Any] = [
'ENCODEC_PRETRAINED_MODEL_ARCHIVE_LIST',
'EncodecModel',
'EncodecPreTrainedModel',
]
if TYPE_CHECKING:
from .configuration_encodec import (
ENCODEC_PRETRAINED_CONFIG_ARCHIVE_MAP,
EncodecConfig,
)
from .feature_extraction_encodec import EncodecFeatureExtractor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_encodec import (
ENCODEC_PRETRAINED_MODEL_ARCHIVE_LIST,
EncodecModel,
EncodecPreTrainedModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
import torch
def snake_case ( ) -> Dict:
if torch.cuda.is_available():
lowerCamelCase : Optional[Any] = torch.cuda.device_count()
else:
lowerCamelCase : Union[str, Any] = 0
print(F'Successfully ran on {num_gpus} GPUs' )
if __name__ == "__main__":
main()
| 713
|
"""simple docstring"""
import inspect
import unittest
from transformers import ConvNextConfig
from transformers.testing_utils import require_torch, require_vision, slow, torch_device
from transformers.utils import cached_property, is_torch_available, is_vision_available
from ...test_backbone_common import BackboneTesterMixin
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import ConvNextBackbone, ConvNextForImageClassification, ConvNextModel
from transformers.models.convnext.modeling_convnext import CONVNEXT_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import AutoImageProcessor
class A__ :
"""simple docstring"""
def __init__( self: Optional[Any] , __a: Union[str, Any] , __a: Optional[Any]=13 , __a: Optional[Any]=32 , __a: Dict=3 , __a: int=4 , __a: Dict=[10, 20, 30, 40] , __a: int=[2, 2, 3, 2] , __a: Any=True , __a: List[Any]=True , __a: Any=37 , __a: Optional[int]="gelu" , __a: List[str]=10 , __a: Optional[int]=0.02 , __a: Dict=["stage2", "stage3", "stage4"] , __a: List[str]=[2, 3, 4] , __a: List[str]=None , )-> Union[str, Any]:
lowerCamelCase : Optional[int] = parent
lowerCamelCase : Optional[int] = batch_size
lowerCamelCase : Any = image_size
lowerCamelCase : Tuple = num_channels
lowerCamelCase : str = num_stages
lowerCamelCase : List[str] = hidden_sizes
lowerCamelCase : str = depths
lowerCamelCase : Dict = is_training
lowerCamelCase : Optional[Any] = use_labels
lowerCamelCase : List[str] = intermediate_size
lowerCamelCase : List[str] = hidden_act
lowerCamelCase : List[str] = num_labels
lowerCamelCase : Union[str, Any] = initializer_range
lowerCamelCase : List[Any] = out_features
lowerCamelCase : Optional[Any] = out_indices
lowerCamelCase : int = scope
def a__ ( self: str )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : Dict = None
if self.use_labels:
lowerCamelCase : Dict = ids_tensor([self.batch_size] , self.num_labels )
lowerCamelCase : Any = self.get_config()
return config, pixel_values, labels
def a__ ( self: Dict )-> Union[str, Any]:
return ConvNextConfig(
num_channels=self.num_channels , hidden_sizes=self.hidden_sizes , depths=self.depths , num_stages=self.num_stages , hidden_act=self.hidden_act , is_decoder=__a , initializer_range=self.initializer_range , out_features=self.out_features , out_indices=self.out_indices , num_labels=self.num_labels , )
def a__ ( self: Optional[Any] , __a: List[Any] , __a: Any , __a: int )-> List[Any]:
lowerCamelCase : Optional[int] = ConvNextModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a )
# expected last hidden states: B, C, H // 32, W // 32
self.parent.assertEqual(
result.last_hidden_state.shape , (self.batch_size, self.hidden_sizes[-1], self.image_size // 32, self.image_size // 32) , )
def a__ ( self: int , __a: Union[str, Any] , __a: List[Any] , __a: Tuple )-> Optional[int]:
lowerCamelCase : str = ConvNextForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) )
def a__ ( self: List[Any] , __a: Any , __a: Optional[int] , __a: Tuple )-> List[str]:
lowerCamelCase : List[str] = ConvNextBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a )
# verify hidden states
self.parent.assertEqual(len(result.feature_maps ) , len(config.out_features ) )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.hidden_sizes[1], 4, 4] )
# verify channels
self.parent.assertEqual(len(model.channels ) , len(config.out_features ) )
self.parent.assertListEqual(model.channels , config.hidden_sizes[1:] )
# verify backbone works with out_features=None
lowerCamelCase : Tuple = None
lowerCamelCase : List[str] = ConvNextBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[Any] = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , 1 )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.hidden_sizes[-1], 1, 1] )
# verify channels
self.parent.assertEqual(len(model.channels ) , 1 )
self.parent.assertListEqual(model.channels , [config.hidden_sizes[-1]] )
def a__ ( self: Optional[Any] )-> Any:
lowerCamelCase : List[Any] = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : Tuple = config_and_inputs
lowerCamelCase : int = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : int =(
(
ConvNextModel,
ConvNextForImageClassification,
ConvNextBackbone,
)
if is_torch_available()
else ()
)
snake_case__ : str =(
{'''feature-extraction''': ConvNextModel, '''image-classification''': ConvNextForImageClassification}
if is_torch_available()
else {}
)
snake_case__ : Union[str, Any] =True
snake_case__ : Optional[int] =False
snake_case__ : Tuple =False
snake_case__ : Union[str, Any] =False
snake_case__ : Tuple =False
def a__ ( self: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Tuple = ConvNextModelTester(self )
lowerCamelCase : List[Any] = ConfigTester(self , config_class=__a , has_text_modality=__a , hidden_size=37 )
def a__ ( self: Optional[int] )-> Dict:
self.create_and_test_config_common_properties()
self.config_tester.create_and_test_config_to_json_string()
self.config_tester.create_and_test_config_to_json_file()
self.config_tester.create_and_test_config_from_and_save_pretrained()
self.config_tester.create_and_test_config_with_num_labels()
self.config_tester.check_config_can_be_init_without_params()
self.config_tester.check_config_arguments_init()
def a__ ( self: Optional[int] )-> Optional[Any]:
return
@unittest.skip(reason="""ConvNext does not use inputs_embeds""" )
def a__ ( self: int )-> Dict:
pass
@unittest.skip(reason="""ConvNext does not support input and output embeddings""" )
def a__ ( self: Dict )-> Optional[Any]:
pass
@unittest.skip(reason="""ConvNext does not use feedforward chunking""" )
def a__ ( self: int )-> List[Any]:
pass
def a__ ( self: Union[str, Any] )-> int:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : Any = model_class(__a )
lowerCamelCase : str = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Optional[Any] = [*signature.parameters.keys()]
lowerCamelCase : List[str] = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: Optional[int] )-> str:
lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: str )-> int:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_backbone(*__a )
def a__ ( self: int )-> Optional[int]:
def check_hidden_states_output(__a: Tuple , __a: int , __a: Tuple ):
lowerCamelCase : str = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : Tuple = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : int = outputs.encoder_hidden_states if config.is_encoder_decoder else outputs.hidden_states
lowerCamelCase : Optional[int] = self.model_tester.num_stages
self.assertEqual(len(__a ) , expected_num_stages + 1 )
# ConvNext's feature maps are of shape (batch_size, num_channels, height, width)
self.assertListEqual(
list(hidden_states[0].shape[-2:] ) , [self.model_tester.image_size // 4, self.model_tester.image_size // 4] , )
lowerCamelCase , lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : List[Any] = True
check_hidden_states_output(__a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : Tuple = True
check_hidden_states_output(__a , __a , __a )
def a__ ( self: Dict )-> Optional[Any]:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_classification(*__a )
@slow
def a__ ( self: Optional[Any] )-> Tuple:
for model_name in CONVNEXT_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : str = ConvNextModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def snake_case ( ) -> Optional[int]:
lowerCamelCase : int = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
return image
@require_torch
@require_vision
class A__ ( unittest.TestCase):
"""simple docstring"""
@cached_property
def a__ ( self: Dict )-> Union[str, Any]:
return AutoImageProcessor.from_pretrained("""facebook/convnext-tiny-224""" ) if is_vision_available() else None
@slow
def a__ ( self: List[str] )-> Dict:
lowerCamelCase : Tuple = ConvNextForImageClassification.from_pretrained("""facebook/convnext-tiny-224""" ).to(__a )
lowerCamelCase : Dict = self.default_image_processor
lowerCamelCase : Union[str, Any] = prepare_img()
lowerCamelCase : Optional[Any] = image_processor(images=__a , return_tensors="""pt""" ).to(__a )
# forward pass
with torch.no_grad():
lowerCamelCase : Any = model(**__a )
# verify the logits
lowerCamelCase : Union[str, Any] = torch.Size((1, 1_000) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : Tuple = torch.tensor([-0.02_60, -0.47_39, 0.19_11] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3] , __a , atol=1e-4 ) )
@require_torch
class A__ ( unittest.TestCase , __lowercase):
"""simple docstring"""
snake_case__ : Union[str, Any] =(ConvNextBackbone,) if is_torch_available() else ()
snake_case__ : Optional[Any] =ConvNextConfig
snake_case__ : Optional[Any] =False
def a__ ( self: List[str] )-> int:
lowerCamelCase : Dict = ConvNextModelTester(self )
| 42
| 0
|
"""simple docstring"""
from __future__ import annotations
import math
def snake_case ( UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : Tuple , UpperCamelCase__ : Tuple , UpperCamelCase__ : Any , UpperCamelCase__ : Union[str, Any] ) -> Any:
if depth < 0:
raise ValueError("""Depth cannot be less than 0""" )
if len(lowerCAmelCase__ ) == 0:
raise ValueError("""Scores cannot be empty""" )
if depth == height:
return scores[node_index]
if is_max:
return max(
minimax(depth + 1 , node_index * 2 , lowerCAmelCase__ , lowerCAmelCase__ , lowerCAmelCase__ ) , minimax(depth + 1 , node_index * 2 + 1 , lowerCAmelCase__ , lowerCAmelCase__ , lowerCAmelCase__ ) , )
return min(
minimax(depth + 1 , node_index * 2 , lowerCAmelCase__ , lowerCAmelCase__ , lowerCAmelCase__ ) , minimax(depth + 1 , node_index * 2 + 1 , lowerCAmelCase__ , lowerCAmelCase__ , lowerCAmelCase__ ) , )
def snake_case ( ) -> Dict:
lowerCamelCase : str = [90, 23, 6, 33, 21, 65, 123, 34423]
lowerCamelCase : Optional[int] = math.log(len(lowerCAmelCase__ ) , 2 )
print("""Optimal value : """ , end="""""" )
print(minimax(0 , 0 , lowerCAmelCase__ , lowerCAmelCase__ , lowerCAmelCase__ ) )
if __name__ == "__main__":
import doctest
doctest.testmod()
main()
| 714
|
"""simple docstring"""
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :Optional[int] = logging.get_logger(__name__)
__lowerCamelCase :List[str] = {
'google/realm-cc-news-pretrained-embedder': (
'https://huggingface.co/google/realm-cc-news-pretrained-embedder/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-encoder': (
'https://huggingface.co/google/realm-cc-news-pretrained-encoder/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-scorer': (
'https://huggingface.co/google/realm-cc-news-pretrained-scorer/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-openqa': (
'https://huggingface.co/google/realm-cc-news-pretrained-openqa/aresolve/main/config.json'
),
'google/realm-orqa-nq-openqa': 'https://huggingface.co/google/realm-orqa-nq-openqa/resolve/main/config.json',
'google/realm-orqa-nq-reader': 'https://huggingface.co/google/realm-orqa-nq-reader/resolve/main/config.json',
'google/realm-orqa-wq-openqa': 'https://huggingface.co/google/realm-orqa-wq-openqa/resolve/main/config.json',
'google/realm-orqa-wq-reader': 'https://huggingface.co/google/realm-orqa-wq-reader/resolve/main/config.json',
# See all REALM models at https://huggingface.co/models?filter=realm
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Optional[Any] ='''realm'''
def __init__( self: Union[str, Any] , __a: List[Any]=30_522 , __a: List[Any]=768 , __a: List[Any]=128 , __a: Union[str, Any]=12 , __a: Union[str, Any]=12 , __a: Optional[Any]=8 , __a: Dict=3_072 , __a: List[Any]="gelu_new" , __a: List[Any]=0.1 , __a: Tuple=0.1 , __a: Optional[Any]=512 , __a: Optional[int]=2 , __a: str=0.02 , __a: int=1e-1_2 , __a: Optional[Any]=256 , __a: Any=10 , __a: Dict=1e-3 , __a: Optional[Any]=5 , __a: Dict=320 , __a: Tuple=13_353_718 , __a: List[Any]=5_000 , __a: Dict=1 , __a: int=0 , __a: Dict=2 , **__a: List[str] , )-> Any:
super().__init__(pad_token_id=__a , bos_token_id=__a , eos_token_id=__a , **__a )
# Common config
lowerCamelCase : Optional[Any] = vocab_size
lowerCamelCase : str = max_position_embeddings
lowerCamelCase : Dict = hidden_size
lowerCamelCase : Dict = retriever_proj_size
lowerCamelCase : Optional[Any] = num_hidden_layers
lowerCamelCase : List[str] = num_attention_heads
lowerCamelCase : Tuple = num_candidates
lowerCamelCase : int = intermediate_size
lowerCamelCase : Dict = hidden_act
lowerCamelCase : List[str] = hidden_dropout_prob
lowerCamelCase : Dict = attention_probs_dropout_prob
lowerCamelCase : Optional[int] = initializer_range
lowerCamelCase : Dict = type_vocab_size
lowerCamelCase : Optional[Any] = layer_norm_eps
# Reader config
lowerCamelCase : List[str] = span_hidden_size
lowerCamelCase : Dict = max_span_width
lowerCamelCase : Optional[Any] = reader_layer_norm_eps
lowerCamelCase : Optional[int] = reader_beam_size
lowerCamelCase : List[Any] = reader_seq_len
# Retrieval config
lowerCamelCase : int = num_block_records
lowerCamelCase : Dict = searcher_beam_size
| 42
| 0
|
"""simple docstring"""
from typing import List, Union
from ..utils import (
add_end_docstrings,
is_tf_available,
is_torch_available,
is_vision_available,
logging,
requires_backends,
)
from .base import PIPELINE_INIT_ARGS, Pipeline
if is_vision_available():
from PIL import Image
from ..image_utils import load_image
if is_tf_available():
import tensorflow as tf
from ..models.auto.modeling_tf_auto import TF_MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING
from ..tf_utils import stable_softmax
if is_torch_available():
from ..models.auto.modeling_auto import MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING
__lowerCamelCase :Optional[Any] = logging.get_logger(__name__)
@add_end_docstrings(__lowerCAmelCase)
class A__ ( __lowerCAmelCase):
"""simple docstring"""
def __init__( self: str , *__a: int , **__a: List[str] )-> Tuple:
super().__init__(*_UpperCamelCase , **_UpperCamelCase )
requires_backends(self , """vision""" )
self.check_model_type(
TF_MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING
if self.framework == """tf"""
else MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING )
def a__ ( self: Dict , __a: List[str]=None )-> Optional[int]:
lowerCamelCase : Optional[int] = {}
if top_k is not None:
lowerCamelCase : str = top_k
return {}, {}, postprocess_params
def __call__( self: Optional[int] , __a: Union[str, List[str], "Image.Image", List["Image.Image"]] , **__a: Optional[int] )-> Dict:
return super().__call__(_UpperCamelCase , **_UpperCamelCase )
def a__ ( self: Optional[int] , __a: Optional[int] )-> str:
lowerCamelCase : Tuple = load_image(_UpperCamelCase )
lowerCamelCase : Any = self.image_processor(images=_UpperCamelCase , return_tensors=self.framework )
return model_inputs
def a__ ( self: str , __a: Union[str, Any] )-> List[str]:
lowerCamelCase : Any = self.model(**_UpperCamelCase )
return model_outputs
def a__ ( self: Optional[Any] , __a: List[str] , __a: List[str]=5 )-> str:
if top_k > self.model.config.num_labels:
lowerCamelCase : Union[str, Any] = self.model.config.num_labels
if self.framework == "pt":
lowerCamelCase : Optional[Any] = model_outputs.logits.softmax(-1 )[0]
lowerCamelCase : Dict = probs.topk(_UpperCamelCase )
elif self.framework == "tf":
lowerCamelCase : List[Any] = stable_softmax(model_outputs.logits , axis=-1 )[0]
lowerCamelCase : List[Any] = tf.math.top_k(_UpperCamelCase , k=_UpperCamelCase )
lowerCamelCase : str = topk.values.numpy(), topk.indices.numpy()
else:
raise ValueError(f'Unsupported framework: {self.framework}' )
lowerCamelCase : str = scores.tolist()
lowerCamelCase : str = ids.tolist()
return [{"score": score, "label": self.model.config.idalabel[_id]} for score, _id in zip(_UpperCamelCase , _UpperCamelCase )]
| 715
|
"""simple docstring"""
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :Tuple = logging.get_logger(__name__)
__lowerCamelCase :Any = {
'vinvino02/glpn-kitti': 'https://huggingface.co/vinvino02/glpn-kitti/resolve/main/config.json',
# See all GLPN models at https://huggingface.co/models?filter=glpn
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Tuple ='''glpn'''
def __init__( self: Dict , __a: List[str]=3 , __a: Optional[int]=4 , __a: Dict=[2, 2, 2, 2] , __a: str=[8, 4, 2, 1] , __a: Optional[int]=[32, 64, 160, 256] , __a: Dict=[7, 3, 3, 3] , __a: Dict=[4, 2, 2, 2] , __a: Optional[Any]=[1, 2, 5, 8] , __a: Tuple=[4, 4, 4, 4] , __a: int="gelu" , __a: Union[str, Any]=0.0 , __a: str=0.0 , __a: Union[str, Any]=0.02 , __a: str=0.1 , __a: Union[str, Any]=1e-6 , __a: Any=64 , __a: Dict=10 , __a: Union[str, Any]=-1 , **__a: Optional[Any] , )-> Dict:
super().__init__(**__a )
lowerCamelCase : Dict = num_channels
lowerCamelCase : Any = num_encoder_blocks
lowerCamelCase : Dict = depths
lowerCamelCase : List[str] = sr_ratios
lowerCamelCase : Dict = hidden_sizes
lowerCamelCase : Tuple = patch_sizes
lowerCamelCase : Optional[int] = strides
lowerCamelCase : Optional[Any] = mlp_ratios
lowerCamelCase : Union[str, Any] = num_attention_heads
lowerCamelCase : List[str] = hidden_act
lowerCamelCase : Any = hidden_dropout_prob
lowerCamelCase : Optional[int] = attention_probs_dropout_prob
lowerCamelCase : List[Any] = initializer_range
lowerCamelCase : Dict = drop_path_rate
lowerCamelCase : Any = layer_norm_eps
lowerCamelCase : Optional[Any] = decoder_hidden_size
lowerCamelCase : Tuple = max_depth
lowerCamelCase : Optional[Any] = head_in_index
| 42
| 0
|
"""simple docstring"""
import logging
from dataclasses import dataclass, field
from pathlib import Path
from typing import Optional, Union
from .generation.configuration_utils import GenerationConfig
from .training_args import TrainingArguments
from .utils import add_start_docstrings
__lowerCamelCase :Union[str, Any] = logging.getLogger(__name__)
@dataclass
@add_start_docstrings(TrainingArguments.__doc__)
class A__ ( a__):
"""simple docstring"""
snake_case__ : Dict =field(default=a__ , metadata={'''help''': '''Whether to use SortishSampler or not.'''})
snake_case__ : Union[str, Any] =field(
default=a__ , metadata={'''help''': '''Whether to use generate to calculate generative metrics (ROUGE, BLEU).'''})
snake_case__ : Any =field(
default=a__ , metadata={
'''help''': (
'''The `max_length` to use on each evaluation loop when `predict_with_generate=True`. Will default '''
'''to the `max_length` value of the model configuration.'''
)
} , )
snake_case__ : Tuple =field(
default=a__ , metadata={
'''help''': (
'''The `num_beams` to use on each evaluation loop when `predict_with_generate=True`. Will default '''
'''to the `num_beams` value of the model configuration.'''
)
} , )
snake_case__ : List[Any] =field(
default=a__ , metadata={
'''help''': '''Model id, file path or url pointing to a GenerationConfig json file, to use during prediction.'''
} , )
def a__ ( self: Union[str, Any] )-> Optional[Any]:
lowerCamelCase : str = super().to_dict()
for k, v in d.items():
if isinstance(_A , _A ):
lowerCamelCase : Tuple = v.to_dict()
return d
| 716
|
"""simple docstring"""
from __future__ import annotations
import math
def snake_case ( UpperCamelCase__ : float , UpperCamelCase__ : int ) -> float:
lowerCamelCase : Dict = u
for i in range(1 , UpperCamelCase__ ):
lowerCamelCase : List[str] = temp * (u - i)
return temp
def snake_case ( ) -> None:
lowerCamelCase : List[Any] = int(input("""enter the numbers of values: """ ) )
lowerCamelCase : list[list[float]] = []
for _ in range(UpperCamelCase__ ):
y.append([] )
for i in range(UpperCamelCase__ ):
for j in range(UpperCamelCase__ ):
y[i].append(UpperCamelCase__ )
lowerCamelCase : Union[str, Any] = 0
print("""enter the values of parameters in a list: """ )
lowerCamelCase : Any = list(map(UpperCamelCase__ , input().split() ) )
print("""enter the values of corresponding parameters: """ )
for i in range(UpperCamelCase__ ):
lowerCamelCase : int = float(input() )
lowerCamelCase : Dict = int(input("""enter the value to interpolate: """ ) )
lowerCamelCase : List[Any] = (value - x[0]) / (x[1] - x[0])
# for calculating forward difference table
for i in range(1 , UpperCamelCase__ ):
for j in range(n - i ):
lowerCamelCase : str = y[j + 1][i - 1] - y[j][i - 1]
lowerCamelCase : Any = y[0][0]
for i in range(1 , UpperCamelCase__ ):
summ += (ucal(UpperCamelCase__ , UpperCamelCase__ ) * y[0][i]) / math.factorial(UpperCamelCase__ )
print(F'the value at {value} is {summ}' )
if __name__ == "__main__":
main()
| 42
| 0
|
"""simple docstring"""
from __future__ import annotations
import requests
__lowerCamelCase :int = set(
'approved_at_utc approved_by author_flair_background_color\nauthor_flair_css_class author_flair_richtext author_flair_template_id author_fullname\nauthor_premium can_mod_post category clicked content_categories created_utc downs\nedited gilded gildings hidden hide_score is_created_from_ads_ui is_meta\nis_original_content is_reddit_media_domain is_video link_flair_css_class\nlink_flair_richtext link_flair_text link_flair_text_color media_embed mod_reason_title\nname permalink pwls quarantine saved score secure_media secure_media_embed selftext\nsubreddit subreddit_name_prefixed subreddit_type thumbnail title top_awarded_type\ntotal_awards_received ups upvote_ratio url user_reports'.split()
)
def snake_case ( UpperCamelCase__ : Tuple , UpperCamelCase__ : Tuple = 1 , UpperCamelCase__ : Tuple = "new" , UpperCamelCase__ : Any = None ) -> Any:
lowerCamelCase : int = wanted_data or []
if invalid_search_terms := ", ".join(sorted(set(_lowerCamelCase ) - valid_terms ) ):
lowerCamelCase : Optional[Any] = F'Invalid search term: {invalid_search_terms}'
raise ValueError(_lowerCamelCase )
lowerCamelCase : List[str] = requests.get(
F'https://reddit.com/r/{subreddit}/{age}.json?limit={limit}' , headers={"""User-agent""": """A random string"""} , )
if response.status_code == 429:
raise requests.HTTPError
lowerCamelCase : Any = response.json()
if not wanted_data:
return {id_: data["data"]["children"][id_] for id_ in range(_lowerCamelCase )}
lowerCamelCase : Tuple = {}
for id_ in range(_lowerCamelCase ):
lowerCamelCase : Union[str, Any] = {
item: data["data"]["children"][id_]["data"][item] for item in wanted_data
}
return data_dict
if __name__ == "__main__":
# If you get Error 429, that means you are rate limited.Try after some time
print(get_subreddit_data('learnpython', wanted_data=['title', 'url', 'selftext']))
| 717
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_sentencepiece_available
__lowerCamelCase :str = {}
try:
if not is_sentencepiece_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[Any] = ['GPTSw3Tokenizer']
if TYPE_CHECKING:
try:
if not is_sentencepiece_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .tokenization_gpt_swa import GPTSwaTokenizer
else:
import sys
__lowerCamelCase :Tuple = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
from ..utils import DummyObject, requires_backends
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : Union[str, Any] =['''flax''']
def __init__( self: Any , *__a: int , **__a: List[Any] )-> Optional[int]:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: Optional[int] , *__a: List[Any] , **__a: str )-> Any:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: int , *__a: Optional[Any] , **__a: str )-> List[str]:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : List[str] =['''flax''']
def __init__( self: Dict , *__a: str , **__a: Tuple )-> Any:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: Any , *__a: str , **__a: Union[str, Any] )-> Optional[int]:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: Union[str, Any] , *__a: Tuple , **__a: List[Any] )-> Optional[int]:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : Optional[Any] =['''flax''']
def __init__( self: int , *__a: Dict , **__a: Optional[Any] )-> int:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: Optional[int] , *__a: Optional[int] , **__a: Any )-> int:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: Dict , *__a: Optional[int] , **__a: Union[str, Any] )-> int:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : int =['''flax''']
def __init__( self: List[str] , *__a: Dict , **__a: Tuple )-> List[str]:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: int , *__a: Optional[int] , **__a: Union[str, Any] )-> Dict:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: Union[str, Any] , *__a: Tuple , **__a: Optional[Any] )-> List[Any]:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : Tuple =['''flax''']
def __init__( self: Optional[Any] , *__a: int , **__a: List[Any] )-> Optional[int]:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: Dict , *__a: List[str] , **__a: Union[str, Any] )-> int:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: Any , *__a: Union[str, Any] , **__a: Dict )-> Optional[Any]:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : Tuple =['''flax''']
def __init__( self: Tuple , *__a: Optional[Any] , **__a: List[Any] )-> int:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: str , *__a: List[Any] , **__a: Dict )-> int:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: Optional[int] , *__a: Union[str, Any] , **__a: Tuple )-> Optional[Any]:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : Union[str, Any] =['''flax''']
def __init__( self: List[Any] , *__a: Optional[Any] , **__a: str )-> int:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: str , *__a: List[Any] , **__a: Union[str, Any] )-> Any:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: Optional[int] , *__a: Optional[Any] , **__a: Any )-> Union[str, Any]:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : int =['''flax''']
def __init__( self: Union[str, Any] , *__a: List[str] , **__a: Any )-> Dict:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: Dict , *__a: Tuple , **__a: Tuple )-> List[str]:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: Any , *__a: Optional[Any] , **__a: Optional[Any] )-> Any:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : Any =['''flax''']
def __init__( self: List[str] , *__a: Optional[int] , **__a: Tuple )-> Any:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: Union[str, Any] , *__a: Optional[int] , **__a: Optional[Any] )-> Union[str, Any]:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: Optional[Any] , *__a: List[str] , **__a: Tuple )-> List[Any]:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : Optional[Any] =['''flax''']
def __init__( self: Dict , *__a: Dict , **__a: int )-> str:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: Any , *__a: Optional[int] , **__a: Union[str, Any] )-> str:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: Optional[int] , *__a: List[Any] , **__a: Any )-> Union[str, Any]:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : int =['''flax''']
def __init__( self: Optional[Any] , *__a: Dict , **__a: List[Any] )-> List[Any]:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: List[str] , *__a: Union[str, Any] , **__a: Any )-> int:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: List[Any] , *__a: str , **__a: Any )-> List[Any]:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : Tuple =['''flax''']
def __init__( self: Optional[int] , *__a: Union[str, Any] , **__a: Optional[Any] )-> str:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: Any , *__a: Optional[Any] , **__a: Optional[int] )-> List[Any]:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: Any , *__a: List[str] , **__a: Tuple )-> Dict:
requires_backends(cls , ["""flax"""] )
class A__ ( metaclass=_UpperCAmelCase):
"""simple docstring"""
snake_case__ : Any =['''flax''']
def __init__( self: Tuple , *__a: str , **__a: Any )-> Dict:
requires_backends(self , ["""flax"""] )
@classmethod
def a__ ( cls: List[Any] , *__a: int , **__a: List[str] )-> Any:
requires_backends(cls , ["""flax"""] )
@classmethod
def a__ ( cls: List[str] , *__a: Dict , **__a: Tuple )-> Union[str, Any]:
requires_backends(cls , ["""flax"""] )
| 718
|
"""simple docstring"""
import argparse
import json
from collections import OrderedDict
from functools import partial
from pathlib import Path
import timm
import torch
from huggingface_hub import hf_hub_download
from transformers import LevitConfig, LevitForImageClassificationWithTeacher, LevitImageProcessor
from transformers.utils import logging
logging.set_verbosity_info()
__lowerCamelCase :Dict = logging.get_logger()
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : str , UpperCamelCase__ : LevitConfig , UpperCamelCase__ : Path , UpperCamelCase__ : bool = True ) -> Dict:
print(F'Converting {name}...' )
with torch.no_grad():
if hidden_sizes == 128:
if name[-1] == "S":
lowerCamelCase : Optional[Any] = timm.create_model("""levit_128s""" , pretrained=UpperCamelCase__ )
else:
lowerCamelCase : Dict = timm.create_model("""levit_128""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 192:
lowerCamelCase : Tuple = timm.create_model("""levit_192""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 256:
lowerCamelCase : Optional[int] = timm.create_model("""levit_256""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 384:
lowerCamelCase : Dict = timm.create_model("""levit_384""" , pretrained=UpperCamelCase__ )
from_model.eval()
lowerCamelCase : Optional[Any] = LevitForImageClassificationWithTeacher(UpperCamelCase__ ).eval()
lowerCamelCase : Tuple = OrderedDict()
lowerCamelCase : Optional[Any] = from_model.state_dict()
lowerCamelCase : str = list(from_model.state_dict().keys() )
lowerCamelCase : List[Any] = list(our_model.state_dict().keys() )
print(len(UpperCamelCase__ ) , len(UpperCamelCase__ ) )
for i in range(len(UpperCamelCase__ ) ):
lowerCamelCase : str = weights[og_keys[i]]
our_model.load_state_dict(UpperCamelCase__ )
lowerCamelCase : int = torch.randn((2, 3, 224, 224) )
lowerCamelCase : Any = from_model(UpperCamelCase__ )
lowerCamelCase : List[Any] = our_model(UpperCamelCase__ ).logits
assert torch.allclose(UpperCamelCase__ , UpperCamelCase__ ), "The model logits don't match the original one."
lowerCamelCase : Dict = name
print(UpperCamelCase__ )
if push_to_hub:
our_model.save_pretrained(save_directory / checkpoint_name )
lowerCamelCase : Optional[int] = LevitImageProcessor()
image_processor.save_pretrained(save_directory / checkpoint_name )
print(F'Pushed {checkpoint_name}' )
def snake_case ( UpperCamelCase__ : Path , UpperCamelCase__ : str = None , UpperCamelCase__ : bool = True ) -> Optional[int]:
lowerCamelCase : Optional[Any] = """imagenet-1k-id2label.json"""
lowerCamelCase : List[Any] = 1000
lowerCamelCase : Dict = (1, num_labels)
lowerCamelCase : List[Any] = """huggingface/label-files"""
lowerCamelCase : Optional[int] = num_labels
lowerCamelCase : List[str] = json.load(open(hf_hub_download(UpperCamelCase__ , UpperCamelCase__ , repo_type="""dataset""" ) , """r""" ) )
lowerCamelCase : Any = {int(UpperCamelCase__ ): v for k, v in idalabel.items()}
lowerCamelCase : List[Any] = idalabel
lowerCamelCase : str = {v: k for k, v in idalabel.items()}
lowerCamelCase : Tuple = partial(UpperCamelCase__ , num_labels=UpperCamelCase__ , idalabel=UpperCamelCase__ , labelaid=UpperCamelCase__ )
lowerCamelCase : Optional[int] = {
"""levit-128S""": 128,
"""levit-128""": 128,
"""levit-192""": 192,
"""levit-256""": 256,
"""levit-384""": 384,
}
lowerCamelCase : List[Any] = {
"""levit-128S""": ImageNetPreTrainedConfig(
hidden_sizes=[128, 256, 384] , num_attention_heads=[4, 6, 8] , depths=[2, 3, 4] , key_dim=[16, 16, 16] , drop_path_rate=0 , ),
"""levit-128""": ImageNetPreTrainedConfig(
hidden_sizes=[128, 256, 384] , num_attention_heads=[4, 8, 12] , depths=[4, 4, 4] , key_dim=[16, 16, 16] , drop_path_rate=0 , ),
"""levit-192""": ImageNetPreTrainedConfig(
hidden_sizes=[192, 288, 384] , num_attention_heads=[3, 5, 6] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0 , ),
"""levit-256""": ImageNetPreTrainedConfig(
hidden_sizes=[256, 384, 512] , num_attention_heads=[4, 6, 8] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0 , ),
"""levit-384""": ImageNetPreTrainedConfig(
hidden_sizes=[384, 512, 768] , num_attention_heads=[6, 9, 12] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0.1 , ),
}
if model_name:
convert_weight_and_push(
names_to_hidden_sizes[model_name] , UpperCamelCase__ , names_to_config[model_name] , UpperCamelCase__ , UpperCamelCase__ )
else:
for model_name, config in names_to_config.items():
convert_weight_and_push(names_to_hidden_sizes[model_name] , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return config, expected_shape
if __name__ == "__main__":
__lowerCamelCase :Union[str, Any] = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--model_name',
default=None,
type=str,
help='The name of the model you wish to convert, it must be one of the supported Levit* architecture,',
)
parser.add_argument(
'--pytorch_dump_folder_path',
default='levit-dump-folder/',
type=Path,
required=False,
help='Path to the output PyTorch model directory.',
)
parser.add_argument('--push_to_hub', action='store_true', help='Push model and image processor to the hub')
parser.add_argument(
'--no-push_to_hub',
dest='push_to_hub',
action='store_false',
help='Do not push model and image processor to the hub',
)
__lowerCamelCase :List[Any] = parser.parse_args()
__lowerCamelCase :Path = args.pytorch_dump_folder_path
pytorch_dump_folder_path.mkdir(exist_ok=True, parents=True)
convert_weights_and_push(pytorch_dump_folder_path, args.model_name, args.push_to_hub)
| 42
| 0
|
"""simple docstring"""
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : List[Any] , UpperCamelCase__ : Optional[Any] ) -> list:
lowerCamelCase : int = len(_A )
lowerCamelCase : Dict = [[0] * n for i in range(_A )]
for i in range(_A ):
lowerCamelCase : List[Any] = y_points[i]
for i in range(2 , _A ):
for j in range(_A , _A ):
lowerCamelCase : Dict = (
(xa - x_points[j - i + 1]) * q[j][i - 1]
- (xa - x_points[j]) * q[j - 1][i - 1]
) / (x_points[j] - x_points[j - i + 1])
return [q[n - 1][n - 1], q]
if __name__ == "__main__":
import doctest
doctest.testmod()
| 719
|
"""simple docstring"""
import torch
from diffusers import KDPMaDiscreteScheduler
from diffusers.utils import torch_device
from .test_schedulers import SchedulerCommonTest
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Tuple =(KDPMaDiscreteScheduler,)
snake_case__ : Tuple =10
def a__ ( self: List[Any] , **__a: Optional[int] )-> Union[str, Any]:
lowerCamelCase : int = {
"""num_train_timesteps""": 1_100,
"""beta_start""": 0.00_01,
"""beta_end""": 0.02,
"""beta_schedule""": """linear""",
}
config.update(**__a )
return config
def a__ ( self: Union[str, Any] )-> Any:
for timesteps in [10, 50, 100, 1_000]:
self.check_over_configs(num_train_timesteps=__a )
def a__ ( self: str )-> int:
for beta_start, beta_end in zip([0.0_00_01, 0.00_01, 0.0_01] , [0.00_02, 0.0_02, 0.02] ):
self.check_over_configs(beta_start=__a , beta_end=__a )
def a__ ( self: int )-> Union[str, Any]:
for schedule in ["linear", "scaled_linear"]:
self.check_over_configs(beta_schedule=__a )
def a__ ( self: List[Any] )-> List[Any]:
for prediction_type in ["epsilon", "v_prediction"]:
self.check_over_configs(prediction_type=__a )
def a__ ( self: Union[str, Any] )-> int:
lowerCamelCase : List[str] = self.scheduler_classes[0]
lowerCamelCase : Union[str, Any] = self.get_scheduler_config(prediction_type="""v_prediction""" )
lowerCamelCase : List[str] = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps )
lowerCamelCase : Dict = self.dummy_model()
lowerCamelCase : List[Any] = self.dummy_sample_deter * scheduler.init_noise_sigma
lowerCamelCase : List[Any] = sample.to(__a )
for i, t in enumerate(scheduler.timesteps ):
lowerCamelCase : Optional[Any] = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[int] = model(__a , __a )
lowerCamelCase : Tuple = scheduler.step(__a , __a , __a )
lowerCamelCase : Optional[Any] = output.prev_sample
lowerCamelCase : List[str] = torch.sum(torch.abs(__a ) )
lowerCamelCase : Tuple = torch.mean(torch.abs(__a ) )
if torch_device in ["cpu", "mps"]:
assert abs(result_sum.item() - 4.6_9_3_4e-0_7 ) < 1e-2
assert abs(result_mean.item() - 6.1_1_1_2e-1_0 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 4.6_9_3_4_2_8_6_5_0_1_7_0_9_7_2e-0_7 ) < 1e-2
assert abs(result_mean.item() - 0.00_02 ) < 1e-3
def a__ ( self: Any )-> Any:
if torch_device == "mps":
return
lowerCamelCase : Dict = self.scheduler_classes[0]
lowerCamelCase : Dict = self.get_scheduler_config()
lowerCamelCase : int = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps )
lowerCamelCase : List[Any] = self.dummy_model()
lowerCamelCase : Optional[Any] = self.dummy_sample_deter * scheduler.init_noise_sigma
lowerCamelCase : Optional[int] = sample.to(__a )
for i, t in enumerate(scheduler.timesteps ):
lowerCamelCase : Dict = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[Any] = model(__a , __a )
lowerCamelCase : Tuple = scheduler.step(__a , __a , __a )
lowerCamelCase : str = output.prev_sample
lowerCamelCase : Tuple = torch.sum(torch.abs(__a ) )
lowerCamelCase : Tuple = torch.mean(torch.abs(__a ) )
if torch_device in ["cpu", "mps"]:
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
def a__ ( self: Optional[Any] )-> List[Any]:
if torch_device == "mps":
return
lowerCamelCase : Any = self.scheduler_classes[0]
lowerCamelCase : Union[str, Any] = self.get_scheduler_config()
lowerCamelCase : Optional[Any] = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps , device=__a )
lowerCamelCase : Union[str, Any] = self.dummy_model()
lowerCamelCase : List[str] = self.dummy_sample_deter.to(__a ) * scheduler.init_noise_sigma
for t in scheduler.timesteps:
lowerCamelCase : Union[str, Any] = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[int] = model(__a , __a )
lowerCamelCase : int = scheduler.step(__a , __a , __a )
lowerCamelCase : int = output.prev_sample
lowerCamelCase : Union[str, Any] = torch.sum(torch.abs(__a ) )
lowerCamelCase : int = torch.mean(torch.abs(__a ) )
if str(__a ).startswith("""cpu""" ):
# The following sum varies between 148 and 156 on mps. Why?
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
| 42
| 0
|
"""simple docstring"""
def snake_case ( UpperCamelCase__ : Any = 1000000 ) -> int:
lowerCamelCase : int = [i - 1 for i in range(limit + 1 )]
for i in range(2 , limit + 1 ):
if phi[i] == i - 1:
for j in range(2 * i , limit + 1 , UpperCamelCase__ ):
phi[j] -= phi[j] // i
return sum(phi[2 : limit + 1] )
if __name__ == "__main__":
print(solution())
| 720
|
"""simple docstring"""
import gc
import random
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTextModelWithProjection, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DiffusionPipeline,
EulerDiscreteScheduler,
StableDiffusionXLImgaImgPipeline,
UNetaDConditionModel,
)
from diffusers.utils import floats_tensor, slow, torch_device
from diffusers.utils.testing_utils import enable_full_determinism, require_torch_gpu
from ..pipeline_params import (
IMAGE_TO_IMAGE_IMAGE_PARAMS,
TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS,
TEXT_GUIDED_IMAGE_VARIATION_PARAMS,
)
from ..test_pipelines_common import PipelineLatentTesterMixin, PipelineTesterMixin
enable_full_determinism()
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : str =StableDiffusionXLImgaImgPipeline
snake_case__ : Any =TEXT_GUIDED_IMAGE_VARIATION_PARAMS - {'''height''', '''width'''}
snake_case__ : Optional[int] =PipelineTesterMixin.required_optional_params - {'''latents'''}
snake_case__ : Dict =TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS
snake_case__ : Tuple =IMAGE_TO_IMAGE_IMAGE_PARAMS
snake_case__ : List[str] =IMAGE_TO_IMAGE_IMAGE_PARAMS
def a__ ( self: List[str] )-> int:
torch.manual_seed(0 )
lowerCamelCase : Any = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=2 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , attention_head_dim=(2, 4) , use_linear_projection=__a , addition_embed_type="""text_time""" , addition_time_embed_dim=8 , transformer_layers_per_block=(1, 2) , projection_class_embeddings_input_dim=80 , cross_attention_dim=64 , )
lowerCamelCase : Any = EulerDiscreteScheduler(
beta_start=0.0_00_85 , beta_end=0.0_12 , steps_offset=1 , beta_schedule="""scaled_linear""" , timestep_spacing="""leading""" , )
torch.manual_seed(0 )
lowerCamelCase : Any = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
lowerCamelCase : Optional[Any] = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-0_5 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1_000 , hidden_act="""gelu""" , projection_dim=32 , )
lowerCamelCase : Dict = CLIPTextModel(__a )
lowerCamelCase : Union[str, Any] = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" , local_files_only=__a )
lowerCamelCase : Dict = CLIPTextModelWithProjection(__a )
lowerCamelCase : Optional[int] = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" , local_files_only=__a )
lowerCamelCase : str = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""text_encoder_2""": text_encoder_a,
"""tokenizer_2""": tokenizer_a,
# "safety_checker": None,
# "feature_extractor": None,
}
return components
def a__ ( self: Any , __a: str , __a: Tuple=0 )-> Union[str, Any]:
lowerCamelCase : List[Any] = floats_tensor((1, 3, 32, 32) , rng=random.Random(__a ) ).to(__a )
lowerCamelCase : Any = image / 2 + 0.5
if str(__a ).startswith("""mps""" ):
lowerCamelCase : Dict = torch.manual_seed(__a )
else:
lowerCamelCase : Tuple = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : Tuple = {
"""prompt""": """A painting of a squirrel eating a burger""",
"""image""": image,
"""generator""": generator,
"""num_inference_steps""": 2,
"""guidance_scale""": 5.0,
"""output_type""": """numpy""",
"""strength""": 0.75,
}
return inputs
def a__ ( self: Dict )-> Optional[Any]:
lowerCamelCase : Any = """cpu""" # ensure determinism for the device-dependent torch.Generator
lowerCamelCase : Union[str, Any] = self.get_dummy_components()
lowerCamelCase : Optional[int] = StableDiffusionXLImgaImgPipeline(**__a )
lowerCamelCase : int = sd_pipe.to(__a )
sd_pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Optional[Any] = self.get_dummy_inputs(__a )
lowerCamelCase : Optional[int] = sd_pipe(**__a ).images
lowerCamelCase : Tuple = image[0, -3:, -3:, -1]
assert image.shape == (1, 32, 32, 3)
lowerCamelCase : Any = np.array([0.46_56, 0.48_40, 0.44_39, 0.66_98, 0.55_74, 0.45_24, 0.57_99, 0.59_43, 0.51_65] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-2
def a__ ( self: Optional[int] )-> Union[str, Any]:
super().test_attention_slicing_forward_pass(expected_max_diff=3e-3 )
def a__ ( self: Optional[Any] )-> str:
super().test_inference_batch_single_identical(expected_max_diff=3e-3 )
def a__ ( self: List[str] )-> Optional[Any]:
pass
def a__ ( self: List[Any] )-> Union[str, Any]:
lowerCamelCase : Tuple = self.get_dummy_components()
lowerCamelCase : Union[str, Any] = StableDiffusionXLImgaImgPipeline(**__a )
lowerCamelCase : str = sd_pipe.to(__a )
lowerCamelCase : Any = sd_pipe.to(__a )
sd_pipe.set_progress_bar_config(disable=__a )
# forward without prompt embeds
lowerCamelCase : Dict = self.get_dummy_inputs(__a )
lowerCamelCase : Any = 3 * ["""this is a negative prompt"""]
lowerCamelCase : Optional[int] = negative_prompt
lowerCamelCase : Tuple = 3 * [inputs["""prompt"""]]
lowerCamelCase : List[Any] = sd_pipe(**__a )
lowerCamelCase : Optional[int] = output.images[0, -3:, -3:, -1]
# forward with prompt embeds
lowerCamelCase : Tuple = self.get_dummy_inputs(__a )
lowerCamelCase : List[Any] = 3 * ["""this is a negative prompt"""]
lowerCamelCase : Tuple = 3 * [inputs.pop("""prompt""" )]
(
(
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) ,
) : Union[str, Any] = sd_pipe.encode_prompt(__a , negative_prompt=__a )
lowerCamelCase : int = sd_pipe(
**__a , prompt_embeds=__a , negative_prompt_embeds=__a , pooled_prompt_embeds=__a , negative_pooled_prompt_embeds=__a , )
lowerCamelCase : Union[str, Any] = output.images[0, -3:, -3:, -1]
# make sure that it's equal
assert np.abs(image_slice_a.flatten() - image_slice_a.flatten() ).max() < 1e-4
@slow
@require_torch_gpu
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: Dict )-> str:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: Union[str, Any] , __a: Any , __a: Any="cpu" , __a: str=torch.floataa , __a: Any=0 )-> Optional[Any]:
lowerCamelCase : Optional[Any] = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : List[Any] = np.random.RandomState(__a ).standard_normal((1, 4, 64, 64) )
lowerCamelCase : List[str] = torch.from_numpy(__a ).to(device=__a , dtype=__a )
lowerCamelCase : int = {
"""prompt""": """a photograph of an astronaut riding a horse""",
"""latents""": latents,
"""generator""": generator,
"""num_inference_steps""": 3,
"""guidance_scale""": 7.5,
"""output_type""": """numpy""",
}
return inputs
def a__ ( self: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = DiffusionPipeline.from_pretrained("""stabilityai/stable-diffusion-2-base""" )
pipe.to(__a )
pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Optional[int] = self.get_inputs(__a )
lowerCamelCase : Optional[Any] = pipe(**__a ).images
lowerCamelCase : Dict = image[0, -3:, -3:, -1].flatten()
assert image.shape == (1, 512, 512, 3)
lowerCamelCase : List[str] = np.array([0.4_94_93, 0.4_78_96, 0.4_07_98, 0.5_42_14, 0.5_32_12, 0.4_82_02, 0.4_76_56, 0.4_63_29, 0.4_85_06] )
assert np.abs(image_slice - expected_slice ).max() < 7e-3
| 42
| 0
|
"""simple docstring"""
from __future__ import annotations
def snake_case ( UpperCamelCase__ : list , UpperCamelCase__ : int | None = None , UpperCamelCase__ : int | None = None ) -> Union[str, Any]:
if start is None:
lowerCamelCase : int = 0
if end is None:
lowerCamelCase : int = len(__UpperCamelCase ) - 1
if start >= end:
return
lowerCamelCase : str = (start + end) // 2
slowsort(__UpperCamelCase , __UpperCamelCase , __UpperCamelCase )
slowsort(__UpperCamelCase , mid + 1 , __UpperCamelCase )
if sequence[end] < sequence[mid]:
lowerCamelCase : str = sequence[mid], sequence[end]
slowsort(__UpperCamelCase , __UpperCamelCase , end - 1 )
if __name__ == "__main__":
from doctest import testmod
testmod()
| 721
|
"""simple docstring"""
import unittest
from pathlib import Path
from tempfile import NamedTemporaryFile, TemporaryDirectory
from transformers import BertConfig, BertTokenizerFast, FeatureExtractionPipeline
from transformers.convert_graph_to_onnx import (
convert,
ensure_valid_input,
generate_identified_filename,
infer_shapes,
quantize,
)
from transformers.testing_utils import require_tf, require_tokenizers, require_torch, slow
class A__ :
"""simple docstring"""
def a__ ( self: Optional[int] , __a: Optional[int] , __a: Tuple , __a: Optional[int] )-> List[str]:
return None
class A__ :
"""simple docstring"""
def a__ ( self: Optional[int] , __a: Tuple , __a: str , __a: str , __a: str )-> Tuple:
return None
class A__ ( unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[Any] =[
# (model_name, model_kwargs)
('''bert-base-cased''', {}),
('''gpt2''', {'''use_cache''': False}), # We don't support exporting GPT2 past keys anymore
]
@require_tf
@slow
def a__ ( self: Optional[Any] )-> int:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
self._test_export(__a , """tf""" , 12 , **__a )
@require_torch
@slow
def a__ ( self: str )-> int:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
self._test_export(__a , """pt""" , 12 , **__a )
@require_torch
@slow
def a__ ( self: Union[str, Any] )-> Dict:
from transformers import BertModel
lowerCamelCase : int = ["""[UNK]""", """[SEP]""", """[CLS]""", """[PAD]""", """[MASK]""", """some""", """other""", """words"""]
with NamedTemporaryFile(mode="""w+t""" ) as vocab_file:
vocab_file.write("""\n""".join(__a ) )
vocab_file.flush()
lowerCamelCase : Dict = BertTokenizerFast(vocab_file.name )
with TemporaryDirectory() as bert_save_dir:
lowerCamelCase : List[str] = BertModel(BertConfig(vocab_size=len(__a ) ) )
model.save_pretrained(__a )
self._test_export(__a , """pt""" , 12 , __a )
@require_tf
@slow
def a__ ( self: Optional[Any] )-> Optional[int]:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
lowerCamelCase : Optional[int] = self._test_export(__a , """tf""" , 12 , **__a )
lowerCamelCase : Tuple = quantize(Path(__a ) )
# Ensure the actual quantized model is not bigger than the original one
if quantized_path.stat().st_size >= Path(__a ).stat().st_size:
self.fail("""Quantized model is bigger than initial ONNX model""" )
@require_torch
@slow
def a__ ( self: Any )-> Optional[int]:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
lowerCamelCase : Any = self._test_export(__a , """pt""" , 12 , **__a )
lowerCamelCase : Dict = quantize(__a )
# Ensure the actual quantized model is not bigger than the original one
if quantized_path.stat().st_size >= Path(__a ).stat().st_size:
self.fail("""Quantized model is bigger than initial ONNX model""" )
def a__ ( self: List[Any] , __a: Optional[Any] , __a: List[Any] , __a: Union[str, Any] , __a: Optional[Any]=None , **__a: Optional[int] )-> Any:
try:
# Compute path
with TemporaryDirectory() as tempdir:
lowerCamelCase : Optional[Any] = Path(__a ).joinpath("""model.onnx""" )
# Remove folder if exists
if path.parent.exists():
path.parent.rmdir()
# Export
convert(__a , __a , __a , __a , __a , **__a )
return path
except Exception as e:
self.fail(__a )
@require_torch
@require_tokenizers
@slow
def a__ ( self: Tuple )-> Dict:
from transformers import BertModel
lowerCamelCase : int = BertModel(BertConfig.from_pretrained("""lysandre/tiny-bert-random""" ) )
lowerCamelCase : List[Any] = BertTokenizerFast.from_pretrained("""lysandre/tiny-bert-random""" )
self._test_infer_dynamic_axis(__a , __a , """pt""" )
@require_tf
@require_tokenizers
@slow
def a__ ( self: Optional[Any] )-> List[Any]:
from transformers import TFBertModel
lowerCamelCase : Union[str, Any] = TFBertModel(BertConfig.from_pretrained("""lysandre/tiny-bert-random""" ) )
lowerCamelCase : str = BertTokenizerFast.from_pretrained("""lysandre/tiny-bert-random""" )
self._test_infer_dynamic_axis(__a , __a , """tf""" )
def a__ ( self: List[str] , __a: str , __a: Optional[Any] , __a: str )-> List[Any]:
lowerCamelCase : List[str] = FeatureExtractionPipeline(__a , __a )
lowerCamelCase : List[str] = ["""input_ids""", """token_type_ids""", """attention_mask""", """output_0""", """output_1"""]
lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase : Tuple = infer_shapes(__a , __a )
# Assert all variables are present
self.assertEqual(len(__a ) , len(__a ) )
self.assertTrue(all(var_name in shapes for var_name in variable_names ) )
self.assertSequenceEqual(variable_names[:3] , __a )
self.assertSequenceEqual(variable_names[3:] , __a )
# Assert inputs are {0: batch, 1: sequence}
for var_name in ["input_ids", "token_type_ids", "attention_mask"]:
self.assertDictEqual(shapes[var_name] , {0: """batch""", 1: """sequence"""} )
# Assert outputs are {0: batch, 1: sequence} and {0: batch}
self.assertDictEqual(shapes["""output_0"""] , {0: """batch""", 1: """sequence"""} )
self.assertDictEqual(shapes["""output_1"""] , {0: """batch"""} )
def a__ ( self: List[Any] )-> int:
lowerCamelCase : List[str] = ["""input_ids""", """attention_mask""", """token_type_ids"""]
lowerCamelCase : str = {"""input_ids""": [1, 2, 3, 4], """attention_mask""": [0, 0, 0, 0], """token_type_ids""": [1, 1, 1, 1]}
lowerCamelCase , lowerCamelCase : List[Any] = ensure_valid_input(FuncContiguousArgs() , __a , __a )
# Should have exactly the same number of args (all are valid)
self.assertEqual(len(__a ) , 3 )
# Should have exactly the same input names
self.assertEqual(set(__a ) , set(__a ) )
# Parameter should be reordered according to their respective place in the function:
# (input_ids, token_type_ids, attention_mask)
self.assertEqual(__a , (tokens["""input_ids"""], tokens["""token_type_ids"""], tokens["""attention_mask"""]) )
# Generated args are interleaved with another args (for instance parameter "past" in GPT2)
lowerCamelCase , lowerCamelCase : List[Any] = ensure_valid_input(FuncNonContiguousArgs() , __a , __a )
# Should have exactly the one arg (all before the one not provided "some_other_args")
self.assertEqual(len(__a ) , 1 )
self.assertEqual(len(__a ) , 1 )
# Should have only "input_ids"
self.assertEqual(inputs_args[0] , tokens["""input_ids"""] )
self.assertEqual(ordered_input_names[0] , """input_ids""" )
def a__ ( self: Tuple )-> Tuple:
lowerCamelCase : Optional[int] = generate_identified_filename(Path("""/home/something/my_fake_model.onnx""" ) , """-test""" )
self.assertEqual("""/home/something/my_fake_model-test.onnx""" , generated.as_posix() )
| 42
| 0
|
"""simple docstring"""
from __future__ import annotations
import copy
import inspect
import json
import math
import os
import tempfile
import unittest
from importlib import import_module
import numpy as np
from transformers import ViTMAEConfig
from transformers.file_utils import cached_property, is_tf_available, is_vision_available
from transformers.testing_utils import require_tf, require_vision, slow
from ...test_configuration_common import ConfigTester
from ...test_modeling_tf_common import TFModelTesterMixin, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_tf_available():
import tensorflow as tf
from transformers import TFViTMAEForPreTraining, TFViTMAEModel
if is_vision_available():
from PIL import Image
from transformers import ViTImageProcessor
class A__ :
"""simple docstring"""
def __init__( self: Dict , __a: Union[str, Any] , __a: List[Any]=13 , __a: Optional[Any]=30 , __a: List[Any]=2 , __a: Dict=3 , __a: str=True , __a: Any=True , __a: List[str]=32 , __a: Optional[Any]=2 , __a: int=4 , __a: Optional[Any]=37 , __a: Union[str, Any]="gelu" , __a: str=0.1 , __a: str=0.1 , __a: Tuple=10 , __a: Tuple=0.02 , __a: Optional[int]=3 , __a: Optional[Any]=0.6 , __a: Dict=None , )-> List[str]:
lowerCamelCase : Any = parent
lowerCamelCase : Optional[int] = batch_size
lowerCamelCase : Dict = image_size
lowerCamelCase : List[str] = patch_size
lowerCamelCase : int = num_channels
lowerCamelCase : str = is_training
lowerCamelCase : Union[str, Any] = use_labels
lowerCamelCase : Union[str, Any] = hidden_size
lowerCamelCase : int = num_hidden_layers
lowerCamelCase : List[str] = num_attention_heads
lowerCamelCase : Optional[Any] = intermediate_size
lowerCamelCase : Any = hidden_act
lowerCamelCase : Tuple = hidden_dropout_prob
lowerCamelCase : Optional[Any] = attention_probs_dropout_prob
lowerCamelCase : List[Any] = type_sequence_label_size
lowerCamelCase : Tuple = initializer_range
lowerCamelCase : Dict = mask_ratio
lowerCamelCase : int = scope
# in ViTMAE, the expected sequence length = (num_patches + 1) * (1 - config.mask_ratio), rounded above
# (we add 1 for the [CLS] token)
lowerCamelCase : Union[str, Any] = (image_size // patch_size) ** 2
lowerCamelCase : int = int(math.ceil((1 - mask_ratio) * (num_patches + 1) ) )
def a__ ( self: int )-> str:
lowerCamelCase : Any = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : Union[str, Any] = None
if self.use_labels:
lowerCamelCase : List[Any] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : Union[str, Any] = self.get_config()
return config, pixel_values, labels
def a__ ( self: List[str] )-> List[str]:
return ViTMAEConfig(
image_size=self.image_size , patch_size=self.patch_size , num_channels=self.num_channels , hidden_size=self.hidden_size , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , decoder_hidden_size=self.hidden_size , decoder_num_hidden_layers=self.num_hidden_layers , decoder_num_attention_heads=self.num_attention_heads , decoder_intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , is_decoder=A_ , initializer_range=self.initializer_range , mask_ratio=self.mask_ratio , )
def a__ ( self: List[str] , __a: int , __a: str , __a: Union[str, Any] )-> Tuple:
lowerCamelCase : List[str] = TFViTMAEModel(config=A_ )
lowerCamelCase : Any = model(A_ , training=A_ )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) )
def a__ ( self: Optional[Any] , __a: Dict , __a: Union[str, Any] , __a: List[str] )-> Optional[int]:
lowerCamelCase : List[str] = TFViTMAEForPreTraining(A_ )
lowerCamelCase : int = model(A_ , training=A_ )
# expected sequence length = num_patches
lowerCamelCase : str = (self.image_size // self.patch_size) ** 2
lowerCamelCase : Tuple = self.patch_size**2 * self.num_channels
self.parent.assertEqual(result.logits.shape , (self.batch_size, num_patches, expected_num_channels) )
# test greyscale images
lowerCamelCase : List[str] = 1
lowerCamelCase : List[str] = TFViTMAEForPreTraining(A_ )
lowerCamelCase : Any = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Union[str, Any] = model(A_ , training=A_ )
lowerCamelCase : Union[str, Any] = self.patch_size**2
self.parent.assertEqual(result.logits.shape , (self.batch_size, num_patches, expected_num_channels) )
def a__ ( self: Dict )-> int:
lowerCamelCase : List[str] = self.prepare_config_and_inputs()
((lowerCamelCase) , (lowerCamelCase) , (lowerCamelCase)) : List[Any] = config_and_inputs
lowerCamelCase : Union[str, Any] = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_tf
class A__ ( _lowercase , _lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : int =(TFViTMAEModel, TFViTMAEForPreTraining) if is_tf_available() else ()
snake_case__ : str ={'''feature-extraction''': TFViTMAEModel} if is_tf_available() else {}
snake_case__ : Any =False
snake_case__ : Any =False
snake_case__ : str =False
snake_case__ : str =False
def a__ ( self: List[Any] )-> Optional[int]:
lowerCamelCase : List[str] = TFViTMAEModelTester(self )
lowerCamelCase : Optional[int] = ConfigTester(self , config_class=A_ , has_text_modality=A_ , hidden_size=37 )
def a__ ( self: int )-> Optional[int]:
self.config_tester.run_common_tests()
@unittest.skip(reason="""ViTMAE does not use inputs_embeds""" )
def a__ ( self: Any )-> Any:
pass
def a__ ( self: List[str] )-> List[str]:
lowerCamelCase , lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : str = model_class(A_ )
self.assertIsInstance(model.get_input_embeddings() , (tf.keras.layers.Layer) )
lowerCamelCase : int = model.get_output_embeddings()
self.assertTrue(x is None or isinstance(A_ , tf.keras.layers.Layer ) )
def a__ ( self: Dict )-> Optional[int]:
lowerCamelCase , lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : str = model_class(A_ )
lowerCamelCase : Optional[int] = inspect.signature(model.call )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Optional[int] = [*signature.parameters.keys()]
lowerCamelCase : Tuple = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , A_ )
def a__ ( self: List[Any] )-> List[str]:
lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*A_ )
def a__ ( self: Optional[Any] )-> int:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_pretraining(*A_ )
def a__ ( self: List[str] )-> Optional[Any]:
# make the mask reproducible
np.random.seed(2 )
lowerCamelCase , lowerCamelCase : Optional[Any] = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : str = int((config.image_size // config.patch_size) ** 2 )
lowerCamelCase : List[Any] = np.random.uniform(size=(self.model_tester.batch_size, num_patches) )
for model_class in self.all_model_classes:
lowerCamelCase : str = model_class(A_ )
lowerCamelCase : Dict = self._prepare_for_class(A_ , A_ )
lowerCamelCase : Optional[int] = model(A_ , noise=A_ )
lowerCamelCase : int = copy.deepcopy(self._prepare_for_class(A_ , A_ ) )
lowerCamelCase : str = model(**A_ , noise=A_ )
lowerCamelCase : Optional[int] = outputs_dict[0].numpy()
lowerCamelCase : Any = outputs_keywords[0].numpy()
self.assertLess(np.sum(np.abs(output_dict - output_keywords ) ) , 1e-6 )
def a__ ( self: List[Any] )-> Dict:
# make the mask reproducible
np.random.seed(2 )
lowerCamelCase , lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Optional[Any] = int((config.image_size // config.patch_size) ** 2 )
lowerCamelCase : str = np.random.uniform(size=(self.model_tester.batch_size, num_patches) )
def prepare_numpy_arrays(__a: Union[str, Any] ):
lowerCamelCase : Union[str, Any] = {}
for k, v in inputs_dict.items():
if tf.is_tensor(A_ ):
lowerCamelCase : Tuple = v.numpy()
else:
lowerCamelCase : int = np.array(A_ )
return inputs_np_dict
for model_class in self.all_model_classes:
lowerCamelCase : List[str] = model_class(A_ )
lowerCamelCase : Dict = self._prepare_for_class(A_ , A_ )
lowerCamelCase : Dict = prepare_numpy_arrays(A_ )
lowerCamelCase : Any = model(A_ , noise=A_ )
lowerCamelCase : int = model(**A_ , noise=A_ )
self.assert_outputs_same(A_ , A_ )
def a__ ( self: Tuple , __a: Dict , __a: Union[str, Any] , __a: Optional[Any] )-> Dict:
# make masks reproducible
np.random.seed(2 )
lowerCamelCase : Optional[Any] = int((tf_model.config.image_size // tf_model.config.patch_size) ** 2 )
lowerCamelCase : Any = np.random.uniform(size=(self.model_tester.batch_size, num_patches) )
lowerCamelCase : Dict = tf.constant(A_ )
# Add `noise` argument.
# PT inputs will be prepared in `super().check_pt_tf_models()` with this added `noise` argument
lowerCamelCase : List[str] = tf_noise
super().check_pt_tf_models(A_ , A_ , A_ )
def a__ ( self: Optional[int] )-> Any:
# make mask reproducible
np.random.seed(2 )
lowerCamelCase , lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Tuple = {
module_member
for model_class in self.all_model_classes
for module in (import_module(model_class.__module__ ),)
for module_member_name in dir(A_ )
if module_member_name.endswith("""MainLayer""" )
# This condition is required, since `modeling_tf_clip.py` has 3 classes whose names end with `MainLayer`.
and module_member_name[: -len("""MainLayer""" )] == model_class.__name__[: -len("""Model""" )]
for module_member in (getattr(A_ , A_ ),)
if isinstance(A_ , A_ )
and tf.keras.layers.Layer in module_member.__bases__
and getattr(A_ , """_keras_serializable""" , A_ )
}
lowerCamelCase : Optional[Any] = int((config.image_size // config.patch_size) ** 2 )
lowerCamelCase : str = np.random.uniform(size=(self.model_tester.batch_size, num_patches) )
lowerCamelCase : Optional[Any] = tf.convert_to_tensor(A_ )
inputs_dict.update({"""noise""": noise} )
for main_layer_class in tf_main_layer_classes:
lowerCamelCase : List[str] = main_layer_class(A_ )
lowerCamelCase : Union[str, Any] = {
name: tf.keras.Input(tensor.shape[1:] , dtype=tensor.dtype ) for name, tensor in inputs_dict.items()
}
lowerCamelCase : List[str] = tf.keras.Model(A_ , outputs=main_layer(A_ ) )
lowerCamelCase : Optional[int] = model(A_ )
with tempfile.TemporaryDirectory() as tmpdirname:
lowerCamelCase : Any = os.path.join(A_ , """keras_model.h5""" )
model.save(A_ )
lowerCamelCase : str = tf.keras.models.load_model(
A_ , custom_objects={main_layer_class.__name__: main_layer_class} )
assert isinstance(A_ , tf.keras.Model )
lowerCamelCase : Any = model(A_ )
self.assert_outputs_same(A_ , A_ )
@slow
def a__ ( self: Any )-> str:
# make mask reproducible
np.random.seed(2 )
lowerCamelCase , lowerCamelCase : Optional[int] = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Tuple = int((config.image_size // config.patch_size) ** 2 )
lowerCamelCase : Optional[Any] = np.random.uniform(size=(self.model_tester.batch_size, num_patches) )
for model_class in self.all_model_classes:
lowerCamelCase : Optional[int] = model_class(A_ )
lowerCamelCase : List[str] = self._prepare_for_class(A_ , A_ )
lowerCamelCase : str = model(A_ , noise=A_ )
if model_class.__name__ == "TFViTMAEModel":
lowerCamelCase : Any = outputs.last_hidden_state.numpy()
lowerCamelCase : List[Any] = 0
else:
lowerCamelCase : Any = outputs.logits.numpy()
lowerCamelCase : List[Any] = 0
with tempfile.TemporaryDirectory() as tmpdirname:
model.save_pretrained(A_ , saved_model=A_ )
lowerCamelCase : Optional[int] = model_class.from_pretrained(A_ )
lowerCamelCase : Tuple = model(A_ , noise=A_ )
if model_class.__name__ == "TFViTMAEModel":
lowerCamelCase : List[Any] = after_outputs["""last_hidden_state"""].numpy()
lowerCamelCase : Dict = 0
else:
lowerCamelCase : Tuple = after_outputs["""logits"""].numpy()
lowerCamelCase : int = 0
lowerCamelCase : Any = np.amax(np.abs(out_a - out_a ) )
self.assertLessEqual(A_ , 1e-5 )
def a__ ( self: str )-> Any:
# make mask reproducible
np.random.seed(2 )
lowerCamelCase , lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Tuple = int((config.image_size // config.patch_size) ** 2 )
lowerCamelCase : List[Any] = np.random.uniform(size=(self.model_tester.batch_size, num_patches) )
for model_class in self.all_model_classes:
lowerCamelCase : List[Any] = model_class(A_ )
lowerCamelCase : Any = self._prepare_for_class(A_ , A_ )
lowerCamelCase : Dict = model(A_ , noise=A_ )
lowerCamelCase : Any = model.get_config()
# make sure that returned config is jsonifiable, which is required by keras
json.dumps(A_ )
lowerCamelCase : Optional[int] = model_class.from_config(model.get_config() )
# make sure it also accepts a normal config
lowerCamelCase : List[Any] = model_class.from_config(model.config )
lowerCamelCase : Optional[int] = new_model(A_ ) # Build model
new_model.set_weights(model.get_weights() )
lowerCamelCase : Any = new_model(A_ , noise=A_ )
self.assert_outputs_same(A_ , A_ )
@unittest.skip(
reason="""ViTMAE returns a random mask + ids_restore in each forward pass. See test_save_load\n to get deterministic results.""" )
def a__ ( self: Optional[int] )-> Optional[Any]:
pass
@unittest.skip(reason="""ViTMAE returns a random mask + ids_restore in each forward pass. See test_save_load""" )
def a__ ( self: List[Any] )-> Dict:
pass
@slow
def a__ ( self: Union[str, Any] )-> List[str]:
lowerCamelCase : Dict = TFViTMAEModel.from_pretrained("""google/vit-base-patch16-224""" )
self.assertIsNotNone(A_ )
def snake_case ( ) -> str:
lowerCamelCase : Tuple = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
return image
@require_tf
@require_vision
class A__ ( unittest.TestCase):
"""simple docstring"""
@cached_property
def a__ ( self: Dict )-> Union[str, Any]:
return ViTImageProcessor.from_pretrained("""facebook/vit-mae-base""" ) if is_vision_available() else None
@slow
def a__ ( self: Union[str, Any] )-> Optional[int]:
# make random mask reproducible across the PT and TF model
np.random.seed(2 )
lowerCamelCase : int = TFViTMAEForPreTraining.from_pretrained("""facebook/vit-mae-base""" )
lowerCamelCase : int = self.default_image_processor
lowerCamelCase : Optional[Any] = prepare_img()
lowerCamelCase : Any = image_processor(images=A_ , return_tensors="""tf""" )
# prepare a noise vector that will be also used for testing the TF model
# (this way we can ensure that the PT and TF models operate on the same inputs)
lowerCamelCase : Optional[Any] = ViTMAEConfig()
lowerCamelCase : int = int((vit_mae_config.image_size // vit_mae_config.patch_size) ** 2 )
lowerCamelCase : Union[str, Any] = np.random.uniform(size=(1, num_patches) )
# forward pass
lowerCamelCase : int = model(**A_ , noise=A_ )
# verify the logits
lowerCamelCase : Optional[Any] = tf.convert_to_tensor([1, 196, 768] )
self.assertEqual(outputs.logits.shape , A_ )
lowerCamelCase : Dict = tf.convert_to_tensor(
[[-0.05_48, -1.70_23, -0.93_25], [0.37_21, -0.56_70, -0.22_33], [0.82_35, -1.38_78, -0.35_24]] )
tf.debugging.assert_near(outputs.logits[0, :3, :3] , A_ , atol=1e-4 )
| 700
|
"""simple docstring"""
import unittest
from knapsack import greedy_knapsack as kp
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: Optional[int] )-> Union[str, Any]:
lowerCamelCase : Tuple = [10, 20, 30, 40, 50, 60]
lowerCamelCase : Union[str, Any] = [2, 4, 6, 8, 10, 12]
lowerCamelCase : Union[str, Any] = 100
self.assertEqual(kp.calc_profit(__a , __a , __a ) , 210 )
def a__ ( self: str )-> str:
self.assertRaisesRegex(__a , """max_weight must greater than zero.""" )
def a__ ( self: str )-> List[Any]:
self.assertRaisesRegex(__a , """Weight can not be negative.""" )
def a__ ( self: Any )-> Dict:
self.assertRaisesRegex(__a , """Profit can not be negative.""" )
def a__ ( self: Optional[Any] )-> List[Any]:
self.assertRaisesRegex(__a , """max_weight must greater than zero.""" )
def a__ ( self: Optional[Any] )-> Tuple:
self.assertRaisesRegex(
__a , """The length of profit and weight must be same.""" )
if __name__ == "__main__":
unittest.main()
| 42
| 0
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ....utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available
__lowerCamelCase :Any = {
'configuration_mctct': ['MCTCT_PRETRAINED_CONFIG_ARCHIVE_MAP', 'MCTCTConfig'],
'feature_extraction_mctct': ['MCTCTFeatureExtractor'],
'processing_mctct': ['MCTCTProcessor'],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :str = [
'MCTCT_PRETRAINED_MODEL_ARCHIVE_LIST',
'MCTCTForCTC',
'MCTCTModel',
'MCTCTPreTrainedModel',
]
if TYPE_CHECKING:
from .configuration_mctct import MCTCT_PRETRAINED_CONFIG_ARCHIVE_MAP, MCTCTConfig
from .feature_extraction_mctct import MCTCTFeatureExtractor
from .processing_mctct import MCTCTProcessor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_mctct import MCTCT_PRETRAINED_MODEL_ARCHIVE_LIST, MCTCTForCTC, MCTCTModel, MCTCTPreTrainedModel
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 701
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_flax_available,
is_tf_available,
is_tokenizers_available,
is_torch_available,
is_vision_available,
)
__lowerCamelCase :List[str] = {
'configuration_owlvit': [
'OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP',
'OwlViTConfig',
'OwlViTOnnxConfig',
'OwlViTTextConfig',
'OwlViTVisionConfig',
],
'processing_owlvit': ['OwlViTProcessor'],
}
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[int] = ['OwlViTFeatureExtractor']
__lowerCamelCase :List[str] = ['OwlViTImageProcessor']
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[Any] = [
'OWLVIT_PRETRAINED_MODEL_ARCHIVE_LIST',
'OwlViTModel',
'OwlViTPreTrainedModel',
'OwlViTTextModel',
'OwlViTVisionModel',
'OwlViTForObjectDetection',
]
if TYPE_CHECKING:
from .configuration_owlvit import (
OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP,
OwlViTConfig,
OwlViTOnnxConfig,
OwlViTTextConfig,
OwlViTVisionConfig,
)
from .processing_owlvit import OwlViTProcessor
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .feature_extraction_owlvit import OwlViTFeatureExtractor
from .image_processing_owlvit import OwlViTImageProcessor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_owlvit import (
OWLVIT_PRETRAINED_MODEL_ARCHIVE_LIST,
OwlViTForObjectDetection,
OwlViTModel,
OwlViTPreTrainedModel,
OwlViTTextModel,
OwlViTVisionModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
import inspect
import unittest
from transformers import SegformerConfig, is_torch_available, is_vision_available
from transformers.models.auto import get_values
from transformers.testing_utils import require_torch, slow, torch_device
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import (
MODEL_MAPPING,
SegformerForImageClassification,
SegformerForSemanticSegmentation,
SegformerModel,
)
from transformers.models.segformer.modeling_segformer import SEGFORMER_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import SegformerImageProcessor
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Tuple )-> int:
lowerCamelCase : Union[str, Any] = self.config_class(**self.inputs_dict )
self.parent.assertTrue(hasattr(__a , """hidden_sizes""" ) )
self.parent.assertTrue(hasattr(__a , """num_attention_heads""" ) )
self.parent.assertTrue(hasattr(__a , """num_encoder_blocks""" ) )
class A__ :
"""simple docstring"""
def __init__( self: Any , __a: Union[str, Any] , __a: Any=13 , __a: Dict=64 , __a: Tuple=3 , __a: List[str]=4 , __a: str=[2, 2, 2, 2] , __a: int=[8, 4, 2, 1] , __a: Optional[Any]=[16, 32, 64, 128] , __a: Optional[int]=[1, 4, 8, 16] , __a: List[Any]=[1, 2, 4, 8] , __a: int=True , __a: Optional[int]=True , __a: List[str]="gelu" , __a: List[str]=0.1 , __a: List[str]=0.1 , __a: List[Any]=0.02 , __a: List[Any]=3 , __a: Optional[Any]=None , )-> List[str]:
lowerCamelCase : Dict = parent
lowerCamelCase : List[str] = batch_size
lowerCamelCase : int = image_size
lowerCamelCase : Any = num_channels
lowerCamelCase : str = num_encoder_blocks
lowerCamelCase : int = sr_ratios
lowerCamelCase : Union[str, Any] = depths
lowerCamelCase : List[str] = hidden_sizes
lowerCamelCase : str = downsampling_rates
lowerCamelCase : str = num_attention_heads
lowerCamelCase : int = is_training
lowerCamelCase : Tuple = use_labels
lowerCamelCase : Optional[int] = hidden_act
lowerCamelCase : int = hidden_dropout_prob
lowerCamelCase : Union[str, Any] = attention_probs_dropout_prob
lowerCamelCase : Tuple = initializer_range
lowerCamelCase : Union[str, Any] = num_labels
lowerCamelCase : Tuple = scope
def a__ ( self: Union[str, Any] )-> Union[str, Any]:
lowerCamelCase : List[Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : Dict = None
if self.use_labels:
lowerCamelCase : List[Any] = ids_tensor([self.batch_size, self.image_size, self.image_size] , self.num_labels )
lowerCamelCase : Union[str, Any] = self.get_config()
return config, pixel_values, labels
def a__ ( self: int )-> List[Any]:
return SegformerConfig(
image_size=self.image_size , num_channels=self.num_channels , num_encoder_blocks=self.num_encoder_blocks , depths=self.depths , hidden_sizes=self.hidden_sizes , num_attention_heads=self.num_attention_heads , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , initializer_range=self.initializer_range , )
def a__ ( self: str , __a: Any , __a: Dict , __a: Union[str, Any] )-> str:
lowerCamelCase : Union[str, Any] = SegformerModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Optional[Any] = model(__a )
lowerCamelCase : Optional[int] = self.image_size // (self.downsampling_rates[-1] * 2)
self.parent.assertEqual(
result.last_hidden_state.shape , (self.batch_size, self.hidden_sizes[-1], expected_height, expected_width) )
def a__ ( self: str , __a: Any , __a: str , __a: Any )-> Union[str, Any]:
lowerCamelCase : Dict = self.num_labels
lowerCamelCase : Any = SegformerForSemanticSegmentation(__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a )
self.parent.assertEqual(
result.logits.shape , (self.batch_size, self.num_labels, self.image_size // 4, self.image_size // 4) )
lowerCamelCase : List[Any] = model(__a , labels=__a )
self.parent.assertEqual(
result.logits.shape , (self.batch_size, self.num_labels, self.image_size // 4, self.image_size // 4) )
self.parent.assertGreater(result.loss , 0.0 )
def a__ ( self: int , __a: Tuple , __a: Union[str, Any] , __a: List[str] )-> Tuple:
lowerCamelCase : Dict = 1
lowerCamelCase : Optional[Any] = SegformerForSemanticSegmentation(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[Any] = torch.randint(0 , 1 , (self.batch_size, self.image_size, self.image_size) ).to(__a )
lowerCamelCase : Optional[int] = model(__a , labels=__a )
self.parent.assertGreater(result.loss , 0.0 )
def a__ ( self: Optional[int] )-> Any:
lowerCamelCase : Any = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : List[Any] = config_and_inputs
lowerCamelCase : Dict = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =(
(
SegformerModel,
SegformerForSemanticSegmentation,
SegformerForImageClassification,
)
if is_torch_available()
else ()
)
snake_case__ : Dict =(
{
'''feature-extraction''': SegformerModel,
'''image-classification''': SegformerForImageClassification,
'''image-segmentation''': SegformerForSemanticSegmentation,
}
if is_torch_available()
else {}
)
snake_case__ : Dict =True
snake_case__ : Dict =False
snake_case__ : Dict =False
snake_case__ : Any =False
def a__ ( self: Union[str, Any] )-> Union[str, Any]:
lowerCamelCase : Any = SegformerModelTester(self )
lowerCamelCase : int = SegformerConfigTester(self , config_class=__a )
def a__ ( self: List[Any] )-> Optional[Any]:
self.config_tester.run_common_tests()
def a__ ( self: Dict )-> int:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: Tuple )-> Any:
lowerCamelCase : str = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_binary_image_segmentation(*__a )
def a__ ( self: Dict )-> List[str]:
lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_segmentation(*__a )
@unittest.skip("""SegFormer does not use inputs_embeds""" )
def a__ ( self: List[Any] )-> Union[str, Any]:
pass
@unittest.skip("""SegFormer does not have get_input_embeddings method and get_output_embeddings methods""" )
def a__ ( self: Union[str, Any] )-> str:
pass
def a__ ( self: Optional[Any] )-> int:
lowerCamelCase , lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : Union[str, Any] = model_class(__a )
lowerCamelCase : Optional[Any] = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : int = [*signature.parameters.keys()]
lowerCamelCase : int = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase , lowerCamelCase : Optional[int] = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : List[str] = True
for model_class in self.all_model_classes:
lowerCamelCase : str = True
lowerCamelCase : Optional[int] = False
lowerCamelCase : Union[str, Any] = True
lowerCamelCase : str = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : Tuple = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : Optional[int] = outputs.attentions
lowerCamelCase : Any = sum(self.model_tester.depths )
self.assertEqual(len(__a ) , __a )
# check that output_attentions also work using config
del inputs_dict["output_attentions"]
lowerCamelCase : str = True
lowerCamelCase : Dict = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : List[str] = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : Union[str, Any] = outputs.attentions
self.assertEqual(len(__a ) , __a )
# verify the first attentions (first block, first layer)
lowerCamelCase : Tuple = (self.model_tester.image_size // 4) ** 2
lowerCamelCase : Union[str, Any] = (self.model_tester.image_size // (4 * self.model_tester.sr_ratios[0])) ** 2
self.assertListEqual(
list(attentions[0].shape[-3:] ) , [self.model_tester.num_attention_heads[0], expected_seq_len, expected_reduced_seq_len] , )
# verify the last attentions (last block, last layer)
lowerCamelCase : Optional[Any] = (self.model_tester.image_size // 32) ** 2
lowerCamelCase : Any = (self.model_tester.image_size // (32 * self.model_tester.sr_ratios[-1])) ** 2
self.assertListEqual(
list(attentions[-1].shape[-3:] ) , [self.model_tester.num_attention_heads[-1], expected_seq_len, expected_reduced_seq_len] , )
lowerCamelCase : Union[str, Any] = len(__a )
# Check attention is always last and order is fine
lowerCamelCase : Any = True
lowerCamelCase : Any = True
lowerCamelCase : Optional[int] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : List[str] = model(**self._prepare_for_class(__a , __a ) )
self.assertEqual(out_len + 1 , len(__a ) )
lowerCamelCase : Optional[int] = outputs.attentions
self.assertEqual(len(__a ) , __a )
# verify the first attentions (first block, first layer)
lowerCamelCase : List[str] = (self.model_tester.image_size // 4) ** 2
lowerCamelCase : Dict = (self.model_tester.image_size // (4 * self.model_tester.sr_ratios[0])) ** 2
self.assertListEqual(
list(self_attentions[0].shape[-3:] ) , [self.model_tester.num_attention_heads[0], expected_seq_len, expected_reduced_seq_len] , )
def a__ ( self: List[Any] )-> int:
def check_hidden_states_output(__a: str , __a: int , __a: Union[str, Any] ):
lowerCamelCase : Union[str, Any] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : int = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : List[Any] = outputs.hidden_states
lowerCamelCase : List[Any] = self.model_tester.num_encoder_blocks
self.assertEqual(len(__a ) , __a )
# verify the first hidden states (first block)
self.assertListEqual(
list(hidden_states[0].shape[-3:] ) , [
self.model_tester.hidden_sizes[0],
self.model_tester.image_size // 4,
self.model_tester.image_size // 4,
] , )
lowerCamelCase , lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : List[str] = True
check_hidden_states_output(__a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : List[str] = True
check_hidden_states_output(__a , __a , __a )
def a__ ( self: List[Any] )-> Optional[Any]:
if not self.model_tester.is_training:
return
lowerCamelCase , lowerCamelCase : Optional[Any] = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Union[str, Any] = True
for model_class in self.all_model_classes:
if model_class in get_values(__a ):
continue
lowerCamelCase : Tuple = model_class(__a )
model.to(__a )
model.train()
lowerCamelCase : Optional[int] = self._prepare_for_class(__a , __a , return_labels=__a )
lowerCamelCase : Dict = model(**__a ).loss
loss.backward()
@unittest.skip("""Will be fixed soon by reducing the size of the model used for common tests.""" )
def a__ ( self: Tuple )-> int:
pass
@slow
def a__ ( self: Union[str, Any] )-> Union[str, Any]:
for model_name in SEGFORMER_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : List[Any] = SegformerModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def snake_case ( ) -> Any:
lowerCamelCase : Dict = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
return image
@require_torch
class A__ ( unittest.TestCase):
"""simple docstring"""
@slow
def a__ ( self: Optional[Any] )-> Union[str, Any]:
# only resize + normalize
lowerCamelCase : Dict = SegformerImageProcessor(
image_scale=(512, 512) , keep_ratio=__a , align=__a , do_random_crop=__a )
lowerCamelCase : Union[str, Any] = SegformerForSemanticSegmentation.from_pretrained("""nvidia/segformer-b0-finetuned-ade-512-512""" ).to(
__a )
lowerCamelCase : Optional[Any] = prepare_img()
lowerCamelCase : List[str] = image_processor(images=__a , return_tensors="""pt""" )
lowerCamelCase : Dict = encoded_inputs.pixel_values.to(__a )
with torch.no_grad():
lowerCamelCase : int = model(__a )
lowerCamelCase : str = torch.Size((1, model.config.num_labels, 128, 128) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : Optional[int] = torch.tensor(
[
[[-4.63_10, -5.52_32, -6.23_56], [-5.19_21, -6.14_44, -6.59_96], [-5.44_24, -6.27_90, -6.75_74]],
[[-12.13_91, -13.31_22, -13.95_54], [-12.87_32, -13.93_52, -14.35_63], [-12.94_38, -13.82_26, -14.25_13]],
[[-12.51_34, -13.46_86, -14.49_15], [-12.86_69, -14.43_43, -14.77_58], [-13.25_23, -14.58_19, -15.06_94]],
] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3, :3, :3] , __a , atol=1e-4 ) )
@slow
def a__ ( self: Union[str, Any] )-> int:
# only resize + normalize
lowerCamelCase : Tuple = SegformerImageProcessor(
image_scale=(512, 512) , keep_ratio=__a , align=__a , do_random_crop=__a )
lowerCamelCase : Any = SegformerForSemanticSegmentation.from_pretrained(
"""nvidia/segformer-b1-finetuned-cityscapes-1024-1024""" ).to(__a )
lowerCamelCase : List[Any] = prepare_img()
lowerCamelCase : Tuple = image_processor(images=__a , return_tensors="""pt""" )
lowerCamelCase : Optional[Any] = encoded_inputs.pixel_values.to(__a )
with torch.no_grad():
lowerCamelCase : int = model(__a )
lowerCamelCase : Union[str, Any] = torch.Size((1, model.config.num_labels, 128, 128) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : Optional[int] = torch.tensor(
[
[[-13.57_48, -13.91_11, -12.65_00], [-14.35_00, -15.36_83, -14.23_28], [-14.75_32, -16.04_24, -15.60_87]],
[[-17.16_51, -15.87_25, -12.96_53], [-17.25_80, -17.37_18, -14.82_23], [-16.60_58, -16.87_83, -16.74_52]],
[[-3.64_56, -3.02_09, -1.42_03], [-3.07_97, -3.19_59, -2.00_00], [-1.87_57, -1.92_17, -1.69_97]],
] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3, :3, :3] , __a , atol=1e-1 ) )
@slow
def a__ ( self: Optional[int] )-> Any:
# only resize + normalize
lowerCamelCase : List[Any] = SegformerImageProcessor(
image_scale=(512, 512) , keep_ratio=__a , align=__a , do_random_crop=__a )
lowerCamelCase : Dict = SegformerForSemanticSegmentation.from_pretrained("""nvidia/segformer-b0-finetuned-ade-512-512""" ).to(
__a )
lowerCamelCase : Dict = prepare_img()
lowerCamelCase : Any = image_processor(images=__a , return_tensors="""pt""" )
lowerCamelCase : Tuple = encoded_inputs.pixel_values.to(__a )
with torch.no_grad():
lowerCamelCase : int = model(__a )
lowerCamelCase : List[Any] = outputs.logits.detach().cpu()
lowerCamelCase : Optional[Any] = image_processor.post_process_semantic_segmentation(outputs=__a , target_sizes=[(500, 300)] )
lowerCamelCase : str = torch.Size((500, 300) )
self.assertEqual(segmentation[0].shape , __a )
lowerCamelCase : List[Any] = image_processor.post_process_semantic_segmentation(outputs=__a )
lowerCamelCase : Optional[int] = torch.Size((128, 128) )
self.assertEqual(segmentation[0].shape , __a )
| 702
|
"""simple docstring"""
import collections
import inspect
import unittest
from transformers import FocalNetConfig
from transformers.testing_utils import require_torch, require_vision, slow, torch_device
from transformers.utils import cached_property, is_torch_available, is_vision_available
from ...test_backbone_common import BackboneTesterMixin
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, _config_zero_init, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from torch import nn
from transformers import (
FocalNetBackbone,
FocalNetForImageClassification,
FocalNetForMaskedImageModeling,
FocalNetModel,
)
from transformers.models.focalnet.modeling_focalnet import FOCALNET_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import AutoImageProcessor
class A__ :
"""simple docstring"""
def __init__( self: List[Any] , __a: List[str] , __a: Optional[int]=13 , __a: List[str]=32 , __a: int=2 , __a: List[str]=3 , __a: Union[str, Any]=16 , __a: int=[32, 64, 128] , __a: Optional[Any]=[1, 2, 1] , __a: Optional[int]=[2, 2, 4] , __a: Tuple=2 , __a: Dict=2.0 , __a: List[str]=True , __a: Optional[Any]=0.0 , __a: Any=0.0 , __a: List[Any]=0.1 , __a: List[str]="gelu" , __a: Tuple=False , __a: Union[str, Any]=True , __a: Optional[int]=0.02 , __a: Tuple=1e-5 , __a: int=True , __a: List[Any]=None , __a: Optional[int]=True , __a: Dict=10 , __a: List[str]=8 , __a: Any=["stage1", "stage2"] , __a: Union[str, Any]=[1, 2] , )-> Dict:
lowerCamelCase : Dict = parent
lowerCamelCase : Optional[Any] = batch_size
lowerCamelCase : Union[str, Any] = image_size
lowerCamelCase : Optional[int] = patch_size
lowerCamelCase : Any = num_channels
lowerCamelCase : Any = embed_dim
lowerCamelCase : Dict = hidden_sizes
lowerCamelCase : List[Any] = depths
lowerCamelCase : Tuple = num_heads
lowerCamelCase : List[Any] = window_size
lowerCamelCase : str = mlp_ratio
lowerCamelCase : str = qkv_bias
lowerCamelCase : str = hidden_dropout_prob
lowerCamelCase : Dict = attention_probs_dropout_prob
lowerCamelCase : Tuple = drop_path_rate
lowerCamelCase : Dict = hidden_act
lowerCamelCase : Tuple = use_absolute_embeddings
lowerCamelCase : List[str] = patch_norm
lowerCamelCase : List[str] = layer_norm_eps
lowerCamelCase : str = initializer_range
lowerCamelCase : Tuple = is_training
lowerCamelCase : int = scope
lowerCamelCase : Union[str, Any] = use_labels
lowerCamelCase : List[str] = type_sequence_label_size
lowerCamelCase : str = encoder_stride
lowerCamelCase : List[str] = out_features
lowerCamelCase : Optional[int] = out_indices
def a__ ( self: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : str = None
if self.use_labels:
lowerCamelCase : List[str] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : str = self.get_config()
return config, pixel_values, labels
def a__ ( self: List[Any] )-> Optional[int]:
return FocalNetConfig(
image_size=self.image_size , patch_size=self.patch_size , num_channels=self.num_channels , embed_dim=self.embed_dim , hidden_sizes=self.hidden_sizes , depths=self.depths , num_heads=self.num_heads , window_size=self.window_size , mlp_ratio=self.mlp_ratio , qkv_bias=self.qkv_bias , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , drop_path_rate=self.drop_path_rate , hidden_act=self.hidden_act , use_absolute_embeddings=self.use_absolute_embeddings , path_norm=self.patch_norm , layer_norm_eps=self.layer_norm_eps , initializer_range=self.initializer_range , encoder_stride=self.encoder_stride , out_features=self.out_features , out_indices=self.out_indices , )
def a__ ( self: Tuple , __a: Optional[int] , __a: Optional[int] , __a: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = FocalNetModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Tuple = model(__a )
lowerCamelCase : Any = ((config.image_size // config.patch_size) ** 2) // (4 ** (len(config.depths ) - 1))
lowerCamelCase : List[Any] = int(config.embed_dim * 2 ** (len(config.depths ) - 1) )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, expected_seq_len, expected_dim) )
def a__ ( self: Optional[int] , __a: Dict , __a: Tuple , __a: List[Any] )-> int:
lowerCamelCase : List[Any] = FocalNetBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Optional[Any] = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , len(config.out_features ) )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.image_size, 8, 8] )
# verify channels
self.parent.assertEqual(len(model.channels ) , len(config.out_features ) )
self.parent.assertListEqual(model.channels , config.hidden_sizes[:-1] )
# verify backbone works with out_features=None
lowerCamelCase : Dict = None
lowerCamelCase : Dict = FocalNetBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , 1 )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.image_size * 2, 4, 4] )
# verify channels
self.parent.assertEqual(len(model.channels ) , 1 )
self.parent.assertListEqual(model.channels , [config.hidden_sizes[-1]] )
def a__ ( self: Optional[int] , __a: Optional[int] , __a: Optional[int] , __a: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = FocalNetForMaskedImageModeling(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[str] = model(__a )
self.parent.assertEqual(
result.reconstruction.shape , (self.batch_size, self.num_channels, self.image_size, self.image_size) )
# test greyscale images
lowerCamelCase : List[str] = 1
lowerCamelCase : Any = FocalNetForMaskedImageModeling(__a )
model.to(__a )
model.eval()
lowerCamelCase : str = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Tuple = model(__a )
self.parent.assertEqual(result.reconstruction.shape , (self.batch_size, 1, self.image_size, self.image_size) )
def a__ ( self: str , __a: Optional[Any] , __a: Optional[Any] , __a: Tuple )-> str:
lowerCamelCase : Optional[Any] = self.type_sequence_label_size
lowerCamelCase : Optional[Any] = FocalNetForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : List[str] = model(__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
# test greyscale images
lowerCamelCase : int = 1
lowerCamelCase : List[Any] = FocalNetForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Optional[Any] = model(__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
def a__ ( self: int )-> Optional[int]:
lowerCamelCase : str = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = config_and_inputs
lowerCamelCase : List[Any] = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =(
(
FocalNetModel,
FocalNetForImageClassification,
FocalNetForMaskedImageModeling,
FocalNetBackbone,
)
if is_torch_available()
else ()
)
snake_case__ : Optional[int] =(
{'''feature-extraction''': FocalNetModel, '''image-classification''': FocalNetForImageClassification}
if is_torch_available()
else {}
)
snake_case__ : Tuple =False
snake_case__ : Dict =False
snake_case__ : Dict =False
snake_case__ : Tuple =False
snake_case__ : Optional[int] =False
def a__ ( self: Union[str, Any] )-> Optional[int]:
lowerCamelCase : List[str] = FocalNetModelTester(self )
lowerCamelCase : Optional[Any] = ConfigTester(self , config_class=__a , embed_dim=37 , has_text_modality=__a )
def a__ ( self: List[str] )-> List[str]:
self.create_and_test_config_common_properties()
self.config_tester.create_and_test_config_to_json_string()
self.config_tester.create_and_test_config_to_json_file()
self.config_tester.create_and_test_config_from_and_save_pretrained()
self.config_tester.create_and_test_config_with_num_labels()
self.config_tester.check_config_can_be_init_without_params()
self.config_tester.check_config_arguments_init()
def a__ ( self: List[str] )-> Union[str, Any]:
return
def a__ ( self: Tuple )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: List[Any] )-> Dict:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_backbone(*__a )
def a__ ( self: List[Any] )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_masked_image_modeling(*__a )
def a__ ( self: List[str] )-> Dict:
lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_classification(*__a )
@unittest.skip(reason="""FocalNet does not use inputs_embeds""" )
def a__ ( self: Optional[Any] )-> str:
pass
@unittest.skip(reason="""FocalNet does not use feedforward chunking""" )
def a__ ( self: Optional[Any] )-> Dict:
pass
def a__ ( self: Optional[Any] )-> Dict:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : Any = model_class(__a )
self.assertIsInstance(model.get_input_embeddings() , (nn.Module) )
lowerCamelCase : Dict = model.get_output_embeddings()
self.assertTrue(x is None or isinstance(__a , nn.Linear ) )
def a__ ( self: Tuple )-> Optional[int]:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : int = model_class(__a )
lowerCamelCase : int = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Any = [*signature.parameters.keys()]
lowerCamelCase : List[Any] = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: str , __a: Union[str, Any] , __a: int , __a: Tuple , __a: List[str] )-> Union[str, Any]:
lowerCamelCase : List[Any] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : List[str] = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : List[str] = outputs.hidden_states
lowerCamelCase : Tuple = getattr(
self.model_tester , """expected_num_hidden_layers""" , len(self.model_tester.depths ) + 1 )
self.assertEqual(len(__a ) , __a )
# FocalNet has a different seq_length
lowerCamelCase : Tuple = (
config.patch_size
if isinstance(config.patch_size , collections.abc.Iterable )
else (config.patch_size, config.patch_size)
)
lowerCamelCase : Dict = (image_size[1] // patch_size[1]) * (image_size[0] // patch_size[0])
self.assertListEqual(
list(hidden_states[0].shape[-2:] ) , [num_patches, self.model_tester.embed_dim] , )
lowerCamelCase : Optional[Any] = outputs.reshaped_hidden_states
self.assertEqual(len(__a ) , __a )
lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = reshaped_hidden_states[0].shape
lowerCamelCase : Tuple = (
reshaped_hidden_states[0].view(__a , __a , height * width ).permute(0 , 2 , 1 )
)
self.assertListEqual(
list(reshaped_hidden_states.shape[-2:] ) , [num_patches, self.model_tester.embed_dim] , )
def a__ ( self: Any )-> Any:
lowerCamelCase , lowerCamelCase : str = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Union[str, Any] = (
self.model_tester.image_size
if isinstance(self.model_tester.image_size , collections.abc.Iterable )
else (self.model_tester.image_size, self.model_tester.image_size)
)
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : List[str] = True
self.check_hidden_states_output(__a , __a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : List[Any] = True
self.check_hidden_states_output(__a , __a , __a , __a )
def a__ ( self: str )-> Union[str, Any]:
lowerCamelCase , lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : List[str] = 3
lowerCamelCase : Any = (
self.model_tester.image_size
if isinstance(self.model_tester.image_size , collections.abc.Iterable )
else (self.model_tester.image_size, self.model_tester.image_size)
)
lowerCamelCase : Optional[int] = (
config.patch_size
if isinstance(config.patch_size , collections.abc.Iterable )
else (config.patch_size, config.patch_size)
)
lowerCamelCase : Optional[Any] = image_size[0] + patch_size[0] - (image_size[0] % patch_size[0])
lowerCamelCase : List[str] = image_size[1] + patch_size[1] - (image_size[1] % patch_size[1])
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : str = True
self.check_hidden_states_output(__a , __a , __a , (padded_height, padded_width) )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : Union[str, Any] = True
self.check_hidden_states_output(__a , __a , __a , (padded_height, padded_width) )
@slow
def a__ ( self: Optional[int] )-> List[Any]:
for model_name in FOCALNET_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : List[str] = FocalNetModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def a__ ( self: str )-> Any:
lowerCamelCase , lowerCamelCase : Optional[Any] = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : int = _config_zero_init(__a )
for model_class in self.all_model_classes:
lowerCamelCase : int = model_class(config=__a )
for name, param in model.named_parameters():
if "embeddings" not in name and param.requires_grad:
self.assertIn(
((param.data.mean() * 1e9).round() / 1e9).item() , [0.0, 1.0] , msg=f'Parameter {name} of model {model_class} seems not properly initialized' , )
@require_vision
@require_torch
class A__ ( unittest.TestCase):
"""simple docstring"""
@cached_property
def a__ ( self: Optional[int] )-> Optional[Any]:
# TODO update organization
return AutoImageProcessor.from_pretrained("""microsoft/focalnet-tiny""" ) if is_vision_available() else None
@slow
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : Tuple = FocalNetForImageClassification.from_pretrained("""microsoft/focalnet-tiny""" ).to(__a )
lowerCamelCase : Any = self.default_image_processor
lowerCamelCase : Union[str, Any] = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
lowerCamelCase : int = image_processor(images=__a , return_tensors="""pt""" ).to(__a )
# forward pass
with torch.no_grad():
lowerCamelCase : Any = model(**__a )
# verify the logits
lowerCamelCase : Tuple = torch.Size((1, 1_000) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : List[str] = torch.tensor([0.21_66, -0.43_68, 0.21_91] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3] , __a , atol=1e-4 ) )
self.assertTrue(outputs.logits.argmax(dim=-1 ).item() , 281 )
@require_torch
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : str =(FocalNetBackbone,) if is_torch_available() else ()
snake_case__ : Optional[int] =FocalNetConfig
snake_case__ : str =False
def a__ ( self: Union[str, Any] )-> Tuple:
lowerCamelCase : str = FocalNetModelTester(self )
| 42
| 0
|
"""simple docstring"""
import tempfile
import unittest
from transformers import SPIECE_UNDERLINE, BatchEncoding, PLBartTokenizer, is_torch_available
from transformers.testing_utils import (
get_tests_dir,
nested_simplify,
require_sentencepiece,
require_tokenizers,
require_torch,
)
from ...test_tokenization_common import TokenizerTesterMixin
__lowerCamelCase :Optional[Any] = get_tests_dir('fixtures/test_sentencepiece.model')
if is_torch_available():
from transformers.models.plbart.modeling_plbart import shift_tokens_right
__lowerCamelCase :str = 50_003
__lowerCamelCase :List[Any] = 50_002
@require_sentencepiece
@require_tokenizers
class A__ ( UpperCamelCase_ , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =PLBartTokenizer
snake_case__ : Any =None
snake_case__ : List[Any] =False
def a__ ( self: int )-> int:
super().setUp()
# We have a SentencePiece fixture for testing
lowerCamelCase : List[Any] = PLBartTokenizer(__A , language_codes="""base""" , keep_accents=__A )
tokenizer.save_pretrained(self.tmpdirname )
def a__ ( self: Tuple )-> Optional[int]:
lowerCamelCase : Tuple = PLBartTokenizer(__A , language_codes="""base""" , keep_accents=__A )
lowerCamelCase : Optional[int] = tokenizer.tokenize("""This is a test""" )
self.assertListEqual(__A , ["""▁This""", """▁is""", """▁a""", """▁t""", """est"""] )
self.assertListEqual(
tokenizer.convert_tokens_to_ids(__A ) , [value + tokenizer.fairseq_offset for value in [285, 46, 10, 170, 382]] , )
lowerCamelCase : Optional[Any] = tokenizer.tokenize("""I was born in 92000, and this is falsé.""" )
self.assertListEqual(
__A , [
SPIECE_UNDERLINE + """I""",
SPIECE_UNDERLINE + """was""",
SPIECE_UNDERLINE + """b""",
"""or""",
"""n""",
SPIECE_UNDERLINE + """in""",
SPIECE_UNDERLINE + """""",
"""9""",
"""2""",
"""0""",
"""0""",
"""0""",
""",""",
SPIECE_UNDERLINE + """and""",
SPIECE_UNDERLINE + """this""",
SPIECE_UNDERLINE + """is""",
SPIECE_UNDERLINE + """f""",
"""al""",
"""s""",
"""é""",
""".""",
] , )
lowerCamelCase : Optional[int] = tokenizer.convert_tokens_to_ids(__A )
self.assertListEqual(
__A , [
value + tokenizer.fairseq_offset
for value in [8, 21, 84, 55, 24, 19, 7, 2, 602, 347, 347, 347, 3, 12, 66, 46, 72, 80, 6, 2, 4]
] , )
lowerCamelCase : Union[str, Any] = tokenizer.convert_ids_to_tokens(__A )
self.assertListEqual(
__A , [
SPIECE_UNDERLINE + """I""",
SPIECE_UNDERLINE + """was""",
SPIECE_UNDERLINE + """b""",
"""or""",
"""n""",
SPIECE_UNDERLINE + """in""",
SPIECE_UNDERLINE + """""",
"""<unk>""",
"""2""",
"""0""",
"""0""",
"""0""",
""",""",
SPIECE_UNDERLINE + """and""",
SPIECE_UNDERLINE + """this""",
SPIECE_UNDERLINE + """is""",
SPIECE_UNDERLINE + """f""",
"""al""",
"""s""",
"""<unk>""",
""".""",
] , )
lowerCamelCase : int = tokenizer.vocab_size
lowerCamelCase : Tuple = [tokenizer.convert_ids_to_tokens(__A ) for x in range(end - 4 , __A )]
self.assertListEqual(__A , ["""__java__""", """__python__""", """__en_XX__""", """<mask>"""] )
lowerCamelCase : Tuple = "java.lang.Exception, python.lang.Exception, javascript, php, ruby, go"
lowerCamelCase : Dict = tokenizer(__A ).input_ids
self.assertEqual(
tokenizer.decode(__A , skip_special_tokens=__A , clean_up_tokenization_spaces=__A ) , __A , )
def a__ ( self: Dict )-> List[str]:
lowerCamelCase : Dict = PLBartTokenizer(__A , language_codes="""multi""" , keep_accents=__A )
lowerCamelCase : Optional[Any] = tokenizer.tokenize("""This is a test""" )
self.assertListEqual(__A , ["""▁This""", """▁is""", """▁a""", """▁t""", """est"""] )
self.assertListEqual(
tokenizer.convert_tokens_to_ids(__A ) , [value + tokenizer.fairseq_offset for value in [285, 46, 10, 170, 382]] , )
lowerCamelCase : Dict = tokenizer.tokenize("""I was born in 92000, and this is falsé.""" )
self.assertListEqual(
__A , [
SPIECE_UNDERLINE + """I""",
SPIECE_UNDERLINE + """was""",
SPIECE_UNDERLINE + """b""",
"""or""",
"""n""",
SPIECE_UNDERLINE + """in""",
SPIECE_UNDERLINE + """""",
"""9""",
"""2""",
"""0""",
"""0""",
"""0""",
""",""",
SPIECE_UNDERLINE + """and""",
SPIECE_UNDERLINE + """this""",
SPIECE_UNDERLINE + """is""",
SPIECE_UNDERLINE + """f""",
"""al""",
"""s""",
"""é""",
""".""",
] , )
lowerCamelCase : Dict = tokenizer.convert_tokens_to_ids(__A )
self.assertListEqual(
__A , [
value + tokenizer.fairseq_offset
for value in [8, 21, 84, 55, 24, 19, 7, 2, 602, 347, 347, 347, 3, 12, 66, 46, 72, 80, 6, 2, 4]
] , )
lowerCamelCase : Optional[Any] = tokenizer.convert_ids_to_tokens(__A )
self.assertListEqual(
__A , [
SPIECE_UNDERLINE + """I""",
SPIECE_UNDERLINE + """was""",
SPIECE_UNDERLINE + """b""",
"""or""",
"""n""",
SPIECE_UNDERLINE + """in""",
SPIECE_UNDERLINE + """""",
"""<unk>""",
"""2""",
"""0""",
"""0""",
"""0""",
""",""",
SPIECE_UNDERLINE + """and""",
SPIECE_UNDERLINE + """this""",
SPIECE_UNDERLINE + """is""",
SPIECE_UNDERLINE + """f""",
"""al""",
"""s""",
"""<unk>""",
""".""",
] , )
lowerCamelCase : List[Any] = tokenizer.vocab_size
lowerCamelCase : Union[str, Any] = [tokenizer.convert_ids_to_tokens(__A ) for x in range(end - 7 , __A )]
self.assertListEqual(
__A , ["""__java__""", """__python__""", """__en_XX__""", """__javascript__""", """__php__""", """__ruby__""", """__go__"""] )
lowerCamelCase : str = "java.lang.Exception, python.lang.Exception, javascript, php, ruby, go"
lowerCamelCase : List[Any] = tokenizer(__A ).input_ids
self.assertEqual(
tokenizer.decode(__A , skip_special_tokens=__A , clean_up_tokenization_spaces=__A ) , __A , )
@require_torch
@require_sentencepiece
@require_tokenizers
class A__ ( unittest.TestCase):
"""simple docstring"""
snake_case__ : int ='''uclanlp/plbart-python-en_XX'''
snake_case__ : List[Any] =[
'''def maximum(a,b,c):NEW_LINE_INDENTreturn max([a,b,c])''',
'''def sum(a,b,c):NEW_LINE_INDENTreturn sum([a,b,c])''',
]
snake_case__ : Tuple =[
'''Returns the maximum value of a b c.''',
'''Sums the values of a b c.''',
]
snake_case__ : Any =[
1_34,
54_52,
3_34_60,
3_34_41,
3_34_63,
3_34_65,
3_34_63,
3_34_49,
9_88,
20,
3_34_56,
19,
3_34_56,
7_71,
39,
42_58,
8_89,
33_18,
3_34_41,
3_34_63,
3_34_65,
3_34_63,
3_34_49,
24_71,
2,
PYTHON_CODE,
]
@classmethod
def a__ ( cls: Optional[int] )-> List[str]:
lowerCamelCase : PLBartTokenizer = PLBartTokenizer.from_pretrained(
cls.checkpoint_name , language_codes="""base""" , src_lang="""python""" , tgt_lang="""en_XX""" )
lowerCamelCase : Optional[Any] = 1
return cls
def a__ ( self: Optional[Any] )-> List[Any]:
self.assertEqual(self.tokenizer.fairseq_tokens_to_ids["""__java__"""] , 50_001 )
self.assertEqual(self.tokenizer.fairseq_tokens_to_ids["""__python__"""] , 50_002 )
self.assertEqual(self.tokenizer.fairseq_tokens_to_ids["""__en_XX__"""] , 50_003 )
def a__ ( self: Optional[Any] )-> List[str]:
lowerCamelCase : str = self.tokenizer.batch_encode_plus(self.src_text ).input_ids[0]
self.assertListEqual(self.expected_src_tokens , __A )
def a__ ( self: Optional[Any] )-> Tuple:
self.assertIn(__A , self.tokenizer.all_special_ids )
lowerCamelCase : List[str] = [EN_CODE, 9_037, 33_442, 57, 752, 153, 14, 56, 18, 9, 2]
lowerCamelCase : List[str] = self.tokenizer.decode(__A , skip_special_tokens=__A )
lowerCamelCase : int = self.tokenizer.decode(generated_ids[1:] , skip_special_tokens=__A )
self.assertEqual(__A , __A )
self.assertNotIn(self.tokenizer.eos_token , __A )
def a__ ( self: Optional[Any] )-> Optional[Any]:
lowerCamelCase : int = ["def sum(a,b,c):NEW_LINE_INDENTreturn sum([a,b,c])" * 20]
self.assertIsInstance(src_text[0] , __A )
lowerCamelCase : Dict = 10
lowerCamelCase : List[Any] = self.tokenizer(__A , max_length=__A , truncation=__A ).input_ids[0]
self.assertEqual(ids[-2] , 2 )
self.assertEqual(ids[-1] , __A )
self.assertEqual(len(__A ) , __A )
def a__ ( self: Optional[Any] )-> Dict:
self.assertListEqual(self.tokenizer.convert_tokens_to_ids(["""<mask>""", """__java__"""] ) , [50_004, 50_001] )
def a__ ( self: Optional[int] )-> List[Any]:
lowerCamelCase : str = tempfile.mkdtemp()
lowerCamelCase : List[Any] = self.tokenizer.fairseq_tokens_to_ids
self.tokenizer.save_pretrained(__A )
lowerCamelCase : List[Any] = PLBartTokenizer.from_pretrained(__A )
self.assertDictEqual(new_tok.fairseq_tokens_to_ids , __A )
@require_torch
def a__ ( self: Optional[Any] )-> Any:
lowerCamelCase : Any = self.tokenizer(self.src_text , text_target=self.tgt_text , padding=__A , return_tensors="""pt""" )
lowerCamelCase : int = shift_tokens_right(batch["""labels"""] , self.tokenizer.pad_token_id )
# fairseq batch: https://gist.github.com/sshleifer/cba08bc2109361a74ac3760a7e30e4f4
self.assertEqual(batch.input_ids[1][-2:].tolist() , [2, PYTHON_CODE] )
self.assertEqual(batch.decoder_input_ids[1][0] , __A )
self.assertEqual(batch.decoder_input_ids[1][-1] , 2 )
self.assertEqual(batch.labels[1][-2:].tolist() , [2, EN_CODE] )
@require_torch
def a__ ( self: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Optional[int] = self.tokenizer(
self.src_text , text_target=self.tgt_text , padding=__A , truncation=__A , max_length=len(self.expected_src_tokens ) , return_tensors="""pt""" , )
lowerCamelCase : Optional[Any] = shift_tokens_right(batch["""labels"""] , self.tokenizer.pad_token_id )
self.assertIsInstance(__A , __A )
self.assertEqual((2, 26) , batch.input_ids.shape )
self.assertEqual((2, 26) , batch.attention_mask.shape )
lowerCamelCase : Optional[int] = batch.input_ids.tolist()[0]
self.assertListEqual(self.expected_src_tokens , __A )
self.assertEqual(2 , batch.decoder_input_ids[0, -1] ) # EOS
# Test that special tokens are reset
self.assertEqual(self.tokenizer.prefix_tokens , [] )
self.assertEqual(self.tokenizer.suffix_tokens , [self.tokenizer.eos_token_id, PYTHON_CODE] )
def a__ ( self: Any )-> List[str]:
lowerCamelCase : List[Any] = self.tokenizer(self.src_text , padding=__A , truncation=__A , max_length=3 , return_tensors="""pt""" )
lowerCamelCase : str = self.tokenizer(
text_target=self.tgt_text , padding=__A , truncation=__A , max_length=10 , return_tensors="""pt""" )
lowerCamelCase : int = targets["input_ids"]
lowerCamelCase : Tuple = shift_tokens_right(__A , self.tokenizer.pad_token_id )
self.assertEqual(batch.input_ids.shape[1] , 3 )
self.assertEqual(batch.decoder_input_ids.shape[1] , 10 )
@require_torch
def a__ ( self: Tuple )-> Optional[int]:
lowerCamelCase : str = self.tokenizer._build_translation_inputs(
"""A test""" , return_tensors="""pt""" , src_lang="""en_XX""" , tgt_lang="""java""" )
self.assertEqual(
nested_simplify(__A ) , {
# A, test, EOS, en_XX
"""input_ids""": [[150, 242, 2, 50_003]],
"""attention_mask""": [[1, 1, 1, 1]],
# java
"""forced_bos_token_id""": 50_001,
} , )
| 703
|
"""simple docstring"""
import os
def snake_case ( ) -> Optional[Any]:
with open(os.path.dirname(UpperCamelCase__ ) + """/grid.txt""" ) as f:
lowerCamelCase : int = [] # noqa: E741
for _ in range(20 ):
l.append([int(UpperCamelCase__ ) for x in f.readline().split()] )
lowerCamelCase : Union[str, Any] = 0
# right
for i in range(20 ):
for j in range(17 ):
lowerCamelCase : Dict = l[i][j] * l[i][j + 1] * l[i][j + 2] * l[i][j + 3]
if temp > maximum:
lowerCamelCase : Tuple = temp
# down
for i in range(17 ):
for j in range(20 ):
lowerCamelCase : Any = l[i][j] * l[i + 1][j] * l[i + 2][j] * l[i + 3][j]
if temp > maximum:
lowerCamelCase : Optional[Any] = temp
# diagonal 1
for i in range(17 ):
for j in range(17 ):
lowerCamelCase : List[Any] = l[i][j] * l[i + 1][j + 1] * l[i + 2][j + 2] * l[i + 3][j + 3]
if temp > maximum:
lowerCamelCase : List[str] = temp
# diagonal 2
for i in range(17 ):
for j in range(3 , 20 ):
lowerCamelCase : List[str] = l[i][j] * l[i + 1][j - 1] * l[i + 2][j - 2] * l[i + 3][j - 3]
if temp > maximum:
lowerCamelCase : List[Any] = temp
return maximum
if __name__ == "__main__":
print(solution())
| 42
| 0
|
"""simple docstring"""
from random import randint
from tempfile import TemporaryFile
import numpy as np
def snake_case ( UpperCamelCase__ : Optional[int] , UpperCamelCase__ : int , UpperCamelCase__ : Union[str, Any] ) -> Optional[Any]:
lowerCamelCase : Optional[Any] = 0
if start < end:
lowerCamelCase : Union[str, Any] = randint(_lowerCamelCase , _lowerCamelCase )
lowerCamelCase : str = a[end]
lowerCamelCase : List[str] = a[pivot]
lowerCamelCase : Tuple = temp
lowerCamelCase : int = _in_place_partition(_lowerCamelCase , _lowerCamelCase , _lowerCamelCase )
count += _in_place_quick_sort(_lowerCamelCase , _lowerCamelCase , p - 1 )
count += _in_place_quick_sort(_lowerCamelCase , p + 1 , _lowerCamelCase )
return count
def snake_case ( UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : Tuple , UpperCamelCase__ : List[Any] ) -> str:
lowerCamelCase : Tuple = 0
lowerCamelCase : Optional[int] = randint(_lowerCamelCase , _lowerCamelCase )
lowerCamelCase : List[Any] = a[end]
lowerCamelCase : Optional[int] = a[pivot]
lowerCamelCase : List[Any] = temp
lowerCamelCase : Union[str, Any] = start - 1
for index in range(_lowerCamelCase , _lowerCamelCase ):
count += 1
if a[index] < a[end]: # check if current val is less than pivot value
lowerCamelCase : Optional[int] = new_pivot_index + 1
lowerCamelCase : Optional[Any] = a[new_pivot_index]
lowerCamelCase : Tuple = a[index]
lowerCamelCase : Any = temp
lowerCamelCase : List[Any] = a[new_pivot_index + 1]
lowerCamelCase : Dict = a[end]
lowerCamelCase : Optional[Any] = temp
return new_pivot_index + 1, count
__lowerCamelCase :str = TemporaryFile()
__lowerCamelCase :List[Any] = 100 # 1000 elements are to be sorted
__lowerCamelCase :Dict = 0, 1 # mean and standard deviation
__lowerCamelCase :str = np.random.normal(mu, sigma, p)
np.save(outfile, X)
print('The array is')
print(X)
outfile.seek(0) # using the same array
__lowerCamelCase :Tuple = np.load(outfile)
__lowerCamelCase :int = len(M) - 1
__lowerCamelCase :Tuple = _in_place_quick_sort(M, 0, r)
print(
'No of Comparisons for 100 elements selected from a standard normal distribution'
'is :'
)
print(z)
| 704
|
"""simple docstring"""
import gc
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DDIMScheduler,
StableDiffusionAttendAndExcitePipeline,
UNetaDConditionModel,
)
from diffusers.utils import load_numpy, skip_mps, slow
from diffusers.utils.testing_utils import require_torch_gpu
from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
from ..test_pipelines_common import PipelineKarrasSchedulerTesterMixin, PipelineLatentTesterMixin, PipelineTesterMixin
__lowerCamelCase :Any = False
@skip_mps
class A__ ( __lowercase , __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[Any] =StableDiffusionAttendAndExcitePipeline
snake_case__ : Any =False
snake_case__ : Dict =TEXT_TO_IMAGE_PARAMS
snake_case__ : Any =TEXT_TO_IMAGE_BATCH_PARAMS.union({'''token_indices'''})
snake_case__ : Dict =TEXT_TO_IMAGE_IMAGE_PARAMS
snake_case__ : str =TEXT_TO_IMAGE_IMAGE_PARAMS
@classmethod
def a__ ( cls: Dict )-> Tuple:
super().setUpClass()
torch.use_deterministic_algorithms(__a )
@classmethod
def a__ ( cls: Union[str, Any] )-> Any:
super().tearDownClass()
torch.use_deterministic_algorithms(__a )
def a__ ( self: Tuple )-> Union[str, Any]:
torch.manual_seed(0 )
lowerCamelCase : str = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=1 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , cross_attention_dim=32 , attention_head_dim=(2, 4) , use_linear_projection=__a , )
lowerCamelCase : Union[str, Any] = DDIMScheduler(
beta_start=0.0_00_85 , beta_end=0.0_12 , beta_schedule="""scaled_linear""" , clip_sample=__a , set_alpha_to_one=__a , )
torch.manual_seed(0 )
lowerCamelCase : Union[str, Any] = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
lowerCamelCase : str = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-0_5 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1_000 , hidden_act="""gelu""" , projection_dim=512 , )
lowerCamelCase : Optional[int] = CLIPTextModel(__a )
lowerCamelCase : str = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" )
lowerCamelCase : List[str] = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""safety_checker""": None,
"""feature_extractor""": None,
}
return components
def a__ ( self: Tuple , __a: int , __a: Union[str, Any]=0 )-> Optional[Any]:
if str(__a ).startswith("""mps""" ):
lowerCamelCase : Tuple = torch.manual_seed(__a )
else:
lowerCamelCase : str = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : Dict = {
"""prompt""": """a cat and a frog""",
"""token_indices""": [2, 5],
"""generator""": generator,
"""num_inference_steps""": 1,
"""guidance_scale""": 6.0,
"""output_type""": """numpy""",
"""max_iter_to_alter""": 2,
"""thresholds""": {0: 0.7},
}
return inputs
def a__ ( self: Dict )-> str:
lowerCamelCase : Tuple = """cpu"""
lowerCamelCase : List[str] = self.get_dummy_components()
lowerCamelCase : List[Any] = self.pipeline_class(**__a )
pipe.to(__a )
pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Any = self.get_dummy_inputs(__a )
lowerCamelCase : Union[str, Any] = pipe(**__a ).images
lowerCamelCase : Tuple = image[0, -3:, -3:, -1]
self.assertEqual(image.shape , (1, 64, 64, 3) )
lowerCamelCase : Optional[Any] = np.array(
[0.63_90_53_64, 0.62_89_73_07, 0.48_59_90_17, 0.5_13_36_24, 0.5_55_00_48, 0.45_76_95_16, 0.50_32_69_73, 0.5_02_31_39, 0.45_38_44_96] )
lowerCamelCase : Optional[Any] = np.abs(image_slice.flatten() - expected_slice ).max()
self.assertLessEqual(__a , 1e-3 )
def a__ ( self: int )-> Optional[Any]:
super().test_cpu_offload_forward_pass(expected_max_diff=5e-4 )
def a__ ( self: Union[str, Any] )-> Optional[int]:
# NOTE: Larger batch sizes cause this test to timeout, only test on smaller batches
self._test_inference_batch_consistent(batch_sizes=[1, 2] )
def a__ ( self: Tuple )-> int:
self._test_inference_batch_single_identical(batch_size=2 , expected_max_diff=7e-4 )
def a__ ( self: Dict )-> List[Any]:
super().test_dict_tuple_outputs_equivalent(expected_max_difference=3e-3 )
def a__ ( self: Optional[int] )-> Dict:
super().test_pt_np_pil_outputs_equivalent(expected_max_diff=5e-4 )
def a__ ( self: Any )-> Tuple:
super().test_save_load_local(expected_max_difference=5e-4 )
def a__ ( self: str )-> str:
super().test_save_load_optional_components(expected_max_difference=4e-4 )
@require_torch_gpu
@slow
class A__ ( unittest.TestCase):
"""simple docstring"""
@classmethod
def a__ ( cls: Any )-> Tuple:
super().setUpClass()
torch.use_deterministic_algorithms(__a )
@classmethod
def a__ ( cls: Dict )-> Optional[int]:
super().tearDownClass()
torch.use_deterministic_algorithms(__a )
def a__ ( self: int )-> Optional[int]:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : List[Any] = torch.manual_seed(51 )
lowerCamelCase : List[str] = StableDiffusionAttendAndExcitePipeline.from_pretrained(
"""CompVis/stable-diffusion-v1-4""" , safety_checker=__a , torch_dtype=torch.floataa )
pipe.to("""cuda""" )
lowerCamelCase : Dict = """a painting of an elephant with glasses"""
lowerCamelCase : Any = [5, 7]
lowerCamelCase : Tuple = pipe(
prompt=__a , token_indices=__a , guidance_scale=7.5 , generator=__a , num_inference_steps=5 , max_iter_to_alter=5 , output_type="""numpy""" , ).images[0]
lowerCamelCase : Union[str, Any] = load_numpy(
"""https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/attend-and-excite/elephant_glasses.npy""" )
assert np.abs((expected_image - image).max() ) < 5e-1
| 42
| 0
|
"""simple docstring"""
def snake_case ( UpperCamelCase__ : int ) -> Dict:
lowerCamelCase : int = generate_pascal_triangle(UpperCamelCase__ )
for row_idx in range(UpperCamelCase__ ):
# Print left spaces
for _ in range(num_rows - row_idx - 1 ):
print(end=""" """ )
# Print row values
for col_idx in range(row_idx + 1 ):
if col_idx != row_idx:
print(triangle[row_idx][col_idx] , end=""" """ )
else:
print(triangle[row_idx][col_idx] , end="""""" )
print()
def snake_case ( UpperCamelCase__ : int ) -> Optional[int]:
if not isinstance(UpperCamelCase__ , UpperCamelCase__ ):
raise TypeError("""The input value of 'num_rows' should be 'int'""" )
if num_rows == 0:
return []
elif num_rows < 0:
raise ValueError(
"""The input value of 'num_rows' should be greater than or equal to 0""" )
lowerCamelCase : list[list[int]] = []
for current_row_idx in range(UpperCamelCase__ ):
lowerCamelCase : Dict = populate_current_row(UpperCamelCase__ , UpperCamelCase__ )
triangle.append(UpperCamelCase__ )
return triangle
def snake_case ( UpperCamelCase__ : list[list[int]] , UpperCamelCase__ : int ) -> Union[str, Any]:
lowerCamelCase : List[Any] = [-1] * (current_row_idx + 1)
# first and last elements of current row are equal to 1
lowerCamelCase : Any = 1, 1
for current_col_idx in range(1 , UpperCamelCase__ ):
calculate_current_element(
UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return current_row
def snake_case ( UpperCamelCase__ : list[list[int]] , UpperCamelCase__ : list[int] , UpperCamelCase__ : int , UpperCamelCase__ : int , ) -> str:
lowerCamelCase : int = triangle[current_row_idx - 1][current_col_idx - 1]
lowerCamelCase : int = triangle[current_row_idx - 1][current_col_idx]
lowerCamelCase : int = above_to_left_elt + above_to_right_elt
def snake_case ( UpperCamelCase__ : int ) -> Tuple:
if not isinstance(UpperCamelCase__ , UpperCamelCase__ ):
raise TypeError("""The input value of 'num_rows' should be 'int'""" )
if num_rows == 0:
return []
elif num_rows < 0:
raise ValueError(
"""The input value of 'num_rows' should be greater than or equal to 0""" )
lowerCamelCase : list[list[int]] = [[1]]
for row_index in range(1 , UpperCamelCase__ ):
lowerCamelCase : Dict = [0] + result[-1] + [0]
lowerCamelCase : Any = row_index + 1
# Calculate the number of distinct elements in a row
lowerCamelCase : Optional[Any] = sum(divmod(UpperCamelCase__ , 2 ) )
lowerCamelCase : Tuple = [
temp_row[i - 1] + temp_row[i] for i in range(1 , distinct_elements + 1 )
]
lowerCamelCase : str = row_first_half[: (row_index + 1) // 2]
row_second_half.reverse()
lowerCamelCase : Optional[int] = row_first_half + row_second_half
result.append(UpperCamelCase__ )
return result
def snake_case ( ) -> List[Any]:
from collections.abc import Callable
from timeit import timeit
def benchmark_a_function(UpperCamelCase__ : Callable , UpperCamelCase__ : int ) -> None:
lowerCamelCase : Union[str, Any] = F'{func.__name__}({value})'
lowerCamelCase : str = timeit(F'__main__.{call}' , setup="""import __main__""" )
# print(f"{call:38} = {func(value)} -- {timing:.4f} seconds")
print(F'{call:38} -- {timing:.4f} seconds' )
for value in range(15 ): # (1, 7, 14):
for func in (generate_pascal_triangle, generate_pascal_triangle_optimized):
benchmark_a_function(UpperCamelCase__ , UpperCamelCase__ )
print()
if __name__ == "__main__":
import doctest
doctest.testmod()
benchmark()
| 705
|
"""simple docstring"""
import unittest
from transformers import EsmConfig, is_torch_available
from transformers.testing_utils import TestCasePlus, require_torch, slow, torch_device
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, ids_tensor, random_attention_mask
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import EsmForMaskedLM, EsmForSequenceClassification, EsmForTokenClassification, EsmModel
from transformers.models.esm.modeling_esm import (
ESM_PRETRAINED_MODEL_ARCHIVE_LIST,
EsmEmbeddings,
create_position_ids_from_input_ids,
)
class A__ :
"""simple docstring"""
def __init__( self: List[str] , __a: List[str] , __a: Dict=13 , __a: Tuple=7 , __a: Dict=False , __a: str=True , __a: List[Any]=False , __a: Dict=True , __a: Any=33 , __a: Optional[Any]=32 , __a: List[Any]=5 , __a: Any=4 , __a: Dict=37 , __a: str="gelu" , __a: str=0.1 , __a: int=0.1 , __a: Optional[int]=512 , __a: List[Any]=16 , __a: int=2 , __a: int=0.02 , __a: Optional[int]=3 , __a: str=4 , __a: Tuple=None , )-> Tuple:
lowerCamelCase : Union[str, Any] = parent
lowerCamelCase : Tuple = batch_size
lowerCamelCase : Any = seq_length
lowerCamelCase : Any = is_training
lowerCamelCase : Tuple = use_input_mask
lowerCamelCase : int = use_token_type_ids
lowerCamelCase : List[str] = use_labels
lowerCamelCase : Optional[int] = vocab_size
lowerCamelCase : Tuple = hidden_size
lowerCamelCase : List[str] = num_hidden_layers
lowerCamelCase : Optional[int] = num_attention_heads
lowerCamelCase : Optional[Any] = intermediate_size
lowerCamelCase : Optional[Any] = hidden_act
lowerCamelCase : Union[str, Any] = hidden_dropout_prob
lowerCamelCase : Optional[Any] = attention_probs_dropout_prob
lowerCamelCase : Any = max_position_embeddings
lowerCamelCase : str = type_vocab_size
lowerCamelCase : List[Any] = type_sequence_label_size
lowerCamelCase : Optional[Any] = initializer_range
lowerCamelCase : Union[str, Any] = num_labels
lowerCamelCase : Optional[Any] = num_choices
lowerCamelCase : Any = scope
def a__ ( self: Optional[int] )-> List[Any]:
lowerCamelCase : List[Any] = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : Dict = None
if self.use_input_mask:
lowerCamelCase : Tuple = random_attention_mask([self.batch_size, self.seq_length] )
lowerCamelCase : Any = None
lowerCamelCase : int = None
lowerCamelCase : Union[str, Any] = None
if self.use_labels:
lowerCamelCase : Optional[Any] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : Optional[int] = ids_tensor([self.batch_size, self.seq_length] , self.num_labels )
lowerCamelCase : Union[str, Any] = ids_tensor([self.batch_size] , self.num_choices )
lowerCamelCase : List[str] = self.get_config()
return config, input_ids, input_mask, sequence_labels, token_labels, choice_labels
def a__ ( self: Tuple )-> Union[str, Any]:
return EsmConfig(
vocab_size=self.vocab_size , hidden_size=self.hidden_size , pad_token_id=1 , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , initializer_range=self.initializer_range , )
def a__ ( self: List[Any] , __a: List[str] , __a: str , __a: Tuple , __a: List[str] , __a: List[str] , __a: str )-> int:
lowerCamelCase : Optional[int] = EsmModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a , attention_mask=__a )
lowerCamelCase : str = model(__a )
lowerCamelCase : Optional[Any] = model(__a )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) )
self.parent.assertEqual(result.pooler_output.shape , (self.batch_size, self.hidden_size) )
def a__ ( self: int , __a: Union[str, Any] , __a: Optional[int] , __a: List[str] , __a: str , __a: List[str] , __a: Tuple )-> int:
lowerCamelCase : str = EsmForMaskedLM(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[Any] = model(__a , attention_mask=__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) )
def a__ ( self: List[str] , __a: List[Any] , __a: List[str] , __a: int , __a: Union[str, Any] , __a: List[Any] , __a: Tuple )-> List[str]:
lowerCamelCase : Tuple = self.num_labels
lowerCamelCase : Dict = EsmForTokenClassification(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a , attention_mask=__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) )
def a__ ( self: Optional[int] )-> Optional[int]:
lowerCamelCase : Any = self.prepare_config_and_inputs()
(
(
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) ,
) : Tuple = config_and_inputs
lowerCamelCase : List[Any] = {"""input_ids""": input_ids, """attention_mask""": input_mask}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Any =False
snake_case__ : Dict =(
(
EsmForMaskedLM,
EsmModel,
EsmForSequenceClassification,
EsmForTokenClassification,
)
if is_torch_available()
else ()
)
snake_case__ : Dict =()
snake_case__ : Optional[int] =(
{
'''feature-extraction''': EsmModel,
'''fill-mask''': EsmForMaskedLM,
'''text-classification''': EsmForSequenceClassification,
'''token-classification''': EsmForTokenClassification,
'''zero-shot''': EsmForSequenceClassification,
}
if is_torch_available()
else {}
)
snake_case__ : Any =True
def a__ ( self: Optional[int] )-> Optional[int]:
lowerCamelCase : Optional[Any] = EsmModelTester(self )
lowerCamelCase : Any = ConfigTester(self , config_class=__a , hidden_size=37 )
def a__ ( self: List[Any] )-> Optional[Any]:
self.config_tester.run_common_tests()
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: Tuple )-> Any:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
for type in ["absolute", "relative_key", "relative_key_query"]:
lowerCamelCase : Tuple = type
self.model_tester.create_and_check_model(*__a )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Optional[int] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_masked_lm(*__a )
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_token_classification(*__a )
@slow
def a__ ( self: Any )-> List[Any]:
for model_name in ESM_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : int = EsmModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def a__ ( self: str )-> List[str]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()[0]
lowerCamelCase : Union[str, Any] = EsmEmbeddings(config=__a )
lowerCamelCase : List[str] = torch.as_tensor([[12, 31, 13, model.padding_idx]] )
lowerCamelCase : Union[str, Any] = torch.as_tensor(
[
[
0 + model.padding_idx + 1,
1 + model.padding_idx + 1,
2 + model.padding_idx + 1,
model.padding_idx,
]
] )
lowerCamelCase : Optional[Any] = create_position_ids_from_input_ids(__a , model.padding_idx )
self.assertEqual(position_ids.shape , expected_positions.shape )
self.assertTrue(torch.all(torch.eq(__a , __a ) ) )
def a__ ( self: Optional[int] )-> int:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()[0]
lowerCamelCase : Any = EsmEmbeddings(config=__a )
lowerCamelCase : Dict = torch.empty(2 , 4 , 30 )
lowerCamelCase : List[Any] = [
0 + embeddings.padding_idx + 1,
1 + embeddings.padding_idx + 1,
2 + embeddings.padding_idx + 1,
3 + embeddings.padding_idx + 1,
]
lowerCamelCase : Any = torch.as_tensor([expected_single_positions, expected_single_positions] )
lowerCamelCase : List[str] = embeddings.create_position_ids_from_inputs_embeds(__a )
self.assertEqual(position_ids.shape , expected_positions.shape )
self.assertTrue(torch.all(torch.eq(__a , __a ) ) )
@unittest.skip("""Esm does not support embedding resizing""" )
def a__ ( self: Any )-> Optional[Any]:
pass
@unittest.skip("""Esm does not support embedding resizing""" )
def a__ ( self: Dict )-> Dict:
pass
@unittest.skip("""Will be fixed soon by reducing the size of the model used for common tests.""" )
def a__ ( self: List[str] )-> Dict:
pass
@require_torch
class A__ ( __lowercase):
"""simple docstring"""
@slow
def a__ ( self: Any )-> Union[str, Any]:
with torch.no_grad():
lowerCamelCase : Union[str, Any] = EsmForMaskedLM.from_pretrained("""facebook/esm2_t6_8M_UR50D""" )
model.eval()
lowerCamelCase : List[str] = torch.tensor([[0, 1, 2, 3, 4, 5]] )
lowerCamelCase : Tuple = model(__a )[0]
lowerCamelCase : Dict = 33
lowerCamelCase : List[str] = torch.Size((1, 6, vocab_size) )
self.assertEqual(output.shape , __a )
lowerCamelCase : Tuple = torch.tensor(
[[[8.92_15, -10.58_98, -6.46_71], [-6.39_67, -13.91_14, -1.12_12], [-7.78_12, -13.95_16, -3.74_06]]] )
self.assertTrue(torch.allclose(output[:, :3, :3] , __a , atol=1e-4 ) )
@slow
def a__ ( self: Dict )-> str:
with torch.no_grad():
lowerCamelCase : Any = EsmModel.from_pretrained("""facebook/esm2_t6_8M_UR50D""" )
model.eval()
lowerCamelCase : Optional[Any] = torch.tensor([[0, 6, 4, 13, 5, 4, 16, 12, 11, 7, 2]] )
lowerCamelCase : Any = model(__a )[0]
# compare the actual values for a slice.
lowerCamelCase : Tuple = torch.tensor(
[[[0.14_44, 0.54_13, 0.32_48], [0.30_34, 0.00_53, 0.31_08], [0.32_28, -0.24_99, 0.34_15]]] )
self.assertTrue(torch.allclose(output[:, :3, :3] , __a , atol=1e-4 ) )
| 42
| 0
|
"""simple docstring"""
from abc import ABC, abstractmethod
from typing import List, Optional
class A__ ( __a):
"""simple docstring"""
def __init__( self: Dict )-> List[str]:
self.test()
def a__ ( self: Optional[Any] )-> Dict:
lowerCamelCase : Tuple = 0
lowerCamelCase : List[Any] = False
while not completed:
if counter == 1:
self.reset()
lowerCamelCase : List[str] = self.advance()
if not self.does_advance(a_ ):
raise Exception(
"""Custom Constraint is not defined correctly. self.does_advance(self.advance()) must be true.""" )
lowerCamelCase : Union[str, Any] = self.update(a_ )
counter += 1
if counter > 10_000:
raise Exception("""update() does not fulfill the constraint.""" )
if self.remaining() != 0:
raise Exception("""Custom Constraint is not defined correctly.""" )
@abstractmethod
def a__ ( self: Optional[int] )-> Union[str, Any]:
raise NotImplementedError(
f'{self.__class__} is an abstract class. Only classes inheriting this class can be called.' )
@abstractmethod
def a__ ( self: Optional[int] , __a: int )-> Union[str, Any]:
raise NotImplementedError(
f'{self.__class__} is an abstract class. Only classes inheriting this class can be called.' )
@abstractmethod
def a__ ( self: Optional[int] , __a: int )-> List[str]:
raise NotImplementedError(
f'{self.__class__} is an abstract class. Only classes inheriting this class can be called.' )
@abstractmethod
def a__ ( self: Tuple )-> Tuple:
raise NotImplementedError(
f'{self.__class__} is an abstract class. Only classes inheriting this class can be called.' )
@abstractmethod
def a__ ( self: int )-> int:
raise NotImplementedError(
f'{self.__class__} is an abstract class. Only classes inheriting this class can be called.' )
@abstractmethod
def a__ ( self: List[str] , __a: Tuple=False )-> str:
raise NotImplementedError(
f'{self.__class__} is an abstract class. Only classes inheriting this class can be called.' )
class A__ ( __a):
"""simple docstring"""
def __init__( self: Tuple , __a: List[int] )-> List[str]:
super(a_ , self ).__init__()
if not isinstance(a_ , a_ ) or len(a_ ) == 0:
raise ValueError(f'`token_ids` has to be a non-empty list, but is {token_ids}.' )
if any((not isinstance(a_ , a_ ) or token_id < 0) for token_id in token_ids ):
raise ValueError(f'Each list in `token_ids` has to be a list of positive integers, but is {token_ids}.' )
lowerCamelCase : Optional[int] = token_ids
lowerCamelCase : Dict = len(self.token_ids )
lowerCamelCase : Union[str, Any] = -1 # the index of the currently fulfilled step
lowerCamelCase : str = False
def a__ ( self: Optional[int] )-> List[str]:
if self.completed:
return None
return self.token_ids[self.fulfilled_idx + 1]
def a__ ( self: List[Any] , __a: int )-> Optional[Any]:
if not isinstance(a_ , a_ ):
raise ValueError(f'`token_id` has to be an `int`, but is {token_id} of type {type(a_ )}' )
if self.completed:
return False
return token_id == self.token_ids[self.fulfilled_idx + 1]
def a__ ( self: List[str] , __a: int )-> Optional[int]:
if not isinstance(a_ , a_ ):
raise ValueError(f'`token_id` has to be an `int`, but is {token_id} of type {type(a_ )}' )
lowerCamelCase : Optional[int] = False
lowerCamelCase : List[str] = False
lowerCamelCase : List[Any] = False
if self.does_advance(a_ ):
self.fulfilled_idx += 1
lowerCamelCase : List[str] = True
if self.fulfilled_idx == (self.seqlen - 1):
lowerCamelCase : str = True
lowerCamelCase : Optional[int] = completed
else:
# failed to make progress.
lowerCamelCase : Union[str, Any] = True
self.reset()
return stepped, completed, reset
def a__ ( self: Union[str, Any] )-> Optional[int]:
lowerCamelCase : int = False
lowerCamelCase : Dict = 0
def a__ ( self: Tuple )-> Optional[int]:
return self.seqlen - (self.fulfilled_idx + 1)
def a__ ( self: Union[str, Any] , __a: str=False )-> Union[str, Any]:
lowerCamelCase : Dict = PhrasalConstraint(self.token_ids )
if stateful:
lowerCamelCase : List[Any] = self.seqlen
lowerCamelCase : str = self.fulfilled_idx
lowerCamelCase : Union[str, Any] = self.completed
return new_constraint
class A__ :
"""simple docstring"""
def __init__( self: int , __a: List[List[int]] , __a: int=True )-> Dict:
lowerCamelCase : List[str] = max([len(a_ ) for one in nested_token_ids] )
lowerCamelCase : int = {}
for token_ids in nested_token_ids:
lowerCamelCase : List[str] = root
for tidx, token_id in enumerate(a_ ):
if token_id not in level:
lowerCamelCase : int = {}
lowerCamelCase : List[Any] = level[token_id]
if no_subsets and self.has_subsets(a_ , a_ ):
raise ValueError(
"""Each list in `nested_token_ids` can't be a complete subset of another list, but is"""
f' {nested_token_ids}.' )
lowerCamelCase : int = root
def a__ ( self: Tuple , __a: Tuple )-> Dict:
lowerCamelCase : Union[str, Any] = self.trie
for current_token in current_seq:
lowerCamelCase : Any = start[current_token]
lowerCamelCase : Any = list(start.keys() )
return next_tokens
def a__ ( self: List[Any] , __a: Optional[int] )-> List[Any]:
lowerCamelCase : Any = self.next_tokens(a_ )
return len(a_ ) == 0
def a__ ( self: Dict , __a: List[Any] )-> Any:
lowerCamelCase : int = list(root.values() )
if len(a_ ) == 0:
return 1
else:
return sum([self.count_leaves(a_ ) for nn in next_nodes] )
def a__ ( self: str , __a: List[Any] , __a: int )-> int:
lowerCamelCase : str = self.count_leaves(a_ )
return len(a_ ) != leaf_count
class A__ ( __a):
"""simple docstring"""
def __init__( self: Tuple , __a: List[List[int]] )-> int:
super(a_ , self ).__init__()
if not isinstance(a_ , a_ ) or len(a_ ) == 0:
raise ValueError(f'`nested_token_ids` has to be a non-empty list, but is {nested_token_ids}.' )
if any(not isinstance(a_ , a_ ) for token_ids in nested_token_ids ):
raise ValueError(f'`nested_token_ids` has to be a list of lists, but is {nested_token_ids}.' )
if any(
any((not isinstance(a_ , a_ ) or token_id < 0) for token_id in token_ids )
for token_ids in nested_token_ids ):
raise ValueError(
f'Each list in `nested_token_ids` has to be a list of positive integers, but is {nested_token_ids}.' )
lowerCamelCase : Optional[Any] = DisjunctiveTrie(a_ )
lowerCamelCase : Optional[int] = nested_token_ids
lowerCamelCase : Tuple = self.trie.max_height
lowerCamelCase : List[str] = []
lowerCamelCase : Union[str, Any] = False
def a__ ( self: Any )-> Any:
lowerCamelCase : List[Any] = self.trie.next_tokens(self.current_seq )
if len(a_ ) == 0:
return None
else:
return token_list
def a__ ( self: Tuple , __a: int )-> List[str]:
if not isinstance(a_ , a_ ):
raise ValueError(f'`token_id` is supposed to be type `int`, but is {token_id} of type {type(a_ )}' )
lowerCamelCase : Optional[Any] = self.trie.next_tokens(self.current_seq )
return token_id in next_tokens
def a__ ( self: Any , __a: int )-> List[Any]:
if not isinstance(a_ , a_ ):
raise ValueError(f'`token_id` is supposed to be type `int`, but is {token_id} of type {type(a_ )}' )
lowerCamelCase : Any = False
lowerCamelCase : Union[str, Any] = False
lowerCamelCase : Tuple = False
if self.does_advance(a_ ):
self.current_seq.append(a_ )
lowerCamelCase : Optional[Any] = True
else:
lowerCamelCase : Tuple = True
self.reset()
lowerCamelCase : Any = self.trie.reached_leaf(self.current_seq )
lowerCamelCase : Tuple = completed
return stepped, completed, reset
def a__ ( self: str )-> int:
lowerCamelCase : Any = False
lowerCamelCase : Tuple = []
def a__ ( self: str )-> Union[str, Any]:
if self.completed:
# since this can be completed without reaching max height
return 0
else:
return self.seqlen - len(self.current_seq )
def a__ ( self: Dict , __a: List[Any]=False )-> List[str]:
lowerCamelCase : Union[str, Any] = DisjunctiveConstraint(self.token_ids )
if stateful:
lowerCamelCase : Tuple = self.seqlen
lowerCamelCase : List[str] = self.current_seq
lowerCamelCase : List[Any] = self.completed
return new_constraint
class A__ :
"""simple docstring"""
def __init__( self: Any , __a: List[Constraint] )-> Optional[Any]:
lowerCamelCase : Any = constraints
# max # of steps required to fulfill a given constraint
lowerCamelCase : Optional[int] = max([c.seqlen for c in constraints] )
lowerCamelCase : Optional[Any] = len(a_ )
lowerCamelCase : List[Any] = False
self.init_state()
def a__ ( self: Optional[int] )-> Optional[Any]:
lowerCamelCase : List[Any] = []
lowerCamelCase : Union[str, Any] = None
lowerCamelCase : Optional[int] = [constraint.copy(stateful=a_ ) for constraint in self.constraints]
def a__ ( self: Union[str, Any] )-> List[str]:
lowerCamelCase : Any = 0
if self.inprogress_constraint:
# extra points for having a constraint mid-fulfilled
add += self.max_seqlen - self.inprogress_constraint.remaining()
return (len(self.complete_constraints ) * self.max_seqlen) + add
def a__ ( self: Optional[int] )-> Any:
lowerCamelCase : Union[str, Any] = []
if self.inprogress_constraint is None:
for constraint in self.pending_constraints: # "pending" == "unfulfilled yet"
lowerCamelCase : str = constraint.advance()
if isinstance(a_ , a_ ):
token_list.append(a_ )
elif isinstance(a_ , a_ ):
token_list.extend(a_ )
else:
lowerCamelCase : List[str] = self.inprogress_constraint.advance()
if isinstance(a_ , a_ ):
token_list.append(a_ )
elif isinstance(a_ , a_ ):
token_list.extend(a_ )
if len(a_ ) == 0:
return None
else:
return token_list
def a__ ( self: int , __a: Optional[List[int]] )-> List[Any]:
self.init_state()
if token_ids is not None:
for token in token_ids:
# completes or steps **one** constraint
lowerCamelCase : int = self.add(a_ )
# the entire list of constraints are fulfilled
if self.completed:
break
def a__ ( self: List[Any] , __a: int )-> int:
if not isinstance(a_ , a_ ):
raise ValueError(f'`token_id` should be an `int`, but is `{token_id}`.' )
lowerCamelCase : List[str] = False, False
if self.completed:
lowerCamelCase : Dict = True
lowerCamelCase : List[Any] = False
return complete, stepped
if self.inprogress_constraint is not None:
# In the middle of fulfilling a constraint. If the `token_id` *does* makes an incremental progress to current
# job, simply update the state
lowerCamelCase : Optional[int] = self.inprogress_constraint.update(a_ )
if reset:
# 1. If the next token breaks the progress, then we must restart.
# e.g. constraint = "I love pies" and sequence so far is "I love" but `token_id` == "books".
# But that doesn't mean we self.init_state(), since we only reset the state for this particular
# constraint, not the full list of constraints.
self.pending_constraints.append(self.inprogress_constraint.copy(stateful=a_ ) )
lowerCamelCase : Union[str, Any] = None
if complete:
# 2. If the next token completes the constraint, move it to completed list, set
# inprogress to None. If there are no pending constraints either, then this full list of constraints
# is complete.
self.complete_constraints.append(self.inprogress_constraint )
lowerCamelCase : Dict = None
if len(self.pending_constraints ) == 0:
# we're done!
lowerCamelCase : Union[str, Any] = True
else:
# Not in the middle of fulfilling a constraint. So does this `token_id` helps us step towards any of our list
# of constraints?
for cidx, pending_constraint in enumerate(self.pending_constraints ):
if pending_constraint.does_advance(a_ ):
lowerCamelCase : Any = pending_constraint.update(a_ )
if not stepped:
raise Exception(
"""`constraint.update(token_id)` is not yielding incremental progress, """
"""even though `constraint.does_advance(token_id)` is true.""" )
if complete:
self.complete_constraints.append(a_ )
lowerCamelCase : Optional[Any] = None
if not complete and stepped:
lowerCamelCase : List[Any] = pending_constraint
if complete or stepped:
# If we made any progress at all, then it's at least not a "pending constraint".
lowerCamelCase : Any = (
self.pending_constraints[:cidx] + self.pending_constraints[cidx + 1 :]
)
if len(self.pending_constraints ) == 0 and self.inprogress_constraint is None:
# If there's no longer any pending after this and no inprogress either, then we must be
# complete.
lowerCamelCase : List[str] = True
break # prevent accidentally stepping through multiple constraints with just one token.
return complete, stepped
def a__ ( self: Dict , __a: int=True )-> Optional[int]:
lowerCamelCase : Dict = ConstraintListState(self.constraints ) # we actually never though self.constraints objects
# throughout this process. So it's at initialization state.
if stateful:
lowerCamelCase : List[Any] = [
constraint.copy(stateful=a_ ) for constraint in self.complete_constraints
]
if self.inprogress_constraint is not None:
lowerCamelCase : Dict = self.inprogress_constraint.copy(stateful=a_ )
lowerCamelCase : List[str] = [constraint.copy() for constraint in self.pending_constraints]
return new_state
| 706
|
"""simple docstring"""
import unittest
from transformers import AlbertTokenizer, AlbertTokenizerFast
from transformers.testing_utils import get_tests_dir, require_sentencepiece, require_tokenizers, slow
from ...test_tokenization_common import TokenizerTesterMixin
__lowerCamelCase :str = get_tests_dir('fixtures/spiece.model')
@require_sentencepiece
@require_tokenizers
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =AlbertTokenizer
snake_case__ : Optional[Any] =AlbertTokenizerFast
snake_case__ : Optional[int] =True
snake_case__ : Any =True
snake_case__ : Optional[int] =True
def a__ ( self: Dict )-> Optional[Any]:
super().setUp()
# We have a SentencePiece fixture for testing
lowerCamelCase : int = AlbertTokenizer(__a )
tokenizer.save_pretrained(self.tmpdirname )
def a__ ( self: Tuple , __a: Tuple )-> Union[str, Any]:
lowerCamelCase : List[str] = """this is a test"""
lowerCamelCase : int = """this is a test"""
return input_text, output_text
def a__ ( self: Any )-> List[Any]:
lowerCamelCase : int = """<pad>"""
lowerCamelCase : Optional[int] = 0
self.assertEqual(self.get_tokenizer()._convert_token_to_id(__a ) , __a )
self.assertEqual(self.get_tokenizer()._convert_id_to_token(__a ) , __a )
def a__ ( self: Tuple )-> str:
lowerCamelCase : Optional[int] = list(self.get_tokenizer().get_vocab().keys() )
self.assertEqual(vocab_keys[0] , """<pad>""" )
self.assertEqual(vocab_keys[1] , """<unk>""" )
self.assertEqual(vocab_keys[-1] , """▁eloquent""" )
self.assertEqual(len(__a ) , 30_000 )
def a__ ( self: List[str] )-> Any:
self.assertEqual(self.get_tokenizer().vocab_size , 30_000 )
def a__ ( self: Optional[Any] )-> Union[str, Any]:
if not self.test_rust_tokenizer:
return
lowerCamelCase : str = self.get_tokenizer()
lowerCamelCase : Tuple = self.get_rust_tokenizer()
lowerCamelCase : Union[str, Any] = """I was born in 92000, and this is falsé."""
lowerCamelCase : List[str] = tokenizer.tokenize(__a )
lowerCamelCase : Tuple = rust_tokenizer.tokenize(__a )
self.assertListEqual(__a , __a )
lowerCamelCase : Dict = tokenizer.encode(__a , add_special_tokens=__a )
lowerCamelCase : List[str] = rust_tokenizer.encode(__a , add_special_tokens=__a )
self.assertListEqual(__a , __a )
lowerCamelCase : Any = self.get_rust_tokenizer()
lowerCamelCase : List[str] = tokenizer.encode(__a )
lowerCamelCase : str = rust_tokenizer.encode(__a )
self.assertListEqual(__a , __a )
def a__ ( self: Tuple )-> List[Any]:
lowerCamelCase : List[str] = AlbertTokenizer(__a , keep_accents=__a )
lowerCamelCase : int = tokenizer.tokenize("""This is a test""" )
self.assertListEqual(__a , ["""▁this""", """▁is""", """▁a""", """▁test"""] )
self.assertListEqual(tokenizer.convert_tokens_to_ids(__a ) , [48, 25, 21, 1_289] )
lowerCamelCase : List[str] = tokenizer.tokenize("""I was born in 92000, and this is falsé.""" )
self.assertListEqual(
__a , ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """é""", """."""] )
lowerCamelCase : List[str] = tokenizer.convert_tokens_to_ids(__a )
self.assertListEqual(__a , [31, 23, 386, 19, 561, 3_050, 15, 17, 48, 25, 8_256, 18, 1, 9] )
lowerCamelCase : Tuple = tokenizer.convert_ids_to_tokens(__a )
self.assertListEqual(
__a , ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """<unk>""", """."""] , )
def a__ ( self: Tuple )-> str:
lowerCamelCase : str = AlbertTokenizer(__a )
lowerCamelCase : Union[str, Any] = tokenizer.encode("""sequence builders""" )
lowerCamelCase : List[Any] = tokenizer.encode("""multi-sequence build""" )
lowerCamelCase : Any = tokenizer.build_inputs_with_special_tokens(__a )
lowerCamelCase : str = tokenizer.build_inputs_with_special_tokens(__a , __a )
assert encoded_sentence == [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id]
assert encoded_pair == [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id] + text_a + [
tokenizer.sep_token_id
]
@slow
def a__ ( self: Any )-> Dict:
# fmt: off
lowerCamelCase : Optional[Any] = {"""attention_mask""": [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """input_ids""": [[2, 21_970, 13, 5, 6_092, 167, 28, 7_103, 2_153, 673, 8, 7_028, 12_051, 18, 17, 7_103, 2_153, 673, 8, 3_515, 18_684, 8, 4_461, 6, 1_927, 297, 8, 12_060, 2_607, 18, 13, 5, 4_461, 15, 10_538, 38, 8, 135, 15, 822, 58, 15, 993, 10_363, 15, 1_460, 8_005, 4_461, 15, 993, 255, 2_328, 9, 9, 9, 6, 26, 1_112, 816, 3_260, 13, 5, 103, 2_377, 6, 17, 1_112, 816, 2_782, 13, 5, 103, 10_641, 6, 29, 84, 2_512, 2_430, 782, 18_684, 2_761, 19, 808, 2_430, 2_556, 17, 855, 1_480, 9_477, 4_091, 128, 11_712, 15, 7_103, 2_153, 673, 17, 24_883, 9_990, 9, 3], [2, 11_502, 25, 1_006, 20, 782, 8, 11_809, 855, 1_732, 19_393, 18_667, 37, 367, 21_018, 69, 1_854, 34, 11_860, 19_124, 27, 156, 225, 17, 193, 4_141, 19, 65, 9_124, 9, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [2, 14, 2_231, 886, 2_385, 17_659, 84, 14, 16_792, 1_952, 9, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """token_type_ids""": [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]} # noqa: E501
# fmt: on
self.tokenizer_integration_test_util(
expected_encoding=__a , model_name="""albert-base-v2""" , revision="""6b6560eaf5ff2e250b00c50f380c5389a9c2d82e""" , )
| 42
| 0
|
from __future__ import annotations
from math import pow, sqrt
def snake_case ( UpperCamelCase__ : Tuple , UpperCamelCase__ : List[Any] , UpperCamelCase__ : Union[str, Any] ) -> dict[str, float]:
if (resistance, reactance, impedance).count(0 ) != 1:
raise ValueError("""One and only one argument must be 0""" )
if resistance == 0:
return {"resistance": sqrt(pow(__snake_case , 2 ) - pow(__snake_case , 2 ) )}
elif reactance == 0:
return {"reactance": sqrt(pow(__snake_case , 2 ) - pow(__snake_case , 2 ) )}
elif impedance == 0:
return {"impedance": sqrt(pow(__snake_case , 2 ) + pow(__snake_case , 2 ) )}
else:
raise ValueError("""Exactly one argument must be 0""" )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 707
|
"""simple docstring"""
__lowerCamelCase :List[Any] = {0: [2, 3], 1: [0], 2: [1], 3: [4], 4: []}
__lowerCamelCase :Union[str, Any] = {0: [1, 2, 3], 1: [2], 2: [0], 3: [4], 4: [5], 5: [3]}
def snake_case ( UpperCamelCase__ : dict[int, list[int]] , UpperCamelCase__ : int , UpperCamelCase__ : list[bool] ) -> list[int]:
lowerCamelCase : Tuple = True
lowerCamelCase : Any = []
for neighbour in graph[vert]:
if not visited[neighbour]:
order += topology_sort(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
order.append(UpperCamelCase__ )
return order
def snake_case ( UpperCamelCase__ : dict[int, list[int]] , UpperCamelCase__ : int , UpperCamelCase__ : list[bool] ) -> list[int]:
lowerCamelCase : List[Any] = True
lowerCamelCase : int = [vert]
for neighbour in reversed_graph[vert]:
if not visited[neighbour]:
component += find_components(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return component
def snake_case ( UpperCamelCase__ : dict[int, list[int]] ) -> list[list[int]]:
lowerCamelCase : int = len(UpperCamelCase__ ) * [False]
lowerCamelCase : dict[int, list[int]] = {vert: [] for vert in range(len(UpperCamelCase__ ) )}
for vert, neighbours in graph.items():
for neighbour in neighbours:
reversed_graph[neighbour].append(UpperCamelCase__ )
lowerCamelCase : int = []
for i, was_visited in enumerate(UpperCamelCase__ ):
if not was_visited:
order += topology_sort(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
lowerCamelCase : Tuple = []
lowerCamelCase : str = len(UpperCamelCase__ ) * [False]
for i in range(len(UpperCamelCase__ ) ):
lowerCamelCase : Any = order[len(UpperCamelCase__ ) - i - 1]
if not visited[vert]:
lowerCamelCase : List[str] = find_components(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
components_list.append(UpperCamelCase__ )
return components_list
| 42
| 0
|
"""simple docstring"""
from __future__ import annotations
def lowerCAmelCase ( UpperCamelCase__ : list[float] ) -> float:
lowerCamelCase : Union[str, Any] = 0.0_0
lowerCamelCase : Optional[int] = 0
for resistor in resistors:
if resistor <= 0:
lowerCamelCase : Dict = F'Resistor at index {index} has a negative or zero value!'
raise ValueError(UpperCAmelCase__ )
first_sum += 1 / float(UpperCAmelCase__ )
index += 1
return 1 / first_sum
def lowerCAmelCase ( UpperCamelCase__ : list[float] ) -> float:
lowerCamelCase : int = 0.0_0
lowerCamelCase : Union[str, Any] = 0
for resistor in resistors:
sum_r += resistor
if resistor < 0:
lowerCamelCase : List[str] = F'Resistor at index {index} has a negative value!'
raise ValueError(UpperCAmelCase__ )
index += 1
return sum_r
if __name__ == "__main__":
import doctest
doctest.testmod()
| 708
|
"""simple docstring"""
from typing import List, Optional, Union
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :str = logging.get_logger(__name__)
__lowerCamelCase :Any = {
'huggingface/time-series-transformer-tourism-monthly': (
'https://huggingface.co/huggingface/time-series-transformer-tourism-monthly/resolve/main/config.json'
),
# See all TimeSeriesTransformer models at https://huggingface.co/models?filter=time_series_transformer
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : List[Any] ='''time_series_transformer'''
snake_case__ : List[Any] ={
'''hidden_size''': '''d_model''',
'''num_attention_heads''': '''encoder_attention_heads''',
'''num_hidden_layers''': '''encoder_layers''',
}
def __init__( self: List[str] , __a: Optional[int] = None , __a: Optional[int] = None , __a: str = "student_t" , __a: str = "nll" , __a: int = 1 , __a: List[int] = [1, 2, 3, 4, 5, 6, 7] , __a: Optional[Union[str, bool]] = "mean" , __a: int = 0 , __a: int = 0 , __a: int = 0 , __a: int = 0 , __a: Optional[List[int]] = None , __a: Optional[List[int]] = None , __a: int = 32 , __a: int = 32 , __a: int = 2 , __a: int = 2 , __a: int = 2 , __a: int = 2 , __a: bool = True , __a: str = "gelu" , __a: int = 64 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: int = 100 , __a: float = 0.02 , __a: Tuple=True , **__a: str , )-> Any:
# time series specific configuration
lowerCamelCase : str = prediction_length
lowerCamelCase : Optional[Any] = context_length or prediction_length
lowerCamelCase : Tuple = distribution_output
lowerCamelCase : Any = loss
lowerCamelCase : List[Any] = input_size
lowerCamelCase : int = num_time_features
lowerCamelCase : Dict = lags_sequence
lowerCamelCase : Optional[int] = scaling
lowerCamelCase : int = num_dynamic_real_features
lowerCamelCase : Tuple = num_static_real_features
lowerCamelCase : Any = num_static_categorical_features
if cardinality and num_static_categorical_features > 0:
if len(__a ) != num_static_categorical_features:
raise ValueError(
"""The cardinality should be a list of the same length as `num_static_categorical_features`""" )
lowerCamelCase : int = cardinality
else:
lowerCamelCase : Dict = [0]
if embedding_dimension and num_static_categorical_features > 0:
if len(__a ) != num_static_categorical_features:
raise ValueError(
"""The embedding dimension should be a list of the same length as `num_static_categorical_features`""" )
lowerCamelCase : str = embedding_dimension
else:
lowerCamelCase : str = [min(50 , (cat + 1) // 2 ) for cat in self.cardinality]
lowerCamelCase : Any = num_parallel_samples
# Transformer architecture configuration
lowerCamelCase : Any = input_size * len(__a ) + self._number_of_features
lowerCamelCase : List[str] = d_model
lowerCamelCase : Tuple = encoder_attention_heads
lowerCamelCase : Optional[int] = decoder_attention_heads
lowerCamelCase : Union[str, Any] = encoder_ffn_dim
lowerCamelCase : str = decoder_ffn_dim
lowerCamelCase : str = encoder_layers
lowerCamelCase : Any = decoder_layers
lowerCamelCase : Optional[int] = dropout
lowerCamelCase : List[str] = attention_dropout
lowerCamelCase : Tuple = activation_dropout
lowerCamelCase : Optional[int] = encoder_layerdrop
lowerCamelCase : int = decoder_layerdrop
lowerCamelCase : Optional[int] = activation_function
lowerCamelCase : Optional[Any] = init_std
lowerCamelCase : Optional[Any] = use_cache
super().__init__(is_encoder_decoder=__a , **__a )
@property
def a__ ( self: int )-> int:
return (
sum(self.embedding_dimension )
+ self.num_dynamic_real_features
+ self.num_time_features
+ self.num_static_real_features
+ self.input_size * 2 # the log1p(abs(loc)) and log(scale) features
)
| 42
| 0
|
"""simple docstring"""
import argparse
import os
from accelerate.test_utils import execute_subprocess_async
def snake_case ( UpperCamelCase__ : str=None ) -> Optional[Any]:
if subparsers is not None:
lowerCamelCase : Dict = subparsers.add_parser("""test""" )
else:
lowerCamelCase : Tuple = argparse.ArgumentParser("""Accelerate test command""" )
parser.add_argument(
"""--config_file""" , default=_lowercase , help=(
"""The path to use to store the config file. Will default to a file named default_config.yaml in the cache """
"""location, which is the content of the environment `HF_HOME` suffixed with 'accelerate', or if you don't have """
"""such an environment variable, your cache directory ('~/.cache' or the content of `XDG_CACHE_HOME`) suffixed """
"""with 'huggingface'."""
) , )
if subparsers is not None:
parser.set_defaults(func=_lowercase )
return parser
def snake_case ( UpperCamelCase__ : str ) -> str:
lowerCamelCase : List[Any] = os.path.sep.join(__file__.split(os.path.sep )[:-2] + ["""test_utils""", """scripts""", """test_script.py"""] )
if args.config_file is None:
lowerCamelCase : int = script_name
else:
lowerCamelCase : int = F'--config_file={args.config_file} {script_name}'
lowerCamelCase : Optional[int] = ["accelerate-launch"] + test_args.split()
lowerCamelCase : Optional[int] = execute_subprocess_async(_lowercase , env=os.environ.copy() )
if result.returncode == 0:
print("""Test is a success! You are ready for your distributed training!""" )
def snake_case ( ) -> Tuple:
lowerCamelCase : Any = test_command_parser()
lowerCamelCase : Union[str, Any] = parser.parse_args()
test_command(_lowercase )
if __name__ == "__main__":
main()
| 709
|
"""simple docstring"""
from __future__ import annotations
__lowerCamelCase :int = 10
def snake_case ( UpperCamelCase__ : list[int] ) -> list[int]:
lowerCamelCase : int = 1
lowerCamelCase : Union[str, Any] = max(UpperCamelCase__ )
while placement <= max_digit:
# declare and initialize empty buckets
lowerCamelCase : list[list] = [[] for _ in range(UpperCamelCase__ )]
# split list_of_ints between the buckets
for i in list_of_ints:
lowerCamelCase : Any = int((i / placement) % RADIX )
buckets[tmp].append(UpperCamelCase__ )
# put each buckets' contents into list_of_ints
lowerCamelCase : Dict = 0
for b in range(UpperCamelCase__ ):
for i in buckets[b]:
lowerCamelCase : List[str] = i
a += 1
# move to next
placement *= RADIX
return list_of_ints
if __name__ == "__main__":
import doctest
doctest.testmod()
| 42
| 0
|
"""simple docstring"""
from ..utils import DummyObject, requires_backends
class A__ ( metaclass=UpperCAmelCase__):
"""simple docstring"""
snake_case__ : Union[str, Any] =['''torch''', '''scipy''']
def __init__( self: Tuple , *__a: Tuple , **__a: Union[str, Any] )-> Optional[Any]:
requires_backends(self , ["""torch""", """scipy"""] )
@classmethod
def a__ ( cls: Union[str, Any] , *__a: str , **__a: Tuple )-> Optional[Any]:
requires_backends(cls , ["""torch""", """scipy"""] )
@classmethod
def a__ ( cls: List[Any] , *__a: Any , **__a: Any )-> Any:
requires_backends(cls , ["""torch""", """scipy"""] )
| 710
|
"""simple docstring"""
import argparse
import pickle
import numpy as np
import torch
from torch import nn
from transformers import ReformerConfig, ReformerModelWithLMHead
from transformers.utils import logging
logging.set_verbosity_info()
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Tuple , UpperCamelCase__ : Optional[Any]=None ) -> Tuple:
# set parameter of one layer
assert torch_layer.weight.shape == weight.shape, F'{torch_layer} layer.weight does not match'
lowerCamelCase : Dict = nn.Parameter(UpperCamelCase__ )
if bias is not None:
assert torch_layer.bias.shape == bias.shape, F'{torch_layer} layer.bias does not match'
lowerCamelCase : Any = nn.Parameter(UpperCamelCase__ )
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : Any , UpperCamelCase__ : Tuple ) -> Union[str, Any]:
# set torch weights for 1-to-1 comparison
lowerCamelCase : Dict = np.asarray(weights[0] )
lowerCamelCase : List[Any] = np.asarray(weights[1] )
lowerCamelCase : List[str] = np.asarray(weights[2] )
set_param(
torch_layer.self_attention.query_key , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.value , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.output.dense , torch.tensor(UpperCamelCase__ ).view(-1 , UpperCamelCase__ ).contiguous().transpose(0 , 1 ) , )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Any ) -> List[Any]:
# set torch weights for 1-to-1 comparison
lowerCamelCase : Tuple = np.asarray(weights[0] )
lowerCamelCase : Any = np.asarray(weights[1] )
lowerCamelCase : List[Any] = np.asarray(weights[2] )
lowerCamelCase : List[str] = np.asarray(weights[3] )
set_param(
torch_layer.self_attention.query , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.key , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.value , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.output.dense , torch.tensor(UpperCamelCase__ ).view(-1 , UpperCamelCase__ ).contiguous().transpose(0 , 1 ) , )
def snake_case ( UpperCamelCase__ : Dict , UpperCamelCase__ : Optional[int] , UpperCamelCase__ : Any ) -> Optional[Any]:
# layernorm 1
lowerCamelCase : str = weights[0][0][0]
lowerCamelCase : Optional[int] = np.asarray(layer_norm_a[0] )
lowerCamelCase : Tuple = np.asarray(layer_norm_a[1] )
set_param(
torch_block.attention.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# lsh weights + output
lowerCamelCase : List[Any] = weights[0][1]
if len(UpperCamelCase__ ) < 4:
set_layer_weights_in_torch_lsh(UpperCamelCase__ , torch_block.attention , UpperCamelCase__ )
else:
set_layer_weights_in_torch_local(UpperCamelCase__ , torch_block.attention , UpperCamelCase__ )
# intermediate weighs
lowerCamelCase : int = weights[2][0][1][2]
# Chunked Feed Forward
if len(UpperCamelCase__ ) == 4:
lowerCamelCase : Dict = intermediate_weights[2]
# layernorm 2
lowerCamelCase : Optional[int] = np.asarray(intermediate_weights[0][0] )
lowerCamelCase : Tuple = np.asarray(intermediate_weights[0][1] )
set_param(
torch_block.feed_forward.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# intermediate dense
lowerCamelCase : Optional[Any] = np.asarray(intermediate_weights[1][0] )
lowerCamelCase : Union[str, Any] = np.asarray(intermediate_weights[1][1] )
set_param(
torch_block.feed_forward.dense.dense , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
# intermediate out
lowerCamelCase : Optional[int] = np.asarray(intermediate_weights[4][0] )
lowerCamelCase : List[Any] = np.asarray(intermediate_weights[4][1] )
set_param(
torch_block.feed_forward.output.dense , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : List[str] , UpperCamelCase__ : int ) -> List[Any]:
# reformer model
lowerCamelCase : List[Any] = torch_model.reformer
# word embeds
lowerCamelCase : Union[str, Any] = np.asarray(weights[1] )
set_param(
torch_model_reformer.embeddings.word_embeddings , torch.tensor(UpperCamelCase__ ) , )
if isinstance(weights[3] , UpperCamelCase__ ):
lowerCamelCase : Union[str, Any] = torch_model_reformer.embeddings.position_embeddings
for emb_idx in range(len(position_embeddings.weights ) ):
lowerCamelCase : str = np.asarray(weights[3][emb_idx][0] )
assert (
position_embeddings.weights[emb_idx].shape == emb_weights.shape
), F'{position_embeddings[emb_idx]} emb does not match'
lowerCamelCase : Dict = nn.Parameter(torch.tensor(UpperCamelCase__ ) )
lowerCamelCase : int = weights[5]
assert len(torch_model_reformer.encoder.layers ) * 4 == len(
UpperCamelCase__ ), "HF and trax model do not have the same number of layers"
for layer_idx, layer in enumerate(torch_model_reformer.encoder.layers ):
lowerCamelCase : Dict = trax_layer_weights[4 * layer_idx : 4 * (layer_idx + 1)]
set_block_weights_in_torch(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
# output layer norm
lowerCamelCase : Any = np.asarray(weights[7][0] )
lowerCamelCase : List[str] = np.asarray(weights[7][1] )
set_param(
torch_model_reformer.encoder.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# output embeddings
lowerCamelCase : List[Any] = np.asarray(weights[9][0] )
lowerCamelCase : Optional[int] = np.asarray(weights[9][1] )
set_param(
torch_model.lm_head.decoder , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
def snake_case ( UpperCamelCase__ : str , UpperCamelCase__ : int , UpperCamelCase__ : str ) -> Optional[int]:
# Initialise PyTorch model
lowerCamelCase : Union[str, Any] = ReformerConfig.from_json_file(UpperCamelCase__ )
print(F'Building PyTorch model from configuration: {config}' )
lowerCamelCase : str = ReformerModelWithLMHead(UpperCamelCase__ )
with open(UpperCamelCase__ , """rb""" ) as f:
lowerCamelCase : str = pickle.load(UpperCamelCase__ )["""weights"""]
set_model_weights_in_torch(UpperCamelCase__ , UpperCamelCase__ , config.hidden_size )
# Save pytorch-model
print(F'Save PyTorch model to {pytorch_dump_path}' )
torch.save(model.state_dict() , UpperCamelCase__ )
if __name__ == "__main__":
__lowerCamelCase :Dict = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--trax_model_pkl_path', default=None, type=str, required=True, help='Path to the TensorFlow checkpoint path.'
)
parser.add_argument(
'--config_file',
default=None,
type=str,
required=True,
help=(
'The config json file corresponding to the pre-trained Reformer model. \n'
'This specifies the model architecture.'
),
)
parser.add_argument(
'--pytorch_dump_path', default=None, type=str, required=True, help='Path to the output PyTorch model.'
)
__lowerCamelCase :Optional[int] = parser.parse_args()
convert_trax_checkpoint_to_pytorch(args.trax_model_pkl_path, args.config_file, args.pytorch_dump_path)
| 42
| 0
|
"""simple docstring"""
import json
import os
import tempfile
import unittest
import unittest.mock as mock
from pathlib import Path
from requests.exceptions import HTTPError
from transformers.utils import (
CONFIG_NAME,
FLAX_WEIGHTS_NAME,
TF2_WEIGHTS_NAME,
TRANSFORMERS_CACHE,
WEIGHTS_NAME,
cached_file,
get_file_from_repo,
has_file,
)
__lowerCamelCase :int = 'hf-internal-testing/tiny-random-bert'
__lowerCamelCase :Tuple = os.path.join(TRANSFORMERS_CACHE, 'models--hf-internal-testing--tiny-random-bert')
__lowerCamelCase :List[str] = '9b8c223d42b2188cb49d29af482996f9d0f3e5a6'
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: List[str] )-> Optional[int]:
lowerCamelCase : List[str] = cached_file(UpperCAmelCase_ , UpperCAmelCase_ )
# Should have downloaded the file in here
self.assertTrue(os.path.isdir(UpperCAmelCase_ ) )
# Cache should contain at least those three subfolders:
for subfolder in ["blobs", "refs", "snapshots"]:
self.assertTrue(os.path.isdir(os.path.join(UpperCAmelCase_ , UpperCAmelCase_ ) ) )
with open(os.path.join(UpperCAmelCase_ , """refs""" , """main""" ) ) as f:
lowerCamelCase : List[str] = f.read()
self.assertEqual(UpperCAmelCase_ , os.path.join(UpperCAmelCase_ , """snapshots""" , UpperCAmelCase_ , UpperCAmelCase_ ) )
self.assertTrue(os.path.isfile(UpperCAmelCase_ ) )
# File is cached at the same place the second time.
lowerCamelCase : str = cached_file(UpperCAmelCase_ , UpperCAmelCase_ )
self.assertEqual(UpperCAmelCase_ , UpperCAmelCase_ )
# Using a specific revision to test the full commit hash.
lowerCamelCase : Dict = cached_file(UpperCAmelCase_ , UpperCAmelCase_ , revision="""9b8c223""" )
self.assertEqual(UpperCAmelCase_ , os.path.join(UpperCAmelCase_ , """snapshots""" , UpperCAmelCase_ , UpperCAmelCase_ ) )
def a__ ( self: Union[str, Any] )-> Optional[int]:
with self.assertRaisesRegex(UpperCAmelCase_ , """is not a valid model identifier""" ):
lowerCamelCase : int = cached_file("""tiny-random-bert""" , UpperCAmelCase_ )
with self.assertRaisesRegex(UpperCAmelCase_ , """is not a valid git identifier""" ):
lowerCamelCase : Tuple = cached_file(UpperCAmelCase_ , UpperCAmelCase_ , revision="""aaaa""" )
with self.assertRaisesRegex(UpperCAmelCase_ , """does not appear to have a file named""" ):
lowerCamelCase : List[str] = cached_file(UpperCAmelCase_ , """conf""" )
def a__ ( self: Union[str, Any] )-> Optional[int]:
with self.assertRaisesRegex(UpperCAmelCase_ , """does not appear to have a file named""" ):
lowerCamelCase : str = cached_file(UpperCAmelCase_ , """conf""" )
with open(os.path.join(UpperCAmelCase_ , """refs""" , """main""" ) ) as f:
lowerCamelCase : Tuple = f.read()
self.assertTrue(os.path.isfile(os.path.join(UpperCAmelCase_ , """.no_exist""" , UpperCAmelCase_ , """conf""" ) ) )
lowerCamelCase : List[str] = cached_file(UpperCAmelCase_ , """conf""" , _raise_exceptions_for_missing_entries=UpperCAmelCase_ )
self.assertIsNone(UpperCAmelCase_ )
lowerCamelCase : Dict = cached_file(UpperCAmelCase_ , """conf""" , local_files_only=UpperCAmelCase_ , _raise_exceptions_for_missing_entries=UpperCAmelCase_ )
self.assertIsNone(UpperCAmelCase_ )
lowerCamelCase : Tuple = mock.Mock()
lowerCamelCase : Any = 500
lowerCamelCase : Dict = {}
lowerCamelCase : Optional[int] = HTTPError
lowerCamelCase : Union[str, Any] = {}
# Under the mock environment we get a 500 error when trying to reach the tokenizer.
with mock.patch("""requests.Session.request""" , return_value=UpperCAmelCase_ ) as mock_head:
lowerCamelCase : str = cached_file(UpperCAmelCase_ , """conf""" , _raise_exceptions_for_connection_errors=UpperCAmelCase_ )
self.assertIsNone(UpperCAmelCase_ )
# This check we did call the fake head request
mock_head.assert_called()
def a__ ( self: str )-> Any:
self.assertTrue(has_file("""hf-internal-testing/tiny-bert-pt-only""" , UpperCAmelCase_ ) )
self.assertFalse(has_file("""hf-internal-testing/tiny-bert-pt-only""" , UpperCAmelCase_ ) )
self.assertFalse(has_file("""hf-internal-testing/tiny-bert-pt-only""" , UpperCAmelCase_ ) )
def a__ ( self: int )-> Dict:
# `get_file_from_repo` returns None if the file does not exist
self.assertIsNone(get_file_from_repo("""bert-base-cased""" , """ahah.txt""" ) )
# The function raises if the repository does not exist.
with self.assertRaisesRegex(UpperCAmelCase_ , """is not a valid model identifier""" ):
get_file_from_repo("""bert-base-case""" , UpperCAmelCase_ )
# The function raises if the revision does not exist.
with self.assertRaisesRegex(UpperCAmelCase_ , """is not a valid git identifier""" ):
get_file_from_repo("""bert-base-cased""" , UpperCAmelCase_ , revision="""ahaha""" )
lowerCamelCase : str = get_file_from_repo("""bert-base-cased""" , UpperCAmelCase_ )
# The name is the cached name which is not very easy to test, so instead we load the content.
lowerCamelCase : int = json.loads(open(UpperCAmelCase_ , """r""" ).read() )
self.assertEqual(config["""hidden_size"""] , 768 )
def a__ ( self: Optional[Any] )-> str:
with tempfile.TemporaryDirectory() as tmp_dir:
lowerCamelCase : int = Path(UpperCAmelCase_ ) / """a.txt"""
filename.touch()
self.assertEqual(get_file_from_repo(UpperCAmelCase_ , """a.txt""" ) , str(UpperCAmelCase_ ) )
self.assertIsNone(get_file_from_repo(UpperCAmelCase_ , """b.txt""" ) )
| 711
|
"""simple docstring"""
import inspect
import unittest
import torch
import torch.nn as nn
from accelerate.hooks import (
AlignDevicesHook,
ModelHook,
SequentialHook,
add_hook_to_module,
attach_align_device_hook,
remove_hook_from_module,
remove_hook_from_submodules,
)
from accelerate.test_utils import require_multi_gpu
class A__ ( nn.Module):
"""simple docstring"""
def __init__( self: Dict )-> Dict:
super().__init__()
lowerCamelCase : Tuple = nn.Linear(3 , 4 )
lowerCamelCase : Optional[Any] = nn.BatchNormad(4 )
lowerCamelCase : Optional[Any] = nn.Linear(4 , 5 )
def a__ ( self: List[str] , __a: List[Any] )-> Optional[Any]:
return self.lineara(self.batchnorm(self.lineara(__a ) ) )
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Tuple , __a: int , *__a: Any , **__a: Tuple )-> Tuple:
return (args[0] + 1,) + args[1:], kwargs
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Optional[int] , __a: List[str] , __a: List[Any] )-> List[str]:
return output + 1
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: int )-> str:
lowerCamelCase : List[str] = ModelForTest()
lowerCamelCase : Dict = ModelHook()
add_hook_to_module(__a , __a )
self.assertEqual(test_model._hf_hook , __a )
self.assertTrue(hasattr(__a , """_old_forward""" ) )
# Check adding the hook did not change the name or the signature
self.assertEqual(test_model.forward.__name__ , """forward""" )
self.assertListEqual(list(inspect.signature(test_model.forward ).parameters ) , ["""x"""] )
remove_hook_from_module(__a )
self.assertFalse(hasattr(__a , """_hf_hook""" ) )
self.assertFalse(hasattr(__a , """_old_forward""" ) )
def a__ ( self: int )-> str:
lowerCamelCase : List[str] = ModelForTest()
lowerCamelCase : Union[str, Any] = ModelHook()
add_hook_to_module(__a , __a )
add_hook_to_module(__a , __a , append=__a )
self.assertEqual(isinstance(test_model._hf_hook , __a ) , __a )
self.assertEqual(len(test_model._hf_hook.hooks ) , 2 )
self.assertTrue(hasattr(__a , """_old_forward""" ) )
# Check adding the hook did not change the name or the signature
self.assertEqual(test_model.forward.__name__ , """forward""" )
self.assertListEqual(list(inspect.signature(test_model.forward ).parameters ) , ["""x"""] )
remove_hook_from_module(__a )
self.assertFalse(hasattr(__a , """_hf_hook""" ) )
self.assertFalse(hasattr(__a , """_old_forward""" ) )
def a__ ( self: List[Any] )-> List[str]:
lowerCamelCase : str = ModelForTest()
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : Union[str, Any] = test_model(x + 1 )
lowerCamelCase : Optional[int] = test_model(x + 2 )
lowerCamelCase : List[Any] = PreForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[int] = test_model(__a )
self.assertTrue(torch.allclose(__a , __a , atol=1e-5 ) )
# Attaching a hook to a model when it already has one replaces, does not chain
lowerCamelCase : Dict = PreForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Tuple = test_model(__a )
self.assertTrue(torch.allclose(__a , __a , atol=1e-5 ) )
# You need to use the sequential hook to chain two or more hooks
lowerCamelCase : Any = SequentialHook(PreForwardHook() , PreForwardHook() )
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[Any] = test_model(__a )
assert torch.allclose(__a , __a , atol=1e-5 )
def a__ ( self: Any )-> Optional[int]:
lowerCamelCase : str = ModelForTest()
lowerCamelCase : List[str] = torch.randn(2 , 3 )
lowerCamelCase : int = test_model(__a )
lowerCamelCase : Dict = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Tuple = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 , atol=1e-5 ) )
# Attaching a hook to a model when it already has one replaces, does not chain
lowerCamelCase : str = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[Any] = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 , atol=1e-5 ) )
# You need to use the sequential hook to chain two or more hooks
lowerCamelCase : Union[str, Any] = SequentialHook(PostForwardHook() , PostForwardHook() )
add_hook_to_module(__a , __a )
lowerCamelCase : str = test_model(__a )
assert torch.allclose(__a , output + 2 , atol=1e-5 )
def a__ ( self: int )-> Dict:
lowerCamelCase : List[Any] = ModelForTest()
lowerCamelCase : Optional[int] = torch.randn(2 , 3 )
lowerCamelCase : List[str] = test_model(__a )
lowerCamelCase : Any = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : str = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 ) )
self.assertTrue(outputa.requires_grad )
lowerCamelCase : Optional[int] = True
lowerCamelCase : Optional[int] = test_model(__a )
self.assertFalse(outputa.requires_grad )
@require_multi_gpu
def a__ ( self: List[str] )-> Union[str, Any]:
lowerCamelCase : int = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
add_hook_to_module(model.lineara , AlignDevicesHook(execution_device=0 ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(execution_device=0 ) )
add_hook_to_module(model.lineara , AlignDevicesHook(execution_device=1 ) )
self.assertEqual(model.lineara.weight.device , torch.device(0 ) )
self.assertEqual(model.batchnorm.weight.device , torch.device(0 ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device(0 ) )
self.assertEqual(model.lineara.weight.device , torch.device(1 ) )
# We can still make a forward pass. The input does not need to be on any particular device
lowerCamelCase : str = torch.randn(2 , 3 )
lowerCamelCase : Dict = model(__a )
self.assertEqual(output.device , torch.device(1 ) )
# We can add a general hook to put back output on same device as input.
add_hook_to_module(__a , AlignDevicesHook(io_same_device=__a ) )
lowerCamelCase : Optional[int] = torch.randn(2 , 3 ).to(0 )
lowerCamelCase : str = model(__a )
self.assertEqual(output.device , torch.device(0 ) )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Union[str, Any] = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : Tuple = {"""execution_device""": 0 if torch.cuda.is_available() else """cpu""", """offload""": True}
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(**__a ) )
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(hook_kwargs["""execution_device"""] )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Optional[Any] = torch.randn(2 , 3 )
lowerCamelCase : Optional[Any] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_module(model.lineara )
remove_hook_from_module(model.batchnorm )
remove_hook_from_module(model.lineara )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
lowerCamelCase : Any = {
"""execution_device""": 0 if torch.cuda.is_available() else """cpu""",
"""offload""": True,
"""offload_buffers""": True,
}
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(**__a ) )
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : int = torch.randn(2 , 3 )
lowerCamelCase : Optional[int] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_module(model.lineara )
remove_hook_from_module(model.batchnorm )
remove_hook_from_module(model.lineara )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
def a__ ( self: Any )-> List[str]:
lowerCamelCase : int = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : int = 0 if torch.cuda.is_available() else """cpu"""
attach_align_device_hook(__a , execution_device=__a , offload=__a )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(__a )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : Optional[Any] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
attach_align_device_hook(__a , execution_device=__a , offload=__a , offload_buffers=__a )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : Optional[int] = torch.randn(2 , 3 )
lowerCamelCase : int = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
def a__ ( self: Optional[Any] )-> List[Any]:
lowerCamelCase : List[Any] = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : Any = 0 if torch.cuda.is_available() else """cpu"""
attach_align_device_hook(
__a , execution_device=__a , offload=__a , weights_map=model.state_dict() )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(__a )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : int = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
attach_align_device_hook(
__a , execution_device=__a , offload=__a , weights_map=model.state_dict() , offload_buffers=__a , )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : Tuple = torch.randn(2 , 3 )
lowerCamelCase : Any = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
| 42
| 0
|
"""simple docstring"""
from typing import Any, Dict, List, Optional, Tuple, Union
import torch
from torch import nn
from torch.utils.data import DistributedSampler, RandomSampler
from transformers import PreTrainedModel, Trainer, logging
from transformers.integrations import is_fairscale_available
from transformers.models.fsmt.configuration_fsmt import FSMTConfig
from transformers.optimization import (
Adafactor,
AdamW,
get_constant_schedule,
get_constant_schedule_with_warmup,
get_cosine_schedule_with_warmup,
get_cosine_with_hard_restarts_schedule_with_warmup,
get_linear_schedule_with_warmup,
get_polynomial_decay_schedule_with_warmup,
)
from transformers.trainer_pt_utils import get_tpu_sampler
from transformers.training_args import ParallelMode
from transformers.utils import is_torch_tpu_available
if is_fairscale_available():
from fairscale.optim import OSS
__lowerCamelCase :Dict = logging.get_logger(__name__)
__lowerCamelCase :Union[str, Any] = {
'linear': get_linear_schedule_with_warmup,
'cosine': get_cosine_schedule_with_warmup,
'cosine_w_restarts': get_cosine_with_hard_restarts_schedule_with_warmup,
'polynomial': get_polynomial_decay_schedule_with_warmup,
'constant': get_constant_schedule,
'constant_w_warmup': get_constant_schedule_with_warmup,
}
class A__ ( UpperCamelCase_):
"""simple docstring"""
def __init__( self: Union[str, Any] , __a: Union[str, Any]=None , __a: Optional[int]=None , *__a: Dict , **__a: str )-> List[Any]:
super().__init__(*UpperCamelCase__ , **UpperCamelCase__ )
if config is None:
assert isinstance(self.model , UpperCamelCase__ ), (
"If no `config` is passed the model to be trained has to be of type `PreTrainedModel`, but is"
f' {self.model.__class__}'
)
lowerCamelCase : List[Any] = self.model.config
else:
lowerCamelCase : Union[str, Any] = config
lowerCamelCase : List[Any] = data_args
lowerCamelCase : str = self.config.tgt_vocab_size if isinstance(self.config , UpperCamelCase__ ) else self.config.vocab_size
if self.args.label_smoothing != 0 or (self.data_args is not None and self.data_args.ignore_pad_token_for_loss):
assert self.config.pad_token_id is not None, (
"Make sure that `config.pad_token_id` is correcly defined when ignoring `pad_token` for loss"
" calculation or doing label smoothing."
)
if self.config.pad_token_id is None and self.config.eos_token_id is not None:
logger.warning(
f'The `config.pad_token_id` is `None`. Using `config.eos_token_id` = {self.config.eos_token_id} for'
""" padding..""" )
if self.args.label_smoothing == 0:
lowerCamelCase : int = torch.nn.CrossEntropyLoss(ignore_index=self.config.pad_token_id )
else:
# dynamically import label_smoothed_nll_loss
from utils import label_smoothed_nll_loss
lowerCamelCase : List[str] = label_smoothed_nll_loss
def a__ ( self: Any , __a: int )-> Optional[Any]:
if self.optimizer is None:
lowerCamelCase : str = ['''bias''', '''LayerNorm.weight''']
lowerCamelCase : Optional[int] = [
{
'''params''': [p for n, p in self.model.named_parameters() if not any(nd in n for nd in no_decay )],
'''weight_decay''': self.args.weight_decay,
},
{
'''params''': [p for n, p in self.model.named_parameters() if any(nd in n for nd in no_decay )],
'''weight_decay''': 0.0,
},
]
lowerCamelCase : int = Adafactor if self.args.adafactor else AdamW
if self.args.adafactor:
lowerCamelCase : str = Adafactor
lowerCamelCase : Optional[Any] = {'''scale_parameter''': False, '''relative_step''': False}
else:
lowerCamelCase : Optional[int] = AdamW
lowerCamelCase : Any = {
'''betas''': (self.args.adam_betaa, self.args.adam_betaa),
'''eps''': self.args.adam_epsilon,
}
lowerCamelCase : Dict = self.args.learning_rate
if self.sharded_ddp:
lowerCamelCase : Any = OSS(
params=UpperCamelCase__ , optim=UpperCamelCase__ , **UpperCamelCase__ , )
else:
lowerCamelCase : Any = optimizer_cls(UpperCamelCase__ , **UpperCamelCase__ )
if self.lr_scheduler is None:
lowerCamelCase : List[Any] = self._get_lr_scheduler(UpperCamelCase__ )
else: # ignoring --lr_scheduler
logger.warning("""scheduler is passed to `Seq2SeqTrainer`, `--lr_scheduler` arg is ignored.""" )
def a__ ( self: Union[str, Any] , __a: List[Any] )-> Optional[int]:
lowerCamelCase : Union[str, Any] = arg_to_scheduler[self.args.lr_scheduler]
if self.args.lr_scheduler == "constant":
lowerCamelCase : List[str] = schedule_func(self.optimizer )
elif self.args.lr_scheduler == "constant_w_warmup":
lowerCamelCase : Optional[Any] = schedule_func(self.optimizer , num_warmup_steps=self.args.warmup_steps )
else:
lowerCamelCase : str = schedule_func(
self.optimizer , num_warmup_steps=self.args.warmup_steps , num_training_steps=UpperCamelCase__ )
return scheduler
def a__ ( self: Any )-> Optional[Any]:
if isinstance(self.train_dataset , torch.utils.data.IterableDataset ):
return None
elif is_torch_tpu_available():
return get_tpu_sampler(self.train_dataset )
else:
if self.args.sortish_sampler:
self.train_dataset.make_sortish_sampler(
self.args.per_device_train_batch_size , distributed=(self.args.parallel_mode == ParallelMode.DISTRIBUTED) , )
return (
RandomSampler(self.train_dataset )
if self.args.local_rank == -1
else DistributedSampler(self.train_dataset )
)
def a__ ( self: Union[str, Any] , __a: Tuple , __a: Union[str, Any] , __a: List[str] )-> str:
if self.args.label_smoothing == 0:
if self.data_args is not None and self.data_args.ignore_pad_token_for_loss:
# force training to ignore pad token
lowerCamelCase : List[str] = model(**UpperCamelCase__ , use_cache=UpperCamelCase__ )[0]
lowerCamelCase : List[str] = self.loss_fn(logits.view(-1 , logits.shape[-1] ) , labels.view(-1 ) )
else:
# compute usual loss via models
lowerCamelCase : Union[str, Any] = model(**UpperCamelCase__ , labels=UpperCamelCase__ , use_cache=UpperCamelCase__ )[:2]
else:
# compute label smoothed loss
lowerCamelCase : List[str] = model(**UpperCamelCase__ , use_cache=UpperCamelCase__ )[0]
lowerCamelCase : int = torch.nn.functional.log_softmax(UpperCamelCase__ , dim=-1 )
lowerCamelCase : int = self.loss_fn(UpperCamelCase__ , UpperCamelCase__ , self.args.label_smoothing , ignore_index=self.config.pad_token_id )
return loss, logits
def a__ ( self: str , __a: Dict , __a: Union[str, Any] )-> List[Any]:
lowerCamelCase : List[str] = inputs.pop("""labels""" )
lowerCamelCase : List[Any] = self._compute_loss(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return loss
def a__ ( self: str , __a: nn.Module , __a: Dict[str, Union[torch.Tensor, Any]] , __a: bool , __a: Optional[List[str]] = None , )-> Union[str, Any]:
lowerCamelCase : Union[str, Any] = self._prepare_inputs(UpperCamelCase__ )
lowerCamelCase : Dict = {
'''max_length''': self.data_args.val_max_target_length
if self.data_args is not None
else self.config.max_length,
'''num_beams''': self.data_args.eval_beams if self.data_args is not None else self.config.num_beams,
}
if self.args.predict_with_generate and not self.args.prediction_loss_only:
lowerCamelCase : Optional[Any] = self.model.generate(
inputs["""input_ids"""] , attention_mask=inputs["""attention_mask"""] , **UpperCamelCase__ , )
# in case the batch is shorter than max length, the output should be padded
if generated_tokens.shape[-1] < gen_kwargs["max_length"]:
lowerCamelCase : Optional[Any] = self._pad_tensors_to_max_len(UpperCamelCase__ , gen_kwargs["""max_length"""] )
lowerCamelCase : Tuple = inputs.pop("""labels""" )
with torch.no_grad():
# compute loss on predict data
lowerCamelCase : int = self._compute_loss(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
lowerCamelCase : Optional[Any] = loss.mean().detach()
if self.args.prediction_loss_only:
return (loss, None, None)
lowerCamelCase : int = generated_tokens if self.args.predict_with_generate else logits
if labels.shape[-1] < gen_kwargs["max_length"]:
lowerCamelCase : Dict = self._pad_tensors_to_max_len(UpperCamelCase__ , gen_kwargs["""max_length"""] )
return (loss, logits, labels)
def a__ ( self: List[str] , __a: Optional[int] , __a: Optional[int] )-> int:
lowerCamelCase : Union[str, Any] = self.config.pad_token_id if self.config.pad_token_id is not None else self.config.eos_token_id
if pad_token_id is None:
raise ValueError(
"""Make sure that either `config.pad_token_id` or `config.eos_token_id` is defined if tensor has to be"""
f' padded to `max_length`={max_length}' )
lowerCamelCase : List[Any] = pad_token_id * torch.ones(
(tensor.shape[0], max_length) , dtype=tensor.dtype , device=tensor.device )
lowerCamelCase : Tuple = tensor
return padded_tensor
| 712
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_torch_available,
)
__lowerCamelCase :Optional[Any] = {
'configuration_encodec': [
'ENCODEC_PRETRAINED_CONFIG_ARCHIVE_MAP',
'EncodecConfig',
],
'feature_extraction_encodec': ['EncodecFeatureExtractor'],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Union[str, Any] = [
'ENCODEC_PRETRAINED_MODEL_ARCHIVE_LIST',
'EncodecModel',
'EncodecPreTrainedModel',
]
if TYPE_CHECKING:
from .configuration_encodec import (
ENCODEC_PRETRAINED_CONFIG_ARCHIVE_MAP,
EncodecConfig,
)
from .feature_extraction_encodec import EncodecFeatureExtractor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_encodec import (
ENCODEC_PRETRAINED_MODEL_ARCHIVE_LIST,
EncodecModel,
EncodecPreTrainedModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
import logging
import os
from typing import Dict, List, Optional, Union
import torch
import torch.nn as nn
from accelerate.utils.imports import (
is_abit_bnb_available,
is_abit_bnb_available,
is_bnb_available,
)
from ..big_modeling import dispatch_model, init_empty_weights
from .dataclasses import BnbQuantizationConfig
from .modeling import (
find_tied_parameters,
get_balanced_memory,
infer_auto_device_map,
load_checkpoint_in_model,
offload_weight,
set_module_tensor_to_device,
)
if is_bnb_available():
import bitsandbytes as bnb
from copy import deepcopy
__lowerCamelCase :Any = logging.getLogger(__name__)
def snake_case ( UpperCamelCase__ : List[Any] , UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : Union[str, Any] = None , UpperCamelCase__ : Optional[Any] = None , UpperCamelCase__ : Optional[int] = None , UpperCamelCase__ : int = None , UpperCamelCase__ : List[str] = None , UpperCamelCase__ : Tuple = False , ) -> str:
lowerCamelCase : List[str] = bnb_quantization_config.load_in_abit
lowerCamelCase : str = bnb_quantization_config.load_in_abit
if load_in_abit and not is_abit_bnb_available():
raise ImportError(
"""You have a version of `bitsandbytes` that is not compatible with 8bit quantization,"""
""" make sure you have the latest version of `bitsandbytes` installed.""" )
if load_in_abit and not is_abit_bnb_available():
raise ValueError(
"""You have a version of `bitsandbytes` that is not compatible with 4bit quantization,"""
"""make sure you have the latest version of `bitsandbytes` installed.""" )
lowerCamelCase : Optional[int] = []
# custom device map
if isinstance(__UpperCamelCase , __UpperCamelCase ) and len(device_map.keys() ) > 1:
lowerCamelCase : Dict = [key for key, value in device_map.items() if value in ["""disk""", """cpu"""]]
# We keep some modules such as the lm_head in their original dtype for numerical stability reasons
if bnb_quantization_config.skip_modules is None:
lowerCamelCase : Optional[int] = get_keys_to_not_convert(__UpperCamelCase )
# add cpu modules to skip modules only for 4-bit modules
if load_in_abit:
bnb_quantization_config.skip_modules.extend(__UpperCamelCase )
lowerCamelCase : Optional[Any] = bnb_quantization_config.skip_modules
# We add the modules we want to keep in full precision
if bnb_quantization_config.keep_in_fpaa_modules is None:
lowerCamelCase : Dict = []
lowerCamelCase : Union[str, Any] = bnb_quantization_config.keep_in_fpaa_modules
modules_to_not_convert.extend(__UpperCamelCase )
# compatibility with peft
lowerCamelCase : Union[str, Any] = load_in_abit
lowerCamelCase : Union[str, Any] = load_in_abit
lowerCamelCase : Union[str, Any] = get_parameter_device(__UpperCamelCase )
if model_device.type != "meta":
# quantization of an already loaded model
logger.warning(
"""It is not recommended to quantize a loaded model. """
"""The model should be instantiated under the `init_empty_weights` context manager.""" )
lowerCamelCase : str = replace_with_bnb_layers(__UpperCamelCase , __UpperCamelCase , modules_to_not_convert=__UpperCamelCase )
# convert param to the right dtype
lowerCamelCase : int = bnb_quantization_config.torch_dtype
for name, param in model.state_dict().items():
if any(module_to_keep_in_fpaa in name for module_to_keep_in_fpaa in keep_in_fpaa_modules ):
param.to(torch.floataa )
if param.dtype != torch.floataa:
lowerCamelCase : Optional[int] = name.replace(""".weight""" , """""" ).replace(""".bias""" , """""" )
lowerCamelCase : Tuple = getattr(__UpperCamelCase , __UpperCamelCase , __UpperCamelCase )
if param is not None:
param.to(torch.floataa )
elif torch.is_floating_point(__UpperCamelCase ):
param.to(__UpperCamelCase )
if model_device.type == "cuda":
# move everything to cpu in the first place because we can't do quantization if the weights are already on cuda
model.cuda(torch.cuda.current_device() )
torch.cuda.empty_cache()
elif torch.cuda.is_available():
model.to(torch.cuda.current_device() )
else:
raise RuntimeError("""No GPU found. A GPU is needed for quantization.""" )
logger.info(
F'The model device type is {model_device.type}. However, cuda is needed for quantization.'
"""We move the model to cuda.""" )
return model
elif weights_location is None:
raise RuntimeError(
F'`weights_location` needs to be the folder path containing the weights of the model, but we found {weights_location} ' )
else:
with init_empty_weights():
lowerCamelCase : List[Any] = replace_with_bnb_layers(
__UpperCamelCase , __UpperCamelCase , modules_to_not_convert=__UpperCamelCase )
lowerCamelCase : Any = get_quantized_model_device_map(
__UpperCamelCase , __UpperCamelCase , __UpperCamelCase , max_memory=__UpperCamelCase , no_split_module_classes=__UpperCamelCase , )
if offload_state_dict is None and device_map is not None and "disk" in device_map.values():
lowerCamelCase : Union[str, Any] = True
lowerCamelCase : int = any(x in list(device_map.values() ) for x in ["""cpu""", """disk"""] )
load_checkpoint_in_model(
__UpperCamelCase , __UpperCamelCase , __UpperCamelCase , dtype=bnb_quantization_config.torch_dtype , offload_folder=__UpperCamelCase , offload_state_dict=__UpperCamelCase , keep_in_fpaa_modules=bnb_quantization_config.keep_in_fpaa_modules , offload_abit_bnb=load_in_abit and offload , )
return dispatch_model(__UpperCamelCase , device_map=__UpperCamelCase , offload_dir=__UpperCamelCase )
def snake_case ( UpperCamelCase__ : Tuple , UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : str=None , UpperCamelCase__ : Union[str, Any]=None , UpperCamelCase__ : int=None ) -> Union[str, Any]:
if device_map is None:
if torch.cuda.is_available():
lowerCamelCase : int = {"""""": torch.cuda.current_device()}
else:
raise RuntimeError("""No GPU found. A GPU is needed for quantization.""" )
logger.info("""The device_map was not initialized.""" """Setting device_map to `{'':torch.cuda.current_device()}`.""" )
if isinstance(__UpperCamelCase , __UpperCamelCase ):
if device_map not in ["auto", "balanced", "balanced_low_0", "sequential"]:
raise ValueError(
"""If passing a string for `device_map`, please choose 'auto', 'balanced', 'balanced_low_0' or """
"""'sequential'.""" )
lowerCamelCase : Any = {}
special_dtypes.update(
{
name: bnb_quantization_config.torch_dtype
for name, _ in model.named_parameters()
if any(m in name for m in bnb_quantization_config.skip_modules )
} )
special_dtypes.update(
{
name: torch.floataa
for name, _ in model.named_parameters()
if any(m in name for m in bnb_quantization_config.keep_in_fpaa_modules )
} )
lowerCamelCase : List[str] = {}
lowerCamelCase : Any = special_dtypes
lowerCamelCase : Optional[Any] = no_split_module_classes
lowerCamelCase : Dict = bnb_quantization_config.target_dtype
# get max_memory for each device.
if device_map != "sequential":
lowerCamelCase : str = get_balanced_memory(
__UpperCamelCase , low_zero=(device_map == """balanced_low_0""") , max_memory=__UpperCamelCase , **__UpperCamelCase , )
lowerCamelCase : Optional[Any] = max_memory
lowerCamelCase : int = infer_auto_device_map(__UpperCamelCase , **__UpperCamelCase )
if isinstance(__UpperCamelCase , __UpperCamelCase ):
# check if don't have any quantized module on the cpu
lowerCamelCase : List[str] = bnb_quantization_config.skip_modules + bnb_quantization_config.keep_in_fpaa_modules
lowerCamelCase : Union[str, Any] = {
key: device_map[key] for key in device_map.keys() if key not in modules_not_to_convert
}
for device in ["cpu", "disk"]:
if device in device_map_without_some_modules.values():
if bnb_quantization_config.load_in_abit:
raise ValueError(
"""
Some modules are dispatched on the CPU or the disk. Make sure you have enough GPU RAM to fit
the quantized model. If you want to dispatch the model on the CPU or the disk while keeping
these modules in `torch_dtype`, you need to pass a custom `device_map` to
`load_and_quantize_model`. Check
https://huggingface.co/docs/accelerate/main/en/usage_guides/quantization#offload-modules-to-cpu-and-disk
for more details.
""" )
else:
logger.info(
"""Some modules are are offloaded to the CPU or the disk. Note that these modules will be converted to 8-bit""" )
del device_map_without_some_modules
return device_map
def snake_case ( UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : Dict , UpperCamelCase__ : Optional[Any]=None , UpperCamelCase__ : Tuple=None ) -> List[Any]:
if modules_to_not_convert is None:
lowerCamelCase : Optional[int] = []
lowerCamelCase , lowerCamelCase : Dict = _replace_with_bnb_layers(
__UpperCamelCase , __UpperCamelCase , __UpperCamelCase , __UpperCamelCase )
if not has_been_replaced:
logger.warning(
"""You are loading your model in 8bit or 4bit but no linear modules were found in your model."""
""" this can happen for some architectures such as gpt2 that uses Conv1D instead of Linear layers."""
""" Please double check your model architecture, or submit an issue on github if you think this is"""
""" a bug.""" )
return model
def snake_case ( UpperCamelCase__ : List[Any] , UpperCamelCase__ : Dict , UpperCamelCase__ : Union[str, Any]=None , UpperCamelCase__ : List[Any]=None , ) -> int:
lowerCamelCase : Optional[Any] = False
for name, module in model.named_children():
if current_key_name is None:
lowerCamelCase : int = []
current_key_name.append(__UpperCamelCase )
if isinstance(__UpperCamelCase , nn.Linear ) and name not in modules_to_not_convert:
# Check if the current key is not in the `modules_to_not_convert`
lowerCamelCase : int = """.""".join(__UpperCamelCase )
lowerCamelCase : Optional[int] = True
for key in modules_to_not_convert:
if (
(key in current_key_name_str) and (key + "." in current_key_name_str)
) or key == current_key_name_str:
lowerCamelCase : Union[str, Any] = False
break
if proceed:
# Load bnb module with empty weight and replace ``nn.Linear` module
if bnb_quantization_config.load_in_abit:
lowerCamelCase : Union[str, Any] = bnb.nn.LinearabitLt(
module.in_features , module.out_features , module.bias is not None , has_fpaa_weights=__UpperCamelCase , threshold=bnb_quantization_config.llm_inta_threshold , )
elif bnb_quantization_config.load_in_abit:
lowerCamelCase : Dict = bnb.nn.Linearabit(
module.in_features , module.out_features , module.bias is not None , bnb_quantization_config.bnb_abit_compute_dtype , compress_statistics=bnb_quantization_config.bnb_abit_use_double_quant , quant_type=bnb_quantization_config.bnb_abit_quant_type , )
else:
raise ValueError("""load_in_8bit and load_in_4bit can't be both False""" )
lowerCamelCase : Dict = module.weight.data
if module.bias is not None:
lowerCamelCase : Tuple = module.bias.data
bnb_module.requires_grad_(__UpperCamelCase )
setattr(__UpperCamelCase , __UpperCamelCase , __UpperCamelCase )
lowerCamelCase : Dict = True
if len(list(module.children() ) ) > 0:
lowerCamelCase , lowerCamelCase : Tuple = _replace_with_bnb_layers(
__UpperCamelCase , __UpperCamelCase , __UpperCamelCase , __UpperCamelCase )
lowerCamelCase : List[str] = has_been_replaced | _has_been_replaced
# Remove the last key for recursion
current_key_name.pop(-1 )
return model, has_been_replaced
def snake_case ( UpperCamelCase__ : List[Any] ) -> Union[str, Any]:
# Create a copy of the model
with init_empty_weights():
lowerCamelCase : str = deepcopy(__UpperCamelCase ) # this has 0 cost since it is done inside `init_empty_weights` context manager`
lowerCamelCase : Any = find_tied_parameters(__UpperCamelCase )
# For compatibility with Accelerate < 0.18
if isinstance(__UpperCamelCase , __UpperCamelCase ):
lowerCamelCase : Tuple = sum(list(tied_params.values() ) , [] ) + list(tied_params.keys() )
else:
lowerCamelCase : int = sum(__UpperCamelCase , [] )
lowerCamelCase : Optional[Any] = len(__UpperCamelCase ) > 0
# Check if it is a base model
lowerCamelCase : int = False
if hasattr(__UpperCamelCase , """base_model_prefix""" ):
lowerCamelCase : Optional[int] = not hasattr(__UpperCamelCase , model.base_model_prefix )
# Ignore this for base models (BertModel, GPT2Model, etc.)
if (not has_tied_params) and is_base_model:
return []
# otherwise they have an attached head
lowerCamelCase : List[str] = list(model.named_children() )
lowerCamelCase : str = [list_modules[-1][0]]
# add last module together with tied weights
lowerCamelCase : Tuple = set(__UpperCamelCase ) - set(__UpperCamelCase )
lowerCamelCase : Tuple = list(set(__UpperCamelCase ) ) + list(__UpperCamelCase )
# remove ".weight" from the keys
lowerCamelCase : Optional[Any] = [""".weight""", """.bias"""]
lowerCamelCase : Tuple = []
for name in list_untouched:
for name_to_remove in names_to_remove:
if name_to_remove in name:
lowerCamelCase : Any = name.replace(__UpperCamelCase , """""" )
filtered_module_names.append(__UpperCamelCase )
return filtered_module_names
def snake_case ( UpperCamelCase__ : Optional[int] ) -> str:
for m in model.modules():
if isinstance(__UpperCamelCase , bnb.nn.Linearabit ):
return True
return False
def snake_case ( UpperCamelCase__ : Dict ) -> Optional[int]:
return next(parameter.parameters() ).device
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Optional[int] , UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : Tuple , UpperCamelCase__ : Dict , UpperCamelCase__ : int , UpperCamelCase__ : Tuple ) -> Union[str, Any]:
# if it is not quantized, we quantize and offload the quantized weights and the SCB stats
if fpaa_statistics is None:
set_module_tensor_to_device(__UpperCamelCase , __UpperCamelCase , 0 , dtype=__UpperCamelCase , value=__UpperCamelCase )
lowerCamelCase : Any = param_name
lowerCamelCase : Any = model
if "." in tensor_name:
lowerCamelCase : Union[str, Any] = tensor_name.split(""".""" )
for split in splits[:-1]:
lowerCamelCase : Tuple = getattr(__UpperCamelCase , __UpperCamelCase )
if new_module is None:
raise ValueError(F'{module} has no attribute {split}.' )
lowerCamelCase : Union[str, Any] = new_module
lowerCamelCase : Dict = splits[-1]
# offload weights
lowerCamelCase : Optional[Any] = False
offload_weight(module._parameters[tensor_name] , __UpperCamelCase , __UpperCamelCase , index=__UpperCamelCase )
if hasattr(module._parameters[tensor_name] , """SCB""" ):
offload_weight(
module._parameters[tensor_name].SCB , param_name.replace("""weight""" , """SCB""" ) , __UpperCamelCase , index=__UpperCamelCase , )
else:
offload_weight(__UpperCamelCase , __UpperCamelCase , __UpperCamelCase , index=__UpperCamelCase )
offload_weight(__UpperCamelCase , param_name.replace("""weight""" , """SCB""" ) , __UpperCamelCase , index=__UpperCamelCase )
set_module_tensor_to_device(__UpperCamelCase , __UpperCamelCase , """meta""" , dtype=__UpperCamelCase , value=torch.empty(*param.size() ) )
| 713
|
"""simple docstring"""
import inspect
import unittest
from transformers import ConvNextConfig
from transformers.testing_utils import require_torch, require_vision, slow, torch_device
from transformers.utils import cached_property, is_torch_available, is_vision_available
from ...test_backbone_common import BackboneTesterMixin
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import ConvNextBackbone, ConvNextForImageClassification, ConvNextModel
from transformers.models.convnext.modeling_convnext import CONVNEXT_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import AutoImageProcessor
class A__ :
"""simple docstring"""
def __init__( self: Optional[Any] , __a: Union[str, Any] , __a: Optional[Any]=13 , __a: Optional[Any]=32 , __a: Dict=3 , __a: int=4 , __a: Dict=[10, 20, 30, 40] , __a: int=[2, 2, 3, 2] , __a: Any=True , __a: List[Any]=True , __a: Any=37 , __a: Optional[int]="gelu" , __a: List[str]=10 , __a: Optional[int]=0.02 , __a: Dict=["stage2", "stage3", "stage4"] , __a: List[str]=[2, 3, 4] , __a: List[str]=None , )-> Union[str, Any]:
lowerCamelCase : Optional[int] = parent
lowerCamelCase : Optional[int] = batch_size
lowerCamelCase : Any = image_size
lowerCamelCase : Tuple = num_channels
lowerCamelCase : str = num_stages
lowerCamelCase : List[str] = hidden_sizes
lowerCamelCase : str = depths
lowerCamelCase : Dict = is_training
lowerCamelCase : Optional[Any] = use_labels
lowerCamelCase : List[str] = intermediate_size
lowerCamelCase : List[str] = hidden_act
lowerCamelCase : List[str] = num_labels
lowerCamelCase : Union[str, Any] = initializer_range
lowerCamelCase : List[Any] = out_features
lowerCamelCase : Optional[Any] = out_indices
lowerCamelCase : int = scope
def a__ ( self: str )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : Dict = None
if self.use_labels:
lowerCamelCase : Dict = ids_tensor([self.batch_size] , self.num_labels )
lowerCamelCase : Any = self.get_config()
return config, pixel_values, labels
def a__ ( self: Dict )-> Union[str, Any]:
return ConvNextConfig(
num_channels=self.num_channels , hidden_sizes=self.hidden_sizes , depths=self.depths , num_stages=self.num_stages , hidden_act=self.hidden_act , is_decoder=__a , initializer_range=self.initializer_range , out_features=self.out_features , out_indices=self.out_indices , num_labels=self.num_labels , )
def a__ ( self: Optional[Any] , __a: List[Any] , __a: Any , __a: int )-> List[Any]:
lowerCamelCase : Optional[int] = ConvNextModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a )
# expected last hidden states: B, C, H // 32, W // 32
self.parent.assertEqual(
result.last_hidden_state.shape , (self.batch_size, self.hidden_sizes[-1], self.image_size // 32, self.image_size // 32) , )
def a__ ( self: int , __a: Union[str, Any] , __a: List[Any] , __a: Tuple )-> Optional[int]:
lowerCamelCase : str = ConvNextForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) )
def a__ ( self: List[Any] , __a: Any , __a: Optional[int] , __a: Tuple )-> List[str]:
lowerCamelCase : List[str] = ConvNextBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a )
# verify hidden states
self.parent.assertEqual(len(result.feature_maps ) , len(config.out_features ) )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.hidden_sizes[1], 4, 4] )
# verify channels
self.parent.assertEqual(len(model.channels ) , len(config.out_features ) )
self.parent.assertListEqual(model.channels , config.hidden_sizes[1:] )
# verify backbone works with out_features=None
lowerCamelCase : Tuple = None
lowerCamelCase : List[str] = ConvNextBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[Any] = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , 1 )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.hidden_sizes[-1], 1, 1] )
# verify channels
self.parent.assertEqual(len(model.channels ) , 1 )
self.parent.assertListEqual(model.channels , [config.hidden_sizes[-1]] )
def a__ ( self: Optional[Any] )-> Any:
lowerCamelCase : List[Any] = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : Tuple = config_and_inputs
lowerCamelCase : int = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : int =(
(
ConvNextModel,
ConvNextForImageClassification,
ConvNextBackbone,
)
if is_torch_available()
else ()
)
snake_case__ : str =(
{'''feature-extraction''': ConvNextModel, '''image-classification''': ConvNextForImageClassification}
if is_torch_available()
else {}
)
snake_case__ : Union[str, Any] =True
snake_case__ : Optional[int] =False
snake_case__ : Tuple =False
snake_case__ : Union[str, Any] =False
snake_case__ : Tuple =False
def a__ ( self: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Tuple = ConvNextModelTester(self )
lowerCamelCase : List[Any] = ConfigTester(self , config_class=__a , has_text_modality=__a , hidden_size=37 )
def a__ ( self: Optional[int] )-> Dict:
self.create_and_test_config_common_properties()
self.config_tester.create_and_test_config_to_json_string()
self.config_tester.create_and_test_config_to_json_file()
self.config_tester.create_and_test_config_from_and_save_pretrained()
self.config_tester.create_and_test_config_with_num_labels()
self.config_tester.check_config_can_be_init_without_params()
self.config_tester.check_config_arguments_init()
def a__ ( self: Optional[int] )-> Optional[Any]:
return
@unittest.skip(reason="""ConvNext does not use inputs_embeds""" )
def a__ ( self: int )-> Dict:
pass
@unittest.skip(reason="""ConvNext does not support input and output embeddings""" )
def a__ ( self: Dict )-> Optional[Any]:
pass
@unittest.skip(reason="""ConvNext does not use feedforward chunking""" )
def a__ ( self: int )-> List[Any]:
pass
def a__ ( self: Union[str, Any] )-> int:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : Any = model_class(__a )
lowerCamelCase : str = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Optional[Any] = [*signature.parameters.keys()]
lowerCamelCase : List[str] = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: Optional[int] )-> str:
lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: str )-> int:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_backbone(*__a )
def a__ ( self: int )-> Optional[int]:
def check_hidden_states_output(__a: Tuple , __a: int , __a: Tuple ):
lowerCamelCase : str = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : Tuple = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : int = outputs.encoder_hidden_states if config.is_encoder_decoder else outputs.hidden_states
lowerCamelCase : Optional[int] = self.model_tester.num_stages
self.assertEqual(len(__a ) , expected_num_stages + 1 )
# ConvNext's feature maps are of shape (batch_size, num_channels, height, width)
self.assertListEqual(
list(hidden_states[0].shape[-2:] ) , [self.model_tester.image_size // 4, self.model_tester.image_size // 4] , )
lowerCamelCase , lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : List[Any] = True
check_hidden_states_output(__a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : Tuple = True
check_hidden_states_output(__a , __a , __a )
def a__ ( self: Dict )-> Optional[Any]:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_classification(*__a )
@slow
def a__ ( self: Optional[Any] )-> Tuple:
for model_name in CONVNEXT_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : str = ConvNextModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def snake_case ( ) -> Optional[int]:
lowerCamelCase : int = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
return image
@require_torch
@require_vision
class A__ ( unittest.TestCase):
"""simple docstring"""
@cached_property
def a__ ( self: Dict )-> Union[str, Any]:
return AutoImageProcessor.from_pretrained("""facebook/convnext-tiny-224""" ) if is_vision_available() else None
@slow
def a__ ( self: List[str] )-> Dict:
lowerCamelCase : Tuple = ConvNextForImageClassification.from_pretrained("""facebook/convnext-tiny-224""" ).to(__a )
lowerCamelCase : Dict = self.default_image_processor
lowerCamelCase : Union[str, Any] = prepare_img()
lowerCamelCase : Optional[Any] = image_processor(images=__a , return_tensors="""pt""" ).to(__a )
# forward pass
with torch.no_grad():
lowerCamelCase : Any = model(**__a )
# verify the logits
lowerCamelCase : Union[str, Any] = torch.Size((1, 1_000) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : Tuple = torch.tensor([-0.02_60, -0.47_39, 0.19_11] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3] , __a , atol=1e-4 ) )
@require_torch
class A__ ( unittest.TestCase , __lowercase):
"""simple docstring"""
snake_case__ : Union[str, Any] =(ConvNextBackbone,) if is_torch_available() else ()
snake_case__ : Optional[Any] =ConvNextConfig
snake_case__ : Optional[Any] =False
def a__ ( self: List[str] )-> int:
lowerCamelCase : Dict = ConvNextModelTester(self )
| 42
| 0
|
"""simple docstring"""
import argparse
import json
import os
import evaluate
import torch
from datasets import load_dataset
from torch.optim import AdamW
from torch.utils.data import DataLoader
from transformers import AutoModelForSequenceClassification, AutoTokenizer, get_linear_schedule_with_warmup, set_seed
from accelerate import Accelerator, DistributedType
from accelerate.utils.deepspeed import DummyOptim, DummyScheduler
__lowerCamelCase :Tuple = 16
__lowerCamelCase :Union[str, Any] = 32
def snake_case ( UpperCamelCase__ : List[str] , UpperCamelCase__ : Dict = 16 , UpperCamelCase__ : str = "bert-base-cased" ) -> List[Any]:
lowerCamelCase : Dict = AutoTokenizer.from_pretrained(UpperCamelCase__ )
lowerCamelCase : Optional[int] = load_dataset("""glue""" , """mrpc""" )
def tokenize_function(UpperCamelCase__ : str ):
# max_length=None => use the model max length (it's actually the default)
lowerCamelCase : List[str] = tokenizer(examples["""sentence1"""] , examples["""sentence2"""] , truncation=UpperCamelCase__ , max_length=UpperCamelCase__ )
return outputs
# Apply the method we just defined to all the examples in all the splits of the dataset
lowerCamelCase : List[str] = datasets.map(
UpperCamelCase__ , batched=UpperCamelCase__ , remove_columns=["""idx""", """sentence1""", """sentence2"""] , load_from_cache_file=UpperCamelCase__ )
# We also rename the 'label' column to 'labels' which is the expected name for labels by the models of the
# transformers library
lowerCamelCase : Optional[Any] = tokenized_datasets.rename_column("""label""" , """labels""" )
def collate_fn(UpperCamelCase__ : List[str] ):
# On TPU it's best to pad everything to the same length or training will be very slow.
if accelerator.distributed_type == DistributedType.TPU:
return tokenizer.pad(UpperCamelCase__ , padding="""max_length""" , max_length=128 , return_tensors="""pt""" )
return tokenizer.pad(UpperCamelCase__ , padding="""longest""" , return_tensors="""pt""" )
# Instantiate dataloaders.
lowerCamelCase : Any = DataLoader(
tokenized_datasets["""train"""] , shuffle=UpperCamelCase__ , collate_fn=UpperCamelCase__ , batch_size=UpperCamelCase__ )
lowerCamelCase : int = DataLoader(
tokenized_datasets["""validation"""] , shuffle=UpperCamelCase__ , collate_fn=UpperCamelCase__ , batch_size=UpperCamelCase__ )
return train_dataloader, eval_dataloader
def snake_case ( UpperCamelCase__ : Optional[int] , UpperCamelCase__ : List[str] ) -> Optional[Any]:
# Initialize accelerator
lowerCamelCase : List[Any] = Accelerator()
# Sample hyper-parameters for learning rate, batch size, seed and a few other HPs
lowerCamelCase : Any = config["""lr"""]
lowerCamelCase : Dict = int(config["""num_epochs"""] )
lowerCamelCase : List[Any] = int(config["""seed"""] )
lowerCamelCase : List[str] = int(config["""batch_size"""] )
lowerCamelCase : List[Any] = args.model_name_or_path
set_seed(UpperCamelCase__ )
lowerCamelCase : List[str] = get_dataloaders(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
# Instantiate the model (we build the model here so that the seed also control new weights initialization)
lowerCamelCase : Tuple = AutoModelForSequenceClassification.from_pretrained(UpperCamelCase__ , return_dict=UpperCamelCase__ )
# Instantiate optimizer
lowerCamelCase : Tuple = (
AdamW
if accelerator.state.deepspeed_plugin is None
or """optimizer""" not in accelerator.state.deepspeed_plugin.deepspeed_config
else DummyOptim
)
lowerCamelCase : List[str] = optimizer_cls(params=model.parameters() , lr=UpperCamelCase__ )
if accelerator.state.deepspeed_plugin is not None:
lowerCamelCase : Any = accelerator.state.deepspeed_plugin.deepspeed_config[
"""gradient_accumulation_steps"""
]
else:
lowerCamelCase : Any = 1
lowerCamelCase : str = (len(UpperCamelCase__ ) * num_epochs) // gradient_accumulation_steps
# Instantiate scheduler
if (
accelerator.state.deepspeed_plugin is None
or "scheduler" not in accelerator.state.deepspeed_plugin.deepspeed_config
):
lowerCamelCase : List[Any] = get_linear_schedule_with_warmup(
optimizer=UpperCamelCase__ , num_warmup_steps=0 , num_training_steps=UpperCamelCase__ , )
else:
lowerCamelCase : Dict = DummyScheduler(UpperCamelCase__ , total_num_steps=UpperCamelCase__ , warmup_num_steps=0 )
# Prepare everything
# There is no specific order to remember, we just need to unpack the objects in the same order we gave them to the
# prepare method.
lowerCamelCase : Dict = accelerator.prepare(
UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
# We need to keep track of how many total steps we have iterated over
lowerCamelCase : str = 0
# We also need to keep track of the stating epoch so files are named properly
lowerCamelCase : Tuple = 0
# Now we train the model
lowerCamelCase : str = evaluate.load("""glue""" , """mrpc""" )
lowerCamelCase : int = 0
lowerCamelCase : Tuple = {}
for epoch in range(UpperCamelCase__ , UpperCamelCase__ ):
model.train()
for step, batch in enumerate(UpperCamelCase__ ):
lowerCamelCase : List[str] = model(**UpperCamelCase__ )
lowerCamelCase : int = outputs.loss
lowerCamelCase : List[Any] = loss / gradient_accumulation_steps
accelerator.backward(UpperCamelCase__ )
if step % gradient_accumulation_steps == 0:
optimizer.step()
lr_scheduler.step()
optimizer.zero_grad()
overall_step += 1
model.eval()
lowerCamelCase : List[Any] = 0
for step, batch in enumerate(UpperCamelCase__ ):
# We could avoid this line since we set the accelerator with `device_placement=True`.
batch.to(accelerator.device )
with torch.no_grad():
lowerCamelCase : Optional[int] = model(**UpperCamelCase__ )
lowerCamelCase : Dict = outputs.logits.argmax(dim=-1 )
# It is slightly faster to call this once, than multiple times
lowerCamelCase : str = accelerator.gather(
(predictions, batch["""labels"""]) ) # If we are in a multiprocess environment, the last batch has duplicates
if accelerator.use_distributed:
if step == len(UpperCamelCase__ ) - 1:
lowerCamelCase : int = predictions[: len(eval_dataloader.dataset ) - samples_seen]
lowerCamelCase : Union[str, Any] = references[: len(eval_dataloader.dataset ) - samples_seen]
else:
samples_seen += references.shape[0]
metric.add_batch(
predictions=UpperCamelCase__ , references=UpperCamelCase__ , )
lowerCamelCase : List[str] = metric.compute()
# Use accelerator.print to print only on the main process.
accelerator.print(F'epoch {epoch}:' , UpperCamelCase__ )
lowerCamelCase : str = eval_metric["""accuracy"""]
if best_performance < eval_metric["accuracy"]:
lowerCamelCase : Dict = eval_metric["""accuracy"""]
if args.performance_lower_bound is not None:
assert (
args.performance_lower_bound <= best_performance
), F'Best performance metric {best_performance} is lower than the lower bound {args.performance_lower_bound}'
accelerator.wait_for_everyone()
if accelerator.is_main_process:
with open(os.path.join(args.output_dir , """all_results.json""" ) , """w""" ) as f:
json.dump(UpperCamelCase__ , UpperCamelCase__ )
def snake_case ( ) -> Optional[int]:
lowerCamelCase : int = argparse.ArgumentParser(description="""Simple example of training script tracking peak GPU memory usage.""" )
parser.add_argument(
"""--model_name_or_path""" , type=UpperCamelCase__ , default="""bert-base-cased""" , help="""Path to pretrained model or model identifier from huggingface.co/models.""" , required=UpperCamelCase__ , )
parser.add_argument(
"""--output_dir""" , type=UpperCamelCase__ , default=""".""" , help="""Optional save directory where all checkpoint folders will be stored. Default is the current working directory.""" , )
parser.add_argument(
"""--performance_lower_bound""" , type=UpperCamelCase__ , default=UpperCamelCase__ , help="""Optional lower bound for the performance metric. If set, the training will throw error when the performance metric drops below this value.""" , )
parser.add_argument(
"""--num_epochs""" , type=UpperCamelCase__ , default=3 , help="""Number of train epochs.""" , )
lowerCamelCase : Optional[int] = parser.parse_args()
lowerCamelCase : Any = {"""lr""": 2E-5, """num_epochs""": args.num_epochs, """seed""": 42, """batch_size""": 16}
training_function(UpperCamelCase__ , UpperCamelCase__ )
if __name__ == "__main__":
main()
| 714
|
"""simple docstring"""
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :Optional[int] = logging.get_logger(__name__)
__lowerCamelCase :List[str] = {
'google/realm-cc-news-pretrained-embedder': (
'https://huggingface.co/google/realm-cc-news-pretrained-embedder/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-encoder': (
'https://huggingface.co/google/realm-cc-news-pretrained-encoder/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-scorer': (
'https://huggingface.co/google/realm-cc-news-pretrained-scorer/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-openqa': (
'https://huggingface.co/google/realm-cc-news-pretrained-openqa/aresolve/main/config.json'
),
'google/realm-orqa-nq-openqa': 'https://huggingface.co/google/realm-orqa-nq-openqa/resolve/main/config.json',
'google/realm-orqa-nq-reader': 'https://huggingface.co/google/realm-orqa-nq-reader/resolve/main/config.json',
'google/realm-orqa-wq-openqa': 'https://huggingface.co/google/realm-orqa-wq-openqa/resolve/main/config.json',
'google/realm-orqa-wq-reader': 'https://huggingface.co/google/realm-orqa-wq-reader/resolve/main/config.json',
# See all REALM models at https://huggingface.co/models?filter=realm
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Optional[Any] ='''realm'''
def __init__( self: Union[str, Any] , __a: List[Any]=30_522 , __a: List[Any]=768 , __a: List[Any]=128 , __a: Union[str, Any]=12 , __a: Union[str, Any]=12 , __a: Optional[Any]=8 , __a: Dict=3_072 , __a: List[Any]="gelu_new" , __a: List[Any]=0.1 , __a: Tuple=0.1 , __a: Optional[Any]=512 , __a: Optional[int]=2 , __a: str=0.02 , __a: int=1e-1_2 , __a: Optional[Any]=256 , __a: Any=10 , __a: Dict=1e-3 , __a: Optional[Any]=5 , __a: Dict=320 , __a: Tuple=13_353_718 , __a: List[Any]=5_000 , __a: Dict=1 , __a: int=0 , __a: Dict=2 , **__a: List[str] , )-> Any:
super().__init__(pad_token_id=__a , bos_token_id=__a , eos_token_id=__a , **__a )
# Common config
lowerCamelCase : Optional[Any] = vocab_size
lowerCamelCase : str = max_position_embeddings
lowerCamelCase : Dict = hidden_size
lowerCamelCase : Dict = retriever_proj_size
lowerCamelCase : Optional[Any] = num_hidden_layers
lowerCamelCase : List[str] = num_attention_heads
lowerCamelCase : Tuple = num_candidates
lowerCamelCase : int = intermediate_size
lowerCamelCase : Dict = hidden_act
lowerCamelCase : List[str] = hidden_dropout_prob
lowerCamelCase : Dict = attention_probs_dropout_prob
lowerCamelCase : Optional[int] = initializer_range
lowerCamelCase : Dict = type_vocab_size
lowerCamelCase : Optional[Any] = layer_norm_eps
# Reader config
lowerCamelCase : List[str] = span_hidden_size
lowerCamelCase : Dict = max_span_width
lowerCamelCase : Optional[Any] = reader_layer_norm_eps
lowerCamelCase : Optional[int] = reader_beam_size
lowerCamelCase : List[Any] = reader_seq_len
# Retrieval config
lowerCamelCase : int = num_block_records
lowerCamelCase : Dict = searcher_beam_size
| 42
| 0
|
"""simple docstring"""
__lowerCamelCase :List[str] = tuple[float, float, float]
__lowerCamelCase :Union[str, Any] = tuple[float, float, float]
def snake_case ( UpperCamelCase__ : Optional[int] , UpperCamelCase__ : List[Any] ) -> int:
lowerCamelCase : Optional[Any] = end_pointa[0] - end_pointa[0]
lowerCamelCase : Optional[int] = end_pointa[1] - end_pointa[1]
lowerCamelCase : int = end_pointa[2] - end_pointa[2]
return (x, y, z)
def snake_case ( UpperCamelCase__ : Optional[int] , UpperCamelCase__ : List[str] ) -> Optional[Any]:
lowerCamelCase : Any = ab[1] * ac[2] - ab[2] * ac[1] # *i
lowerCamelCase : List[Any] = (ab[0] * ac[2] - ab[2] * ac[0]) * -1 # *j
lowerCamelCase : Tuple = ab[0] * ac[1] - ab[1] * ac[0] # *k
return (x, y, z)
def snake_case ( UpperCamelCase__ : Any , UpperCamelCase__ : Union[str, Any] ) -> str:
return tuple(round(SCREAMING_SNAKE_CASE_ , SCREAMING_SNAKE_CASE_ ) for x in vector ) == (0, 0, 0)
def snake_case ( UpperCamelCase__ : Optional[int] , UpperCamelCase__ : List[str] , UpperCamelCase__ : int , UpperCamelCase__ : Optional[Any] = 10 ) -> Tuple:
lowerCamelCase : Tuple = create_vector(SCREAMING_SNAKE_CASE_ , SCREAMING_SNAKE_CASE_ )
lowerCamelCase : List[str] = create_vector(SCREAMING_SNAKE_CASE_ , SCREAMING_SNAKE_CASE_ )
return is_zero_vector(get_ad_vectors_cross(SCREAMING_SNAKE_CASE_ , SCREAMING_SNAKE_CASE_ ) , SCREAMING_SNAKE_CASE_ )
| 715
|
"""simple docstring"""
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :Tuple = logging.get_logger(__name__)
__lowerCamelCase :Any = {
'vinvino02/glpn-kitti': 'https://huggingface.co/vinvino02/glpn-kitti/resolve/main/config.json',
# See all GLPN models at https://huggingface.co/models?filter=glpn
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Tuple ='''glpn'''
def __init__( self: Dict , __a: List[str]=3 , __a: Optional[int]=4 , __a: Dict=[2, 2, 2, 2] , __a: str=[8, 4, 2, 1] , __a: Optional[int]=[32, 64, 160, 256] , __a: Dict=[7, 3, 3, 3] , __a: Dict=[4, 2, 2, 2] , __a: Optional[Any]=[1, 2, 5, 8] , __a: Tuple=[4, 4, 4, 4] , __a: int="gelu" , __a: Union[str, Any]=0.0 , __a: str=0.0 , __a: Union[str, Any]=0.02 , __a: str=0.1 , __a: Union[str, Any]=1e-6 , __a: Any=64 , __a: Dict=10 , __a: Union[str, Any]=-1 , **__a: Optional[Any] , )-> Dict:
super().__init__(**__a )
lowerCamelCase : Dict = num_channels
lowerCamelCase : Any = num_encoder_blocks
lowerCamelCase : Dict = depths
lowerCamelCase : List[str] = sr_ratios
lowerCamelCase : Dict = hidden_sizes
lowerCamelCase : Tuple = patch_sizes
lowerCamelCase : Optional[int] = strides
lowerCamelCase : Optional[Any] = mlp_ratios
lowerCamelCase : Union[str, Any] = num_attention_heads
lowerCamelCase : List[str] = hidden_act
lowerCamelCase : Any = hidden_dropout_prob
lowerCamelCase : Optional[int] = attention_probs_dropout_prob
lowerCamelCase : List[Any] = initializer_range
lowerCamelCase : Dict = drop_path_rate
lowerCamelCase : Any = layer_norm_eps
lowerCamelCase : Optional[Any] = decoder_hidden_size
lowerCamelCase : Tuple = max_depth
lowerCamelCase : Optional[Any] = head_in_index
| 42
| 0
|
"""simple docstring"""
from __future__ import annotations
def snake_case ( UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : List[str] ) -> bool:
lowerCamelCase : Optional[int] = get_failure_array(_lowerCAmelCase )
# 2) Step through text searching for pattern
lowerCamelCase : Optional[Any] = 0, 0 # index into text, pattern
while i < len(_lowerCAmelCase ):
if pattern[j] == text[i]:
if j == (len(_lowerCAmelCase ) - 1):
return True
j += 1
# if this is a prefix in our pattern
# just go back far enough to continue
elif j > 0:
lowerCamelCase : str = failure[j - 1]
continue
i += 1
return False
def snake_case ( UpperCamelCase__ : str ) -> list[int]:
lowerCamelCase : List[Any] = [0]
lowerCamelCase : List[Any] = 0
lowerCamelCase : Union[str, Any] = 1
while j < len(_lowerCAmelCase ):
if pattern[i] == pattern[j]:
i += 1
elif i > 0:
lowerCamelCase : Optional[int] = failure[i - 1]
continue
j += 1
failure.append(_lowerCAmelCase )
return failure
if __name__ == "__main__":
# Test 1)
__lowerCamelCase :Optional[Any] = """abc1abc12"""
__lowerCamelCase :int = """alskfjaldsabc1abc1abc12k23adsfabcabc"""
__lowerCamelCase :Tuple = """alskfjaldsk23adsfabcabc"""
assert kmp(pattern, texta) and not kmp(pattern, texta)
# Test 2)
__lowerCamelCase :List[Any] = """ABABX"""
__lowerCamelCase :Dict = """ABABZABABYABABX"""
assert kmp(pattern, text)
# Test 3)
__lowerCamelCase :Optional[int] = """AAAB"""
__lowerCamelCase :Any = """ABAAAAAB"""
assert kmp(pattern, text)
# Test 4)
__lowerCamelCase :Optional[Any] = """abcdabcy"""
__lowerCamelCase :Optional[Any] = """abcxabcdabxabcdabcdabcy"""
assert kmp(pattern, text)
# Test 5)
__lowerCamelCase :str = """aabaabaaa"""
assert get_failure_array(pattern) == [0, 1, 0, 1, 2, 3, 4, 5, 2]
| 716
|
"""simple docstring"""
from __future__ import annotations
import math
def snake_case ( UpperCamelCase__ : float , UpperCamelCase__ : int ) -> float:
lowerCamelCase : Dict = u
for i in range(1 , UpperCamelCase__ ):
lowerCamelCase : List[str] = temp * (u - i)
return temp
def snake_case ( ) -> None:
lowerCamelCase : List[Any] = int(input("""enter the numbers of values: """ ) )
lowerCamelCase : list[list[float]] = []
for _ in range(UpperCamelCase__ ):
y.append([] )
for i in range(UpperCamelCase__ ):
for j in range(UpperCamelCase__ ):
y[i].append(UpperCamelCase__ )
lowerCamelCase : Union[str, Any] = 0
print("""enter the values of parameters in a list: """ )
lowerCamelCase : Any = list(map(UpperCamelCase__ , input().split() ) )
print("""enter the values of corresponding parameters: """ )
for i in range(UpperCamelCase__ ):
lowerCamelCase : int = float(input() )
lowerCamelCase : Dict = int(input("""enter the value to interpolate: """ ) )
lowerCamelCase : List[Any] = (value - x[0]) / (x[1] - x[0])
# for calculating forward difference table
for i in range(1 , UpperCamelCase__ ):
for j in range(n - i ):
lowerCamelCase : str = y[j + 1][i - 1] - y[j][i - 1]
lowerCamelCase : Any = y[0][0]
for i in range(1 , UpperCamelCase__ ):
summ += (ucal(UpperCamelCase__ , UpperCamelCase__ ) * y[0][i]) / math.factorial(UpperCamelCase__ )
print(F'the value at {value} is {summ}' )
if __name__ == "__main__":
main()
| 42
| 0
|
"""simple docstring"""
from dataclasses import dataclass, field
from typing import Tuple
from ..utils import cached_property, is_tf_available, logging, requires_backends
from .benchmark_args_utils import BenchmarkArguments
if is_tf_available():
import tensorflow as tf
__lowerCamelCase :Any = logging.get_logger(__name__)
@dataclass
class A__ ( UpperCamelCase__):
"""simple docstring"""
snake_case__ : int =[
"""no_inference""",
"""no_cuda""",
"""no_tpu""",
"""no_speed""",
"""no_memory""",
"""no_env_print""",
"""no_multi_process""",
]
def __init__( self: Union[str, Any] , **__a: Optional[int] )-> Union[str, Any]:
for deprecated_arg in self.deprecated_args:
if deprecated_arg in kwargs:
lowerCamelCase : Any = deprecated_arg[3:]
lowerCamelCase : Tuple = not kwargs.pop(__a )
logger.warning(
f'{deprecated_arg} is depreciated. Please use --no-{positive_arg} or'
f' {positive_arg}={kwargs[positive_arg]}' )
lowerCamelCase : Tuple = kwargs.pop("""tpu_name""" , self.tpu_name )
lowerCamelCase : Dict = kwargs.pop("""device_idx""" , self.device_idx )
lowerCamelCase : List[str] = kwargs.pop("""eager_mode""" , self.eager_mode )
lowerCamelCase : Optional[int] = kwargs.pop("""use_xla""" , self.use_xla )
super().__init__(**__a )
snake_case__ : str =field(
default=UpperCamelCase__ , metadata={'''help''': '''Name of TPU'''} , )
snake_case__ : int =field(
default=0 , metadata={'''help''': '''CPU / GPU device index. Defaults to 0.'''} , )
snake_case__ : bool =field(default=UpperCamelCase__ , metadata={'''help''': '''Benchmark models in eager model.'''})
snake_case__ : bool =field(
default=UpperCamelCase__ , metadata={
'''help''': '''Benchmark models using XLA JIT compilation. Note that `eager_model` has to be set to `False`.'''
} , )
@cached_property
def a__ ( self: Optional[Any] )-> Tuple["tf.distribute.cluster_resolver.TPUClusterResolver"]:
requires_backends(self , ["""tf"""] )
lowerCamelCase : Tuple = None
if self.tpu:
try:
if self.tpu_name:
lowerCamelCase : Any = tf.distribute.cluster_resolver.TPUClusterResolver(self.tpu_name )
else:
lowerCamelCase : List[str] = tf.distribute.cluster_resolver.TPUClusterResolver()
except ValueError:
lowerCamelCase : List[str] = None
return tpu
@cached_property
def a__ ( self: Tuple )-> Tuple["tf.distribute.Strategy", "tf.distribute.cluster_resolver.TPUClusterResolver"]:
requires_backends(self , ["""tf"""] )
if self.is_tpu:
tf.config.experimental_connect_to_cluster(self._setup_tpu )
tf.tpu.experimental.initialize_tpu_system(self._setup_tpu )
lowerCamelCase : List[Any] = tf.distribute.TPUStrategy(self._setup_tpu )
else:
# currently no multi gpu is allowed
if self.is_gpu:
# TODO: Currently only single GPU is supported
tf.config.set_visible_devices(self.gpu_list[self.device_idx] , """GPU""" )
lowerCamelCase : str = tf.distribute.OneDeviceStrategy(device=f'/gpu:{self.device_idx}' )
else:
tf.config.set_visible_devices([] , """GPU""" ) # disable GPU
lowerCamelCase : Optional[int] = tf.distribute.OneDeviceStrategy(device=f'/cpu:{self.device_idx}' )
return strategy
@property
def a__ ( self: List[Any] )-> bool:
requires_backends(self , ["""tf"""] )
return self._setup_tpu is not None
@property
def a__ ( self: Tuple )-> "tf.distribute.Strategy":
requires_backends(self , ["""tf"""] )
return self._setup_strategy
@property
def a__ ( self: int )-> Union[str, Any]:
requires_backends(self , ["""tf"""] )
return tf.config.list_physical_devices("""GPU""" )
@property
def a__ ( self: Union[str, Any] )-> int:
requires_backends(self , ["""tf"""] )
if self.cuda:
return len(self.gpu_list )
return 0
@property
def a__ ( self: Tuple )-> bool:
return self.n_gpu > 0
| 717
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_sentencepiece_available
__lowerCamelCase :str = {}
try:
if not is_sentencepiece_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[Any] = ['GPTSw3Tokenizer']
if TYPE_CHECKING:
try:
if not is_sentencepiece_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .tokenization_gpt_swa import GPTSwaTokenizer
else:
import sys
__lowerCamelCase :Tuple = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
def snake_case ( UpperCamelCase__ : Dict ) -> Union[str, Any]:
if not all(char in """01""" for char in bin_string ):
raise ValueError("""Non-binary value was passed to the function""" )
if not bin_string:
raise ValueError("""Empty string was passed to the function""" )
lowerCamelCase : List[Any] = ""
while len(__lowerCAmelCase ) % 3 != 0:
lowerCamelCase : Tuple = "0" + bin_string
lowerCamelCase : Dict = [
bin_string[index : index + 3]
for index in range(len(__lowerCAmelCase ) )
if index % 3 == 0
]
for bin_group in bin_string_in_3_list:
lowerCamelCase : int = 0
for index, val in enumerate(__lowerCAmelCase ):
oct_val += int(2 ** (2 - index) * int(__lowerCAmelCase ) )
oct_string += str(__lowerCAmelCase )
return oct_string
if __name__ == "__main__":
from doctest import testmod
testmod()
| 718
|
"""simple docstring"""
import argparse
import json
from collections import OrderedDict
from functools import partial
from pathlib import Path
import timm
import torch
from huggingface_hub import hf_hub_download
from transformers import LevitConfig, LevitForImageClassificationWithTeacher, LevitImageProcessor
from transformers.utils import logging
logging.set_verbosity_info()
__lowerCamelCase :Dict = logging.get_logger()
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : str , UpperCamelCase__ : LevitConfig , UpperCamelCase__ : Path , UpperCamelCase__ : bool = True ) -> Dict:
print(F'Converting {name}...' )
with torch.no_grad():
if hidden_sizes == 128:
if name[-1] == "S":
lowerCamelCase : Optional[Any] = timm.create_model("""levit_128s""" , pretrained=UpperCamelCase__ )
else:
lowerCamelCase : Dict = timm.create_model("""levit_128""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 192:
lowerCamelCase : Tuple = timm.create_model("""levit_192""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 256:
lowerCamelCase : Optional[int] = timm.create_model("""levit_256""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 384:
lowerCamelCase : Dict = timm.create_model("""levit_384""" , pretrained=UpperCamelCase__ )
from_model.eval()
lowerCamelCase : Optional[Any] = LevitForImageClassificationWithTeacher(UpperCamelCase__ ).eval()
lowerCamelCase : Tuple = OrderedDict()
lowerCamelCase : Optional[Any] = from_model.state_dict()
lowerCamelCase : str = list(from_model.state_dict().keys() )
lowerCamelCase : List[Any] = list(our_model.state_dict().keys() )
print(len(UpperCamelCase__ ) , len(UpperCamelCase__ ) )
for i in range(len(UpperCamelCase__ ) ):
lowerCamelCase : str = weights[og_keys[i]]
our_model.load_state_dict(UpperCamelCase__ )
lowerCamelCase : int = torch.randn((2, 3, 224, 224) )
lowerCamelCase : Any = from_model(UpperCamelCase__ )
lowerCamelCase : List[Any] = our_model(UpperCamelCase__ ).logits
assert torch.allclose(UpperCamelCase__ , UpperCamelCase__ ), "The model logits don't match the original one."
lowerCamelCase : Dict = name
print(UpperCamelCase__ )
if push_to_hub:
our_model.save_pretrained(save_directory / checkpoint_name )
lowerCamelCase : Optional[int] = LevitImageProcessor()
image_processor.save_pretrained(save_directory / checkpoint_name )
print(F'Pushed {checkpoint_name}' )
def snake_case ( UpperCamelCase__ : Path , UpperCamelCase__ : str = None , UpperCamelCase__ : bool = True ) -> Optional[int]:
lowerCamelCase : Optional[Any] = """imagenet-1k-id2label.json"""
lowerCamelCase : List[Any] = 1000
lowerCamelCase : Dict = (1, num_labels)
lowerCamelCase : List[Any] = """huggingface/label-files"""
lowerCamelCase : Optional[int] = num_labels
lowerCamelCase : List[str] = json.load(open(hf_hub_download(UpperCamelCase__ , UpperCamelCase__ , repo_type="""dataset""" ) , """r""" ) )
lowerCamelCase : Any = {int(UpperCamelCase__ ): v for k, v in idalabel.items()}
lowerCamelCase : List[Any] = idalabel
lowerCamelCase : str = {v: k for k, v in idalabel.items()}
lowerCamelCase : Tuple = partial(UpperCamelCase__ , num_labels=UpperCamelCase__ , idalabel=UpperCamelCase__ , labelaid=UpperCamelCase__ )
lowerCamelCase : Optional[int] = {
"""levit-128S""": 128,
"""levit-128""": 128,
"""levit-192""": 192,
"""levit-256""": 256,
"""levit-384""": 384,
}
lowerCamelCase : List[Any] = {
"""levit-128S""": ImageNetPreTrainedConfig(
hidden_sizes=[128, 256, 384] , num_attention_heads=[4, 6, 8] , depths=[2, 3, 4] , key_dim=[16, 16, 16] , drop_path_rate=0 , ),
"""levit-128""": ImageNetPreTrainedConfig(
hidden_sizes=[128, 256, 384] , num_attention_heads=[4, 8, 12] , depths=[4, 4, 4] , key_dim=[16, 16, 16] , drop_path_rate=0 , ),
"""levit-192""": ImageNetPreTrainedConfig(
hidden_sizes=[192, 288, 384] , num_attention_heads=[3, 5, 6] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0 , ),
"""levit-256""": ImageNetPreTrainedConfig(
hidden_sizes=[256, 384, 512] , num_attention_heads=[4, 6, 8] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0 , ),
"""levit-384""": ImageNetPreTrainedConfig(
hidden_sizes=[384, 512, 768] , num_attention_heads=[6, 9, 12] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0.1 , ),
}
if model_name:
convert_weight_and_push(
names_to_hidden_sizes[model_name] , UpperCamelCase__ , names_to_config[model_name] , UpperCamelCase__ , UpperCamelCase__ )
else:
for model_name, config in names_to_config.items():
convert_weight_and_push(names_to_hidden_sizes[model_name] , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return config, expected_shape
if __name__ == "__main__":
__lowerCamelCase :Union[str, Any] = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--model_name',
default=None,
type=str,
help='The name of the model you wish to convert, it must be one of the supported Levit* architecture,',
)
parser.add_argument(
'--pytorch_dump_folder_path',
default='levit-dump-folder/',
type=Path,
required=False,
help='Path to the output PyTorch model directory.',
)
parser.add_argument('--push_to_hub', action='store_true', help='Push model and image processor to the hub')
parser.add_argument(
'--no-push_to_hub',
dest='push_to_hub',
action='store_false',
help='Do not push model and image processor to the hub',
)
__lowerCamelCase :List[Any] = parser.parse_args()
__lowerCamelCase :Path = args.pytorch_dump_folder_path
pytorch_dump_folder_path.mkdir(exist_ok=True, parents=True)
convert_weights_and_push(pytorch_dump_folder_path, args.model_name, args.push_to_hub)
| 42
| 0
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_tf_available, is_torch_available
__lowerCamelCase :int = {
'configuration_tapas': ['TAPAS_PRETRAINED_CONFIG_ARCHIVE_MAP', 'TapasConfig'],
'tokenization_tapas': ['TapasTokenizer'],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Any = [
'TAPAS_PRETRAINED_MODEL_ARCHIVE_LIST',
'TapasForMaskedLM',
'TapasForQuestionAnswering',
'TapasForSequenceClassification',
'TapasModel',
'TapasPreTrainedModel',
'load_tf_weights_in_tapas',
]
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :List[Any] = [
'TF_TAPAS_PRETRAINED_MODEL_ARCHIVE_LIST',
'TFTapasForMaskedLM',
'TFTapasForQuestionAnswering',
'TFTapasForSequenceClassification',
'TFTapasModel',
'TFTapasPreTrainedModel',
]
if TYPE_CHECKING:
from .configuration_tapas import TAPAS_PRETRAINED_CONFIG_ARCHIVE_MAP, TapasConfig
from .tokenization_tapas import TapasTokenizer
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_tapas import (
TAPAS_PRETRAINED_MODEL_ARCHIVE_LIST,
TapasForMaskedLM,
TapasForQuestionAnswering,
TapasForSequenceClassification,
TapasModel,
TapasPreTrainedModel,
load_tf_weights_in_tapas,
)
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_tf_tapas import (
TF_TAPAS_PRETRAINED_MODEL_ARCHIVE_LIST,
TFTapasForMaskedLM,
TFTapasForQuestionAnswering,
TFTapasForSequenceClassification,
TFTapasModel,
TFTapasPreTrainedModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 719
|
"""simple docstring"""
import torch
from diffusers import KDPMaDiscreteScheduler
from diffusers.utils import torch_device
from .test_schedulers import SchedulerCommonTest
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Tuple =(KDPMaDiscreteScheduler,)
snake_case__ : Tuple =10
def a__ ( self: List[Any] , **__a: Optional[int] )-> Union[str, Any]:
lowerCamelCase : int = {
"""num_train_timesteps""": 1_100,
"""beta_start""": 0.00_01,
"""beta_end""": 0.02,
"""beta_schedule""": """linear""",
}
config.update(**__a )
return config
def a__ ( self: Union[str, Any] )-> Any:
for timesteps in [10, 50, 100, 1_000]:
self.check_over_configs(num_train_timesteps=__a )
def a__ ( self: str )-> int:
for beta_start, beta_end in zip([0.0_00_01, 0.00_01, 0.0_01] , [0.00_02, 0.0_02, 0.02] ):
self.check_over_configs(beta_start=__a , beta_end=__a )
def a__ ( self: int )-> Union[str, Any]:
for schedule in ["linear", "scaled_linear"]:
self.check_over_configs(beta_schedule=__a )
def a__ ( self: List[Any] )-> List[Any]:
for prediction_type in ["epsilon", "v_prediction"]:
self.check_over_configs(prediction_type=__a )
def a__ ( self: Union[str, Any] )-> int:
lowerCamelCase : List[str] = self.scheduler_classes[0]
lowerCamelCase : Union[str, Any] = self.get_scheduler_config(prediction_type="""v_prediction""" )
lowerCamelCase : List[str] = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps )
lowerCamelCase : Dict = self.dummy_model()
lowerCamelCase : List[Any] = self.dummy_sample_deter * scheduler.init_noise_sigma
lowerCamelCase : List[Any] = sample.to(__a )
for i, t in enumerate(scheduler.timesteps ):
lowerCamelCase : Optional[Any] = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[int] = model(__a , __a )
lowerCamelCase : Tuple = scheduler.step(__a , __a , __a )
lowerCamelCase : Optional[Any] = output.prev_sample
lowerCamelCase : List[str] = torch.sum(torch.abs(__a ) )
lowerCamelCase : Tuple = torch.mean(torch.abs(__a ) )
if torch_device in ["cpu", "mps"]:
assert abs(result_sum.item() - 4.6_9_3_4e-0_7 ) < 1e-2
assert abs(result_mean.item() - 6.1_1_1_2e-1_0 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 4.6_9_3_4_2_8_6_5_0_1_7_0_9_7_2e-0_7 ) < 1e-2
assert abs(result_mean.item() - 0.00_02 ) < 1e-3
def a__ ( self: Any )-> Any:
if torch_device == "mps":
return
lowerCamelCase : Dict = self.scheduler_classes[0]
lowerCamelCase : Dict = self.get_scheduler_config()
lowerCamelCase : int = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps )
lowerCamelCase : List[Any] = self.dummy_model()
lowerCamelCase : Optional[Any] = self.dummy_sample_deter * scheduler.init_noise_sigma
lowerCamelCase : Optional[int] = sample.to(__a )
for i, t in enumerate(scheduler.timesteps ):
lowerCamelCase : Dict = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[Any] = model(__a , __a )
lowerCamelCase : Tuple = scheduler.step(__a , __a , __a )
lowerCamelCase : str = output.prev_sample
lowerCamelCase : Tuple = torch.sum(torch.abs(__a ) )
lowerCamelCase : Tuple = torch.mean(torch.abs(__a ) )
if torch_device in ["cpu", "mps"]:
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
def a__ ( self: Optional[Any] )-> List[Any]:
if torch_device == "mps":
return
lowerCamelCase : Any = self.scheduler_classes[0]
lowerCamelCase : Union[str, Any] = self.get_scheduler_config()
lowerCamelCase : Optional[Any] = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps , device=__a )
lowerCamelCase : Union[str, Any] = self.dummy_model()
lowerCamelCase : List[str] = self.dummy_sample_deter.to(__a ) * scheduler.init_noise_sigma
for t in scheduler.timesteps:
lowerCamelCase : Union[str, Any] = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[int] = model(__a , __a )
lowerCamelCase : int = scheduler.step(__a , __a , __a )
lowerCamelCase : int = output.prev_sample
lowerCamelCase : Union[str, Any] = torch.sum(torch.abs(__a ) )
lowerCamelCase : int = torch.mean(torch.abs(__a ) )
if str(__a ).startswith("""cpu""" ):
# The following sum varies between 148 and 156 on mps. Why?
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
| 42
| 0
|
"""simple docstring"""
import math
from numpy import inf
from scipy.integrate import quad
def snake_case ( UpperCamelCase__ : float ) -> List[str]:
if num <= 0:
raise ValueError("""math domain error""" )
return quad(_UpperCamelCase , 0 , _UpperCamelCase , args=(_UpperCamelCase) )[0]
def snake_case ( UpperCamelCase__ : float , UpperCamelCase__ : float ) -> str:
return math.pow(_UpperCamelCase , z - 1 ) * math.exp(-x )
if __name__ == "__main__":
from doctest import testmod
testmod()
| 720
|
"""simple docstring"""
import gc
import random
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTextModelWithProjection, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DiffusionPipeline,
EulerDiscreteScheduler,
StableDiffusionXLImgaImgPipeline,
UNetaDConditionModel,
)
from diffusers.utils import floats_tensor, slow, torch_device
from diffusers.utils.testing_utils import enable_full_determinism, require_torch_gpu
from ..pipeline_params import (
IMAGE_TO_IMAGE_IMAGE_PARAMS,
TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS,
TEXT_GUIDED_IMAGE_VARIATION_PARAMS,
)
from ..test_pipelines_common import PipelineLatentTesterMixin, PipelineTesterMixin
enable_full_determinism()
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : str =StableDiffusionXLImgaImgPipeline
snake_case__ : Any =TEXT_GUIDED_IMAGE_VARIATION_PARAMS - {'''height''', '''width'''}
snake_case__ : Optional[int] =PipelineTesterMixin.required_optional_params - {'''latents'''}
snake_case__ : Dict =TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS
snake_case__ : Tuple =IMAGE_TO_IMAGE_IMAGE_PARAMS
snake_case__ : List[str] =IMAGE_TO_IMAGE_IMAGE_PARAMS
def a__ ( self: List[str] )-> int:
torch.manual_seed(0 )
lowerCamelCase : Any = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=2 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , attention_head_dim=(2, 4) , use_linear_projection=__a , addition_embed_type="""text_time""" , addition_time_embed_dim=8 , transformer_layers_per_block=(1, 2) , projection_class_embeddings_input_dim=80 , cross_attention_dim=64 , )
lowerCamelCase : Any = EulerDiscreteScheduler(
beta_start=0.0_00_85 , beta_end=0.0_12 , steps_offset=1 , beta_schedule="""scaled_linear""" , timestep_spacing="""leading""" , )
torch.manual_seed(0 )
lowerCamelCase : Any = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
lowerCamelCase : Optional[Any] = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-0_5 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1_000 , hidden_act="""gelu""" , projection_dim=32 , )
lowerCamelCase : Dict = CLIPTextModel(__a )
lowerCamelCase : Union[str, Any] = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" , local_files_only=__a )
lowerCamelCase : Dict = CLIPTextModelWithProjection(__a )
lowerCamelCase : Optional[int] = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" , local_files_only=__a )
lowerCamelCase : str = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""text_encoder_2""": text_encoder_a,
"""tokenizer_2""": tokenizer_a,
# "safety_checker": None,
# "feature_extractor": None,
}
return components
def a__ ( self: Any , __a: str , __a: Tuple=0 )-> Union[str, Any]:
lowerCamelCase : List[Any] = floats_tensor((1, 3, 32, 32) , rng=random.Random(__a ) ).to(__a )
lowerCamelCase : Any = image / 2 + 0.5
if str(__a ).startswith("""mps""" ):
lowerCamelCase : Dict = torch.manual_seed(__a )
else:
lowerCamelCase : Tuple = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : Tuple = {
"""prompt""": """A painting of a squirrel eating a burger""",
"""image""": image,
"""generator""": generator,
"""num_inference_steps""": 2,
"""guidance_scale""": 5.0,
"""output_type""": """numpy""",
"""strength""": 0.75,
}
return inputs
def a__ ( self: Dict )-> Optional[Any]:
lowerCamelCase : Any = """cpu""" # ensure determinism for the device-dependent torch.Generator
lowerCamelCase : Union[str, Any] = self.get_dummy_components()
lowerCamelCase : Optional[int] = StableDiffusionXLImgaImgPipeline(**__a )
lowerCamelCase : int = sd_pipe.to(__a )
sd_pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Optional[Any] = self.get_dummy_inputs(__a )
lowerCamelCase : Optional[int] = sd_pipe(**__a ).images
lowerCamelCase : Tuple = image[0, -3:, -3:, -1]
assert image.shape == (1, 32, 32, 3)
lowerCamelCase : Any = np.array([0.46_56, 0.48_40, 0.44_39, 0.66_98, 0.55_74, 0.45_24, 0.57_99, 0.59_43, 0.51_65] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-2
def a__ ( self: Optional[int] )-> Union[str, Any]:
super().test_attention_slicing_forward_pass(expected_max_diff=3e-3 )
def a__ ( self: Optional[Any] )-> str:
super().test_inference_batch_single_identical(expected_max_diff=3e-3 )
def a__ ( self: List[str] )-> Optional[Any]:
pass
def a__ ( self: List[Any] )-> Union[str, Any]:
lowerCamelCase : Tuple = self.get_dummy_components()
lowerCamelCase : Union[str, Any] = StableDiffusionXLImgaImgPipeline(**__a )
lowerCamelCase : str = sd_pipe.to(__a )
lowerCamelCase : Any = sd_pipe.to(__a )
sd_pipe.set_progress_bar_config(disable=__a )
# forward without prompt embeds
lowerCamelCase : Dict = self.get_dummy_inputs(__a )
lowerCamelCase : Any = 3 * ["""this is a negative prompt"""]
lowerCamelCase : Optional[int] = negative_prompt
lowerCamelCase : Tuple = 3 * [inputs["""prompt"""]]
lowerCamelCase : List[Any] = sd_pipe(**__a )
lowerCamelCase : Optional[int] = output.images[0, -3:, -3:, -1]
# forward with prompt embeds
lowerCamelCase : Tuple = self.get_dummy_inputs(__a )
lowerCamelCase : List[Any] = 3 * ["""this is a negative prompt"""]
lowerCamelCase : Tuple = 3 * [inputs.pop("""prompt""" )]
(
(
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) ,
) : Union[str, Any] = sd_pipe.encode_prompt(__a , negative_prompt=__a )
lowerCamelCase : int = sd_pipe(
**__a , prompt_embeds=__a , negative_prompt_embeds=__a , pooled_prompt_embeds=__a , negative_pooled_prompt_embeds=__a , )
lowerCamelCase : Union[str, Any] = output.images[0, -3:, -3:, -1]
# make sure that it's equal
assert np.abs(image_slice_a.flatten() - image_slice_a.flatten() ).max() < 1e-4
@slow
@require_torch_gpu
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: Dict )-> str:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: Union[str, Any] , __a: Any , __a: Any="cpu" , __a: str=torch.floataa , __a: Any=0 )-> Optional[Any]:
lowerCamelCase : Optional[Any] = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : List[Any] = np.random.RandomState(__a ).standard_normal((1, 4, 64, 64) )
lowerCamelCase : List[str] = torch.from_numpy(__a ).to(device=__a , dtype=__a )
lowerCamelCase : int = {
"""prompt""": """a photograph of an astronaut riding a horse""",
"""latents""": latents,
"""generator""": generator,
"""num_inference_steps""": 3,
"""guidance_scale""": 7.5,
"""output_type""": """numpy""",
}
return inputs
def a__ ( self: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = DiffusionPipeline.from_pretrained("""stabilityai/stable-diffusion-2-base""" )
pipe.to(__a )
pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Optional[int] = self.get_inputs(__a )
lowerCamelCase : Optional[Any] = pipe(**__a ).images
lowerCamelCase : Dict = image[0, -3:, -3:, -1].flatten()
assert image.shape == (1, 512, 512, 3)
lowerCamelCase : List[str] = np.array([0.4_94_93, 0.4_78_96, 0.4_07_98, 0.5_42_14, 0.5_32_12, 0.4_82_02, 0.4_76_56, 0.4_63_29, 0.4_85_06] )
assert np.abs(image_slice - expected_slice ).max() < 7e-3
| 42
| 0
|
"""simple docstring"""
import argparse
import torch
from transformers import GPTaConfig, GPTaModel, load_tf_weights_in_gpta
from transformers.utils import CONFIG_NAME, WEIGHTS_NAME, logging
logging.set_verbosity_info()
def snake_case ( UpperCamelCase__ : Tuple , UpperCamelCase__ : str , UpperCamelCase__ : Any ) -> Union[str, Any]:
# Construct model
if gpta_config_file == "":
lowerCamelCase : Dict = GPTaConfig()
else:
lowerCamelCase : Optional[int] = GPTaConfig.from_json_file(UpperCamelCase__ )
lowerCamelCase : Optional[int] = GPTaModel(UpperCamelCase__ )
# Load weights from numpy
load_tf_weights_in_gpta(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
# Save pytorch-model
lowerCamelCase : Union[str, Any] = pytorch_dump_folder_path + """/""" + WEIGHTS_NAME
lowerCamelCase : Optional[int] = pytorch_dump_folder_path + """/""" + CONFIG_NAME
print(F'Save PyTorch model to {pytorch_weights_dump_path}' )
torch.save(model.state_dict() , UpperCamelCase__ )
print(F'Save configuration file to {pytorch_config_dump_path}' )
with open(UpperCamelCase__ , """w""" , encoding="""utf-8""" ) as f:
f.write(config.to_json_string() )
if __name__ == "__main__":
__lowerCamelCase :Optional[Any] = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--gpt2_checkpoint_path', default=None, type=str, required=True, help='Path to the TensorFlow checkpoint path.'
)
parser.add_argument(
'--pytorch_dump_folder_path', default=None, type=str, required=True, help='Path to the output PyTorch model.'
)
parser.add_argument(
'--gpt2_config_file',
default='',
type=str,
help=(
'An optional config json file corresponding to the pre-trained OpenAI model. \n'
'This specifies the model architecture.'
),
)
__lowerCamelCase :Tuple = parser.parse_args()
convert_gpta_checkpoint_to_pytorch(args.gpta_checkpoint_path, args.gpta_config_file, args.pytorch_dump_folder_path)
| 721
|
"""simple docstring"""
import unittest
from pathlib import Path
from tempfile import NamedTemporaryFile, TemporaryDirectory
from transformers import BertConfig, BertTokenizerFast, FeatureExtractionPipeline
from transformers.convert_graph_to_onnx import (
convert,
ensure_valid_input,
generate_identified_filename,
infer_shapes,
quantize,
)
from transformers.testing_utils import require_tf, require_tokenizers, require_torch, slow
class A__ :
"""simple docstring"""
def a__ ( self: Optional[int] , __a: Optional[int] , __a: Tuple , __a: Optional[int] )-> List[str]:
return None
class A__ :
"""simple docstring"""
def a__ ( self: Optional[int] , __a: Tuple , __a: str , __a: str , __a: str )-> Tuple:
return None
class A__ ( unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[Any] =[
# (model_name, model_kwargs)
('''bert-base-cased''', {}),
('''gpt2''', {'''use_cache''': False}), # We don't support exporting GPT2 past keys anymore
]
@require_tf
@slow
def a__ ( self: Optional[Any] )-> int:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
self._test_export(__a , """tf""" , 12 , **__a )
@require_torch
@slow
def a__ ( self: str )-> int:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
self._test_export(__a , """pt""" , 12 , **__a )
@require_torch
@slow
def a__ ( self: Union[str, Any] )-> Dict:
from transformers import BertModel
lowerCamelCase : int = ["""[UNK]""", """[SEP]""", """[CLS]""", """[PAD]""", """[MASK]""", """some""", """other""", """words"""]
with NamedTemporaryFile(mode="""w+t""" ) as vocab_file:
vocab_file.write("""\n""".join(__a ) )
vocab_file.flush()
lowerCamelCase : Dict = BertTokenizerFast(vocab_file.name )
with TemporaryDirectory() as bert_save_dir:
lowerCamelCase : List[str] = BertModel(BertConfig(vocab_size=len(__a ) ) )
model.save_pretrained(__a )
self._test_export(__a , """pt""" , 12 , __a )
@require_tf
@slow
def a__ ( self: Optional[Any] )-> Optional[int]:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
lowerCamelCase : Optional[int] = self._test_export(__a , """tf""" , 12 , **__a )
lowerCamelCase : Tuple = quantize(Path(__a ) )
# Ensure the actual quantized model is not bigger than the original one
if quantized_path.stat().st_size >= Path(__a ).stat().st_size:
self.fail("""Quantized model is bigger than initial ONNX model""" )
@require_torch
@slow
def a__ ( self: Any )-> Optional[int]:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
lowerCamelCase : Any = self._test_export(__a , """pt""" , 12 , **__a )
lowerCamelCase : Dict = quantize(__a )
# Ensure the actual quantized model is not bigger than the original one
if quantized_path.stat().st_size >= Path(__a ).stat().st_size:
self.fail("""Quantized model is bigger than initial ONNX model""" )
def a__ ( self: List[Any] , __a: Optional[Any] , __a: List[Any] , __a: Union[str, Any] , __a: Optional[Any]=None , **__a: Optional[int] )-> Any:
try:
# Compute path
with TemporaryDirectory() as tempdir:
lowerCamelCase : Optional[Any] = Path(__a ).joinpath("""model.onnx""" )
# Remove folder if exists
if path.parent.exists():
path.parent.rmdir()
# Export
convert(__a , __a , __a , __a , __a , **__a )
return path
except Exception as e:
self.fail(__a )
@require_torch
@require_tokenizers
@slow
def a__ ( self: Tuple )-> Dict:
from transformers import BertModel
lowerCamelCase : int = BertModel(BertConfig.from_pretrained("""lysandre/tiny-bert-random""" ) )
lowerCamelCase : List[Any] = BertTokenizerFast.from_pretrained("""lysandre/tiny-bert-random""" )
self._test_infer_dynamic_axis(__a , __a , """pt""" )
@require_tf
@require_tokenizers
@slow
def a__ ( self: Optional[Any] )-> List[Any]:
from transformers import TFBertModel
lowerCamelCase : Union[str, Any] = TFBertModel(BertConfig.from_pretrained("""lysandre/tiny-bert-random""" ) )
lowerCamelCase : str = BertTokenizerFast.from_pretrained("""lysandre/tiny-bert-random""" )
self._test_infer_dynamic_axis(__a , __a , """tf""" )
def a__ ( self: List[str] , __a: str , __a: Optional[Any] , __a: str )-> List[Any]:
lowerCamelCase : List[str] = FeatureExtractionPipeline(__a , __a )
lowerCamelCase : List[str] = ["""input_ids""", """token_type_ids""", """attention_mask""", """output_0""", """output_1"""]
lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase : Tuple = infer_shapes(__a , __a )
# Assert all variables are present
self.assertEqual(len(__a ) , len(__a ) )
self.assertTrue(all(var_name in shapes for var_name in variable_names ) )
self.assertSequenceEqual(variable_names[:3] , __a )
self.assertSequenceEqual(variable_names[3:] , __a )
# Assert inputs are {0: batch, 1: sequence}
for var_name in ["input_ids", "token_type_ids", "attention_mask"]:
self.assertDictEqual(shapes[var_name] , {0: """batch""", 1: """sequence"""} )
# Assert outputs are {0: batch, 1: sequence} and {0: batch}
self.assertDictEqual(shapes["""output_0"""] , {0: """batch""", 1: """sequence"""} )
self.assertDictEqual(shapes["""output_1"""] , {0: """batch"""} )
def a__ ( self: List[Any] )-> int:
lowerCamelCase : List[str] = ["""input_ids""", """attention_mask""", """token_type_ids"""]
lowerCamelCase : str = {"""input_ids""": [1, 2, 3, 4], """attention_mask""": [0, 0, 0, 0], """token_type_ids""": [1, 1, 1, 1]}
lowerCamelCase , lowerCamelCase : List[Any] = ensure_valid_input(FuncContiguousArgs() , __a , __a )
# Should have exactly the same number of args (all are valid)
self.assertEqual(len(__a ) , 3 )
# Should have exactly the same input names
self.assertEqual(set(__a ) , set(__a ) )
# Parameter should be reordered according to their respective place in the function:
# (input_ids, token_type_ids, attention_mask)
self.assertEqual(__a , (tokens["""input_ids"""], tokens["""token_type_ids"""], tokens["""attention_mask"""]) )
# Generated args are interleaved with another args (for instance parameter "past" in GPT2)
lowerCamelCase , lowerCamelCase : List[Any] = ensure_valid_input(FuncNonContiguousArgs() , __a , __a )
# Should have exactly the one arg (all before the one not provided "some_other_args")
self.assertEqual(len(__a ) , 1 )
self.assertEqual(len(__a ) , 1 )
# Should have only "input_ids"
self.assertEqual(inputs_args[0] , tokens["""input_ids"""] )
self.assertEqual(ordered_input_names[0] , """input_ids""" )
def a__ ( self: Tuple )-> Tuple:
lowerCamelCase : Optional[int] = generate_identified_filename(Path("""/home/something/my_fake_model.onnx""" ) , """-test""" )
self.assertEqual("""/home/something/my_fake_model-test.onnx""" , generated.as_posix() )
| 42
| 0
|
"""simple docstring"""
from __future__ import annotations
def snake_case ( UpperCamelCase__ : Union[str, Any] ) -> int:
if not nums:
return 0
lowerCamelCase : Union[str, Any] = nums[0]
lowerCamelCase : Tuple = 0
for num in nums[1:]:
lowerCamelCase , lowerCamelCase : str = (
max_excluding + num,
max(_SCREAMING_SNAKE_CASE , _SCREAMING_SNAKE_CASE ),
)
return max(_SCREAMING_SNAKE_CASE , _SCREAMING_SNAKE_CASE )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 700
|
"""simple docstring"""
import unittest
from knapsack import greedy_knapsack as kp
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: Optional[int] )-> Union[str, Any]:
lowerCamelCase : Tuple = [10, 20, 30, 40, 50, 60]
lowerCamelCase : Union[str, Any] = [2, 4, 6, 8, 10, 12]
lowerCamelCase : Union[str, Any] = 100
self.assertEqual(kp.calc_profit(__a , __a , __a ) , 210 )
def a__ ( self: str )-> str:
self.assertRaisesRegex(__a , """max_weight must greater than zero.""" )
def a__ ( self: str )-> List[Any]:
self.assertRaisesRegex(__a , """Weight can not be negative.""" )
def a__ ( self: Any )-> Dict:
self.assertRaisesRegex(__a , """Profit can not be negative.""" )
def a__ ( self: Optional[Any] )-> List[Any]:
self.assertRaisesRegex(__a , """max_weight must greater than zero.""" )
def a__ ( self: Optional[Any] )-> Tuple:
self.assertRaisesRegex(
__a , """The length of profit and weight must be same.""" )
if __name__ == "__main__":
unittest.main()
| 42
| 0
|
"""simple docstring"""
from collections.abc import Callable
from math import pi, sqrt
from random import uniform
from statistics import mean
def snake_case ( UpperCamelCase__ : List[str] ) -> List[Any]:
# A local function to see if a dot lands in the circle.
def is_in_circle(UpperCamelCase__ : str , UpperCamelCase__ : Any ) -> bool:
lowerCamelCase : str = sqrt((x**2) + (y**2) )
# Our circle has a radius of 1, so a distance
# greater than 1 would land outside the circle.
return distance_from_centre <= 1
# The proportion of guesses that landed in the circle
lowerCamelCase : Dict = mean(
int(is_in_circle(uniform(-1.0 , 1.0 ) , uniform(-1.0 , 1.0 ) ) )
for _ in range(lowercase__ ) )
# The ratio of the area for circle to square is pi/4.
lowerCamelCase : Tuple = proportion * 4
print(F'The estimated value of pi is {pi_estimate}' )
print(F'The numpy value of pi is {pi}' )
print(F'The total error is {abs(pi - pi_estimate )}' )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Any , UpperCamelCase__ : Optional[Any] = 0.0 , UpperCamelCase__ : List[str] = 1.0 , ) -> Optional[int]:
return mean(
function_to_integrate(uniform(lowercase__ , lowercase__ ) ) for _ in range(lowercase__ ) ) * (max_value - min_value)
def snake_case ( UpperCamelCase__ : List[Any] , UpperCamelCase__ : List[Any] = 0.0 , UpperCamelCase__ : Any = 1.0 ) -> Dict:
def identity_function(UpperCamelCase__ : List[str] ) -> float:
return x
lowerCamelCase : Union[str, Any] = area_under_curve_estimator(
lowercase__ , lowercase__ , lowercase__ , lowercase__ )
lowerCamelCase : Union[str, Any] = (max_value * max_value - min_value * min_value) / 2
print("""******************""" )
print(F'Estimating area under y=x where x varies from {min_value} to {max_value}' )
print(F'Estimated value is {estimated_value}' )
print(F'Expected value is {expected_value}' )
print(F'Total error is {abs(estimated_value - expected_value )}' )
print("""******************""" )
def snake_case ( UpperCamelCase__ : List[Any] ) -> Dict:
def function_to_integrate(UpperCamelCase__ : Dict ) -> float:
return sqrt(4.0 - x * x )
lowerCamelCase : Any = area_under_curve_estimator(
lowercase__ , lowercase__ , 0.0 , 2.0 )
print("""******************""" )
print("""Estimating pi using area_under_curve_estimator""" )
print(F'Estimated value is {estimated_value}' )
print(F'Expected value is {pi}' )
print(F'Total error is {abs(estimated_value - pi )}' )
print("""******************""" )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 701
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_flax_available,
is_tf_available,
is_tokenizers_available,
is_torch_available,
is_vision_available,
)
__lowerCamelCase :List[str] = {
'configuration_owlvit': [
'OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP',
'OwlViTConfig',
'OwlViTOnnxConfig',
'OwlViTTextConfig',
'OwlViTVisionConfig',
],
'processing_owlvit': ['OwlViTProcessor'],
}
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[int] = ['OwlViTFeatureExtractor']
__lowerCamelCase :List[str] = ['OwlViTImageProcessor']
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[Any] = [
'OWLVIT_PRETRAINED_MODEL_ARCHIVE_LIST',
'OwlViTModel',
'OwlViTPreTrainedModel',
'OwlViTTextModel',
'OwlViTVisionModel',
'OwlViTForObjectDetection',
]
if TYPE_CHECKING:
from .configuration_owlvit import (
OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP,
OwlViTConfig,
OwlViTOnnxConfig,
OwlViTTextConfig,
OwlViTVisionConfig,
)
from .processing_owlvit import OwlViTProcessor
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .feature_extraction_owlvit import OwlViTFeatureExtractor
from .image_processing_owlvit import OwlViTImageProcessor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_owlvit import (
OWLVIT_PRETRAINED_MODEL_ARCHIVE_LIST,
OwlViTForObjectDetection,
OwlViTModel,
OwlViTPreTrainedModel,
OwlViTTextModel,
OwlViTVisionModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
import argparse
import csv
import logging
import os
import random
import numpy as np
import torch
from torch.utils.data import DataLoader, RandomSampler, SequentialSampler, TensorDataset
from tqdm import tqdm, trange
from transformers import (
CONFIG_NAME,
WEIGHTS_NAME,
AdamW,
OpenAIGPTDoubleHeadsModel,
OpenAIGPTTokenizer,
get_linear_schedule_with_warmup,
)
logging.basicConfig(
format='%(asctime)s - %(levelname)s - %(name)s - %(message)s', datefmt='%m/%d/%Y %H:%M:%S', level=logging.INFO
)
__lowerCamelCase :List[str] = logging.getLogger(__name__)
def snake_case ( UpperCamelCase__ : str , UpperCamelCase__ : int ) -> List[str]:
lowerCamelCase : Optional[int] = np.argmax(lowerCAmelCase_ , axis=1 )
return np.sum(outputs == labels )
def snake_case ( UpperCamelCase__ : Dict ) -> Any:
with open(lowerCAmelCase_ , encoding="""utf_8""" ) as f:
lowerCamelCase : Dict = csv.reader(lowerCAmelCase_ )
lowerCamelCase : Tuple = []
next(lowerCAmelCase_ ) # skip the first line
for line in tqdm(lowerCAmelCase_ ):
output.append((""" """.join(line[1:5] ), line[5], line[6], int(line[-1] ) - 1) )
return output
def snake_case ( UpperCamelCase__ : List[Any] , UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Dict , UpperCamelCase__ : Dict , UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : int ) -> Optional[int]:
lowerCamelCase : Union[str, Any] = []
for dataset in encoded_datasets:
lowerCamelCase : List[Any] = len(lowerCAmelCase_ )
lowerCamelCase : int = np.zeros((n_batch, 2, input_len) , dtype=np.intaa )
lowerCamelCase : Optional[Any] = np.zeros((n_batch, 2) , dtype=np.intaa )
lowerCamelCase : Optional[Any] = np.full((n_batch, 2, input_len) , fill_value=-100 , dtype=np.intaa )
lowerCamelCase : int = np.zeros((n_batch,) , dtype=np.intaa )
for (
i,
(story, conta, conta, mc_label),
) in enumerate(lowerCAmelCase_ ):
lowerCamelCase : Any = [start_token] + story[:cap_length] + [delimiter_token] + conta[:cap_length] + [clf_token]
lowerCamelCase : List[Any] = [start_token] + story[:cap_length] + [delimiter_token] + conta[:cap_length] + [clf_token]
lowerCamelCase : List[Any] = with_conta
lowerCamelCase : Dict = with_conta
lowerCamelCase : str = len(lowerCAmelCase_ ) - 1
lowerCamelCase : List[Any] = len(lowerCAmelCase_ ) - 1
lowerCamelCase : Optional[Any] = with_conta
lowerCamelCase : List[Any] = with_conta
lowerCamelCase : Optional[Any] = mc_label
lowerCamelCase : Optional[Any] = (input_ids, mc_token_ids, lm_labels, mc_labels)
tensor_datasets.append(tuple(torch.tensor(lowerCAmelCase_ ) for t in all_inputs ) )
return tensor_datasets
def snake_case ( ) -> List[str]:
lowerCamelCase : List[Any] = argparse.ArgumentParser()
parser.add_argument("""--model_name""" , type=lowerCAmelCase_ , default="""openai-gpt""" , help="""pretrained model name""" )
parser.add_argument("""--do_train""" , action="""store_true""" , help="""Whether to run training.""" )
parser.add_argument("""--do_eval""" , action="""store_true""" , help="""Whether to run eval on the dev set.""" )
parser.add_argument(
"""--output_dir""" , default=lowerCAmelCase_ , type=lowerCAmelCase_ , required=lowerCAmelCase_ , help="""The output directory where the model predictions and checkpoints will be written.""" , )
parser.add_argument("""--train_dataset""" , type=lowerCAmelCase_ , default="""""" )
parser.add_argument("""--eval_dataset""" , type=lowerCAmelCase_ , default="""""" )
parser.add_argument("""--seed""" , type=lowerCAmelCase_ , default=42 )
parser.add_argument("""--num_train_epochs""" , type=lowerCAmelCase_ , default=3 )
parser.add_argument("""--train_batch_size""" , type=lowerCAmelCase_ , default=8 )
parser.add_argument("""--eval_batch_size""" , type=lowerCAmelCase_ , default=16 )
parser.add_argument("""--adam_epsilon""" , default=1E-8 , type=lowerCAmelCase_ , help="""Epsilon for Adam optimizer.""" )
parser.add_argument("""--max_grad_norm""" , type=lowerCAmelCase_ , default=1 )
parser.add_argument(
"""--max_steps""" , default=-1 , type=lowerCAmelCase_ , help=(
"""If > 0: set total number of training steps to perform. Override num_train_epochs."""
) , )
parser.add_argument(
"""--gradient_accumulation_steps""" , type=lowerCAmelCase_ , default=1 , help="""Number of updates steps to accumulate before performing a backward/update pass.""" , )
parser.add_argument("""--learning_rate""" , type=lowerCAmelCase_ , default=6.25E-5 )
parser.add_argument("""--warmup_steps""" , default=0 , type=lowerCAmelCase_ , help="""Linear warmup over warmup_steps.""" )
parser.add_argument("""--lr_schedule""" , type=lowerCAmelCase_ , default="""warmup_linear""" )
parser.add_argument("""--weight_decay""" , type=lowerCAmelCase_ , default=0.0_1 )
parser.add_argument("""--lm_coef""" , type=lowerCAmelCase_ , default=0.9 )
parser.add_argument("""--n_valid""" , type=lowerCAmelCase_ , default=374 )
parser.add_argument("""--server_ip""" , type=lowerCAmelCase_ , default="""""" , help="""Can be used for distant debugging.""" )
parser.add_argument("""--server_port""" , type=lowerCAmelCase_ , default="""""" , help="""Can be used for distant debugging.""" )
lowerCamelCase : Optional[int] = parser.parse_args()
print(lowerCAmelCase_ )
if args.server_ip and args.server_port:
# Distant debugging - see https://code.visualstudio.com/docs/python/debugging#_attach-to-a-local-script
import ptvsd
print("""Waiting for debugger attach""" )
ptvsd.enable_attach(address=(args.server_ip, args.server_port) , redirect_output=lowerCAmelCase_ )
ptvsd.wait_for_attach()
random.seed(args.seed )
np.random.seed(args.seed )
torch.manual_seed(args.seed )
torch.cuda.manual_seed_all(args.seed )
lowerCamelCase : Optional[Any] = torch.device("""cuda""" if torch.cuda.is_available() else """cpu""" )
lowerCamelCase : Dict = torch.cuda.device_count()
logger.info("""device: {}, n_gpu {}""".format(lowerCAmelCase_ , lowerCAmelCase_ ) )
if not args.do_train and not args.do_eval:
raise ValueError("""At least one of `do_train` or `do_eval` must be True.""" )
if not os.path.exists(args.output_dir ):
os.makedirs(args.output_dir )
# Load tokenizer and model
# This loading functions also add new tokens and embeddings called `special tokens`
# These new embeddings will be fine-tuned on the RocStories dataset
lowerCamelCase : Optional[Any] = ['''_start_''', '''_delimiter_''', '''_classify_''']
lowerCamelCase : Optional[int] = OpenAIGPTTokenizer.from_pretrained(args.model_name )
tokenizer.add_tokens(lowerCAmelCase_ )
lowerCamelCase : Optional[Any] = tokenizer.convert_tokens_to_ids(lowerCAmelCase_ )
lowerCamelCase : List[str] = OpenAIGPTDoubleHeadsModel.from_pretrained(args.model_name )
model.resize_token_embeddings(len(lowerCAmelCase_ ) )
model.to(lowerCAmelCase_ )
# Load and encode the datasets
def tokenize_and_encode(UpperCamelCase__ : Dict ):
if isinstance(lowerCAmelCase_ , lowerCAmelCase_ ):
return tokenizer.convert_tokens_to_ids(tokenizer.tokenize(lowerCAmelCase_ ) )
elif isinstance(lowerCAmelCase_ , lowerCAmelCase_ ):
return obj
return [tokenize_and_encode(lowerCAmelCase_ ) for o in obj]
logger.info("""Encoding dataset...""" )
lowerCamelCase : Optional[Any] = load_rocstories_dataset(args.train_dataset )
lowerCamelCase : List[Any] = load_rocstories_dataset(args.eval_dataset )
lowerCamelCase : Dict = (train_dataset, eval_dataset)
lowerCamelCase : Optional[Any] = tokenize_and_encode(lowerCAmelCase_ )
# Compute the max input length for the Transformer
lowerCamelCase : Tuple = model.config.n_positions // 2 - 2
lowerCamelCase : Optional[Any] = max(
len(story[:max_length] ) + max(len(conta[:max_length] ) , len(conta[:max_length] ) ) + 3
for dataset in encoded_datasets
for story, conta, conta, _ in dataset )
lowerCamelCase : Optional[int] = min(lowerCAmelCase_ , model.config.n_positions ) # Max size of input for the pre-trained model
# Prepare inputs tensors and dataloaders
lowerCamelCase : int = pre_process_datasets(lowerCAmelCase_ , lowerCAmelCase_ , lowerCAmelCase_ , *lowerCAmelCase_ )
lowerCamelCase : Dict = tensor_datasets[0], tensor_datasets[1]
lowerCamelCase : str = TensorDataset(*lowerCAmelCase_ )
lowerCamelCase : List[Any] = RandomSampler(lowerCAmelCase_ )
lowerCamelCase : List[Any] = DataLoader(lowerCAmelCase_ , sampler=lowerCAmelCase_ , batch_size=args.train_batch_size )
lowerCamelCase : str = TensorDataset(*lowerCAmelCase_ )
lowerCamelCase : Any = SequentialSampler(lowerCAmelCase_ )
lowerCamelCase : str = DataLoader(lowerCAmelCase_ , sampler=lowerCAmelCase_ , batch_size=args.eval_batch_size )
# Prepare optimizer
if args.do_train:
if args.max_steps > 0:
lowerCamelCase : Optional[int] = args.max_steps
lowerCamelCase : Optional[int] = args.max_steps // (len(lowerCAmelCase_ ) // args.gradient_accumulation_steps) + 1
else:
lowerCamelCase : Dict = len(lowerCAmelCase_ ) // args.gradient_accumulation_steps * args.num_train_epochs
lowerCamelCase : Union[str, Any] = list(model.named_parameters() )
lowerCamelCase : str = ['''bias''', '''LayerNorm.bias''', '''LayerNorm.weight''']
lowerCamelCase : Union[str, Any] = [
{
'''params''': [p for n, p in param_optimizer if not any(nd in n for nd in no_decay )],
'''weight_decay''': args.weight_decay,
},
{'''params''': [p for n, p in param_optimizer if any(nd in n for nd in no_decay )], '''weight_decay''': 0.0},
]
lowerCamelCase : Any = AdamW(lowerCAmelCase_ , lr=args.learning_rate , eps=args.adam_epsilon )
lowerCamelCase : str = get_linear_schedule_with_warmup(
lowerCAmelCase_ , num_warmup_steps=args.warmup_steps , num_training_steps=lowerCAmelCase_ )
if args.do_train:
lowerCamelCase : Optional[int] = 0, 0, None
model.train()
for _ in trange(int(args.num_train_epochs ) , desc="""Epoch""" ):
lowerCamelCase : str = 0
lowerCamelCase : List[str] = 0
lowerCamelCase : Tuple = tqdm(lowerCAmelCase_ , desc="""Training""" )
for step, batch in enumerate(lowerCAmelCase_ ):
lowerCamelCase : Any = tuple(t.to(lowerCAmelCase_ ) for t in batch )
lowerCamelCase : Optional[Any] = batch
lowerCamelCase : Union[str, Any] = model(lowerCAmelCase_ , mc_token_ids=lowerCAmelCase_ , lm_labels=lowerCAmelCase_ , mc_labels=lowerCAmelCase_ )
lowerCamelCase : List[Any] = args.lm_coef * losses[0] + losses[1]
loss.backward()
optimizer.step()
scheduler.step()
optimizer.zero_grad()
tr_loss += loss.item()
lowerCamelCase : Optional[Any] = (
loss.item() if exp_average_loss is None else 0.7 * exp_average_loss + 0.3 * loss.item()
)
nb_tr_steps += 1
lowerCamelCase : int = '''Training loss: {:.2e} lr: {:.2e}'''.format(lowerCAmelCase_ , scheduler.get_lr()[0] )
# Save a trained model
if args.do_train:
# Save a trained model, configuration and tokenizer
lowerCamelCase : Optional[int] = model.module if hasattr(lowerCAmelCase_ , """module""" ) else model # Only save the model itself
# If we save using the predefined names, we can load using `from_pretrained`
lowerCamelCase : Dict = os.path.join(args.output_dir , lowerCAmelCase_ )
lowerCamelCase : Dict = os.path.join(args.output_dir , lowerCAmelCase_ )
torch.save(model_to_save.state_dict() , lowerCAmelCase_ )
model_to_save.config.to_json_file(lowerCAmelCase_ )
tokenizer.save_vocabulary(args.output_dir )
# Load a trained model and vocabulary that you have fine-tuned
lowerCamelCase : Dict = OpenAIGPTDoubleHeadsModel.from_pretrained(args.output_dir )
lowerCamelCase : Optional[int] = OpenAIGPTTokenizer.from_pretrained(args.output_dir )
model.to(lowerCAmelCase_ )
if args.do_eval:
model.eval()
lowerCamelCase : Tuple = 0, 0
lowerCamelCase : str = 0, 0
for batch in tqdm(lowerCAmelCase_ , desc="""Evaluating""" ):
lowerCamelCase : Any = tuple(t.to(lowerCAmelCase_ ) for t in batch )
lowerCamelCase : Dict = batch
with torch.no_grad():
lowerCamelCase : List[Any] = model(
lowerCAmelCase_ , mc_token_ids=lowerCAmelCase_ , lm_labels=lowerCAmelCase_ , mc_labels=lowerCAmelCase_ )
lowerCamelCase : int = mc_logits.detach().cpu().numpy()
lowerCamelCase : Any = mc_labels.to("""cpu""" ).numpy()
lowerCamelCase : Optional[Any] = accuracy(lowerCAmelCase_ , lowerCAmelCase_ )
eval_loss += mc_loss.mean().item()
eval_accuracy += tmp_eval_accuracy
nb_eval_examples += input_ids.size(0 )
nb_eval_steps += 1
lowerCamelCase : Optional[int] = eval_loss / nb_eval_steps
lowerCamelCase : Optional[int] = eval_accuracy / nb_eval_examples
lowerCamelCase : List[Any] = tr_loss / nb_tr_steps if args.do_train else None
lowerCamelCase : Tuple = {'''eval_loss''': eval_loss, '''eval_accuracy''': eval_accuracy, '''train_loss''': train_loss}
lowerCamelCase : str = os.path.join(args.output_dir , """eval_results.txt""" )
with open(lowerCAmelCase_ , """w""" ) as writer:
logger.info("""***** Eval results *****""" )
for key in sorted(result.keys() ):
logger.info(""" %s = %s""" , lowerCAmelCase_ , str(result[key] ) )
writer.write("""%s = %s\n""" % (key, str(result[key] )) )
if __name__ == "__main__":
main()
| 702
|
"""simple docstring"""
import collections
import inspect
import unittest
from transformers import FocalNetConfig
from transformers.testing_utils import require_torch, require_vision, slow, torch_device
from transformers.utils import cached_property, is_torch_available, is_vision_available
from ...test_backbone_common import BackboneTesterMixin
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, _config_zero_init, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from torch import nn
from transformers import (
FocalNetBackbone,
FocalNetForImageClassification,
FocalNetForMaskedImageModeling,
FocalNetModel,
)
from transformers.models.focalnet.modeling_focalnet import FOCALNET_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import AutoImageProcessor
class A__ :
"""simple docstring"""
def __init__( self: List[Any] , __a: List[str] , __a: Optional[int]=13 , __a: List[str]=32 , __a: int=2 , __a: List[str]=3 , __a: Union[str, Any]=16 , __a: int=[32, 64, 128] , __a: Optional[Any]=[1, 2, 1] , __a: Optional[int]=[2, 2, 4] , __a: Tuple=2 , __a: Dict=2.0 , __a: List[str]=True , __a: Optional[Any]=0.0 , __a: Any=0.0 , __a: List[Any]=0.1 , __a: List[str]="gelu" , __a: Tuple=False , __a: Union[str, Any]=True , __a: Optional[int]=0.02 , __a: Tuple=1e-5 , __a: int=True , __a: List[Any]=None , __a: Optional[int]=True , __a: Dict=10 , __a: List[str]=8 , __a: Any=["stage1", "stage2"] , __a: Union[str, Any]=[1, 2] , )-> Dict:
lowerCamelCase : Dict = parent
lowerCamelCase : Optional[Any] = batch_size
lowerCamelCase : Union[str, Any] = image_size
lowerCamelCase : Optional[int] = patch_size
lowerCamelCase : Any = num_channels
lowerCamelCase : Any = embed_dim
lowerCamelCase : Dict = hidden_sizes
lowerCamelCase : List[Any] = depths
lowerCamelCase : Tuple = num_heads
lowerCamelCase : List[Any] = window_size
lowerCamelCase : str = mlp_ratio
lowerCamelCase : str = qkv_bias
lowerCamelCase : str = hidden_dropout_prob
lowerCamelCase : Dict = attention_probs_dropout_prob
lowerCamelCase : Tuple = drop_path_rate
lowerCamelCase : Dict = hidden_act
lowerCamelCase : Tuple = use_absolute_embeddings
lowerCamelCase : List[str] = patch_norm
lowerCamelCase : List[str] = layer_norm_eps
lowerCamelCase : str = initializer_range
lowerCamelCase : Tuple = is_training
lowerCamelCase : int = scope
lowerCamelCase : Union[str, Any] = use_labels
lowerCamelCase : List[str] = type_sequence_label_size
lowerCamelCase : str = encoder_stride
lowerCamelCase : List[str] = out_features
lowerCamelCase : Optional[int] = out_indices
def a__ ( self: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : str = None
if self.use_labels:
lowerCamelCase : List[str] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : str = self.get_config()
return config, pixel_values, labels
def a__ ( self: List[Any] )-> Optional[int]:
return FocalNetConfig(
image_size=self.image_size , patch_size=self.patch_size , num_channels=self.num_channels , embed_dim=self.embed_dim , hidden_sizes=self.hidden_sizes , depths=self.depths , num_heads=self.num_heads , window_size=self.window_size , mlp_ratio=self.mlp_ratio , qkv_bias=self.qkv_bias , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , drop_path_rate=self.drop_path_rate , hidden_act=self.hidden_act , use_absolute_embeddings=self.use_absolute_embeddings , path_norm=self.patch_norm , layer_norm_eps=self.layer_norm_eps , initializer_range=self.initializer_range , encoder_stride=self.encoder_stride , out_features=self.out_features , out_indices=self.out_indices , )
def a__ ( self: Tuple , __a: Optional[int] , __a: Optional[int] , __a: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = FocalNetModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Tuple = model(__a )
lowerCamelCase : Any = ((config.image_size // config.patch_size) ** 2) // (4 ** (len(config.depths ) - 1))
lowerCamelCase : List[Any] = int(config.embed_dim * 2 ** (len(config.depths ) - 1) )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, expected_seq_len, expected_dim) )
def a__ ( self: Optional[int] , __a: Dict , __a: Tuple , __a: List[Any] )-> int:
lowerCamelCase : List[Any] = FocalNetBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Optional[Any] = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , len(config.out_features ) )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.image_size, 8, 8] )
# verify channels
self.parent.assertEqual(len(model.channels ) , len(config.out_features ) )
self.parent.assertListEqual(model.channels , config.hidden_sizes[:-1] )
# verify backbone works with out_features=None
lowerCamelCase : Dict = None
lowerCamelCase : Dict = FocalNetBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , 1 )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.image_size * 2, 4, 4] )
# verify channels
self.parent.assertEqual(len(model.channels ) , 1 )
self.parent.assertListEqual(model.channels , [config.hidden_sizes[-1]] )
def a__ ( self: Optional[int] , __a: Optional[int] , __a: Optional[int] , __a: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = FocalNetForMaskedImageModeling(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[str] = model(__a )
self.parent.assertEqual(
result.reconstruction.shape , (self.batch_size, self.num_channels, self.image_size, self.image_size) )
# test greyscale images
lowerCamelCase : List[str] = 1
lowerCamelCase : Any = FocalNetForMaskedImageModeling(__a )
model.to(__a )
model.eval()
lowerCamelCase : str = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Tuple = model(__a )
self.parent.assertEqual(result.reconstruction.shape , (self.batch_size, 1, self.image_size, self.image_size) )
def a__ ( self: str , __a: Optional[Any] , __a: Optional[Any] , __a: Tuple )-> str:
lowerCamelCase : Optional[Any] = self.type_sequence_label_size
lowerCamelCase : Optional[Any] = FocalNetForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : List[str] = model(__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
# test greyscale images
lowerCamelCase : int = 1
lowerCamelCase : List[Any] = FocalNetForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Optional[Any] = model(__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
def a__ ( self: int )-> Optional[int]:
lowerCamelCase : str = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = config_and_inputs
lowerCamelCase : List[Any] = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =(
(
FocalNetModel,
FocalNetForImageClassification,
FocalNetForMaskedImageModeling,
FocalNetBackbone,
)
if is_torch_available()
else ()
)
snake_case__ : Optional[int] =(
{'''feature-extraction''': FocalNetModel, '''image-classification''': FocalNetForImageClassification}
if is_torch_available()
else {}
)
snake_case__ : Tuple =False
snake_case__ : Dict =False
snake_case__ : Dict =False
snake_case__ : Tuple =False
snake_case__ : Optional[int] =False
def a__ ( self: Union[str, Any] )-> Optional[int]:
lowerCamelCase : List[str] = FocalNetModelTester(self )
lowerCamelCase : Optional[Any] = ConfigTester(self , config_class=__a , embed_dim=37 , has_text_modality=__a )
def a__ ( self: List[str] )-> List[str]:
self.create_and_test_config_common_properties()
self.config_tester.create_and_test_config_to_json_string()
self.config_tester.create_and_test_config_to_json_file()
self.config_tester.create_and_test_config_from_and_save_pretrained()
self.config_tester.create_and_test_config_with_num_labels()
self.config_tester.check_config_can_be_init_without_params()
self.config_tester.check_config_arguments_init()
def a__ ( self: List[str] )-> Union[str, Any]:
return
def a__ ( self: Tuple )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: List[Any] )-> Dict:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_backbone(*__a )
def a__ ( self: List[Any] )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_masked_image_modeling(*__a )
def a__ ( self: List[str] )-> Dict:
lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_classification(*__a )
@unittest.skip(reason="""FocalNet does not use inputs_embeds""" )
def a__ ( self: Optional[Any] )-> str:
pass
@unittest.skip(reason="""FocalNet does not use feedforward chunking""" )
def a__ ( self: Optional[Any] )-> Dict:
pass
def a__ ( self: Optional[Any] )-> Dict:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : Any = model_class(__a )
self.assertIsInstance(model.get_input_embeddings() , (nn.Module) )
lowerCamelCase : Dict = model.get_output_embeddings()
self.assertTrue(x is None or isinstance(__a , nn.Linear ) )
def a__ ( self: Tuple )-> Optional[int]:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : int = model_class(__a )
lowerCamelCase : int = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Any = [*signature.parameters.keys()]
lowerCamelCase : List[Any] = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: str , __a: Union[str, Any] , __a: int , __a: Tuple , __a: List[str] )-> Union[str, Any]:
lowerCamelCase : List[Any] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : List[str] = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : List[str] = outputs.hidden_states
lowerCamelCase : Tuple = getattr(
self.model_tester , """expected_num_hidden_layers""" , len(self.model_tester.depths ) + 1 )
self.assertEqual(len(__a ) , __a )
# FocalNet has a different seq_length
lowerCamelCase : Tuple = (
config.patch_size
if isinstance(config.patch_size , collections.abc.Iterable )
else (config.patch_size, config.patch_size)
)
lowerCamelCase : Dict = (image_size[1] // patch_size[1]) * (image_size[0] // patch_size[0])
self.assertListEqual(
list(hidden_states[0].shape[-2:] ) , [num_patches, self.model_tester.embed_dim] , )
lowerCamelCase : Optional[Any] = outputs.reshaped_hidden_states
self.assertEqual(len(__a ) , __a )
lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = reshaped_hidden_states[0].shape
lowerCamelCase : Tuple = (
reshaped_hidden_states[0].view(__a , __a , height * width ).permute(0 , 2 , 1 )
)
self.assertListEqual(
list(reshaped_hidden_states.shape[-2:] ) , [num_patches, self.model_tester.embed_dim] , )
def a__ ( self: Any )-> Any:
lowerCamelCase , lowerCamelCase : str = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Union[str, Any] = (
self.model_tester.image_size
if isinstance(self.model_tester.image_size , collections.abc.Iterable )
else (self.model_tester.image_size, self.model_tester.image_size)
)
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : List[str] = True
self.check_hidden_states_output(__a , __a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : List[Any] = True
self.check_hidden_states_output(__a , __a , __a , __a )
def a__ ( self: str )-> Union[str, Any]:
lowerCamelCase , lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : List[str] = 3
lowerCamelCase : Any = (
self.model_tester.image_size
if isinstance(self.model_tester.image_size , collections.abc.Iterable )
else (self.model_tester.image_size, self.model_tester.image_size)
)
lowerCamelCase : Optional[int] = (
config.patch_size
if isinstance(config.patch_size , collections.abc.Iterable )
else (config.patch_size, config.patch_size)
)
lowerCamelCase : Optional[Any] = image_size[0] + patch_size[0] - (image_size[0] % patch_size[0])
lowerCamelCase : List[str] = image_size[1] + patch_size[1] - (image_size[1] % patch_size[1])
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : str = True
self.check_hidden_states_output(__a , __a , __a , (padded_height, padded_width) )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : Union[str, Any] = True
self.check_hidden_states_output(__a , __a , __a , (padded_height, padded_width) )
@slow
def a__ ( self: Optional[int] )-> List[Any]:
for model_name in FOCALNET_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : List[str] = FocalNetModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def a__ ( self: str )-> Any:
lowerCamelCase , lowerCamelCase : Optional[Any] = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : int = _config_zero_init(__a )
for model_class in self.all_model_classes:
lowerCamelCase : int = model_class(config=__a )
for name, param in model.named_parameters():
if "embeddings" not in name and param.requires_grad:
self.assertIn(
((param.data.mean() * 1e9).round() / 1e9).item() , [0.0, 1.0] , msg=f'Parameter {name} of model {model_class} seems not properly initialized' , )
@require_vision
@require_torch
class A__ ( unittest.TestCase):
"""simple docstring"""
@cached_property
def a__ ( self: Optional[int] )-> Optional[Any]:
# TODO update organization
return AutoImageProcessor.from_pretrained("""microsoft/focalnet-tiny""" ) if is_vision_available() else None
@slow
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : Tuple = FocalNetForImageClassification.from_pretrained("""microsoft/focalnet-tiny""" ).to(__a )
lowerCamelCase : Any = self.default_image_processor
lowerCamelCase : Union[str, Any] = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
lowerCamelCase : int = image_processor(images=__a , return_tensors="""pt""" ).to(__a )
# forward pass
with torch.no_grad():
lowerCamelCase : Any = model(**__a )
# verify the logits
lowerCamelCase : Tuple = torch.Size((1, 1_000) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : List[str] = torch.tensor([0.21_66, -0.43_68, 0.21_91] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3] , __a , atol=1e-4 ) )
self.assertTrue(outputs.logits.argmax(dim=-1 ).item() , 281 )
@require_torch
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : str =(FocalNetBackbone,) if is_torch_available() else ()
snake_case__ : Optional[int] =FocalNetConfig
snake_case__ : str =False
def a__ ( self: Union[str, Any] )-> Tuple:
lowerCamelCase : str = FocalNetModelTester(self )
| 42
| 0
|
"""simple docstring"""
import uuid
from typing import Any, Dict, List, Optional, Union
from ..utils import add_end_docstrings, is_tf_available, is_torch_available, logging
from .base import PIPELINE_INIT_ARGS, Pipeline
if is_tf_available():
import tensorflow as tf
if is_torch_available():
import torch
__lowerCamelCase :Any = logging.get_logger(__name__)
class A__ :
"""simple docstring"""
def __init__( self: Optional[int] , __a: Optional[int] = None , __a: Union[str, Any] = None , __a: Optional[int]=None , __a: Tuple=None )-> Tuple:
if not conversation_id:
lowerCamelCase : int = uuid.uuida()
if past_user_inputs is None:
lowerCamelCase : int = []
if generated_responses is None:
lowerCamelCase : List[Any] = []
lowerCamelCase : uuid.UUID = conversation_id
lowerCamelCase : List[str] = past_user_inputs
lowerCamelCase : List[str] = generated_responses
lowerCamelCase : Optional[str] = text
def __eq__( self: str , __a: List[str] )-> Union[str, Any]:
if not isinstance(lowercase__ , lowercase__ ):
return False
if self.uuid == other.uuid:
return True
return (
self.new_user_input == other.new_user_input
and self.past_user_inputs == other.past_user_inputs
and self.generated_responses == other.generated_responses
)
def a__ ( self: Optional[Any] , __a: List[str] , __a: str = False )-> Any:
if self.new_user_input:
if overwrite:
logger.warning(
f'User input added while unprocessed input was existing: "{self.new_user_input}" was overwritten '
f'with: "{text}".' )
lowerCamelCase : Any = text
else:
logger.warning(
f'User input added while unprocessed input was existing: "{self.new_user_input}" new input '
f'ignored: "{text}". Set `overwrite` to True to overwrite unprocessed user input' )
else:
lowerCamelCase : List[Any] = text
def a__ ( self: List[Any] )-> Any:
if self.new_user_input:
self.past_user_inputs.append(self.new_user_input )
lowerCamelCase : Dict = None
def a__ ( self: List[Any] , __a: Optional[int] )-> List[str]:
self.generated_responses.append(lowercase__ )
def a__ ( self: List[Any] )-> Any:
for user_input, generated_response in zip(self.past_user_inputs , self.generated_responses ):
yield True, user_input
yield False, generated_response
if self.new_user_input:
yield True, self.new_user_input
def __repr__( self: str )-> Dict:
lowerCamelCase : List[str] = f'Conversation id: {self.uuid} \n'
for is_user, text in self.iter_texts():
lowerCamelCase : Optional[Any] = """user""" if is_user else """bot"""
output += f'{name} >> {text} \n'
return output
@add_end_docstrings(
_UpperCAmelCase , R'''
min_length_for_response (`int`, *optional*, defaults to 32):
The minimum length (in number of tokens) for a response.
minimum_tokens (`int`, *optional*, defaults to 10):
The minimum length of tokens to leave for a response.
''' , )
class A__ ( _UpperCAmelCase):
"""simple docstring"""
def __init__( self: Optional[Any] , *__a: Tuple , **__a: int )-> Dict:
super().__init__(*lowercase__ , **lowercase__ )
if self.tokenizer.pad_token_id is None:
lowerCamelCase : Union[str, Any] = self.tokenizer.eos_token
def a__ ( self: Optional[Any] , __a: Optional[Any]=None , __a: List[str]=None , __a: str=None , **__a: Any )-> int:
lowerCamelCase : Tuple = {}
lowerCamelCase : Optional[Any] = {}
lowerCamelCase : Optional[Any] = {}
if min_length_for_response is not None:
lowerCamelCase : Dict = min_length_for_response
if minimum_tokens is not None:
lowerCamelCase : Union[str, Any] = minimum_tokens
if "max_length" in generate_kwargs:
lowerCamelCase : List[str] = generate_kwargs["""max_length"""]
# self.max_length = generate_kwargs.get("max_length", self.model.config.max_length)
if clean_up_tokenization_spaces is not None:
lowerCamelCase : Union[str, Any] = clean_up_tokenization_spaces
if generate_kwargs:
forward_params.update(lowercase__ )
return preprocess_params, forward_params, postprocess_params
def __call__( self: Union[str, Any] , __a: Any , __a: Any=0 , **__a: Dict )-> List[Any]:
lowerCamelCase : List[str] = super().__call__(lowercase__ , num_workers=lowercase__ , **lowercase__ )
if isinstance(lowercase__ , lowercase__ ) and len(lowercase__ ) == 1:
return outputs[0]
return outputs
def a__ ( self: Optional[int] , __a: Union[str, Any] , __a: str=32 )-> str:
if not isinstance(lowercase__ , lowercase__ ):
raise ValueError("""ConversationalPipeline, expects Conversation as inputs""" )
if conversation.new_user_input is None:
raise ValueError(
f'Conversation with UUID {type(conversation.uuid )} does not contain new user input to process. '
"""Add user inputs with the conversation's `add_user_input` method""" )
if hasattr(self.tokenizer , """_build_conversation_input_ids""" ):
lowerCamelCase : str = self.tokenizer._build_conversation_input_ids(lowercase__ )
else:
# If the tokenizer cannot handle conversations, we default to only the old version
lowerCamelCase : Any = self._legacy_parse_and_tokenize(lowercase__ )
if self.framework == "pt":
lowerCamelCase : Optional[Any] = torch.LongTensor([input_ids] )
elif self.framework == "tf":
lowerCamelCase : Dict = tf.constant([input_ids] )
return {"input_ids": input_ids, "conversation": conversation}
def a__ ( self: Tuple , __a: List[str] , __a: str=10 , **__a: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Optional[int] = generate_kwargs.get("""max_length""" , self.model.config.max_length )
lowerCamelCase : str = model_inputs["""input_ids"""].shape[1]
if max_length - minimum_tokens < n:
logger.warning(f'Conversation input is to long ({n}), trimming it to ({max_length} - {minimum_tokens})' )
lowerCamelCase : int = max_length - minimum_tokens
lowerCamelCase : List[Any] = model_inputs["""input_ids"""][:, -trim:]
if "attention_mask" in model_inputs:
lowerCamelCase : Union[str, Any] = model_inputs["""attention_mask"""][:, -trim:]
lowerCamelCase : int = model_inputs.pop("""conversation""" )
lowerCamelCase : Dict = max_length
lowerCamelCase : Tuple = self.model.generate(**lowercase__ , **lowercase__ )
if self.model.config.is_encoder_decoder:
lowerCamelCase : str = 1
else:
lowerCamelCase : List[Any] = n
return {"output_ids": output_ids[:, start_position:], "conversation": conversation}
def a__ ( self: Optional[int] , __a: str , __a: Any=True )-> Any:
lowerCamelCase : int = model_outputs["""output_ids"""]
lowerCamelCase : List[Any] = self.tokenizer.decode(
output_ids[0] , skip_special_tokens=lowercase__ , clean_up_tokenization_spaces=lowercase__ , )
lowerCamelCase : Optional[Any] = model_outputs["""conversation"""]
conversation.mark_processed()
conversation.append_response(lowercase__ )
return conversation
def a__ ( self: Dict , __a: Dict )-> Union[str, Any]:
lowerCamelCase : Dict = self.tokenizer.eos_token_id
lowerCamelCase : Optional[int] = []
for is_user, text in conversation.iter_texts():
if eos_token_id is not None:
input_ids.extend(self.tokenizer.encode(lowercase__ , add_special_tokens=lowercase__ ) + [eos_token_id] )
else:
input_ids.extend(self.tokenizer.encode(lowercase__ , add_special_tokens=lowercase__ ) )
if len(lowercase__ ) > self.tokenizer.model_max_length:
lowerCamelCase : int = input_ids[-self.tokenizer.model_max_length :]
return input_ids
| 703
|
"""simple docstring"""
import os
def snake_case ( ) -> Optional[Any]:
with open(os.path.dirname(UpperCamelCase__ ) + """/grid.txt""" ) as f:
lowerCamelCase : int = [] # noqa: E741
for _ in range(20 ):
l.append([int(UpperCamelCase__ ) for x in f.readline().split()] )
lowerCamelCase : Union[str, Any] = 0
# right
for i in range(20 ):
for j in range(17 ):
lowerCamelCase : Dict = l[i][j] * l[i][j + 1] * l[i][j + 2] * l[i][j + 3]
if temp > maximum:
lowerCamelCase : Tuple = temp
# down
for i in range(17 ):
for j in range(20 ):
lowerCamelCase : Any = l[i][j] * l[i + 1][j] * l[i + 2][j] * l[i + 3][j]
if temp > maximum:
lowerCamelCase : Optional[Any] = temp
# diagonal 1
for i in range(17 ):
for j in range(17 ):
lowerCamelCase : List[Any] = l[i][j] * l[i + 1][j + 1] * l[i + 2][j + 2] * l[i + 3][j + 3]
if temp > maximum:
lowerCamelCase : List[str] = temp
# diagonal 2
for i in range(17 ):
for j in range(3 , 20 ):
lowerCamelCase : List[str] = l[i][j] * l[i + 1][j - 1] * l[i + 2][j - 2] * l[i + 3][j - 3]
if temp > maximum:
lowerCamelCase : List[Any] = temp
return maximum
if __name__ == "__main__":
print(solution())
| 42
| 0
|
"""simple docstring"""
import numpy as np
from PIL import Image
def snake_case ( UpperCamelCase__ : np.ndarray , UpperCamelCase__ : int , UpperCamelCase__ : int ) -> Optional[Any]:
lowerCamelCase : Optional[int] = np.array(a_ )
if arr.shape[0] != arr.shape[1]:
raise ValueError("""The input array is not a square matrix""" )
lowerCamelCase : Dict = 0
lowerCamelCase : int = 0
lowerCamelCase : List[Any] = 0
lowerCamelCase : Dict = 0
# compute the shape of the output matrix
lowerCamelCase : Optional[Any] = (arr.shape[0] - size) // stride + 1
# initialize the output matrix with zeros of shape maxpool_shape
lowerCamelCase : Union[str, Any] = np.zeros((maxpool_shape, maxpool_shape) )
while i < arr.shape[0]:
if i + size > arr.shape[0]:
# if the end of the matrix is reached, break
break
while j < arr.shape[1]:
# if the end of the matrix is reached, break
if j + size > arr.shape[1]:
break
# compute the maximum of the pooling matrix
lowerCamelCase : Union[str, Any] = np.max(arr[i : i + size, j : j + size] )
# shift the pooling matrix by stride of column pixels
j += stride
mat_j += 1
# shift the pooling matrix by stride of row pixels
i += stride
mat_i += 1
# reset the column index to 0
lowerCamelCase : Dict = 0
lowerCamelCase : int = 0
return updated_arr
def snake_case ( UpperCamelCase__ : np.ndarray , UpperCamelCase__ : int , UpperCamelCase__ : int ) -> Tuple:
lowerCamelCase : Optional[int] = np.array(a_ )
if arr.shape[0] != arr.shape[1]:
raise ValueError("""The input array is not a square matrix""" )
lowerCamelCase : Any = 0
lowerCamelCase : Any = 0
lowerCamelCase : Any = 0
lowerCamelCase : List[Any] = 0
# compute the shape of the output matrix
lowerCamelCase : List[Any] = (arr.shape[0] - size) // stride + 1
# initialize the output matrix with zeros of shape avgpool_shape
lowerCamelCase : Union[str, Any] = np.zeros((avgpool_shape, avgpool_shape) )
while i < arr.shape[0]:
# if the end of the matrix is reached, break
if i + size > arr.shape[0]:
break
while j < arr.shape[1]:
# if the end of the matrix is reached, break
if j + size > arr.shape[1]:
break
# compute the average of the pooling matrix
lowerCamelCase : Optional[int] = int(np.average(arr[i : i + size, j : j + size] ) )
# shift the pooling matrix by stride of column pixels
j += stride
mat_j += 1
# shift the pooling matrix by stride of row pixels
i += stride
mat_i += 1
# reset the column index to 0
lowerCamelCase : str = 0
lowerCamelCase : List[str] = 0
return updated_arr
# Main Function
if __name__ == "__main__":
from doctest import testmod
testmod(name='avgpooling', verbose=True)
# Loading the image
__lowerCamelCase :List[Any] = Image.open('path_to_image')
# Converting the image to numpy array and maxpooling, displaying the result
# Ensure that the image is a square matrix
Image.fromarray(maxpooling(np.array(image), size=3, stride=2)).show()
# Converting the image to numpy array and averagepooling, displaying the result
# Ensure that the image is a square matrix
Image.fromarray(avgpooling(np.array(image), size=3, stride=2)).show()
| 704
|
"""simple docstring"""
import gc
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DDIMScheduler,
StableDiffusionAttendAndExcitePipeline,
UNetaDConditionModel,
)
from diffusers.utils import load_numpy, skip_mps, slow
from diffusers.utils.testing_utils import require_torch_gpu
from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
from ..test_pipelines_common import PipelineKarrasSchedulerTesterMixin, PipelineLatentTesterMixin, PipelineTesterMixin
__lowerCamelCase :Any = False
@skip_mps
class A__ ( __lowercase , __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[Any] =StableDiffusionAttendAndExcitePipeline
snake_case__ : Any =False
snake_case__ : Dict =TEXT_TO_IMAGE_PARAMS
snake_case__ : Any =TEXT_TO_IMAGE_BATCH_PARAMS.union({'''token_indices'''})
snake_case__ : Dict =TEXT_TO_IMAGE_IMAGE_PARAMS
snake_case__ : str =TEXT_TO_IMAGE_IMAGE_PARAMS
@classmethod
def a__ ( cls: Dict )-> Tuple:
super().setUpClass()
torch.use_deterministic_algorithms(__a )
@classmethod
def a__ ( cls: Union[str, Any] )-> Any:
super().tearDownClass()
torch.use_deterministic_algorithms(__a )
def a__ ( self: Tuple )-> Union[str, Any]:
torch.manual_seed(0 )
lowerCamelCase : str = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=1 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , cross_attention_dim=32 , attention_head_dim=(2, 4) , use_linear_projection=__a , )
lowerCamelCase : Union[str, Any] = DDIMScheduler(
beta_start=0.0_00_85 , beta_end=0.0_12 , beta_schedule="""scaled_linear""" , clip_sample=__a , set_alpha_to_one=__a , )
torch.manual_seed(0 )
lowerCamelCase : Union[str, Any] = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
lowerCamelCase : str = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-0_5 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1_000 , hidden_act="""gelu""" , projection_dim=512 , )
lowerCamelCase : Optional[int] = CLIPTextModel(__a )
lowerCamelCase : str = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" )
lowerCamelCase : List[str] = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""safety_checker""": None,
"""feature_extractor""": None,
}
return components
def a__ ( self: Tuple , __a: int , __a: Union[str, Any]=0 )-> Optional[Any]:
if str(__a ).startswith("""mps""" ):
lowerCamelCase : Tuple = torch.manual_seed(__a )
else:
lowerCamelCase : str = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : Dict = {
"""prompt""": """a cat and a frog""",
"""token_indices""": [2, 5],
"""generator""": generator,
"""num_inference_steps""": 1,
"""guidance_scale""": 6.0,
"""output_type""": """numpy""",
"""max_iter_to_alter""": 2,
"""thresholds""": {0: 0.7},
}
return inputs
def a__ ( self: Dict )-> str:
lowerCamelCase : Tuple = """cpu"""
lowerCamelCase : List[str] = self.get_dummy_components()
lowerCamelCase : List[Any] = self.pipeline_class(**__a )
pipe.to(__a )
pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Any = self.get_dummy_inputs(__a )
lowerCamelCase : Union[str, Any] = pipe(**__a ).images
lowerCamelCase : Tuple = image[0, -3:, -3:, -1]
self.assertEqual(image.shape , (1, 64, 64, 3) )
lowerCamelCase : Optional[Any] = np.array(
[0.63_90_53_64, 0.62_89_73_07, 0.48_59_90_17, 0.5_13_36_24, 0.5_55_00_48, 0.45_76_95_16, 0.50_32_69_73, 0.5_02_31_39, 0.45_38_44_96] )
lowerCamelCase : Optional[Any] = np.abs(image_slice.flatten() - expected_slice ).max()
self.assertLessEqual(__a , 1e-3 )
def a__ ( self: int )-> Optional[Any]:
super().test_cpu_offload_forward_pass(expected_max_diff=5e-4 )
def a__ ( self: Union[str, Any] )-> Optional[int]:
# NOTE: Larger batch sizes cause this test to timeout, only test on smaller batches
self._test_inference_batch_consistent(batch_sizes=[1, 2] )
def a__ ( self: Tuple )-> int:
self._test_inference_batch_single_identical(batch_size=2 , expected_max_diff=7e-4 )
def a__ ( self: Dict )-> List[Any]:
super().test_dict_tuple_outputs_equivalent(expected_max_difference=3e-3 )
def a__ ( self: Optional[int] )-> Dict:
super().test_pt_np_pil_outputs_equivalent(expected_max_diff=5e-4 )
def a__ ( self: Any )-> Tuple:
super().test_save_load_local(expected_max_difference=5e-4 )
def a__ ( self: str )-> str:
super().test_save_load_optional_components(expected_max_difference=4e-4 )
@require_torch_gpu
@slow
class A__ ( unittest.TestCase):
"""simple docstring"""
@classmethod
def a__ ( cls: Any )-> Tuple:
super().setUpClass()
torch.use_deterministic_algorithms(__a )
@classmethod
def a__ ( cls: Dict )-> Optional[int]:
super().tearDownClass()
torch.use_deterministic_algorithms(__a )
def a__ ( self: int )-> Optional[int]:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : List[Any] = torch.manual_seed(51 )
lowerCamelCase : List[str] = StableDiffusionAttendAndExcitePipeline.from_pretrained(
"""CompVis/stable-diffusion-v1-4""" , safety_checker=__a , torch_dtype=torch.floataa )
pipe.to("""cuda""" )
lowerCamelCase : Dict = """a painting of an elephant with glasses"""
lowerCamelCase : Any = [5, 7]
lowerCamelCase : Tuple = pipe(
prompt=__a , token_indices=__a , guidance_scale=7.5 , generator=__a , num_inference_steps=5 , max_iter_to_alter=5 , output_type="""numpy""" , ).images[0]
lowerCamelCase : Union[str, Any] = load_numpy(
"""https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/attend-and-excite/elephant_glasses.npy""" )
assert np.abs((expected_image - image).max() ) < 5e-1
| 42
| 0
|
"""simple docstring"""
import argparse
import json
import re
from pathlib import Path
import requests
import torch
from huggingface_hub import hf_hub_download
from PIL import Image
from transformers import (
MobileNetVaConfig,
MobileNetVaForImageClassification,
MobileNetVaImageProcessor,
load_tf_weights_in_mobilenet_va,
)
from transformers.utils import logging
logging.set_verbosity_info()
__lowerCamelCase :Union[str, Any] = logging.get_logger(__name__)
def snake_case ( UpperCamelCase__ : List[str] ) -> str:
lowerCamelCase : List[Any] = MobileNetVaConfig(layer_norm_eps=0.0_0_1 )
if "_quant" in model_name:
raise ValueError("""Quantized models are not supported.""" )
lowerCamelCase : Any = re.match(R"""^mobilenet_v1_([^_]*)_([^_]*)$""" , UpperCamelCase__ )
if matches:
lowerCamelCase : int = float(matches[1] )
lowerCamelCase : Union[str, Any] = int(matches[2] )
# The TensorFlow version of MobileNetV1 predicts 1001 classes instead of
# the usual 1000. The first class (index 0) is "background".
lowerCamelCase : Dict = 1001
lowerCamelCase : Dict = """imagenet-1k-id2label.json"""
lowerCamelCase : int = """huggingface/label-files"""
lowerCamelCase : Union[str, Any] = json.load(open(hf_hub_download(UpperCamelCase__ , UpperCamelCase__ , repo_type="""dataset""" ) , """r""" ) )
lowerCamelCase : Optional[Any] = {int(UpperCamelCase__ ) + 1: v for k, v in idalabel.items()}
lowerCamelCase : Union[str, Any] = """background"""
lowerCamelCase : List[Any] = idalabel
lowerCamelCase : Any = {v: k for k, v in idalabel.items()}
return config
def snake_case ( ) -> str:
lowerCamelCase : List[Any] = """http://images.cocodataset.org/val2017/000000039769.jpg"""
lowerCamelCase : List[Any] = Image.open(requests.get(UpperCamelCase__ , stream=UpperCamelCase__ ).raw )
return im
@torch.no_grad()
def snake_case ( UpperCamelCase__ : str , UpperCamelCase__ : str , UpperCamelCase__ : str , UpperCamelCase__ : Union[str, Any]=False ) -> Union[str, Any]:
lowerCamelCase : int = get_mobilenet_va_config(UpperCamelCase__ )
# Load 🤗 model
lowerCamelCase : int = MobileNetVaForImageClassification(UpperCamelCase__ ).eval()
# Load weights from TensorFlow checkpoint
load_tf_weights_in_mobilenet_va(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
# Check outputs on an image, prepared by MobileNetV1ImageProcessor
lowerCamelCase : Dict = MobileNetVaImageProcessor(
crop_size={"""width""": config.image_size, """height""": config.image_size} , size={"""shortest_edge""": config.image_size + 32} , )
lowerCamelCase : Tuple = image_processor(images=prepare_img() , return_tensors="""pt""" )
lowerCamelCase : Dict = model(**UpperCamelCase__ )
lowerCamelCase : Union[str, Any] = outputs.logits
assert logits.shape == (1, 1001)
if model_name == "mobilenet_v1_1.0_224":
lowerCamelCase : Dict = torch.tensor([-4.1_7_3_9, -1.1_2_3_3, 3.1_2_0_5] )
elif model_name == "mobilenet_v1_0.75_192":
lowerCamelCase : Dict = torch.tensor([-3.9_4_4_0, -2.3_1_4_1, -0.3_3_3_3] )
else:
lowerCamelCase : Dict = None
if expected_logits is not None:
assert torch.allclose(logits[0, :3] , UpperCamelCase__ , atol=1E-4 )
Path(UpperCamelCase__ ).mkdir(exist_ok=UpperCamelCase__ )
print(F'Saving model {model_name} to {pytorch_dump_folder_path}' )
model.save_pretrained(UpperCamelCase__ )
print(F'Saving image processor to {pytorch_dump_folder_path}' )
image_processor.save_pretrained(UpperCamelCase__ )
if push_to_hub:
print("""Pushing to the hub...""" )
lowerCamelCase : Optional[Any] = """google/""" + model_name
image_processor.push_to_hub(UpperCamelCase__ )
model.push_to_hub(UpperCamelCase__ )
if __name__ == "__main__":
__lowerCamelCase :str = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--model_name',
default='mobilenet_v1_1.0_224',
type=str,
help='Name of the MobileNetV1 model you\'d like to convert. Should in the form \'mobilenet_v1_<depth>_<size>\'.',
)
parser.add_argument(
'--checkpoint_path', required=True, type=str, help='Path to the original TensorFlow checkpoint (.ckpt file).'
)
parser.add_argument(
'--pytorch_dump_folder_path', required=True, type=str, help='Path to the output PyTorch model directory.'
)
parser.add_argument(
'--push_to_hub', action='store_true', help='Whether or not to push the converted model to the 🤗 hub.'
)
__lowerCamelCase :Optional[int] = parser.parse_args()
convert_movilevit_checkpoint(
args.model_name, args.checkpoint_path, args.pytorch_dump_folder_path, args.push_to_hub
)
| 705
|
"""simple docstring"""
import unittest
from transformers import EsmConfig, is_torch_available
from transformers.testing_utils import TestCasePlus, require_torch, slow, torch_device
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, ids_tensor, random_attention_mask
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import EsmForMaskedLM, EsmForSequenceClassification, EsmForTokenClassification, EsmModel
from transformers.models.esm.modeling_esm import (
ESM_PRETRAINED_MODEL_ARCHIVE_LIST,
EsmEmbeddings,
create_position_ids_from_input_ids,
)
class A__ :
"""simple docstring"""
def __init__( self: List[str] , __a: List[str] , __a: Dict=13 , __a: Tuple=7 , __a: Dict=False , __a: str=True , __a: List[Any]=False , __a: Dict=True , __a: Any=33 , __a: Optional[Any]=32 , __a: List[Any]=5 , __a: Any=4 , __a: Dict=37 , __a: str="gelu" , __a: str=0.1 , __a: int=0.1 , __a: Optional[int]=512 , __a: List[Any]=16 , __a: int=2 , __a: int=0.02 , __a: Optional[int]=3 , __a: str=4 , __a: Tuple=None , )-> Tuple:
lowerCamelCase : Union[str, Any] = parent
lowerCamelCase : Tuple = batch_size
lowerCamelCase : Any = seq_length
lowerCamelCase : Any = is_training
lowerCamelCase : Tuple = use_input_mask
lowerCamelCase : int = use_token_type_ids
lowerCamelCase : List[str] = use_labels
lowerCamelCase : Optional[int] = vocab_size
lowerCamelCase : Tuple = hidden_size
lowerCamelCase : List[str] = num_hidden_layers
lowerCamelCase : Optional[int] = num_attention_heads
lowerCamelCase : Optional[Any] = intermediate_size
lowerCamelCase : Optional[Any] = hidden_act
lowerCamelCase : Union[str, Any] = hidden_dropout_prob
lowerCamelCase : Optional[Any] = attention_probs_dropout_prob
lowerCamelCase : Any = max_position_embeddings
lowerCamelCase : str = type_vocab_size
lowerCamelCase : List[Any] = type_sequence_label_size
lowerCamelCase : Optional[Any] = initializer_range
lowerCamelCase : Union[str, Any] = num_labels
lowerCamelCase : Optional[Any] = num_choices
lowerCamelCase : Any = scope
def a__ ( self: Optional[int] )-> List[Any]:
lowerCamelCase : List[Any] = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : Dict = None
if self.use_input_mask:
lowerCamelCase : Tuple = random_attention_mask([self.batch_size, self.seq_length] )
lowerCamelCase : Any = None
lowerCamelCase : int = None
lowerCamelCase : Union[str, Any] = None
if self.use_labels:
lowerCamelCase : Optional[Any] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : Optional[int] = ids_tensor([self.batch_size, self.seq_length] , self.num_labels )
lowerCamelCase : Union[str, Any] = ids_tensor([self.batch_size] , self.num_choices )
lowerCamelCase : List[str] = self.get_config()
return config, input_ids, input_mask, sequence_labels, token_labels, choice_labels
def a__ ( self: Tuple )-> Union[str, Any]:
return EsmConfig(
vocab_size=self.vocab_size , hidden_size=self.hidden_size , pad_token_id=1 , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , initializer_range=self.initializer_range , )
def a__ ( self: List[Any] , __a: List[str] , __a: str , __a: Tuple , __a: List[str] , __a: List[str] , __a: str )-> int:
lowerCamelCase : Optional[int] = EsmModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a , attention_mask=__a )
lowerCamelCase : str = model(__a )
lowerCamelCase : Optional[Any] = model(__a )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) )
self.parent.assertEqual(result.pooler_output.shape , (self.batch_size, self.hidden_size) )
def a__ ( self: int , __a: Union[str, Any] , __a: Optional[int] , __a: List[str] , __a: str , __a: List[str] , __a: Tuple )-> int:
lowerCamelCase : str = EsmForMaskedLM(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[Any] = model(__a , attention_mask=__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) )
def a__ ( self: List[str] , __a: List[Any] , __a: List[str] , __a: int , __a: Union[str, Any] , __a: List[Any] , __a: Tuple )-> List[str]:
lowerCamelCase : Tuple = self.num_labels
lowerCamelCase : Dict = EsmForTokenClassification(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a , attention_mask=__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) )
def a__ ( self: Optional[int] )-> Optional[int]:
lowerCamelCase : Any = self.prepare_config_and_inputs()
(
(
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) ,
) : Tuple = config_and_inputs
lowerCamelCase : List[Any] = {"""input_ids""": input_ids, """attention_mask""": input_mask}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Any =False
snake_case__ : Dict =(
(
EsmForMaskedLM,
EsmModel,
EsmForSequenceClassification,
EsmForTokenClassification,
)
if is_torch_available()
else ()
)
snake_case__ : Dict =()
snake_case__ : Optional[int] =(
{
'''feature-extraction''': EsmModel,
'''fill-mask''': EsmForMaskedLM,
'''text-classification''': EsmForSequenceClassification,
'''token-classification''': EsmForTokenClassification,
'''zero-shot''': EsmForSequenceClassification,
}
if is_torch_available()
else {}
)
snake_case__ : Any =True
def a__ ( self: Optional[int] )-> Optional[int]:
lowerCamelCase : Optional[Any] = EsmModelTester(self )
lowerCamelCase : Any = ConfigTester(self , config_class=__a , hidden_size=37 )
def a__ ( self: List[Any] )-> Optional[Any]:
self.config_tester.run_common_tests()
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: Tuple )-> Any:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
for type in ["absolute", "relative_key", "relative_key_query"]:
lowerCamelCase : Tuple = type
self.model_tester.create_and_check_model(*__a )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Optional[int] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_masked_lm(*__a )
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_token_classification(*__a )
@slow
def a__ ( self: Any )-> List[Any]:
for model_name in ESM_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : int = EsmModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def a__ ( self: str )-> List[str]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()[0]
lowerCamelCase : Union[str, Any] = EsmEmbeddings(config=__a )
lowerCamelCase : List[str] = torch.as_tensor([[12, 31, 13, model.padding_idx]] )
lowerCamelCase : Union[str, Any] = torch.as_tensor(
[
[
0 + model.padding_idx + 1,
1 + model.padding_idx + 1,
2 + model.padding_idx + 1,
model.padding_idx,
]
] )
lowerCamelCase : Optional[Any] = create_position_ids_from_input_ids(__a , model.padding_idx )
self.assertEqual(position_ids.shape , expected_positions.shape )
self.assertTrue(torch.all(torch.eq(__a , __a ) ) )
def a__ ( self: Optional[int] )-> int:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()[0]
lowerCamelCase : Any = EsmEmbeddings(config=__a )
lowerCamelCase : Dict = torch.empty(2 , 4 , 30 )
lowerCamelCase : List[Any] = [
0 + embeddings.padding_idx + 1,
1 + embeddings.padding_idx + 1,
2 + embeddings.padding_idx + 1,
3 + embeddings.padding_idx + 1,
]
lowerCamelCase : Any = torch.as_tensor([expected_single_positions, expected_single_positions] )
lowerCamelCase : List[str] = embeddings.create_position_ids_from_inputs_embeds(__a )
self.assertEqual(position_ids.shape , expected_positions.shape )
self.assertTrue(torch.all(torch.eq(__a , __a ) ) )
@unittest.skip("""Esm does not support embedding resizing""" )
def a__ ( self: Any )-> Optional[Any]:
pass
@unittest.skip("""Esm does not support embedding resizing""" )
def a__ ( self: Dict )-> Dict:
pass
@unittest.skip("""Will be fixed soon by reducing the size of the model used for common tests.""" )
def a__ ( self: List[str] )-> Dict:
pass
@require_torch
class A__ ( __lowercase):
"""simple docstring"""
@slow
def a__ ( self: Any )-> Union[str, Any]:
with torch.no_grad():
lowerCamelCase : Union[str, Any] = EsmForMaskedLM.from_pretrained("""facebook/esm2_t6_8M_UR50D""" )
model.eval()
lowerCamelCase : List[str] = torch.tensor([[0, 1, 2, 3, 4, 5]] )
lowerCamelCase : Tuple = model(__a )[0]
lowerCamelCase : Dict = 33
lowerCamelCase : List[str] = torch.Size((1, 6, vocab_size) )
self.assertEqual(output.shape , __a )
lowerCamelCase : Tuple = torch.tensor(
[[[8.92_15, -10.58_98, -6.46_71], [-6.39_67, -13.91_14, -1.12_12], [-7.78_12, -13.95_16, -3.74_06]]] )
self.assertTrue(torch.allclose(output[:, :3, :3] , __a , atol=1e-4 ) )
@slow
def a__ ( self: Dict )-> str:
with torch.no_grad():
lowerCamelCase : Any = EsmModel.from_pretrained("""facebook/esm2_t6_8M_UR50D""" )
model.eval()
lowerCamelCase : Optional[Any] = torch.tensor([[0, 6, 4, 13, 5, 4, 16, 12, 11, 7, 2]] )
lowerCamelCase : Any = model(__a )[0]
# compare the actual values for a slice.
lowerCamelCase : Tuple = torch.tensor(
[[[0.14_44, 0.54_13, 0.32_48], [0.30_34, 0.00_53, 0.31_08], [0.32_28, -0.24_99, 0.34_15]]] )
self.assertTrue(torch.allclose(output[:, :3, :3] , __a , atol=1e-4 ) )
| 42
| 0
|
"""simple docstring"""
import json
import os
from functools import lru_cache
from typing import Dict, List, Optional, Tuple, Union
import regex as re
from ...tokenization_utils import AddedToken, PreTrainedTokenizer
from ...tokenization_utils_base import BatchEncoding, EncodedInput
from ...utils import PaddingStrategy, logging
__lowerCamelCase :Optional[Any] = logging.get_logger(__name__)
__lowerCamelCase :Union[str, Any] = {'vocab_file': 'vocab.json', 'merges_file': 'merges.txt'}
# See all LED models at https://huggingface.co/models?filter=LED
__lowerCamelCase :Union[str, Any] = {
'vocab_file': {
'allenai/led-base-16384': 'https://huggingface.co/allenai/led-base-16384/resolve/main/vocab.json',
},
'merges_file': {
'allenai/led-base-16384': 'https://huggingface.co/allenai/led-base-16384/resolve/main/merges.txt',
},
'tokenizer_file': {
'allenai/led-base-16384': 'https://huggingface.co/allenai/led-base-16384/resolve/main/tokenizer.json',
},
}
__lowerCamelCase :Union[str, Any] = {
'allenai/led-base-16384': 16_384,
}
@lru_cache()
# Copied from transformers.models.bart.tokenization_bart.bytes_to_unicode
def snake_case ( ) -> Union[str, Any]:
lowerCamelCase : Union[str, Any] = (
list(range(ord("""!""" ) , ord("""~""" ) + 1 ) ) + list(range(ord("""¡""" ) , ord("""¬""" ) + 1 ) ) + list(range(ord("""®""" ) , ord("""ÿ""" ) + 1 ) )
)
lowerCamelCase : List[str] = bs[:]
lowerCamelCase : str = 0
for b in range(2**8 ):
if b not in bs:
bs.append(lowercase_ )
cs.append(2**8 + n )
n += 1
lowerCamelCase : str = [chr(lowercase_ ) for n in cs]
return dict(zip(lowercase_ , lowercase_ ) )
def snake_case ( UpperCamelCase__ : int ) -> str:
lowerCamelCase : Dict = set()
lowerCamelCase : Optional[Any] = word[0]
for char in word[1:]:
pairs.add((prev_char, char) )
lowerCamelCase : Dict = char
return pairs
class A__ ( _UpperCamelCase):
"""simple docstring"""
snake_case__ : Optional[int] =VOCAB_FILES_NAMES
snake_case__ : Union[str, Any] =PRETRAINED_VOCAB_FILES_MAP
snake_case__ : Tuple =PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
snake_case__ : Optional[int] =["input_ids", "attention_mask"]
def __init__( self: Optional[Any] , __a: int , __a: Optional[int] , __a: Optional[int]="replace" , __a: Any="<s>" , __a: Dict="</s>" , __a: List[str]="</s>" , __a: int="<s>" , __a: Any="<unk>" , __a: str="<pad>" , __a: Any="<mask>" , __a: List[Any]=False , **__a: List[str] , )-> str:
lowerCamelCase : Tuple = AddedToken(__a , lstrip=__a , rstrip=__a ) if isinstance(__a , __a ) else bos_token
lowerCamelCase : List[str] = AddedToken(__a , lstrip=__a , rstrip=__a ) if isinstance(__a , __a ) else eos_token
lowerCamelCase : Tuple = AddedToken(__a , lstrip=__a , rstrip=__a ) if isinstance(__a , __a ) else sep_token
lowerCamelCase : str = AddedToken(__a , lstrip=__a , rstrip=__a ) if isinstance(__a , __a ) else cls_token
lowerCamelCase : Any = AddedToken(__a , lstrip=__a , rstrip=__a ) if isinstance(__a , __a ) else unk_token
lowerCamelCase : List[Any] = AddedToken(__a , lstrip=__a , rstrip=__a ) if isinstance(__a , __a ) else pad_token
# Mask token behave like a normal word, i.e. include the space before it
lowerCamelCase : int = AddedToken(__a , lstrip=__a , rstrip=__a ) if isinstance(__a , __a ) else mask_token
super().__init__(
errors=__a , bos_token=__a , eos_token=__a , unk_token=__a , sep_token=__a , cls_token=__a , pad_token=__a , mask_token=__a , add_prefix_space=__a , **__a , )
with open(__a , encoding="""utf-8""" ) as vocab_handle:
lowerCamelCase : int = json.load(__a )
lowerCamelCase : str = {v: k for k, v in self.encoder.items()}
lowerCamelCase : int = errors # how to handle errors in decoding
lowerCamelCase : str = bytes_to_unicode()
lowerCamelCase : Optional[Any] = {v: k for k, v in self.byte_encoder.items()}
with open(__a , encoding="""utf-8""" ) as merges_handle:
lowerCamelCase : List[Any] = merges_handle.read().split("""\n""" )[1:-1]
lowerCamelCase : Optional[int] = [tuple(merge.split() ) for merge in bpe_merges]
lowerCamelCase : List[Any] = dict(zip(__a , range(len(__a ) ) ) )
lowerCamelCase : Any = {}
lowerCamelCase : List[Any] = add_prefix_space
# Should have added re.IGNORECASE so BPE merges can happen for capitalized versions of contractions
lowerCamelCase : Optional[Any] = re.compile(r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""" )
@property
# Copied from transformers.models.bart.tokenization_bart.BartTokenizer.vocab_size
def a__ ( self: Dict )-> Optional[Any]:
return len(self.encoder )
def a__ ( self: List[str] )-> Tuple:
return dict(self.encoder , **self.added_tokens_encoder )
def a__ ( self: List[Any] , __a: Optional[Any] )-> Tuple:
if token in self.cache:
return self.cache[token]
lowerCamelCase : Optional[int] = tuple(__a )
lowerCamelCase : List[Any] = get_pairs(__a )
if not pairs:
return token
while True:
lowerCamelCase : List[str] = min(__a , key=lambda __a : self.bpe_ranks.get(__a , float("""inf""" ) ) )
if bigram not in self.bpe_ranks:
break
lowerCamelCase : Union[str, Any] = bigram
lowerCamelCase : Dict = []
lowerCamelCase : Optional[Any] = 0
while i < len(__a ):
try:
lowerCamelCase : Dict = word.index(__a , __a )
except ValueError:
new_word.extend(word[i:] )
break
else:
new_word.extend(word[i:j] )
lowerCamelCase : str = j
if word[i] == first and i < len(__a ) - 1 and word[i + 1] == second:
new_word.append(first + second )
i += 2
else:
new_word.append(word[i] )
i += 1
lowerCamelCase : str = tuple(__a )
lowerCamelCase : Optional[int] = new_word
if len(__a ) == 1:
break
else:
lowerCamelCase : Any = get_pairs(__a )
lowerCamelCase : str = " ".join(__a )
lowerCamelCase : Dict = word
return word
def a__ ( self: List[str] , __a: List[str] )-> Tuple:
lowerCamelCase : str = []
for token in re.findall(self.pat , __a ):
lowerCamelCase : List[str] = "".join(
self.byte_encoder[b] for b in token.encode("""utf-8""" ) ) # Maps all our bytes to unicode strings, avoiding control tokens of the BPE (spaces in our case)
bpe_tokens.extend(bpe_token for bpe_token in self.bpe(__a ).split(""" """ ) )
return bpe_tokens
def a__ ( self: List[str] , __a: int )-> Tuple:
return self.encoder.get(__a , self.encoder.get(self.unk_token ) )
def a__ ( self: List[Any] , __a: Dict )-> Union[str, Any]:
return self.decoder.get(__a )
def a__ ( self: Dict , __a: Tuple )-> str:
lowerCamelCase : Tuple = "".join(__a )
lowerCamelCase : List[Any] = bytearray([self.byte_decoder[c] for c in text] ).decode("""utf-8""" , errors=self.errors )
return text
def a__ ( self: Any , __a: str , __a: Optional[str] = None )-> Tuple[str]:
if not os.path.isdir(__a ):
logger.error(f'Vocabulary path ({save_directory}) should be a directory' )
return
lowerCamelCase : Tuple = os.path.join(
__a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""vocab_file"""] )
lowerCamelCase : Dict = os.path.join(
__a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""merges_file"""] )
with open(__a , """w""" , encoding="""utf-8""" ) as f:
f.write(json.dumps(self.encoder , indent=2 , sort_keys=__a , ensure_ascii=__a ) + """\n""" )
lowerCamelCase : str = 0
with open(__a , """w""" , encoding="""utf-8""" ) as writer:
writer.write("""#version: 0.2\n""" )
for bpe_tokens, token_index in sorted(self.bpe_ranks.items() , key=lambda __a : kv[1] ):
if index != token_index:
logger.warning(
f'Saving vocabulary to {merge_file}: BPE merge indices are not consecutive.'
""" Please check that the tokenizer is not corrupted!""" )
lowerCamelCase : Union[str, Any] = token_index
writer.write(""" """.join(__a ) + """\n""" )
index += 1
return vocab_file, merge_file
def a__ ( self: int , __a: List[int] , __a: Optional[List[int]] = None )-> List[int]:
if token_ids_a is None:
return [self.cls_token_id] + token_ids_a + [self.sep_token_id]
lowerCamelCase : Tuple = [self.cls_token_id]
lowerCamelCase : List[Any] = [self.sep_token_id]
return cls + token_ids_a + sep + sep + token_ids_a + sep
def a__ ( self: List[Any] , __a: List[int] , __a: Optional[List[int]] = None , __a: bool = False )-> List[int]:
if already_has_special_tokens:
return super().get_special_tokens_mask(
token_ids_a=__a , token_ids_a=__a , already_has_special_tokens=__a )
if token_ids_a is None:
return [1] + ([0] * len(__a )) + [1]
return [1] + ([0] * len(__a )) + [1, 1] + ([0] * len(__a )) + [1]
def a__ ( self: str , __a: List[int] , __a: Optional[List[int]] = None )-> List[int]:
lowerCamelCase : Optional[int] = [self.sep_token_id]
lowerCamelCase : List[Any] = [self.cls_token_id]
if token_ids_a is None:
return len(cls + token_ids_a + sep ) * [0]
return len(cls + token_ids_a + sep + sep + token_ids_a + sep ) * [0]
def a__ ( self: Any , __a: str , __a: int=False , **__a: List[Any] )-> List[Any]:
lowerCamelCase : Optional[Any] = kwargs.pop("""add_prefix_space""" , self.add_prefix_space )
if (is_split_into_words or add_prefix_space) and (len(__a ) > 0 and not text[0].isspace()):
lowerCamelCase : List[Any] = " " + text
return (text, kwargs)
def a__ ( self: Dict , __a: Union[Dict[str, EncodedInput], BatchEncoding] , __a: Optional[int] = None , __a: PaddingStrategy = PaddingStrategy.DO_NOT_PAD , __a: Optional[int] = None , __a: Optional[bool] = None , )-> dict:
lowerCamelCase : Any = super()._pad(
encoded_inputs=__a , max_length=__a , padding_strategy=__a , pad_to_multiple_of=__a , return_attention_mask=__a , )
# Load from model defaults
if return_attention_mask is None:
lowerCamelCase : str = "attention_mask" in self.model_input_names
if return_attention_mask and "global_attention_mask" in encoded_inputs:
lowerCamelCase : Dict = encoded_inputs[self.model_input_names[0]]
# `global_attention_mask` need to have the same length as other (sequential) inputs.
lowerCamelCase : str = len(encoded_inputs["""global_attention_mask"""] ) != len(__a )
if needs_to_be_padded:
lowerCamelCase : List[str] = len(__a ) - len(encoded_inputs["""global_attention_mask"""] )
if self.padding_side == "right":
# Use `-1` since `0` in `global_attention_mask` means `local attention` instead of `not to attend`
lowerCamelCase : str = (
encoded_inputs["global_attention_mask"] + [-1] * difference
)
elif self.padding_side == "left":
lowerCamelCase : List[Any] = [-1] * difference + encoded_inputs[
"global_attention_mask"
]
else:
raise ValueError("""Invalid padding strategy:""" + str(self.padding_side ) )
return encoded_inputs
| 706
|
"""simple docstring"""
import unittest
from transformers import AlbertTokenizer, AlbertTokenizerFast
from transformers.testing_utils import get_tests_dir, require_sentencepiece, require_tokenizers, slow
from ...test_tokenization_common import TokenizerTesterMixin
__lowerCamelCase :str = get_tests_dir('fixtures/spiece.model')
@require_sentencepiece
@require_tokenizers
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =AlbertTokenizer
snake_case__ : Optional[Any] =AlbertTokenizerFast
snake_case__ : Optional[int] =True
snake_case__ : Any =True
snake_case__ : Optional[int] =True
def a__ ( self: Dict )-> Optional[Any]:
super().setUp()
# We have a SentencePiece fixture for testing
lowerCamelCase : int = AlbertTokenizer(__a )
tokenizer.save_pretrained(self.tmpdirname )
def a__ ( self: Tuple , __a: Tuple )-> Union[str, Any]:
lowerCamelCase : List[str] = """this is a test"""
lowerCamelCase : int = """this is a test"""
return input_text, output_text
def a__ ( self: Any )-> List[Any]:
lowerCamelCase : int = """<pad>"""
lowerCamelCase : Optional[int] = 0
self.assertEqual(self.get_tokenizer()._convert_token_to_id(__a ) , __a )
self.assertEqual(self.get_tokenizer()._convert_id_to_token(__a ) , __a )
def a__ ( self: Tuple )-> str:
lowerCamelCase : Optional[int] = list(self.get_tokenizer().get_vocab().keys() )
self.assertEqual(vocab_keys[0] , """<pad>""" )
self.assertEqual(vocab_keys[1] , """<unk>""" )
self.assertEqual(vocab_keys[-1] , """▁eloquent""" )
self.assertEqual(len(__a ) , 30_000 )
def a__ ( self: List[str] )-> Any:
self.assertEqual(self.get_tokenizer().vocab_size , 30_000 )
def a__ ( self: Optional[Any] )-> Union[str, Any]:
if not self.test_rust_tokenizer:
return
lowerCamelCase : str = self.get_tokenizer()
lowerCamelCase : Tuple = self.get_rust_tokenizer()
lowerCamelCase : Union[str, Any] = """I was born in 92000, and this is falsé."""
lowerCamelCase : List[str] = tokenizer.tokenize(__a )
lowerCamelCase : Tuple = rust_tokenizer.tokenize(__a )
self.assertListEqual(__a , __a )
lowerCamelCase : Dict = tokenizer.encode(__a , add_special_tokens=__a )
lowerCamelCase : List[str] = rust_tokenizer.encode(__a , add_special_tokens=__a )
self.assertListEqual(__a , __a )
lowerCamelCase : Any = self.get_rust_tokenizer()
lowerCamelCase : List[str] = tokenizer.encode(__a )
lowerCamelCase : str = rust_tokenizer.encode(__a )
self.assertListEqual(__a , __a )
def a__ ( self: Tuple )-> List[Any]:
lowerCamelCase : List[str] = AlbertTokenizer(__a , keep_accents=__a )
lowerCamelCase : int = tokenizer.tokenize("""This is a test""" )
self.assertListEqual(__a , ["""▁this""", """▁is""", """▁a""", """▁test"""] )
self.assertListEqual(tokenizer.convert_tokens_to_ids(__a ) , [48, 25, 21, 1_289] )
lowerCamelCase : List[str] = tokenizer.tokenize("""I was born in 92000, and this is falsé.""" )
self.assertListEqual(
__a , ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """é""", """."""] )
lowerCamelCase : List[str] = tokenizer.convert_tokens_to_ids(__a )
self.assertListEqual(__a , [31, 23, 386, 19, 561, 3_050, 15, 17, 48, 25, 8_256, 18, 1, 9] )
lowerCamelCase : Tuple = tokenizer.convert_ids_to_tokens(__a )
self.assertListEqual(
__a , ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """<unk>""", """."""] , )
def a__ ( self: Tuple )-> str:
lowerCamelCase : str = AlbertTokenizer(__a )
lowerCamelCase : Union[str, Any] = tokenizer.encode("""sequence builders""" )
lowerCamelCase : List[Any] = tokenizer.encode("""multi-sequence build""" )
lowerCamelCase : Any = tokenizer.build_inputs_with_special_tokens(__a )
lowerCamelCase : str = tokenizer.build_inputs_with_special_tokens(__a , __a )
assert encoded_sentence == [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id]
assert encoded_pair == [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id] + text_a + [
tokenizer.sep_token_id
]
@slow
def a__ ( self: Any )-> Dict:
# fmt: off
lowerCamelCase : Optional[Any] = {"""attention_mask""": [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """input_ids""": [[2, 21_970, 13, 5, 6_092, 167, 28, 7_103, 2_153, 673, 8, 7_028, 12_051, 18, 17, 7_103, 2_153, 673, 8, 3_515, 18_684, 8, 4_461, 6, 1_927, 297, 8, 12_060, 2_607, 18, 13, 5, 4_461, 15, 10_538, 38, 8, 135, 15, 822, 58, 15, 993, 10_363, 15, 1_460, 8_005, 4_461, 15, 993, 255, 2_328, 9, 9, 9, 6, 26, 1_112, 816, 3_260, 13, 5, 103, 2_377, 6, 17, 1_112, 816, 2_782, 13, 5, 103, 10_641, 6, 29, 84, 2_512, 2_430, 782, 18_684, 2_761, 19, 808, 2_430, 2_556, 17, 855, 1_480, 9_477, 4_091, 128, 11_712, 15, 7_103, 2_153, 673, 17, 24_883, 9_990, 9, 3], [2, 11_502, 25, 1_006, 20, 782, 8, 11_809, 855, 1_732, 19_393, 18_667, 37, 367, 21_018, 69, 1_854, 34, 11_860, 19_124, 27, 156, 225, 17, 193, 4_141, 19, 65, 9_124, 9, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [2, 14, 2_231, 886, 2_385, 17_659, 84, 14, 16_792, 1_952, 9, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """token_type_ids""": [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]} # noqa: E501
# fmt: on
self.tokenizer_integration_test_util(
expected_encoding=__a , model_name="""albert-base-v2""" , revision="""6b6560eaf5ff2e250b00c50f380c5389a9c2d82e""" , )
| 42
| 0
|
import string
from math import logaa
def snake_case ( UpperCamelCase__ : Tuple , UpperCamelCase__ : List[str] ) -> int:
lowerCamelCase : Union[str, Any] = document.translate(
str.maketrans("""""" , """""" , string.punctuation ) ).replace("""\n""" , """""" )
lowerCamelCase : Any = document_without_punctuation.split(""" """ ) # word tokenization
return len([word for word in tokenize_document if word.lower() == term.lower()] )
def snake_case ( UpperCamelCase__ : Any , UpperCamelCase__ : Dict ) -> tuple[int, int]:
lowerCamelCase : Tuple = corpus.lower().translate(
str.maketrans("""""" , """""" , string.punctuation ) ) # strip all punctuation and replace it with ''
lowerCamelCase : Dict = corpus_without_punctuation.split("""\n""" )
lowerCamelCase : Tuple = term.lower()
return (len([doc for doc in docs if term in doc] ), len(__UpperCamelCase ))
def snake_case ( UpperCamelCase__ : List[str] , UpperCamelCase__ : Any , UpperCamelCase__ : Optional[int]=False ) -> float:
if smoothing:
if n == 0:
raise ValueError("""log10(0) is undefined.""" )
return round(1 + logaa(n / (1 + df) ) , 3 )
if df == 0:
raise ZeroDivisionError("""df must be > 0""" )
elif n == 0:
raise ValueError("""log10(0) is undefined.""" )
return round(logaa(n / df ) , 3 )
def snake_case ( UpperCamelCase__ : str , UpperCamelCase__ : Union[str, Any] ) -> float:
return round(tf * idf , 3 )
| 707
|
"""simple docstring"""
__lowerCamelCase :List[Any] = {0: [2, 3], 1: [0], 2: [1], 3: [4], 4: []}
__lowerCamelCase :Union[str, Any] = {0: [1, 2, 3], 1: [2], 2: [0], 3: [4], 4: [5], 5: [3]}
def snake_case ( UpperCamelCase__ : dict[int, list[int]] , UpperCamelCase__ : int , UpperCamelCase__ : list[bool] ) -> list[int]:
lowerCamelCase : Tuple = True
lowerCamelCase : Any = []
for neighbour in graph[vert]:
if not visited[neighbour]:
order += topology_sort(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
order.append(UpperCamelCase__ )
return order
def snake_case ( UpperCamelCase__ : dict[int, list[int]] , UpperCamelCase__ : int , UpperCamelCase__ : list[bool] ) -> list[int]:
lowerCamelCase : List[Any] = True
lowerCamelCase : int = [vert]
for neighbour in reversed_graph[vert]:
if not visited[neighbour]:
component += find_components(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return component
def snake_case ( UpperCamelCase__ : dict[int, list[int]] ) -> list[list[int]]:
lowerCamelCase : int = len(UpperCamelCase__ ) * [False]
lowerCamelCase : dict[int, list[int]] = {vert: [] for vert in range(len(UpperCamelCase__ ) )}
for vert, neighbours in graph.items():
for neighbour in neighbours:
reversed_graph[neighbour].append(UpperCamelCase__ )
lowerCamelCase : int = []
for i, was_visited in enumerate(UpperCamelCase__ ):
if not was_visited:
order += topology_sort(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
lowerCamelCase : Tuple = []
lowerCamelCase : str = len(UpperCamelCase__ ) * [False]
for i in range(len(UpperCamelCase__ ) ):
lowerCamelCase : Any = order[len(UpperCamelCase__ ) - i - 1]
if not visited[vert]:
lowerCamelCase : List[str] = find_components(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
components_list.append(UpperCamelCase__ )
return components_list
| 42
| 0
|
"""simple docstring"""
import argparse
from typing import List
import evaluate
import numpy as np
import torch
from datasets import DatasetDict, load_dataset
# New Code #
# We'll be using StratifiedKFold for this example
from sklearn.model_selection import StratifiedKFold
from torch.optim import AdamW
from torch.utils.data import DataLoader
from transformers import AutoModelForSequenceClassification, AutoTokenizer, get_linear_schedule_with_warmup, set_seed
from accelerate import Accelerator, DistributedType
########################################################################
# This is a fully working simple example to use Accelerate,
# specifically showcasing how to perform Cross Validation,
# and builds off the `nlp_example.py` script.
#
# This example trains a Bert base model on GLUE MRPC
# in any of the following settings (with the same script):
# - single CPU or single GPU
# - multi GPUS (using PyTorch distributed mode)
# - (multi) TPUs
# - fp16 (mixed-precision) or fp32 (normal precision)
#
# To help focus on the differences in the code, building `DataLoaders`
# was refactored into its own function.
# New additions from the base script can be found quickly by
# looking for the # New Code # tags
#
# To run it in each of these various modes, follow the instructions
# in the readme for examples:
# https://github.com/huggingface/accelerate/tree/main/examples
#
########################################################################
__lowerCamelCase :int = 16
__lowerCamelCase :int = 32
def lowerCAmelCase ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Any , UpperCamelCase__ : List[str] , UpperCamelCase__ : List[Any] , UpperCamelCase__ : Optional[int] = 16 ) -> Tuple:
lowerCamelCase : Tuple = AutoTokenizer.from_pretrained("""bert-base-cased""" )
lowerCamelCase : Any = DatasetDict(
{
"""train""": dataset["""train"""].select(UpperCamelCase__ ),
"""validation""": dataset["""train"""].select(UpperCamelCase__ ),
"""test""": dataset["""validation"""],
} )
def tokenize_function(UpperCamelCase__ : Tuple ):
# max_length=None => use the model max length (it's actually the default)
lowerCamelCase : Tuple = tokenizer(examples["""sentence1"""] , examples["""sentence2"""] , truncation=UpperCamelCase__ , max_length=UpperCamelCase__ )
return outputs
# Apply the method we just defined to all the examples in all the splits of the dataset
# starting with the main process first:
with accelerator.main_process_first():
lowerCamelCase : str = datasets.map(
UpperCamelCase__ , batched=UpperCamelCase__ , remove_columns=["""idx""", """sentence1""", """sentence2"""] , )
# We also rename the 'label' column to 'labels' which is the expected name for labels by the models of the
# transformers library
lowerCamelCase : int = tokenized_datasets.rename_column("""label""" , """labels""" )
def collate_fn(UpperCamelCase__ : List[str] ):
# On TPU it's best to pad everything to the same length or training will be very slow.
lowerCamelCase : Tuple = 128 if accelerator.distributed_type == DistributedType.TPU else None
# When using mixed precision we want round multiples of 8/16
if accelerator.mixed_precision == "fp8":
lowerCamelCase : Any = 16
elif accelerator.mixed_precision != "no":
lowerCamelCase : str = 8
else:
lowerCamelCase : Tuple = None
return tokenizer.pad(
UpperCamelCase__ , padding="""longest""" , max_length=UpperCamelCase__ , pad_to_multiple_of=UpperCamelCase__ , return_tensors="""pt""" , )
# Instantiate dataloaders.
lowerCamelCase : Optional[int] = DataLoader(
tokenized_datasets["""train"""] , shuffle=UpperCamelCase__ , collate_fn=UpperCamelCase__ , batch_size=UpperCamelCase__ )
lowerCamelCase : Optional[int] = DataLoader(
tokenized_datasets["""validation"""] , shuffle=UpperCamelCase__ , collate_fn=UpperCamelCase__ , batch_size=UpperCamelCase__ )
lowerCamelCase : Dict = DataLoader(
tokenized_datasets["""test"""] , shuffle=UpperCamelCase__ , collate_fn=UpperCamelCase__ , batch_size=UpperCamelCase__ )
return train_dataloader, eval_dataloader, test_dataloader
def lowerCAmelCase ( UpperCamelCase__ : List[str] , UpperCamelCase__ : Tuple ) -> Optional[Any]:
lowerCamelCase : List[str] = []
# Download the dataset
lowerCamelCase : str = load_dataset("""glue""" , """mrpc""" )
# Create our splits
lowerCamelCase : List[str] = StratifiedKFold(n_splits=int(args.num_folds ) )
# Initialize accelerator
lowerCamelCase : int = Accelerator(cpu=args.cpu , mixed_precision=args.mixed_precision )
# Sample hyper-parameters for learning rate, batch size, seed and a few other HPs
lowerCamelCase : Tuple = config["""lr"""]
lowerCamelCase : Tuple = int(config["""num_epochs"""] )
lowerCamelCase : List[Any] = int(config["""seed"""] )
lowerCamelCase : List[Any] = int(config["""batch_size"""] )
lowerCamelCase : str = evaluate.load("""glue""" , """mrpc""" )
# If the batch size is too big we use gradient accumulation
lowerCamelCase : List[Any] = 1
if batch_size > MAX_GPU_BATCH_SIZE and accelerator.distributed_type != DistributedType.TPU:
lowerCamelCase : Optional[int] = batch_size // MAX_GPU_BATCH_SIZE
lowerCamelCase : Tuple = MAX_GPU_BATCH_SIZE
set_seed(UpperCamelCase__ )
# New Code #
# Create our folds:
lowerCamelCase : Any = kfold.split(np.zeros(datasets["""train"""].num_rows ) , datasets["""train"""]["""label"""] )
lowerCamelCase : Tuple = []
# Iterate over them
for i, (train_idxs, valid_idxs) in enumerate(UpperCamelCase__ ):
lowerCamelCase , lowerCamelCase , lowerCamelCase : List[str] = get_fold_dataloaders(
UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , )
# Instantiate the model (we build the model here so that the seed also control new weights initialization)
lowerCamelCase : int = AutoModelForSequenceClassification.from_pretrained("""bert-base-cased""" , return_dict=UpperCamelCase__ )
# We could avoid this line since the accelerator is set with `device_placement=True` (default value).
# Note that if you are placing tensors on devices manually, this line absolutely needs to be before the optimizer
# creation otherwise training will not work on TPU (`accelerate` will kindly throw an error to make us aware of that).
lowerCamelCase : Optional[int] = model.to(accelerator.device )
# Instantiate optimizer
lowerCamelCase : Tuple = AdamW(params=model.parameters() , lr=UpperCamelCase__ )
# Instantiate scheduler
lowerCamelCase : Any = get_linear_schedule_with_warmup(
optimizer=UpperCamelCase__ , num_warmup_steps=100 , num_training_steps=(len(UpperCamelCase__ ) * num_epochs) // gradient_accumulation_steps , )
# Prepare everything
# There is no specific order to remember, we just need to unpack the objects in the same order we gave them to the
# prepare method.
lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = accelerator.prepare(
UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
# Now we train the model
for epoch in range(UpperCamelCase__ ):
model.train()
for step, batch in enumerate(UpperCamelCase__ ):
# We could avoid this line since we set the accelerator with `device_placement=True`.
batch.to(accelerator.device )
lowerCamelCase : int = model(**UpperCamelCase__ )
lowerCamelCase : Any = outputs.loss
lowerCamelCase : Tuple = loss / gradient_accumulation_steps
accelerator.backward(UpperCamelCase__ )
if step % gradient_accumulation_steps == 0:
optimizer.step()
lr_scheduler.step()
optimizer.zero_grad()
model.eval()
for step, batch in enumerate(UpperCamelCase__ ):
# We could avoid this line since we set the accelerator with `device_placement=True`.
batch.to(accelerator.device )
with torch.no_grad():
lowerCamelCase : int = model(**UpperCamelCase__ )
lowerCamelCase : Union[str, Any] = outputs.logits.argmax(dim=-1 )
lowerCamelCase , lowerCamelCase : Optional[int] = accelerator.gather_for_metrics((predictions, batch["""labels"""]) )
metric.add_batch(
predictions=UpperCamelCase__ , references=UpperCamelCase__ , )
lowerCamelCase : Union[str, Any] = metric.compute()
# Use accelerator.print to print only on the main process.
accelerator.print(F'epoch {epoch}:' , UpperCamelCase__ )
# New Code #
# We also run predictions on the test set at the very end
lowerCamelCase : int = []
for step, batch in enumerate(UpperCamelCase__ ):
# We could avoid this line since we set the accelerator with `device_placement=True`.
batch.to(accelerator.device )
with torch.no_grad():
lowerCamelCase : List[Any] = model(**UpperCamelCase__ )
lowerCamelCase : Dict = outputs.logits
lowerCamelCase , lowerCamelCase : Tuple = accelerator.gather_for_metrics((predictions, batch["""labels"""]) )
fold_predictions.append(predictions.cpu() )
if i == 0:
# We need all of the test predictions
test_references.append(references.cpu() )
# Use accelerator.print to print only on the main process.
test_predictions.append(torch.cat(UpperCamelCase__ , dim=0 ) )
# We now need to release all our memory and get rid of the current model, optimizer, etc
accelerator.free_memory()
# New Code #
# Finally we check the accuracy of our folded results:
lowerCamelCase : int = torch.cat(UpperCamelCase__ , dim=0 )
lowerCamelCase : Optional[Any] = torch.stack(UpperCamelCase__ , dim=0 ).sum(dim=0 ).div(int(args.num_folds ) ).argmax(dim=-1 )
lowerCamelCase : int = metric.compute(predictions=UpperCamelCase__ , references=UpperCamelCase__ )
accelerator.print("""Average test metrics from all folds:""" , UpperCamelCase__ )
def lowerCAmelCase ( ) -> Any:
lowerCamelCase : List[Any] = argparse.ArgumentParser(description="""Simple example of training script.""" )
parser.add_argument(
"""--mixed_precision""" , type=UpperCamelCase__ , default=UpperCamelCase__ , choices=["""no""", """fp16""", """bf16""", """fp8"""] , help="""Whether to use mixed precision. Choose"""
"""between fp16 and bf16 (bfloat16). Bf16 requires PyTorch >= 1.10."""
"""and an Nvidia Ampere GPU.""" , )
parser.add_argument("""--cpu""" , action="""store_true""" , help="""If passed, will train on the CPU.""" )
# New Code #
parser.add_argument("""--num_folds""" , type=UpperCamelCase__ , default=3 , help="""The number of splits to perform across the dataset""" )
lowerCamelCase : List[Any] = parser.parse_args()
lowerCamelCase : Tuple = {"""lr""": 2E-5, """num_epochs""": 3, """seed""": 42, """batch_size""": 16}
training_function(UpperCamelCase__ , UpperCamelCase__ )
if __name__ == "__main__":
main()
| 708
|
"""simple docstring"""
from typing import List, Optional, Union
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :str = logging.get_logger(__name__)
__lowerCamelCase :Any = {
'huggingface/time-series-transformer-tourism-monthly': (
'https://huggingface.co/huggingface/time-series-transformer-tourism-monthly/resolve/main/config.json'
),
# See all TimeSeriesTransformer models at https://huggingface.co/models?filter=time_series_transformer
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : List[Any] ='''time_series_transformer'''
snake_case__ : List[Any] ={
'''hidden_size''': '''d_model''',
'''num_attention_heads''': '''encoder_attention_heads''',
'''num_hidden_layers''': '''encoder_layers''',
}
def __init__( self: List[str] , __a: Optional[int] = None , __a: Optional[int] = None , __a: str = "student_t" , __a: str = "nll" , __a: int = 1 , __a: List[int] = [1, 2, 3, 4, 5, 6, 7] , __a: Optional[Union[str, bool]] = "mean" , __a: int = 0 , __a: int = 0 , __a: int = 0 , __a: int = 0 , __a: Optional[List[int]] = None , __a: Optional[List[int]] = None , __a: int = 32 , __a: int = 32 , __a: int = 2 , __a: int = 2 , __a: int = 2 , __a: int = 2 , __a: bool = True , __a: str = "gelu" , __a: int = 64 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: int = 100 , __a: float = 0.02 , __a: Tuple=True , **__a: str , )-> Any:
# time series specific configuration
lowerCamelCase : str = prediction_length
lowerCamelCase : Optional[Any] = context_length or prediction_length
lowerCamelCase : Tuple = distribution_output
lowerCamelCase : Any = loss
lowerCamelCase : List[Any] = input_size
lowerCamelCase : int = num_time_features
lowerCamelCase : Dict = lags_sequence
lowerCamelCase : Optional[int] = scaling
lowerCamelCase : int = num_dynamic_real_features
lowerCamelCase : Tuple = num_static_real_features
lowerCamelCase : Any = num_static_categorical_features
if cardinality and num_static_categorical_features > 0:
if len(__a ) != num_static_categorical_features:
raise ValueError(
"""The cardinality should be a list of the same length as `num_static_categorical_features`""" )
lowerCamelCase : int = cardinality
else:
lowerCamelCase : Dict = [0]
if embedding_dimension and num_static_categorical_features > 0:
if len(__a ) != num_static_categorical_features:
raise ValueError(
"""The embedding dimension should be a list of the same length as `num_static_categorical_features`""" )
lowerCamelCase : str = embedding_dimension
else:
lowerCamelCase : str = [min(50 , (cat + 1) // 2 ) for cat in self.cardinality]
lowerCamelCase : Any = num_parallel_samples
# Transformer architecture configuration
lowerCamelCase : Any = input_size * len(__a ) + self._number_of_features
lowerCamelCase : List[str] = d_model
lowerCamelCase : Tuple = encoder_attention_heads
lowerCamelCase : Optional[int] = decoder_attention_heads
lowerCamelCase : Union[str, Any] = encoder_ffn_dim
lowerCamelCase : str = decoder_ffn_dim
lowerCamelCase : str = encoder_layers
lowerCamelCase : Any = decoder_layers
lowerCamelCase : Optional[int] = dropout
lowerCamelCase : List[str] = attention_dropout
lowerCamelCase : Tuple = activation_dropout
lowerCamelCase : Optional[int] = encoder_layerdrop
lowerCamelCase : int = decoder_layerdrop
lowerCamelCase : Optional[int] = activation_function
lowerCamelCase : Optional[Any] = init_std
lowerCamelCase : Optional[Any] = use_cache
super().__init__(is_encoder_decoder=__a , **__a )
@property
def a__ ( self: int )-> int:
return (
sum(self.embedding_dimension )
+ self.num_dynamic_real_features
+ self.num_time_features
+ self.num_static_real_features
+ self.input_size * 2 # the log1p(abs(loc)) and log(scale) features
)
| 42
| 0
|
"""simple docstring"""
from scipy.stats import spearmanr
import datasets
__lowerCamelCase :Any = '\nThe Spearman rank-order correlation coefficient is a measure of the\nrelationship between two datasets. Like other correlation coefficients,\nthis one varies between -1 and +1 with 0 implying no correlation.\nPositive correlations imply that as data in dataset x increases, so\ndoes data in dataset y. Negative correlations imply that as x increases,\ny decreases. Correlations of -1 or +1 imply an exact monotonic relationship.\n\nUnlike the Pearson correlation, the Spearman correlation does not\nassume that both datasets are normally distributed.\n\nThe p-value roughly indicates the probability of an uncorrelated system\nproducing datasets that have a Spearman correlation at least as extreme\nas the one computed from these datasets. The p-values are not entirely\nreliable but are probably reasonable for datasets larger than 500 or so.\n'
__lowerCamelCase :Dict = '\nArgs:\n predictions (`List[float]`): Predicted labels, as returned by a model.\n references (`List[float]`): Ground truth labels.\n return_pvalue (`bool`): If `True`, returns the p-value. If `False`, returns\n only the spearmanr score. Defaults to `False`.\nReturns:\n spearmanr (`float`): Spearman correlation coefficient.\n p-value (`float`): p-value. **Note**: is only returned if `return_pvalue=True` is input.\nExamples:\n Example 1:\n >>> spearmanr_metric = datasets.load_metric(\"spearmanr\")\n >>> results = spearmanr_metric.compute(references=[1, 2, 3, 4, 5], predictions=[10, 9, 2.5, 6, 4])\n >>> print(results)\n {\'spearmanr\': -0.7}\n\n Example 2:\n >>> spearmanr_metric = datasets.load_metric(\"spearmanr\")\n >>> results = spearmanr_metric.compute(references=[1, 2, 3, 4, 5],\n ... predictions=[10, 9, 2.5, 6, 4],\n ... return_pvalue=True)\n >>> print(results[\'spearmanr\'])\n -0.7\n >>> print(round(results[\'spearmanr_pvalue\'], 2))\n 0.19\n'
__lowerCamelCase :Any = r'\\n@book{kokoska2000crc,\n title={CRC standard probability and statistics tables and formulae},\n author={Kokoska, Stephen and Zwillinger, Daniel},\n year={2000},\n publisher={Crc Press}\n}\n@article{2020SciPy-NMeth,\n author = {Virtanen, Pauli and Gommers, Ralf and Oliphant, Travis E. and\n Haberland, Matt and Reddy, Tyler and Cournapeau, David and\n Burovski, Evgeni and Peterson, Pearu and Weckesser, Warren and\n Bright, Jonathan and {van der Walt}, St{\'e}fan J. and\n Brett, Matthew and Wilson, Joshua and Millman, K. Jarrod and\n Mayorov, Nikolay and Nelson, Andrew R. J. and Jones, Eric and\n Kern, Robert and Larson, Eric and Carey, C J and\n Polat, {\.I}lhan and Feng, Yu and Moore, Eric W. and\n {VanderPlas}, Jake and Laxalde, Denis and Perktold, Josef and\n Cimrman, Robert and Henriksen, Ian and Quintero, E. A. and\n Harris, Charles R. and Archibald, Anne M. and\n Ribeiro, Ant{\^o}nio H. and Pedregosa, Fabian and\n {van Mulbregt}, Paul and {SciPy 1.0 Contributors}},\n title = {{{SciPy} 1.0: Fundamental Algorithms for Scientific\n Computing in Python}},\n journal = {Nature Methods},\n year = {2020},\n volume = {17},\n pages = {261--272},\n adsurl = {https://rdcu.be/b08Wh},\n doi = {10.1038/s41592-019-0686-2},\n}\n'
@datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION , _KWARGS_DESCRIPTION)
class A__ ( datasets.Metric):
"""simple docstring"""
def a__ ( self: List[Any] )-> str:
return datasets.MetricInfo(
description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features(
{
"""predictions""": datasets.Value("""float""" ),
"""references""": datasets.Value("""float""" ),
} ) , reference_urls=["""https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.spearmanr.html"""] , )
def a__ ( self: Dict , __a: Any , __a: Tuple , __a: Union[str, Any]=False )-> Dict:
lowerCamelCase : int = spearmanr(_a , _a )
if return_pvalue:
return {"spearmanr": results[0], "spearmanr_pvalue": results[1]}
else:
return {"spearmanr": results[0]}
| 709
|
"""simple docstring"""
from __future__ import annotations
__lowerCamelCase :int = 10
def snake_case ( UpperCamelCase__ : list[int] ) -> list[int]:
lowerCamelCase : int = 1
lowerCamelCase : Union[str, Any] = max(UpperCamelCase__ )
while placement <= max_digit:
# declare and initialize empty buckets
lowerCamelCase : list[list] = [[] for _ in range(UpperCamelCase__ )]
# split list_of_ints between the buckets
for i in list_of_ints:
lowerCamelCase : Any = int((i / placement) % RADIX )
buckets[tmp].append(UpperCamelCase__ )
# put each buckets' contents into list_of_ints
lowerCamelCase : Dict = 0
for b in range(UpperCamelCase__ ):
for i in buckets[b]:
lowerCamelCase : List[str] = i
a += 1
# move to next
placement *= RADIX
return list_of_ints
if __name__ == "__main__":
import doctest
doctest.testmod()
| 42
| 0
|
"""simple docstring"""
import re
from flax.core.frozen_dict import freeze
from flax.traverse_util import flatten_dict, unflatten_dict
from jax.experimental import PartitionSpec as P
# Sentinels
__lowerCamelCase :Union[str, Any] = object()
# For specifying empty leaf dict `{}`
__lowerCamelCase :Optional[Any] = object()
def snake_case ( UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : int ) -> List[str]:
lowerCamelCase : Optional[int] = tuple((re.compile(x + """$""" ) for x in qs) )
for i in range(len(__UpperCamelCase ) - len(__UpperCamelCase ) + 1 ):
lowerCamelCase : Tuple = [x.match(__UpperCamelCase ) for x, y in zip(__UpperCamelCase , ks[i:] )]
if matches and all(__UpperCamelCase ):
return True
return False
def snake_case ( UpperCamelCase__ : Optional[int] ) -> Optional[int]:
def replace(UpperCamelCase__ : Any , UpperCamelCase__ : Tuple ):
for rule, replacement in rules:
if _match(__UpperCamelCase , __UpperCamelCase ):
return replacement
return val
return replace
def snake_case ( ) -> Dict:
return [
# embeddings
(("transformer", "wpe", "embedding"), P("""mp""" , __UpperCamelCase )),
(("transformer", "wte", "embedding"), P("""mp""" , __UpperCamelCase )),
# atention
(("attention", "(q_proj|k_proj|v_proj)", "kernel"), P(__UpperCamelCase , """mp""" )),
(("attention", "out_proj", "kernel"), P("""mp""" , __UpperCamelCase )),
(("attention", "out_proj", "bias"), None),
# mlp
(("mlp", "c_fc", "kernel"), P(__UpperCamelCase , """mp""" )),
(("mlp", "c_fc", "bias"), P("""mp""" )),
(("mlp", "c_proj", "kernel"), P("""mp""" , __UpperCamelCase )),
(("mlp", "c_proj", "bias"), None),
# layer norms
((r"ln_\d+", "bias"), None),
((r"\d+", r"ln_\d+", "scale"), None),
(("ln_f", "bias"), None),
(("ln_f", "scale"), None),
]
def snake_case ( UpperCamelCase__ : List[str] ) -> Any:
lowerCamelCase : Optional[Any] = _get_partition_rules()
lowerCamelCase : Dict = _replacement_rules(__UpperCamelCase )
lowerCamelCase : Optional[int] = {k: _unmatched for k in flatten_dict(__UpperCamelCase )}
lowerCamelCase : Tuple = {k: replace(__UpperCamelCase , __UpperCamelCase ) for k, v in initd.items()}
assert _unmatched not in result.values(), "Incomplete partition spec."
return freeze(unflatten_dict(__UpperCamelCase ) )
| 710
|
"""simple docstring"""
import argparse
import pickle
import numpy as np
import torch
from torch import nn
from transformers import ReformerConfig, ReformerModelWithLMHead
from transformers.utils import logging
logging.set_verbosity_info()
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Tuple , UpperCamelCase__ : Optional[Any]=None ) -> Tuple:
# set parameter of one layer
assert torch_layer.weight.shape == weight.shape, F'{torch_layer} layer.weight does not match'
lowerCamelCase : Dict = nn.Parameter(UpperCamelCase__ )
if bias is not None:
assert torch_layer.bias.shape == bias.shape, F'{torch_layer} layer.bias does not match'
lowerCamelCase : Any = nn.Parameter(UpperCamelCase__ )
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : Any , UpperCamelCase__ : Tuple ) -> Union[str, Any]:
# set torch weights for 1-to-1 comparison
lowerCamelCase : Dict = np.asarray(weights[0] )
lowerCamelCase : List[Any] = np.asarray(weights[1] )
lowerCamelCase : List[str] = np.asarray(weights[2] )
set_param(
torch_layer.self_attention.query_key , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.value , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.output.dense , torch.tensor(UpperCamelCase__ ).view(-1 , UpperCamelCase__ ).contiguous().transpose(0 , 1 ) , )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Any ) -> List[Any]:
# set torch weights for 1-to-1 comparison
lowerCamelCase : Tuple = np.asarray(weights[0] )
lowerCamelCase : Any = np.asarray(weights[1] )
lowerCamelCase : List[Any] = np.asarray(weights[2] )
lowerCamelCase : List[str] = np.asarray(weights[3] )
set_param(
torch_layer.self_attention.query , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.key , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.value , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.output.dense , torch.tensor(UpperCamelCase__ ).view(-1 , UpperCamelCase__ ).contiguous().transpose(0 , 1 ) , )
def snake_case ( UpperCamelCase__ : Dict , UpperCamelCase__ : Optional[int] , UpperCamelCase__ : Any ) -> Optional[Any]:
# layernorm 1
lowerCamelCase : str = weights[0][0][0]
lowerCamelCase : Optional[int] = np.asarray(layer_norm_a[0] )
lowerCamelCase : Tuple = np.asarray(layer_norm_a[1] )
set_param(
torch_block.attention.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# lsh weights + output
lowerCamelCase : List[Any] = weights[0][1]
if len(UpperCamelCase__ ) < 4:
set_layer_weights_in_torch_lsh(UpperCamelCase__ , torch_block.attention , UpperCamelCase__ )
else:
set_layer_weights_in_torch_local(UpperCamelCase__ , torch_block.attention , UpperCamelCase__ )
# intermediate weighs
lowerCamelCase : int = weights[2][0][1][2]
# Chunked Feed Forward
if len(UpperCamelCase__ ) == 4:
lowerCamelCase : Dict = intermediate_weights[2]
# layernorm 2
lowerCamelCase : Optional[int] = np.asarray(intermediate_weights[0][0] )
lowerCamelCase : Tuple = np.asarray(intermediate_weights[0][1] )
set_param(
torch_block.feed_forward.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# intermediate dense
lowerCamelCase : Optional[Any] = np.asarray(intermediate_weights[1][0] )
lowerCamelCase : Union[str, Any] = np.asarray(intermediate_weights[1][1] )
set_param(
torch_block.feed_forward.dense.dense , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
# intermediate out
lowerCamelCase : Optional[int] = np.asarray(intermediate_weights[4][0] )
lowerCamelCase : List[Any] = np.asarray(intermediate_weights[4][1] )
set_param(
torch_block.feed_forward.output.dense , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : List[str] , UpperCamelCase__ : int ) -> List[Any]:
# reformer model
lowerCamelCase : List[Any] = torch_model.reformer
# word embeds
lowerCamelCase : Union[str, Any] = np.asarray(weights[1] )
set_param(
torch_model_reformer.embeddings.word_embeddings , torch.tensor(UpperCamelCase__ ) , )
if isinstance(weights[3] , UpperCamelCase__ ):
lowerCamelCase : Union[str, Any] = torch_model_reformer.embeddings.position_embeddings
for emb_idx in range(len(position_embeddings.weights ) ):
lowerCamelCase : str = np.asarray(weights[3][emb_idx][0] )
assert (
position_embeddings.weights[emb_idx].shape == emb_weights.shape
), F'{position_embeddings[emb_idx]} emb does not match'
lowerCamelCase : Dict = nn.Parameter(torch.tensor(UpperCamelCase__ ) )
lowerCamelCase : int = weights[5]
assert len(torch_model_reformer.encoder.layers ) * 4 == len(
UpperCamelCase__ ), "HF and trax model do not have the same number of layers"
for layer_idx, layer in enumerate(torch_model_reformer.encoder.layers ):
lowerCamelCase : Dict = trax_layer_weights[4 * layer_idx : 4 * (layer_idx + 1)]
set_block_weights_in_torch(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
# output layer norm
lowerCamelCase : Any = np.asarray(weights[7][0] )
lowerCamelCase : List[str] = np.asarray(weights[7][1] )
set_param(
torch_model_reformer.encoder.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# output embeddings
lowerCamelCase : List[Any] = np.asarray(weights[9][0] )
lowerCamelCase : Optional[int] = np.asarray(weights[9][1] )
set_param(
torch_model.lm_head.decoder , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
def snake_case ( UpperCamelCase__ : str , UpperCamelCase__ : int , UpperCamelCase__ : str ) -> Optional[int]:
# Initialise PyTorch model
lowerCamelCase : Union[str, Any] = ReformerConfig.from_json_file(UpperCamelCase__ )
print(F'Building PyTorch model from configuration: {config}' )
lowerCamelCase : str = ReformerModelWithLMHead(UpperCamelCase__ )
with open(UpperCamelCase__ , """rb""" ) as f:
lowerCamelCase : str = pickle.load(UpperCamelCase__ )["""weights"""]
set_model_weights_in_torch(UpperCamelCase__ , UpperCamelCase__ , config.hidden_size )
# Save pytorch-model
print(F'Save PyTorch model to {pytorch_dump_path}' )
torch.save(model.state_dict() , UpperCamelCase__ )
if __name__ == "__main__":
__lowerCamelCase :Dict = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--trax_model_pkl_path', default=None, type=str, required=True, help='Path to the TensorFlow checkpoint path.'
)
parser.add_argument(
'--config_file',
default=None,
type=str,
required=True,
help=(
'The config json file corresponding to the pre-trained Reformer model. \n'
'This specifies the model architecture.'
),
)
parser.add_argument(
'--pytorch_dump_path', default=None, type=str, required=True, help='Path to the output PyTorch model.'
)
__lowerCamelCase :Optional[int] = parser.parse_args()
convert_trax_checkpoint_to_pytorch(args.trax_model_pkl_path, args.config_file, args.pytorch_dump_path)
| 42
| 0
|
"""simple docstring"""
import json
import os
import tempfile
import unittest
import numpy as np
from datasets import load_dataset
from transformers.testing_utils import require_torch, require_vision, slow
from transformers.utils import is_torch_available, is_vision_available
from ...test_image_processing_common import ImageProcessingSavingTestMixin
if is_torch_available():
import torch
if is_vision_available():
from PIL import Image
from transformers import ImageGPTImageProcessor
class A__ ( unittest.TestCase):
"""simple docstring"""
def __init__( self: Dict , __a: Optional[Any] , __a: Any=7 , __a: List[Any]=3 , __a: str=18 , __a: List[str]=30 , __a: Tuple=400 , __a: str=True , __a: Any=None , __a: Optional[int]=True , )-> Dict:
lowerCamelCase : Union[str, Any] = size if size is not None else {"""height""": 18, """width""": 18}
lowerCamelCase : List[Any] = parent
lowerCamelCase : Optional[int] = batch_size
lowerCamelCase : Union[str, Any] = num_channels
lowerCamelCase : List[Any] = image_size
lowerCamelCase : Optional[Any] = min_resolution
lowerCamelCase : str = max_resolution
lowerCamelCase : int = do_resize
lowerCamelCase : Union[str, Any] = size
lowerCamelCase : Any = do_normalize
def a__ ( self: Dict )-> Dict:
return {
# here we create 2 clusters for the sake of simplicity
"clusters": np.asarray(
[
[0.88_66_44_36_34_03_32_03, 0.66_18_82_93_69_54_49_83, 0.38_91_74_64_01_78_68_04],
[-0.60_42_55_91_46_88_11_04, -0.0_22_95_00_88_60_52_84_69, 0.54_23_79_73_69_00_32_96],
] ),
"do_resize": self.do_resize,
"size": self.size,
"do_normalize": self.do_normalize,
}
@require_torch
@require_vision
class A__ ( lowerCAmelCase__ , unittest.TestCase):
"""simple docstring"""
snake_case__ : Any =ImageGPTImageProcessor if is_vision_available() else None
def a__ ( self: List[str] )-> str:
lowerCamelCase : Union[str, Any] = ImageGPTImageProcessingTester(self )
@property
def a__ ( self: Optional[int] )-> List[str]:
return self.image_processor_tester.prepare_image_processor_dict()
def a__ ( self: Optional[int] )-> int:
lowerCamelCase : Any = self.image_processing_class(**self.image_processor_dict )
self.assertTrue(hasattr(_lowerCamelCase , """clusters""" ) )
self.assertTrue(hasattr(_lowerCamelCase , """do_resize""" ) )
self.assertTrue(hasattr(_lowerCamelCase , """size""" ) )
self.assertTrue(hasattr(_lowerCamelCase , """do_normalize""" ) )
def a__ ( self: Optional[Any] )-> Optional[int]:
lowerCamelCase : Union[str, Any] = self.image_processing_class.from_dict(self.image_processor_dict )
self.assertEqual(image_processor.size , {"""height""": 18, """width""": 18} )
lowerCamelCase : Union[str, Any] = self.image_processing_class.from_dict(self.image_processor_dict , size=42 )
self.assertEqual(image_processor.size , {"""height""": 42, """width""": 42} )
def a__ ( self: List[Any] )-> int:
lowerCamelCase : Dict = self.image_processing_class(**self.image_processor_dict )
lowerCamelCase : Tuple = json.loads(image_processor.to_json_string() )
for key, value in self.image_processor_dict.items():
if key == "clusters":
self.assertTrue(np.array_equal(_lowerCamelCase , obj[key] ) )
else:
self.assertEqual(obj[key] , _lowerCamelCase )
def a__ ( self: Dict )-> List[str]:
lowerCamelCase : Union[str, Any] = self.image_processing_class(**self.image_processor_dict )
with tempfile.TemporaryDirectory() as tmpdirname:
lowerCamelCase : List[str] = os.path.join(_lowerCamelCase , """image_processor.json""" )
image_processor_first.to_json_file(_lowerCamelCase )
lowerCamelCase : Tuple = self.image_processing_class.from_json_file(_lowerCamelCase ).to_dict()
lowerCamelCase : str = image_processor_first.to_dict()
for key, value in image_processor_first.items():
if key == "clusters":
self.assertTrue(np.array_equal(_lowerCamelCase , image_processor_second[key] ) )
else:
self.assertEqual(image_processor_first[key] , _lowerCamelCase )
def a__ ( self: Tuple )-> Optional[Any]:
lowerCamelCase : List[str] = self.image_processing_class(**self.image_processor_dict )
with tempfile.TemporaryDirectory() as tmpdirname:
image_processor_first.save_pretrained(_lowerCamelCase )
lowerCamelCase : List[Any] = self.image_processing_class.from_pretrained(_lowerCamelCase ).to_dict()
lowerCamelCase : Union[str, Any] = image_processor_first.to_dict()
for key, value in image_processor_first.items():
if key == "clusters":
self.assertTrue(np.array_equal(_lowerCamelCase , image_processor_second[key] ) )
else:
self.assertEqual(image_processor_first[key] , _lowerCamelCase )
@unittest.skip("""ImageGPT requires clusters at initialization""" )
def a__ ( self: Optional[int] )-> List[str]:
pass
def snake_case ( ) -> Optional[int]:
lowerCamelCase : Optional[int] = load_dataset("""hf-internal-testing/fixtures_image_utils""" , split="""test""" )
lowerCamelCase : Any = Image.open(dataset[4]["""file"""] )
lowerCamelCase : Optional[int] = Image.open(dataset[5]["""file"""] )
lowerCamelCase : List[str] = [imagea, imagea]
return images
@require_vision
@require_torch
class A__ ( unittest.TestCase):
"""simple docstring"""
@slow
def a__ ( self: Optional[int] )-> List[Any]:
lowerCamelCase : Optional[Any] = ImageGPTImageProcessor.from_pretrained("""openai/imagegpt-small""" )
lowerCamelCase : Tuple = prepare_images()
# test non-batched
lowerCamelCase : Any = image_processing(images[0] , return_tensors="""pt""" )
self.assertIsInstance(encoding.input_ids , torch.LongTensor )
self.assertEqual(encoding.input_ids.shape , (1, 1_024) )
lowerCamelCase : Any = [306, 191, 191]
self.assertEqual(encoding.input_ids[0, :3].tolist() , _lowerCamelCase )
# test batched
lowerCamelCase : Union[str, Any] = image_processing(_lowerCamelCase , return_tensors="""pt""" )
self.assertIsInstance(encoding.input_ids , torch.LongTensor )
self.assertEqual(encoding.input_ids.shape , (2, 1_024) )
lowerCamelCase : Tuple = [303, 13, 13]
self.assertEqual(encoding.input_ids[1, -3:].tolist() , _lowerCamelCase )
| 711
|
"""simple docstring"""
import inspect
import unittest
import torch
import torch.nn as nn
from accelerate.hooks import (
AlignDevicesHook,
ModelHook,
SequentialHook,
add_hook_to_module,
attach_align_device_hook,
remove_hook_from_module,
remove_hook_from_submodules,
)
from accelerate.test_utils import require_multi_gpu
class A__ ( nn.Module):
"""simple docstring"""
def __init__( self: Dict )-> Dict:
super().__init__()
lowerCamelCase : Tuple = nn.Linear(3 , 4 )
lowerCamelCase : Optional[Any] = nn.BatchNormad(4 )
lowerCamelCase : Optional[Any] = nn.Linear(4 , 5 )
def a__ ( self: List[str] , __a: List[Any] )-> Optional[Any]:
return self.lineara(self.batchnorm(self.lineara(__a ) ) )
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Tuple , __a: int , *__a: Any , **__a: Tuple )-> Tuple:
return (args[0] + 1,) + args[1:], kwargs
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Optional[int] , __a: List[str] , __a: List[Any] )-> List[str]:
return output + 1
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: int )-> str:
lowerCamelCase : List[str] = ModelForTest()
lowerCamelCase : Dict = ModelHook()
add_hook_to_module(__a , __a )
self.assertEqual(test_model._hf_hook , __a )
self.assertTrue(hasattr(__a , """_old_forward""" ) )
# Check adding the hook did not change the name or the signature
self.assertEqual(test_model.forward.__name__ , """forward""" )
self.assertListEqual(list(inspect.signature(test_model.forward ).parameters ) , ["""x"""] )
remove_hook_from_module(__a )
self.assertFalse(hasattr(__a , """_hf_hook""" ) )
self.assertFalse(hasattr(__a , """_old_forward""" ) )
def a__ ( self: int )-> str:
lowerCamelCase : List[str] = ModelForTest()
lowerCamelCase : Union[str, Any] = ModelHook()
add_hook_to_module(__a , __a )
add_hook_to_module(__a , __a , append=__a )
self.assertEqual(isinstance(test_model._hf_hook , __a ) , __a )
self.assertEqual(len(test_model._hf_hook.hooks ) , 2 )
self.assertTrue(hasattr(__a , """_old_forward""" ) )
# Check adding the hook did not change the name or the signature
self.assertEqual(test_model.forward.__name__ , """forward""" )
self.assertListEqual(list(inspect.signature(test_model.forward ).parameters ) , ["""x"""] )
remove_hook_from_module(__a )
self.assertFalse(hasattr(__a , """_hf_hook""" ) )
self.assertFalse(hasattr(__a , """_old_forward""" ) )
def a__ ( self: List[Any] )-> List[str]:
lowerCamelCase : str = ModelForTest()
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : Union[str, Any] = test_model(x + 1 )
lowerCamelCase : Optional[int] = test_model(x + 2 )
lowerCamelCase : List[Any] = PreForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[int] = test_model(__a )
self.assertTrue(torch.allclose(__a , __a , atol=1e-5 ) )
# Attaching a hook to a model when it already has one replaces, does not chain
lowerCamelCase : Dict = PreForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Tuple = test_model(__a )
self.assertTrue(torch.allclose(__a , __a , atol=1e-5 ) )
# You need to use the sequential hook to chain two or more hooks
lowerCamelCase : Any = SequentialHook(PreForwardHook() , PreForwardHook() )
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[Any] = test_model(__a )
assert torch.allclose(__a , __a , atol=1e-5 )
def a__ ( self: Any )-> Optional[int]:
lowerCamelCase : str = ModelForTest()
lowerCamelCase : List[str] = torch.randn(2 , 3 )
lowerCamelCase : int = test_model(__a )
lowerCamelCase : Dict = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Tuple = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 , atol=1e-5 ) )
# Attaching a hook to a model when it already has one replaces, does not chain
lowerCamelCase : str = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[Any] = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 , atol=1e-5 ) )
# You need to use the sequential hook to chain two or more hooks
lowerCamelCase : Union[str, Any] = SequentialHook(PostForwardHook() , PostForwardHook() )
add_hook_to_module(__a , __a )
lowerCamelCase : str = test_model(__a )
assert torch.allclose(__a , output + 2 , atol=1e-5 )
def a__ ( self: int )-> Dict:
lowerCamelCase : List[Any] = ModelForTest()
lowerCamelCase : Optional[int] = torch.randn(2 , 3 )
lowerCamelCase : List[str] = test_model(__a )
lowerCamelCase : Any = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : str = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 ) )
self.assertTrue(outputa.requires_grad )
lowerCamelCase : Optional[int] = True
lowerCamelCase : Optional[int] = test_model(__a )
self.assertFalse(outputa.requires_grad )
@require_multi_gpu
def a__ ( self: List[str] )-> Union[str, Any]:
lowerCamelCase : int = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
add_hook_to_module(model.lineara , AlignDevicesHook(execution_device=0 ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(execution_device=0 ) )
add_hook_to_module(model.lineara , AlignDevicesHook(execution_device=1 ) )
self.assertEqual(model.lineara.weight.device , torch.device(0 ) )
self.assertEqual(model.batchnorm.weight.device , torch.device(0 ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device(0 ) )
self.assertEqual(model.lineara.weight.device , torch.device(1 ) )
# We can still make a forward pass. The input does not need to be on any particular device
lowerCamelCase : str = torch.randn(2 , 3 )
lowerCamelCase : Dict = model(__a )
self.assertEqual(output.device , torch.device(1 ) )
# We can add a general hook to put back output on same device as input.
add_hook_to_module(__a , AlignDevicesHook(io_same_device=__a ) )
lowerCamelCase : Optional[int] = torch.randn(2 , 3 ).to(0 )
lowerCamelCase : str = model(__a )
self.assertEqual(output.device , torch.device(0 ) )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Union[str, Any] = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : Tuple = {"""execution_device""": 0 if torch.cuda.is_available() else """cpu""", """offload""": True}
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(**__a ) )
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(hook_kwargs["""execution_device"""] )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Optional[Any] = torch.randn(2 , 3 )
lowerCamelCase : Optional[Any] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_module(model.lineara )
remove_hook_from_module(model.batchnorm )
remove_hook_from_module(model.lineara )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
lowerCamelCase : Any = {
"""execution_device""": 0 if torch.cuda.is_available() else """cpu""",
"""offload""": True,
"""offload_buffers""": True,
}
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(**__a ) )
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : int = torch.randn(2 , 3 )
lowerCamelCase : Optional[int] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_module(model.lineara )
remove_hook_from_module(model.batchnorm )
remove_hook_from_module(model.lineara )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
def a__ ( self: Any )-> List[str]:
lowerCamelCase : int = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : int = 0 if torch.cuda.is_available() else """cpu"""
attach_align_device_hook(__a , execution_device=__a , offload=__a )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(__a )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : Optional[Any] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
attach_align_device_hook(__a , execution_device=__a , offload=__a , offload_buffers=__a )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : Optional[int] = torch.randn(2 , 3 )
lowerCamelCase : int = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
def a__ ( self: Optional[Any] )-> List[Any]:
lowerCamelCase : List[Any] = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : Any = 0 if torch.cuda.is_available() else """cpu"""
attach_align_device_hook(
__a , execution_device=__a , offload=__a , weights_map=model.state_dict() )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(__a )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : int = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
attach_align_device_hook(
__a , execution_device=__a , offload=__a , weights_map=model.state_dict() , offload_buffers=__a , )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : Tuple = torch.randn(2 , 3 )
lowerCamelCase : Any = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
| 42
| 0
|
"""simple docstring"""
import numpy
# List of input, output pairs
__lowerCamelCase :int = (
((5, 2, 3), 15),
((6, 5, 9), 25),
((11, 12, 13), 41),
((1, 1, 1), 8),
((11, 12, 13), 41),
)
__lowerCamelCase :int = (((515, 22, 13), 555), ((61, 35, 49), 150))
__lowerCamelCase :List[Any] = [2, 4, 1, 5]
__lowerCamelCase :Optional[Any] = len(train_data)
__lowerCamelCase :Tuple = 0.009
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : int="train" ) -> Any:
return calculate_hypothesis_value(__snake_case , __snake_case ) - output(
__snake_case , __snake_case )
def snake_case ( UpperCamelCase__ : int ) -> Tuple:
lowerCamelCase : Union[str, Any] = 0
for i in range(len(__snake_case ) - 1 ):
hyp_val += data_input_tuple[i] * parameter_vector[i + 1]
hyp_val += parameter_vector[0]
return hyp_val
def snake_case ( UpperCamelCase__ : Any , UpperCamelCase__ : Tuple ) -> str:
if data_set == "train":
return train_data[example_no][1]
elif data_set == "test":
return test_data[example_no][1]
return None
def snake_case ( UpperCamelCase__ : str , UpperCamelCase__ : List[str] ) -> Union[str, Any]:
if data_set == "train":
return _hypothesis_value(train_data[example_no][0] )
elif data_set == "test":
return _hypothesis_value(test_data[example_no][0] )
return None
def snake_case ( UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : Union[str, Any]=m ) -> Tuple:
lowerCamelCase : Union[str, Any] = 0
for i in range(__snake_case ):
if index == -1:
summation_value += _error(__snake_case )
else:
summation_value += _error(__snake_case ) * train_data[i][0][index]
return summation_value
def snake_case ( UpperCamelCase__ : List[str] ) -> Any:
lowerCamelCase : str = summation_of_cost_derivative(__snake_case , __snake_case ) / m
return cost_derivative_value
def snake_case ( ) -> Union[str, Any]:
global parameter_vector
# Tune these values to set a tolerance value for predicted output
lowerCamelCase : Dict = 0.0_0_0_0_0_2
lowerCamelCase : List[str] = 0
lowerCamelCase : Tuple = 0
while True:
j += 1
lowerCamelCase : str = [0, 0, 0, 0]
for i in range(0 , len(__snake_case ) ):
lowerCamelCase : Any = get_cost_derivative(i - 1 )
lowerCamelCase : str = (
parameter_vector[i] - LEARNING_RATE * cost_derivative
)
if numpy.allclose(
__snake_case , __snake_case , atol=__snake_case , rtol=__snake_case , ):
break
lowerCamelCase : List[str] = temp_parameter_vector
print(("""Number of iterations:""", j) )
def snake_case ( ) -> Union[str, Any]:
for i in range(len(__snake_case ) ):
print(("""Actual output value:""", output(__snake_case , """test""" )) )
print(("""Hypothesis output:""", calculate_hypothesis_value(__snake_case , """test""" )) )
if __name__ == "__main__":
run_gradient_descent()
print('\nTesting gradient descent for a linear hypothesis function.\n')
test_gradient_descent()
| 712
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_torch_available,
)
__lowerCamelCase :Optional[Any] = {
'configuration_encodec': [
'ENCODEC_PRETRAINED_CONFIG_ARCHIVE_MAP',
'EncodecConfig',
],
'feature_extraction_encodec': ['EncodecFeatureExtractor'],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Union[str, Any] = [
'ENCODEC_PRETRAINED_MODEL_ARCHIVE_LIST',
'EncodecModel',
'EncodecPreTrainedModel',
]
if TYPE_CHECKING:
from .configuration_encodec import (
ENCODEC_PRETRAINED_CONFIG_ARCHIVE_MAP,
EncodecConfig,
)
from .feature_extraction_encodec import EncodecFeatureExtractor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_encodec import (
ENCODEC_PRETRAINED_MODEL_ARCHIVE_LIST,
EncodecModel,
EncodecPreTrainedModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
from typing import List
import jiwer
import jiwer.transforms as tr
from packaging import version
import datasets
from datasets.config import PY_VERSION
if PY_VERSION < version.parse('3.8'):
import importlib_metadata
else:
import importlib.metadata as importlib_metadata
__lowerCamelCase :Tuple = ''
if version.parse(importlib_metadata.version('jiwer')) < version.parse('2.3.0'):
class A__ ( tr.AbstractTransform):
"""simple docstring"""
def __init__( self: str , __a: Dict = " " )-> str:
lowerCamelCase : str = sentence_delimiter
def a__ ( self: Optional[int] , __a: str )-> Optional[Any]:
return list(__A )
def a__ ( self: List[Any] , __a: Tuple )-> Any:
lowerCamelCase : Any = []
for sent_idx, sentence in enumerate(__A ):
chars.extend(self.process_string(__A ) )
if self.sentence_delimiter is not None and self.sentence_delimiter != "" and sent_idx < len(__A ) - 1:
chars.append(self.sentence_delimiter )
return chars
__lowerCamelCase :List[str] = tr.Compose(
[tr.RemoveMultipleSpaces(), tr.Strip(), SentencesToListOfCharacters(SENTENCE_DELIMITER)]
)
else:
__lowerCamelCase :Union[str, Any] = tr.Compose(
[
tr.RemoveMultipleSpaces(),
tr.Strip(),
tr.ReduceToSingleSentence(SENTENCE_DELIMITER),
tr.ReduceToListOfListOfChars(),
]
)
__lowerCamelCase :int = '\\n@inproceedings{inproceedings,\n author = {Morris, Andrew and Maier, Viktoria and Green, Phil},\n year = {2004},\n month = {01},\n pages = {},\n title = {From WER and RIL to MER and WIL: improved evaluation measures for connected speech recognition.}\n}\n'
__lowerCamelCase :Union[str, Any] = '\\nCharacter error rate (CER) is a common metric of the performance of an automatic speech recognition system.\n\nCER is similar to Word Error Rate (WER), but operates on character instead of word. Please refer to docs of WER for further information.\n\nCharacter error rate can be computed as:\n\nCER = (S + D + I) / N = (S + D + I) / (S + D + C)\n\nwhere\n\nS is the number of substitutions,\nD is the number of deletions,\nI is the number of insertions,\nC is the number of correct characters,\nN is the number of characters in the reference (N=S+D+C).\n\nCER\'s output is not always a number between 0 and 1, in particular when there is a high number of insertions. This value is often associated to the percentage of characters that were incorrectly predicted. The lower the value, the better the\nperformance of the ASR system with a CER of 0 being a perfect score.\n'
__lowerCamelCase :Optional[Any] = '\nComputes CER score of transcribed segments against references.\nArgs:\n references: list of references for each speech input.\n predictions: list of transcribtions to score.\n concatenate_texts: Whether or not to concatenate sentences before evaluation, set to True for more accurate result.\nReturns:\n (float): the character error rate\n\nExamples:\n\n >>> predictions = ["this is the prediction", "there is an other sample"]\n >>> references = ["this is the reference", "there is another one"]\n >>> cer = datasets.load_metric("cer")\n >>> cer_score = cer.compute(predictions=predictions, references=references)\n >>> print(cer_score)\n 0.34146341463414637\n'
@datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION , _KWARGS_DESCRIPTION)
class A__ ( datasets.Metric):
"""simple docstring"""
def a__ ( self: Tuple )-> Optional[int]:
return datasets.MetricInfo(
description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features(
{
"""predictions""": datasets.Value("""string""" , id="""sequence""" ),
"""references""": datasets.Value("""string""" , id="""sequence""" ),
} ) , codebase_urls=["""https://github.com/jitsi/jiwer/"""] , reference_urls=[
"""https://en.wikipedia.org/wiki/Word_error_rate""",
"""https://sites.google.com/site/textdigitisation/qualitymeasures/computingerrorrates""",
] , )
def a__ ( self: Optional[Any] , __a: Tuple , __a: List[str] , __a: List[str]=False )-> str:
if concatenate_texts:
return jiwer.compute_measures(
__A , __A , truth_transform=__A , hypothesis_transform=__A , )["wer"]
lowerCamelCase : Any = 0
lowerCamelCase : Dict = 0
for prediction, reference in zip(__A , __A ):
lowerCamelCase : str = jiwer.compute_measures(
__A , __A , truth_transform=__A , hypothesis_transform=__A , )
incorrect += measures["substitutions"] + measures["deletions"] + measures["insertions"]
total += measures["substitutions"] + measures["deletions"] + measures["hits"]
return incorrect / total
| 713
|
"""simple docstring"""
import inspect
import unittest
from transformers import ConvNextConfig
from transformers.testing_utils import require_torch, require_vision, slow, torch_device
from transformers.utils import cached_property, is_torch_available, is_vision_available
from ...test_backbone_common import BackboneTesterMixin
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import ConvNextBackbone, ConvNextForImageClassification, ConvNextModel
from transformers.models.convnext.modeling_convnext import CONVNEXT_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import AutoImageProcessor
class A__ :
"""simple docstring"""
def __init__( self: Optional[Any] , __a: Union[str, Any] , __a: Optional[Any]=13 , __a: Optional[Any]=32 , __a: Dict=3 , __a: int=4 , __a: Dict=[10, 20, 30, 40] , __a: int=[2, 2, 3, 2] , __a: Any=True , __a: List[Any]=True , __a: Any=37 , __a: Optional[int]="gelu" , __a: List[str]=10 , __a: Optional[int]=0.02 , __a: Dict=["stage2", "stage3", "stage4"] , __a: List[str]=[2, 3, 4] , __a: List[str]=None , )-> Union[str, Any]:
lowerCamelCase : Optional[int] = parent
lowerCamelCase : Optional[int] = batch_size
lowerCamelCase : Any = image_size
lowerCamelCase : Tuple = num_channels
lowerCamelCase : str = num_stages
lowerCamelCase : List[str] = hidden_sizes
lowerCamelCase : str = depths
lowerCamelCase : Dict = is_training
lowerCamelCase : Optional[Any] = use_labels
lowerCamelCase : List[str] = intermediate_size
lowerCamelCase : List[str] = hidden_act
lowerCamelCase : List[str] = num_labels
lowerCamelCase : Union[str, Any] = initializer_range
lowerCamelCase : List[Any] = out_features
lowerCamelCase : Optional[Any] = out_indices
lowerCamelCase : int = scope
def a__ ( self: str )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : Dict = None
if self.use_labels:
lowerCamelCase : Dict = ids_tensor([self.batch_size] , self.num_labels )
lowerCamelCase : Any = self.get_config()
return config, pixel_values, labels
def a__ ( self: Dict )-> Union[str, Any]:
return ConvNextConfig(
num_channels=self.num_channels , hidden_sizes=self.hidden_sizes , depths=self.depths , num_stages=self.num_stages , hidden_act=self.hidden_act , is_decoder=__a , initializer_range=self.initializer_range , out_features=self.out_features , out_indices=self.out_indices , num_labels=self.num_labels , )
def a__ ( self: Optional[Any] , __a: List[Any] , __a: Any , __a: int )-> List[Any]:
lowerCamelCase : Optional[int] = ConvNextModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a )
# expected last hidden states: B, C, H // 32, W // 32
self.parent.assertEqual(
result.last_hidden_state.shape , (self.batch_size, self.hidden_sizes[-1], self.image_size // 32, self.image_size // 32) , )
def a__ ( self: int , __a: Union[str, Any] , __a: List[Any] , __a: Tuple )-> Optional[int]:
lowerCamelCase : str = ConvNextForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) )
def a__ ( self: List[Any] , __a: Any , __a: Optional[int] , __a: Tuple )-> List[str]:
lowerCamelCase : List[str] = ConvNextBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a )
# verify hidden states
self.parent.assertEqual(len(result.feature_maps ) , len(config.out_features ) )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.hidden_sizes[1], 4, 4] )
# verify channels
self.parent.assertEqual(len(model.channels ) , len(config.out_features ) )
self.parent.assertListEqual(model.channels , config.hidden_sizes[1:] )
# verify backbone works with out_features=None
lowerCamelCase : Tuple = None
lowerCamelCase : List[str] = ConvNextBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[Any] = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , 1 )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.hidden_sizes[-1], 1, 1] )
# verify channels
self.parent.assertEqual(len(model.channels ) , 1 )
self.parent.assertListEqual(model.channels , [config.hidden_sizes[-1]] )
def a__ ( self: Optional[Any] )-> Any:
lowerCamelCase : List[Any] = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : Tuple = config_and_inputs
lowerCamelCase : int = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : int =(
(
ConvNextModel,
ConvNextForImageClassification,
ConvNextBackbone,
)
if is_torch_available()
else ()
)
snake_case__ : str =(
{'''feature-extraction''': ConvNextModel, '''image-classification''': ConvNextForImageClassification}
if is_torch_available()
else {}
)
snake_case__ : Union[str, Any] =True
snake_case__ : Optional[int] =False
snake_case__ : Tuple =False
snake_case__ : Union[str, Any] =False
snake_case__ : Tuple =False
def a__ ( self: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Tuple = ConvNextModelTester(self )
lowerCamelCase : List[Any] = ConfigTester(self , config_class=__a , has_text_modality=__a , hidden_size=37 )
def a__ ( self: Optional[int] )-> Dict:
self.create_and_test_config_common_properties()
self.config_tester.create_and_test_config_to_json_string()
self.config_tester.create_and_test_config_to_json_file()
self.config_tester.create_and_test_config_from_and_save_pretrained()
self.config_tester.create_and_test_config_with_num_labels()
self.config_tester.check_config_can_be_init_without_params()
self.config_tester.check_config_arguments_init()
def a__ ( self: Optional[int] )-> Optional[Any]:
return
@unittest.skip(reason="""ConvNext does not use inputs_embeds""" )
def a__ ( self: int )-> Dict:
pass
@unittest.skip(reason="""ConvNext does not support input and output embeddings""" )
def a__ ( self: Dict )-> Optional[Any]:
pass
@unittest.skip(reason="""ConvNext does not use feedforward chunking""" )
def a__ ( self: int )-> List[Any]:
pass
def a__ ( self: Union[str, Any] )-> int:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : Any = model_class(__a )
lowerCamelCase : str = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Optional[Any] = [*signature.parameters.keys()]
lowerCamelCase : List[str] = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: Optional[int] )-> str:
lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: str )-> int:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_backbone(*__a )
def a__ ( self: int )-> Optional[int]:
def check_hidden_states_output(__a: Tuple , __a: int , __a: Tuple ):
lowerCamelCase : str = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : Tuple = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : int = outputs.encoder_hidden_states if config.is_encoder_decoder else outputs.hidden_states
lowerCamelCase : Optional[int] = self.model_tester.num_stages
self.assertEqual(len(__a ) , expected_num_stages + 1 )
# ConvNext's feature maps are of shape (batch_size, num_channels, height, width)
self.assertListEqual(
list(hidden_states[0].shape[-2:] ) , [self.model_tester.image_size // 4, self.model_tester.image_size // 4] , )
lowerCamelCase , lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : List[Any] = True
check_hidden_states_output(__a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : Tuple = True
check_hidden_states_output(__a , __a , __a )
def a__ ( self: Dict )-> Optional[Any]:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_classification(*__a )
@slow
def a__ ( self: Optional[Any] )-> Tuple:
for model_name in CONVNEXT_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : str = ConvNextModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def snake_case ( ) -> Optional[int]:
lowerCamelCase : int = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
return image
@require_torch
@require_vision
class A__ ( unittest.TestCase):
"""simple docstring"""
@cached_property
def a__ ( self: Dict )-> Union[str, Any]:
return AutoImageProcessor.from_pretrained("""facebook/convnext-tiny-224""" ) if is_vision_available() else None
@slow
def a__ ( self: List[str] )-> Dict:
lowerCamelCase : Tuple = ConvNextForImageClassification.from_pretrained("""facebook/convnext-tiny-224""" ).to(__a )
lowerCamelCase : Dict = self.default_image_processor
lowerCamelCase : Union[str, Any] = prepare_img()
lowerCamelCase : Optional[Any] = image_processor(images=__a , return_tensors="""pt""" ).to(__a )
# forward pass
with torch.no_grad():
lowerCamelCase : Any = model(**__a )
# verify the logits
lowerCamelCase : Union[str, Any] = torch.Size((1, 1_000) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : Tuple = torch.tensor([-0.02_60, -0.47_39, 0.19_11] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3] , __a , atol=1e-4 ) )
@require_torch
class A__ ( unittest.TestCase , __lowercase):
"""simple docstring"""
snake_case__ : Union[str, Any] =(ConvNextBackbone,) if is_torch_available() else ()
snake_case__ : Optional[Any] =ConvNextConfig
snake_case__ : Optional[Any] =False
def a__ ( self: List[str] )-> int:
lowerCamelCase : Dict = ConvNextModelTester(self )
| 42
| 0
|
"""simple docstring"""
import tempfile
import unittest
import numpy as np
import transformers
from transformers import GPTaTokenizer, GPTJConfig, is_flax_available, is_torch_available
from transformers.testing_utils import is_pt_flax_cross_test, require_flax, tooslow
from ...generation.test_flax_utils import FlaxGenerationTesterMixin
from ...test_modeling_flax_common import FlaxModelTesterMixin, ids_tensor, random_attention_mask
if is_flax_available():
import jax
import jax.numpy as jnp
from transformers.modeling_flax_pytorch_utils import (
convert_pytorch_state_dict_to_flax,
load_flax_weights_in_pytorch_model,
)
from transformers.models.gptj.modeling_flax_gptj import FlaxGPTJForCausalLM, FlaxGPTJModel
if is_torch_available():
import torch
class A__ :
"""simple docstring"""
def __init__( self: str , __a: List[Any] , __a: Any=14 , __a: Optional[Any]=7 , __a: Optional[int]=True , __a: List[Any]=True , __a: int=False , __a: str=True , __a: Optional[int]=99 , __a: Union[str, Any]=32 , __a: Union[str, Any]=4 , __a: List[Any]=4 , __a: int=4 , __a: List[str]=37 , __a: int="gelu" , __a: List[Any]=0.1 , __a: List[str]=0.1 , __a: Optional[Any]=512 , __a: int=0.02 , )-> List[Any]:
lowerCamelCase : List[Any] = parent
lowerCamelCase : Optional[int] = batch_size
lowerCamelCase : int = seq_length
lowerCamelCase : int = is_training
lowerCamelCase : int = use_input_mask
lowerCamelCase : Union[str, Any] = use_token_type_ids
lowerCamelCase : Dict = use_labels
lowerCamelCase : Any = vocab_size
lowerCamelCase : int = hidden_size
lowerCamelCase : Dict = rotary_dim
lowerCamelCase : Optional[Any] = num_hidden_layers
lowerCamelCase : List[str] = num_attention_heads
lowerCamelCase : List[Any] = intermediate_size
lowerCamelCase : int = hidden_act
lowerCamelCase : List[Any] = hidden_dropout_prob
lowerCamelCase : Tuple = attention_probs_dropout_prob
lowerCamelCase : str = max_position_embeddings
lowerCamelCase : int = initializer_range
lowerCamelCase : Tuple = None
lowerCamelCase : List[Any] = vocab_size - 1
lowerCamelCase : Dict = vocab_size - 1
lowerCamelCase : Tuple = vocab_size - 1
def a__ ( self: Optional[Any] )-> Optional[int]:
lowerCamelCase : Optional[int] = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : str = None
if self.use_input_mask:
lowerCamelCase : Any = random_attention_mask([self.batch_size, self.seq_length] )
lowerCamelCase : List[Any] = GPTJConfig(
vocab_size=self.vocab_size , n_embd=self.hidden_size , n_layer=self.num_hidden_layers , n_head=self.num_attention_heads , n_positions=self.max_position_embeddings , use_cache=__A , bos_token_id=self.bos_token_id , eos_token_id=self.eos_token_id , pad_token_id=self.pad_token_id , rotary_dim=self.rotary_dim , )
return (config, input_ids, input_mask)
def a__ ( self: Optional[int] )-> List[Any]:
lowerCamelCase : str = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : Union[str, Any] = config_and_inputs
lowerCamelCase : Tuple = {"""input_ids""": input_ids, """attention_mask""": attention_mask}
return config, inputs_dict
def a__ ( self: int , __a: int , __a: Any , __a: Any , __a: Optional[Any] )-> Dict:
lowerCamelCase : str = 20
lowerCamelCase : str = model_class_name(__A )
lowerCamelCase : Optional[int] = model.init_cache(input_ids.shape[0] , __A )
lowerCamelCase : Dict = jnp.ones((input_ids.shape[0], max_decoder_length) , dtype="""i4""" )
lowerCamelCase : Any = jnp.broadcast_to(
jnp.arange(input_ids.shape[-1] - 1 )[None, :] , (input_ids.shape[0], input_ids.shape[-1] - 1) )
lowerCamelCase : Tuple = model(
input_ids[:, :-1] , attention_mask=__A , past_key_values=__A , position_ids=__A , )
lowerCamelCase : str = jnp.array(input_ids.shape[0] * [[input_ids.shape[-1] - 1]] , dtype="""i4""" )
lowerCamelCase : Any = model(
input_ids[:, -1:] , attention_mask=__A , past_key_values=outputs_cache.past_key_values , position_ids=__A , )
lowerCamelCase : Any = model(__A )
lowerCamelCase : Tuple = np.max(np.abs((outputs_cache_next[0][:, -1, :5] - outputs[0][:, -1, :5]) ) )
self.parent.assertTrue(diff < 1e-3 , msg=f'Max diff is {diff}' )
def a__ ( self: Dict , __a: List[Any] , __a: List[str] , __a: Optional[Any] , __a: Union[str, Any] )-> Optional[int]:
lowerCamelCase : str = 20
lowerCamelCase : int = model_class_name(__A )
lowerCamelCase : str = jnp.concatenate(
[attention_mask, jnp.zeros((attention_mask.shape[0], max_decoder_length - attention_mask.shape[1]) )] , axis=-1 , )
lowerCamelCase : Tuple = model.init_cache(input_ids.shape[0] , __A )
lowerCamelCase : Any = jnp.broadcast_to(
jnp.arange(input_ids.shape[-1] - 1 )[None, :] , (input_ids.shape[0], input_ids.shape[-1] - 1) )
lowerCamelCase : List[str] = model(
input_ids[:, :-1] , attention_mask=__A , past_key_values=__A , position_ids=__A , )
lowerCamelCase : List[str] = jnp.array(input_ids.shape[0] * [[input_ids.shape[-1] - 1]] , dtype="""i4""" )
lowerCamelCase : Tuple = model(
input_ids[:, -1:] , past_key_values=outputs_cache.past_key_values , attention_mask=__A , position_ids=__A , )
lowerCamelCase : Tuple = model(__A , attention_mask=__A )
lowerCamelCase : Union[str, Any] = np.max(np.abs((outputs_cache_next[0][:, -1, :5] - outputs[0][:, -1, :5]) ) )
self.parent.assertTrue(diff < 1e-3 , msg=f'Max diff is {diff}' )
@require_flax
class A__ ( UpperCamelCase__ , UpperCamelCase__ , unittest.TestCase):
"""simple docstring"""
snake_case__ : Union[str, Any] =(FlaxGPTJModel, FlaxGPTJForCausalLM) if is_flax_available() else ()
snake_case__ : Optional[Any] =(FlaxGPTJForCausalLM,) if is_flax_available() else ()
def a__ ( self: Tuple )-> Union[str, Any]:
lowerCamelCase : Optional[int] = FlaxGPTJModelTester(self )
def a__ ( self: Union[str, Any] )-> int:
for model_class_name in self.all_model_classes:
lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = self.model_tester.prepare_config_and_inputs()
self.model_tester.check_use_cache_forward(__A , __A , __A , __A )
def a__ ( self: Dict )-> Dict:
for model_class_name in self.all_model_classes:
lowerCamelCase , lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
self.model_tester.check_use_cache_forward_with_attn_mask(
__A , __A , __A , __A )
@tooslow
def a__ ( self: List[str] )-> int:
lowerCamelCase : Any = GPTaTokenizer.from_pretrained("""gpt2""" , pad_token="""<|endoftext|>""" , padding_side="""left""" )
lowerCamelCase : Union[str, Any] = tokenizer(["""Hello this is a long string""", """Hey"""] , return_tensors="""np""" , padding=__A , truncation=__A )
lowerCamelCase : int = FlaxGPTJForCausalLM.from_pretrained("""EleutherAI/gpt-j-6B""" )
lowerCamelCase : int = False
lowerCamelCase : Optional[Any] = model.config.eos_token_id
lowerCamelCase : Union[str, Any] = jax.jit(model.generate )
lowerCamelCase : str = jit_generate(
inputs["""input_ids"""] , attention_mask=inputs["""attention_mask"""] , pad_token_id=tokenizer.pad_token_id ).sequences
lowerCamelCase : Union[str, Any] = tokenizer.batch_decode(__A , skip_special_tokens=__A )
lowerCamelCase : List[str] = [
"""Hello this is a long string of text.\n\nI'm trying to get the text of the""",
"""Hey, I'm a little late to the party. I'm going to""",
]
self.assertListEqual(__A , __A )
@is_pt_flax_cross_test
def a__ ( self: Any )-> str:
lowerCamelCase , lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
with self.subTest(model_class.__name__ ):
# prepare inputs
lowerCamelCase : Any = self._prepare_for_class(__A , __A )
lowerCamelCase : List[Any] = {k: torch.tensor(v.tolist() ) for k, v in prepared_inputs_dict.items()}
# load corresponding PyTorch class
lowerCamelCase : List[Any] = model_class.__name__[4:] # Skip the "Flax" at the beginning
lowerCamelCase : int = getattr(__A , __A )
lowerCamelCase , lowerCamelCase : Optional[int] = pt_inputs["""input_ids"""].shape
lowerCamelCase : List[str] = np.random.randint(0 , seq_length - 1 , size=(batch_size,) )
for batch_idx, start_index in enumerate(__A ):
lowerCamelCase : Dict = 0
lowerCamelCase : List[str] = 1
lowerCamelCase : Any = 0
lowerCamelCase : List[str] = 1
lowerCamelCase : Optional[Any] = pt_model_class(__A ).eval()
lowerCamelCase : Optional[Any] = model_class(__A , dtype=jnp.floataa )
lowerCamelCase : int = convert_pytorch_state_dict_to_flax(pt_model.state_dict() , __A )
lowerCamelCase : Any = fx_state
with torch.no_grad():
lowerCamelCase : int = pt_model(**__A ).to_tuple()
lowerCamelCase : Any = fx_model(**__A ).to_tuple()
self.assertEqual(len(__A ) , len(__A ) , """Output lengths differ between Flax and PyTorch""" )
for fx_output, pt_output in zip(__A , __A ):
self.assert_almost_equals(fx_output[:, -1] , pt_output[:, -1].numpy() , 4e-2 )
with tempfile.TemporaryDirectory() as tmpdirname:
pt_model.save_pretrained(__A )
lowerCamelCase : Any = model_class.from_pretrained(__A , from_pt=__A )
lowerCamelCase : int = fx_model_loaded(**__A ).to_tuple()
self.assertEqual(
len(__A ) , len(__A ) , """Output lengths differ between Flax and PyTorch""" )
for fx_output_loaded, pt_output in zip(__A , __A ):
self.assert_almost_equals(fx_output_loaded[:, -1] , pt_output[:, -1].numpy() , 4e-2 )
@is_pt_flax_cross_test
def a__ ( self: Any )-> Dict:
lowerCamelCase , lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
with self.subTest(model_class.__name__ ):
# prepare inputs
lowerCamelCase : List[Any] = self._prepare_for_class(__A , __A )
lowerCamelCase : List[Any] = {k: torch.tensor(v.tolist() ) for k, v in prepared_inputs_dict.items()}
# load corresponding PyTorch class
lowerCamelCase : List[Any] = model_class.__name__[4:] # Skip the "Flax" at the beginning
lowerCamelCase : Union[str, Any] = getattr(__A , __A )
lowerCamelCase : List[str] = pt_model_class(__A ).eval()
lowerCamelCase : Tuple = model_class(__A , dtype=jnp.floataa )
lowerCamelCase : Optional[Any] = load_flax_weights_in_pytorch_model(__A , fx_model.params )
lowerCamelCase , lowerCamelCase : str = pt_inputs["""input_ids"""].shape
lowerCamelCase : int = np.random.randint(0 , seq_length - 1 , size=(batch_size,) )
for batch_idx, start_index in enumerate(__A ):
lowerCamelCase : Union[str, Any] = 0
lowerCamelCase : int = 1
lowerCamelCase : Dict = 0
lowerCamelCase : str = 1
# make sure weights are tied in PyTorch
pt_model.tie_weights()
with torch.no_grad():
lowerCamelCase : Union[str, Any] = pt_model(**__A ).to_tuple()
lowerCamelCase : Optional[int] = fx_model(**__A ).to_tuple()
self.assertEqual(len(__A ) , len(__A ) , """Output lengths differ between Flax and PyTorch""" )
for fx_output, pt_output in zip(__A , __A ):
self.assert_almost_equals(fx_output[:, -1] , pt_output[:, -1].numpy() , 4e-2 )
with tempfile.TemporaryDirectory() as tmpdirname:
fx_model.save_pretrained(__A )
lowerCamelCase : Dict = pt_model_class.from_pretrained(__A , from_flax=__A )
with torch.no_grad():
lowerCamelCase : int = pt_model_loaded(**__A ).to_tuple()
self.assertEqual(
len(__A ) , len(__A ) , """Output lengths differ between Flax and PyTorch""" )
for fx_output, pt_output in zip(__A , __A ):
self.assert_almost_equals(fx_output[:, -1] , pt_output[:, -1].numpy() , 4e-2 )
@tooslow
def a__ ( self: Union[str, Any] )-> Tuple:
for model_class_name in self.all_model_classes:
lowerCamelCase : Union[str, Any] = model_class_name.from_pretrained("""EleutherAI/gpt-j-6B""" )
lowerCamelCase : Optional[int] = model(np.ones((1, 1) ) )
self.assertIsNotNone(__A )
| 714
|
"""simple docstring"""
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :Optional[int] = logging.get_logger(__name__)
__lowerCamelCase :List[str] = {
'google/realm-cc-news-pretrained-embedder': (
'https://huggingface.co/google/realm-cc-news-pretrained-embedder/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-encoder': (
'https://huggingface.co/google/realm-cc-news-pretrained-encoder/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-scorer': (
'https://huggingface.co/google/realm-cc-news-pretrained-scorer/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-openqa': (
'https://huggingface.co/google/realm-cc-news-pretrained-openqa/aresolve/main/config.json'
),
'google/realm-orqa-nq-openqa': 'https://huggingface.co/google/realm-orqa-nq-openqa/resolve/main/config.json',
'google/realm-orqa-nq-reader': 'https://huggingface.co/google/realm-orqa-nq-reader/resolve/main/config.json',
'google/realm-orqa-wq-openqa': 'https://huggingface.co/google/realm-orqa-wq-openqa/resolve/main/config.json',
'google/realm-orqa-wq-reader': 'https://huggingface.co/google/realm-orqa-wq-reader/resolve/main/config.json',
# See all REALM models at https://huggingface.co/models?filter=realm
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Optional[Any] ='''realm'''
def __init__( self: Union[str, Any] , __a: List[Any]=30_522 , __a: List[Any]=768 , __a: List[Any]=128 , __a: Union[str, Any]=12 , __a: Union[str, Any]=12 , __a: Optional[Any]=8 , __a: Dict=3_072 , __a: List[Any]="gelu_new" , __a: List[Any]=0.1 , __a: Tuple=0.1 , __a: Optional[Any]=512 , __a: Optional[int]=2 , __a: str=0.02 , __a: int=1e-1_2 , __a: Optional[Any]=256 , __a: Any=10 , __a: Dict=1e-3 , __a: Optional[Any]=5 , __a: Dict=320 , __a: Tuple=13_353_718 , __a: List[Any]=5_000 , __a: Dict=1 , __a: int=0 , __a: Dict=2 , **__a: List[str] , )-> Any:
super().__init__(pad_token_id=__a , bos_token_id=__a , eos_token_id=__a , **__a )
# Common config
lowerCamelCase : Optional[Any] = vocab_size
lowerCamelCase : str = max_position_embeddings
lowerCamelCase : Dict = hidden_size
lowerCamelCase : Dict = retriever_proj_size
lowerCamelCase : Optional[Any] = num_hidden_layers
lowerCamelCase : List[str] = num_attention_heads
lowerCamelCase : Tuple = num_candidates
lowerCamelCase : int = intermediate_size
lowerCamelCase : Dict = hidden_act
lowerCamelCase : List[str] = hidden_dropout_prob
lowerCamelCase : Dict = attention_probs_dropout_prob
lowerCamelCase : Optional[int] = initializer_range
lowerCamelCase : Dict = type_vocab_size
lowerCamelCase : Optional[Any] = layer_norm_eps
# Reader config
lowerCamelCase : List[str] = span_hidden_size
lowerCamelCase : Dict = max_span_width
lowerCamelCase : Optional[Any] = reader_layer_norm_eps
lowerCamelCase : Optional[int] = reader_beam_size
lowerCamelCase : List[Any] = reader_seq_len
# Retrieval config
lowerCamelCase : int = num_block_records
lowerCamelCase : Dict = searcher_beam_size
| 42
| 0
|
"""simple docstring"""
import argparse
from typing import Dict
import tensorflow as tf
import torch
from tqdm import tqdm
from transformers import BigBirdPegasusConfig, BigBirdPegasusForConditionalGeneration
__lowerCamelCase :str = [
# tf -> hf
('''/''', '''.'''),
('''layer_''', '''layers.'''),
('''kernel''', '''weight'''),
('''beta''', '''bias'''),
('''gamma''', '''weight'''),
('''pegasus''', '''model'''),
]
__lowerCamelCase :Optional[Any] = [
('''.output.dense''', '''.fc2'''),
('''intermediate.LayerNorm''', '''final_layer_norm'''),
('''intermediate.dense''', '''fc1'''),
]
__lowerCamelCase :Optional[Any] = (
INIT_COMMON
+ [
('''attention.self.LayerNorm''', '''self_attn_layer_norm'''),
('''attention.output.dense''', '''self_attn.out_proj'''),
('''attention.self''', '''self_attn'''),
('''attention.encdec.LayerNorm''', '''encoder_attn_layer_norm'''),
('''attention.encdec_output.dense''', '''encoder_attn.out_proj'''),
('''attention.encdec''', '''encoder_attn'''),
('''key''', '''k_proj'''),
('''value''', '''v_proj'''),
('''query''', '''q_proj'''),
('''decoder.LayerNorm''', '''decoder.layernorm_embedding'''),
]
+ END_COMMON
)
__lowerCamelCase :Optional[Any] = (
INIT_COMMON
+ [
('''embeddings.word_embeddings''', '''shared.weight'''),
('''embeddings.position_embeddings''', '''embed_positions.weight'''),
('''attention.self.LayerNorm''', '''self_attn_layer_norm'''),
('''attention.output.dense''', '''self_attn.output'''),
('''attention.self''', '''self_attn.self'''),
('''encoder.LayerNorm''', '''encoder.layernorm_embedding'''),
]
+ END_COMMON
)
__lowerCamelCase :Tuple = [
'''encdec/key/bias''',
'''encdec/query/bias''',
'''encdec/value/bias''',
'''self/key/bias''',
'''self/query/bias''',
'''self/value/bias''',
'''encdec_output/dense/bias''',
'''attention/output/dense/bias''',
]
def snake_case ( UpperCamelCase__ : Dict , UpperCamelCase__ : List[str] ) -> Optional[Any]:
for tf_name, hf_name in patterns:
lowerCamelCase : Optional[Any] = k.replace(_UpperCAmelCase , _UpperCAmelCase )
return k
def snake_case ( UpperCamelCase__ : Any , UpperCamelCase__ : Optional[int] ) -> List[Any]:
lowerCamelCase : Optional[int] = BigBirdPegasusConfig(**_UpperCAmelCase )
lowerCamelCase : List[Any] = BigBirdPegasusForConditionalGeneration(_UpperCAmelCase )
lowerCamelCase : Any = torch_model.state_dict()
lowerCamelCase : int = {}
# separating decoder weights
lowerCamelCase : Optional[Any] = {k: tf_weights[k] for k in tf_weights if k.startswith("""pegasus/decoder""" )}
lowerCamelCase : int = {k: tf_weights[k] for k in tf_weights if not k.startswith("""pegasus/decoder""" )}
for k, v in tqdm(decoder_weights.items() , """tf -> hf conversion""" ):
lowerCamelCase : Any = [k.endswith(_UpperCAmelCase ) for ending in KEYS_TO_IGNORE]
if any(_UpperCAmelCase ):
continue
lowerCamelCase : List[str] = DECODER_PATTERNS
lowerCamelCase : Union[str, Any] = rename_state_dict_key(_UpperCAmelCase , _UpperCAmelCase )
if new_k not in state_dict:
raise ValueError(F'could not find new key {new_k} in state dict. (converted from {k})' )
if any(True if i in k else False for i in ["""dense""", """query""", """key""", """value"""] ):
lowerCamelCase : Union[str, Any] = v.T
lowerCamelCase : Tuple = torch.from_numpy(_UpperCAmelCase )
assert v.shape == state_dict[new_k].shape, F'{new_k}, {k}, {v.shape}, {state_dict[new_k].shape}'
for k, v in tqdm(remaining_weights.items() , """tf -> hf conversion""" ):
lowerCamelCase : str = [k.endswith(_UpperCAmelCase ) for ending in KEYS_TO_IGNORE]
if any(_UpperCAmelCase ):
continue
lowerCamelCase : str = REMAINING_PATTERNS
lowerCamelCase : Union[str, Any] = rename_state_dict_key(_UpperCAmelCase , _UpperCAmelCase )
if new_k not in state_dict and k != "pegasus/embeddings/position_embeddings":
raise ValueError(F'could not find new key {new_k} in state dict. (converted from {k})' )
if any(True if i in k else False for i in ["""dense""", """query""", """key""", """value"""] ):
lowerCamelCase : int = v.T
lowerCamelCase : List[str] = torch.from_numpy(_UpperCAmelCase )
if k != "pegasus/embeddings/position_embeddings":
assert v.shape == state_dict[new_k].shape, F'{new_k}, {k}, {v.shape}, {state_dict[new_k].shape}'
lowerCamelCase : Union[str, Any] = mapping["""model.embed_positions.weight"""]
lowerCamelCase : str = mapping.pop("""model.embed_positions.weight""" )
lowerCamelCase , lowerCamelCase : int = torch_model.load_state_dict(_UpperCAmelCase , strict=_UpperCAmelCase )
lowerCamelCase : Optional[Any] = [
k
for k in missing
if k
not in [
"""final_logits_bias""",
"""model.encoder.embed_tokens.weight""",
"""model.decoder.embed_tokens.weight""",
"""lm_head.weight""",
]
]
assert unexpected_missing == [], F'no matches found for the following torch keys {unexpected_missing}'
assert extra == [], F'no matches found for the following tf keys {extra}'
return torch_model
def snake_case ( UpperCamelCase__ : List[Any] ) -> Dict:
lowerCamelCase : List[str] = tf.train.list_variables(_UpperCAmelCase )
lowerCamelCase : Optional[Any] = {}
lowerCamelCase : List[str] = ["""global_step"""]
for name, shape in tqdm(_UpperCAmelCase , desc="""converting tf checkpoint to dict""" ):
lowerCamelCase : Dict = any(pat in name for pat in ignore_name )
if skip_key:
continue
lowerCamelCase : Optional[int] = tf.train.load_variable(_UpperCAmelCase , _UpperCAmelCase )
lowerCamelCase : int = array
return tf_weights
def snake_case ( UpperCamelCase__ : Tuple , UpperCamelCase__ : Dict , UpperCamelCase__ : Dict ) -> Dict:
lowerCamelCase : str = get_tf_weights_as_numpy(_UpperCAmelCase )
lowerCamelCase : Optional[Any] = convert_bigbird_pegasus(_UpperCAmelCase , _UpperCAmelCase )
torch_model.save_pretrained(_UpperCAmelCase )
if __name__ == "__main__":
__lowerCamelCase :Optional[Any] = argparse.ArgumentParser()
parser.add_argument('--tf_ckpt_path', type=str, help='passed to tf.train.list_variables')
parser.add_argument('--save_dir', default=None, type=str, help='Path to the output PyTorch model.')
__lowerCamelCase :int = parser.parse_args()
__lowerCamelCase :str = {}
convert_bigbird_pegasus_ckpt_to_pytorch(args.tf_ckpt_path, args.save_dir, config_update=config_update)
| 715
|
"""simple docstring"""
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :Tuple = logging.get_logger(__name__)
__lowerCamelCase :Any = {
'vinvino02/glpn-kitti': 'https://huggingface.co/vinvino02/glpn-kitti/resolve/main/config.json',
# See all GLPN models at https://huggingface.co/models?filter=glpn
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Tuple ='''glpn'''
def __init__( self: Dict , __a: List[str]=3 , __a: Optional[int]=4 , __a: Dict=[2, 2, 2, 2] , __a: str=[8, 4, 2, 1] , __a: Optional[int]=[32, 64, 160, 256] , __a: Dict=[7, 3, 3, 3] , __a: Dict=[4, 2, 2, 2] , __a: Optional[Any]=[1, 2, 5, 8] , __a: Tuple=[4, 4, 4, 4] , __a: int="gelu" , __a: Union[str, Any]=0.0 , __a: str=0.0 , __a: Union[str, Any]=0.02 , __a: str=0.1 , __a: Union[str, Any]=1e-6 , __a: Any=64 , __a: Dict=10 , __a: Union[str, Any]=-1 , **__a: Optional[Any] , )-> Dict:
super().__init__(**__a )
lowerCamelCase : Dict = num_channels
lowerCamelCase : Any = num_encoder_blocks
lowerCamelCase : Dict = depths
lowerCamelCase : List[str] = sr_ratios
lowerCamelCase : Dict = hidden_sizes
lowerCamelCase : Tuple = patch_sizes
lowerCamelCase : Optional[int] = strides
lowerCamelCase : Optional[Any] = mlp_ratios
lowerCamelCase : Union[str, Any] = num_attention_heads
lowerCamelCase : List[str] = hidden_act
lowerCamelCase : Any = hidden_dropout_prob
lowerCamelCase : Optional[int] = attention_probs_dropout_prob
lowerCamelCase : List[Any] = initializer_range
lowerCamelCase : Dict = drop_path_rate
lowerCamelCase : Any = layer_norm_eps
lowerCamelCase : Optional[Any] = decoder_hidden_size
lowerCamelCase : Tuple = max_depth
lowerCamelCase : Optional[Any] = head_in_index
| 42
| 0
|
"""simple docstring"""
import cva
import numpy as np
class A__ :
"""simple docstring"""
def __init__( self: Tuple , __a: Dict , __a: int )-> int:
if k in (0.04, 0.06):
lowerCamelCase : List[str] = k
lowerCamelCase : str = window_size
else:
raise ValueError("""invalid k value""" )
def __str__( self: List[str] )-> str:
return str(self.k )
def a__ ( self: Optional[Any] , __a: Dict )-> tuple[cva.Mat, list[list[int]]]:
lowerCamelCase : Any = cva.imread(__a , 0 )
lowerCamelCase : Optional[Any] = img.shape
lowerCamelCase : list[list[int]] = []
lowerCamelCase : str = img.copy()
lowerCamelCase : Optional[int] = cva.cvtColor(__a , cva.COLOR_GRAY2RGB )
lowerCamelCase : List[str] = np.gradient(__a )
lowerCamelCase : Optional[Any] = dx**2
lowerCamelCase : int = dy**2
lowerCamelCase : List[str] = dx * dy
lowerCamelCase : Any = 0.04
lowerCamelCase : List[Any] = self.window_size // 2
for y in range(__a , h - offset ):
for x in range(__a , w - offset ):
lowerCamelCase : Union[str, Any] = ixx[
y - offset : y + offset + 1, x - offset : x + offset + 1
].sum()
lowerCamelCase : int = iyy[
y - offset : y + offset + 1, x - offset : x + offset + 1
].sum()
lowerCamelCase : List[Any] = ixy[
y - offset : y + offset + 1, x - offset : x + offset + 1
].sum()
lowerCamelCase : Dict = (wxx * wyy) - (wxy**2)
lowerCamelCase : List[str] = wxx + wyy
lowerCamelCase : Tuple = det - k * (trace**2)
# Can change the value
if r > 0.5:
corner_list.append([x, y, r] )
color_img.itemset((y, x, 0) , 0 )
color_img.itemset((y, x, 1) , 0 )
color_img.itemset((y, x, 2) , 255 )
return color_img, corner_list
if __name__ == "__main__":
__lowerCamelCase :str = HarrisCorner(0.04, 3)
__lowerCamelCase , __lowerCamelCase :int = edge_detect.detect('path_to_image')
cva.imwrite('detect.png', color_img)
| 716
|
"""simple docstring"""
from __future__ import annotations
import math
def snake_case ( UpperCamelCase__ : float , UpperCamelCase__ : int ) -> float:
lowerCamelCase : Dict = u
for i in range(1 , UpperCamelCase__ ):
lowerCamelCase : List[str] = temp * (u - i)
return temp
def snake_case ( ) -> None:
lowerCamelCase : List[Any] = int(input("""enter the numbers of values: """ ) )
lowerCamelCase : list[list[float]] = []
for _ in range(UpperCamelCase__ ):
y.append([] )
for i in range(UpperCamelCase__ ):
for j in range(UpperCamelCase__ ):
y[i].append(UpperCamelCase__ )
lowerCamelCase : Union[str, Any] = 0
print("""enter the values of parameters in a list: """ )
lowerCamelCase : Any = list(map(UpperCamelCase__ , input().split() ) )
print("""enter the values of corresponding parameters: """ )
for i in range(UpperCamelCase__ ):
lowerCamelCase : int = float(input() )
lowerCamelCase : Dict = int(input("""enter the value to interpolate: """ ) )
lowerCamelCase : List[Any] = (value - x[0]) / (x[1] - x[0])
# for calculating forward difference table
for i in range(1 , UpperCamelCase__ ):
for j in range(n - i ):
lowerCamelCase : str = y[j + 1][i - 1] - y[j][i - 1]
lowerCamelCase : Any = y[0][0]
for i in range(1 , UpperCamelCase__ ):
summ += (ucal(UpperCamelCase__ , UpperCamelCase__ ) * y[0][i]) / math.factorial(UpperCamelCase__ )
print(F'the value at {value} is {summ}' )
if __name__ == "__main__":
main()
| 42
| 0
|
"""simple docstring"""
def snake_case ( UpperCamelCase__ : str ) -> Any:
if any(not isinstance(_lowercase , _lowercase ) or x < 0 for x in sequence ):
raise TypeError("""Sequence must be list of non-negative integers""" )
for _ in range(len(_lowercase ) ):
for i, (rod_upper, rod_lower) in enumerate(zip(_lowercase , sequence[1:] ) ):
if rod_upper > rod_lower:
sequence[i] -= rod_upper - rod_lower
sequence[i + 1] += rod_upper - rod_lower
return sequence
if __name__ == "__main__":
assert bead_sort([5, 4, 3, 2, 1]) == [1, 2, 3, 4, 5]
assert bead_sort([7, 9, 4, 3, 5]) == [3, 4, 5, 7, 9]
| 717
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_sentencepiece_available
__lowerCamelCase :str = {}
try:
if not is_sentencepiece_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[Any] = ['GPTSw3Tokenizer']
if TYPE_CHECKING:
try:
if not is_sentencepiece_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .tokenization_gpt_swa import GPTSwaTokenizer
else:
import sys
__lowerCamelCase :Tuple = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
def snake_case ( UpperCamelCase__ : Any , UpperCamelCase__ : List[str] ) -> Optional[Any]:
lowerCamelCase : List[str] = 0
lowerCamelCase : Tuple = len(_lowerCamelCase ) - 1
while left <= right:
# avoid divided by 0 during interpolation
if sorted_collection[left] == sorted_collection[right]:
if sorted_collection[left] == item:
return left
else:
return None
lowerCamelCase : str = left + ((item - sorted_collection[left]) * (right - left)) // (
sorted_collection[right] - sorted_collection[left]
)
# out of range check
if point < 0 or point >= len(_lowerCamelCase ):
return None
lowerCamelCase : List[Any] = sorted_collection[point]
if current_item == item:
return point
else:
if point < left:
lowerCamelCase : List[Any] = left
lowerCamelCase : Any = point
elif point > right:
lowerCamelCase : List[Any] = right
lowerCamelCase : Tuple = point
else:
if item < current_item:
lowerCamelCase : int = point - 1
else:
lowerCamelCase : str = point + 1
return None
def snake_case ( UpperCamelCase__ : Union[str, Any] , UpperCamelCase__ : Optional[int] , UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : int ) -> Optional[int]:
# avoid divided by 0 during interpolation
if sorted_collection[left] == sorted_collection[right]:
if sorted_collection[left] == item:
return left
else:
return None
lowerCamelCase : List[str] = left + ((item - sorted_collection[left]) * (right - left)) // (
sorted_collection[right] - sorted_collection[left]
)
# out of range check
if point < 0 or point >= len(_lowerCamelCase ):
return None
if sorted_collection[point] == item:
return point
elif point < left:
return interpolation_search_by_recursion(_lowerCamelCase , _lowerCamelCase , _lowerCamelCase , _lowerCamelCase )
elif point > right:
return interpolation_search_by_recursion(_lowerCamelCase , _lowerCamelCase , _lowerCamelCase , _lowerCamelCase )
else:
if sorted_collection[point] > item:
return interpolation_search_by_recursion(
_lowerCamelCase , _lowerCamelCase , _lowerCamelCase , point - 1 )
else:
return interpolation_search_by_recursion(
_lowerCamelCase , _lowerCamelCase , point + 1 , _lowerCamelCase )
def snake_case ( UpperCamelCase__ : Optional[Any] ) -> List[Any]:
if collection != sorted(_lowerCamelCase ):
raise ValueError("""Collection must be ascending sorted""" )
return True
if __name__ == "__main__":
import sys
__lowerCamelCase :str = 0
if debug == 1:
__lowerCamelCase :Optional[Any] = [10, 30, 40, 45, 50, 66, 77, 93]
try:
__assert_sorted(collection)
except ValueError:
sys.exit('Sequence must be ascending sorted to apply interpolation search')
__lowerCamelCase :Dict = 67
__lowerCamelCase :Dict = interpolation_search(collection, target)
if result is not None:
print(F"""{target} found at positions: {result}""")
else:
print('Not found')
| 718
|
"""simple docstring"""
import argparse
import json
from collections import OrderedDict
from functools import partial
from pathlib import Path
import timm
import torch
from huggingface_hub import hf_hub_download
from transformers import LevitConfig, LevitForImageClassificationWithTeacher, LevitImageProcessor
from transformers.utils import logging
logging.set_verbosity_info()
__lowerCamelCase :Dict = logging.get_logger()
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : str , UpperCamelCase__ : LevitConfig , UpperCamelCase__ : Path , UpperCamelCase__ : bool = True ) -> Dict:
print(F'Converting {name}...' )
with torch.no_grad():
if hidden_sizes == 128:
if name[-1] == "S":
lowerCamelCase : Optional[Any] = timm.create_model("""levit_128s""" , pretrained=UpperCamelCase__ )
else:
lowerCamelCase : Dict = timm.create_model("""levit_128""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 192:
lowerCamelCase : Tuple = timm.create_model("""levit_192""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 256:
lowerCamelCase : Optional[int] = timm.create_model("""levit_256""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 384:
lowerCamelCase : Dict = timm.create_model("""levit_384""" , pretrained=UpperCamelCase__ )
from_model.eval()
lowerCamelCase : Optional[Any] = LevitForImageClassificationWithTeacher(UpperCamelCase__ ).eval()
lowerCamelCase : Tuple = OrderedDict()
lowerCamelCase : Optional[Any] = from_model.state_dict()
lowerCamelCase : str = list(from_model.state_dict().keys() )
lowerCamelCase : List[Any] = list(our_model.state_dict().keys() )
print(len(UpperCamelCase__ ) , len(UpperCamelCase__ ) )
for i in range(len(UpperCamelCase__ ) ):
lowerCamelCase : str = weights[og_keys[i]]
our_model.load_state_dict(UpperCamelCase__ )
lowerCamelCase : int = torch.randn((2, 3, 224, 224) )
lowerCamelCase : Any = from_model(UpperCamelCase__ )
lowerCamelCase : List[Any] = our_model(UpperCamelCase__ ).logits
assert torch.allclose(UpperCamelCase__ , UpperCamelCase__ ), "The model logits don't match the original one."
lowerCamelCase : Dict = name
print(UpperCamelCase__ )
if push_to_hub:
our_model.save_pretrained(save_directory / checkpoint_name )
lowerCamelCase : Optional[int] = LevitImageProcessor()
image_processor.save_pretrained(save_directory / checkpoint_name )
print(F'Pushed {checkpoint_name}' )
def snake_case ( UpperCamelCase__ : Path , UpperCamelCase__ : str = None , UpperCamelCase__ : bool = True ) -> Optional[int]:
lowerCamelCase : Optional[Any] = """imagenet-1k-id2label.json"""
lowerCamelCase : List[Any] = 1000
lowerCamelCase : Dict = (1, num_labels)
lowerCamelCase : List[Any] = """huggingface/label-files"""
lowerCamelCase : Optional[int] = num_labels
lowerCamelCase : List[str] = json.load(open(hf_hub_download(UpperCamelCase__ , UpperCamelCase__ , repo_type="""dataset""" ) , """r""" ) )
lowerCamelCase : Any = {int(UpperCamelCase__ ): v for k, v in idalabel.items()}
lowerCamelCase : List[Any] = idalabel
lowerCamelCase : str = {v: k for k, v in idalabel.items()}
lowerCamelCase : Tuple = partial(UpperCamelCase__ , num_labels=UpperCamelCase__ , idalabel=UpperCamelCase__ , labelaid=UpperCamelCase__ )
lowerCamelCase : Optional[int] = {
"""levit-128S""": 128,
"""levit-128""": 128,
"""levit-192""": 192,
"""levit-256""": 256,
"""levit-384""": 384,
}
lowerCamelCase : List[Any] = {
"""levit-128S""": ImageNetPreTrainedConfig(
hidden_sizes=[128, 256, 384] , num_attention_heads=[4, 6, 8] , depths=[2, 3, 4] , key_dim=[16, 16, 16] , drop_path_rate=0 , ),
"""levit-128""": ImageNetPreTrainedConfig(
hidden_sizes=[128, 256, 384] , num_attention_heads=[4, 8, 12] , depths=[4, 4, 4] , key_dim=[16, 16, 16] , drop_path_rate=0 , ),
"""levit-192""": ImageNetPreTrainedConfig(
hidden_sizes=[192, 288, 384] , num_attention_heads=[3, 5, 6] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0 , ),
"""levit-256""": ImageNetPreTrainedConfig(
hidden_sizes=[256, 384, 512] , num_attention_heads=[4, 6, 8] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0 , ),
"""levit-384""": ImageNetPreTrainedConfig(
hidden_sizes=[384, 512, 768] , num_attention_heads=[6, 9, 12] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0.1 , ),
}
if model_name:
convert_weight_and_push(
names_to_hidden_sizes[model_name] , UpperCamelCase__ , names_to_config[model_name] , UpperCamelCase__ , UpperCamelCase__ )
else:
for model_name, config in names_to_config.items():
convert_weight_and_push(names_to_hidden_sizes[model_name] , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return config, expected_shape
if __name__ == "__main__":
__lowerCamelCase :Union[str, Any] = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--model_name',
default=None,
type=str,
help='The name of the model you wish to convert, it must be one of the supported Levit* architecture,',
)
parser.add_argument(
'--pytorch_dump_folder_path',
default='levit-dump-folder/',
type=Path,
required=False,
help='Path to the output PyTorch model directory.',
)
parser.add_argument('--push_to_hub', action='store_true', help='Push model and image processor to the hub')
parser.add_argument(
'--no-push_to_hub',
dest='push_to_hub',
action='store_false',
help='Do not push model and image processor to the hub',
)
__lowerCamelCase :List[Any] = parser.parse_args()
__lowerCamelCase :Path = args.pytorch_dump_folder_path
pytorch_dump_folder_path.mkdir(exist_ok=True, parents=True)
convert_weights_and_push(pytorch_dump_folder_path, args.model_name, args.push_to_hub)
| 42
| 0
|
"""simple docstring"""
from pathlib import Path
import cva
import numpy as np
from matplotlib import pyplot as plt
def snake_case ( UpperCamelCase__ : np.ndarray , UpperCamelCase__ : np.ndarray , UpperCamelCase__ : np.ndarray , UpperCamelCase__ : int , UpperCamelCase__ : int ) -> np.ndarray:
lowerCamelCase : Any = cva.getAffineTransform(lowercase_ , lowercase_ )
return cva.warpAffine(lowercase_ , lowercase_ , (rows, cols) )
if __name__ == "__main__":
# read original image
__lowerCamelCase :List[Any] = cva.imread(
str(Path(__file__).resolve().parent.parent / 'image_data' / 'lena.jpg')
)
# turn image in gray scale value
__lowerCamelCase :List[Any] = cva.cvtColor(image, cva.COLOR_BGR2GRAY)
# get image shape
__lowerCamelCase :Optional[Any] = gray_img.shape
# set different points to rotate image
__lowerCamelCase :str = np.array([[50, 50], [200, 50], [50, 200]], np.floataa)
__lowerCamelCase :Dict = np.array([[10, 100], [200, 50], [100, 250]], np.floataa)
__lowerCamelCase :Optional[int] = np.array([[50, 50], [150, 50], [120, 200]], np.floataa)
__lowerCamelCase :Optional[int] = np.array([[10, 100], [80, 50], [180, 250]], np.floataa)
# add all rotated images in a list
__lowerCamelCase :Dict = [
gray_img,
get_rotation(gray_img, ptsa, ptsa, img_rows, img_cols),
get_rotation(gray_img, ptsa, ptsa, img_rows, img_cols),
get_rotation(gray_img, ptsa, ptsa, img_rows, img_cols),
]
# plot different image rotations
__lowerCamelCase :Union[str, Any] = plt.figure(1)
__lowerCamelCase :Union[str, Any] = ['Original', 'Rotation 1', 'Rotation 2', 'Rotation 3']
for i, image in enumerate(images):
plt.subplot(2, 2, i + 1), plt.imshow(image, 'gray')
plt.title(titles[i])
plt.axis('off')
plt.subplots_adjust(left=0.0, bottom=0.05, right=1.0, top=0.95)
plt.show()
| 719
|
"""simple docstring"""
import torch
from diffusers import KDPMaDiscreteScheduler
from diffusers.utils import torch_device
from .test_schedulers import SchedulerCommonTest
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Tuple =(KDPMaDiscreteScheduler,)
snake_case__ : Tuple =10
def a__ ( self: List[Any] , **__a: Optional[int] )-> Union[str, Any]:
lowerCamelCase : int = {
"""num_train_timesteps""": 1_100,
"""beta_start""": 0.00_01,
"""beta_end""": 0.02,
"""beta_schedule""": """linear""",
}
config.update(**__a )
return config
def a__ ( self: Union[str, Any] )-> Any:
for timesteps in [10, 50, 100, 1_000]:
self.check_over_configs(num_train_timesteps=__a )
def a__ ( self: str )-> int:
for beta_start, beta_end in zip([0.0_00_01, 0.00_01, 0.0_01] , [0.00_02, 0.0_02, 0.02] ):
self.check_over_configs(beta_start=__a , beta_end=__a )
def a__ ( self: int )-> Union[str, Any]:
for schedule in ["linear", "scaled_linear"]:
self.check_over_configs(beta_schedule=__a )
def a__ ( self: List[Any] )-> List[Any]:
for prediction_type in ["epsilon", "v_prediction"]:
self.check_over_configs(prediction_type=__a )
def a__ ( self: Union[str, Any] )-> int:
lowerCamelCase : List[str] = self.scheduler_classes[0]
lowerCamelCase : Union[str, Any] = self.get_scheduler_config(prediction_type="""v_prediction""" )
lowerCamelCase : List[str] = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps )
lowerCamelCase : Dict = self.dummy_model()
lowerCamelCase : List[Any] = self.dummy_sample_deter * scheduler.init_noise_sigma
lowerCamelCase : List[Any] = sample.to(__a )
for i, t in enumerate(scheduler.timesteps ):
lowerCamelCase : Optional[Any] = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[int] = model(__a , __a )
lowerCamelCase : Tuple = scheduler.step(__a , __a , __a )
lowerCamelCase : Optional[Any] = output.prev_sample
lowerCamelCase : List[str] = torch.sum(torch.abs(__a ) )
lowerCamelCase : Tuple = torch.mean(torch.abs(__a ) )
if torch_device in ["cpu", "mps"]:
assert abs(result_sum.item() - 4.6_9_3_4e-0_7 ) < 1e-2
assert abs(result_mean.item() - 6.1_1_1_2e-1_0 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 4.6_9_3_4_2_8_6_5_0_1_7_0_9_7_2e-0_7 ) < 1e-2
assert abs(result_mean.item() - 0.00_02 ) < 1e-3
def a__ ( self: Any )-> Any:
if torch_device == "mps":
return
lowerCamelCase : Dict = self.scheduler_classes[0]
lowerCamelCase : Dict = self.get_scheduler_config()
lowerCamelCase : int = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps )
lowerCamelCase : List[Any] = self.dummy_model()
lowerCamelCase : Optional[Any] = self.dummy_sample_deter * scheduler.init_noise_sigma
lowerCamelCase : Optional[int] = sample.to(__a )
for i, t in enumerate(scheduler.timesteps ):
lowerCamelCase : Dict = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[Any] = model(__a , __a )
lowerCamelCase : Tuple = scheduler.step(__a , __a , __a )
lowerCamelCase : str = output.prev_sample
lowerCamelCase : Tuple = torch.sum(torch.abs(__a ) )
lowerCamelCase : Tuple = torch.mean(torch.abs(__a ) )
if torch_device in ["cpu", "mps"]:
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
def a__ ( self: Optional[Any] )-> List[Any]:
if torch_device == "mps":
return
lowerCamelCase : Any = self.scheduler_classes[0]
lowerCamelCase : Union[str, Any] = self.get_scheduler_config()
lowerCamelCase : Optional[Any] = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps , device=__a )
lowerCamelCase : Union[str, Any] = self.dummy_model()
lowerCamelCase : List[str] = self.dummy_sample_deter.to(__a ) * scheduler.init_noise_sigma
for t in scheduler.timesteps:
lowerCamelCase : Union[str, Any] = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[int] = model(__a , __a )
lowerCamelCase : int = scheduler.step(__a , __a , __a )
lowerCamelCase : int = output.prev_sample
lowerCamelCase : Union[str, Any] = torch.sum(torch.abs(__a ) )
lowerCamelCase : int = torch.mean(torch.abs(__a ) )
if str(__a ).startswith("""cpu""" ):
# The following sum varies between 148 and 156 on mps. Why?
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
| 42
| 0
|
"""simple docstring"""
import os
from shutil import copyfile
from typing import List, Optional, Tuple
from ...tokenization_utils import AddedToken
from ...tokenization_utils_fast import PreTrainedTokenizerFast
from ...utils import is_sentencepiece_available, logging
if is_sentencepiece_available():
from .tokenization_camembert import CamembertTokenizer
else:
__lowerCamelCase :Any = None
__lowerCamelCase :Union[str, Any] = logging.get_logger(__name__)
__lowerCamelCase :str = {'vocab_file': 'sentencepiece.bpe.model', 'tokenizer_file': 'tokenizer.json'}
__lowerCamelCase :List[str] = {
'vocab_file': {
'camembert-base': 'https://huggingface.co/camembert-base/resolve/main/sentencepiece.bpe.model',
},
'tokenizer_file': {
'camembert-base': 'https://huggingface.co/camembert-base/resolve/main/tokenizer.json',
},
}
__lowerCamelCase :List[str] = {
'camembert-base': 512,
}
__lowerCamelCase :Tuple = '▁'
class A__ ( __SCREAMING_SNAKE_CASE):
"""simple docstring"""
snake_case__ : int =VOCAB_FILES_NAMES
snake_case__ : Union[str, Any] =PRETRAINED_VOCAB_FILES_MAP
snake_case__ : Optional[Any] =PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
snake_case__ : Dict =['input_ids', 'attention_mask']
snake_case__ : int =CamembertTokenizer
def __init__( self: List[str] , __a: Optional[Any]=None , __a: Any=None , __a: Optional[Any]="<s>" , __a: Tuple="</s>" , __a: int="</s>" , __a: Union[str, Any]="<s>" , __a: Tuple="<unk>" , __a: Union[str, Any]="<pad>" , __a: List[Any]="<mask>" , __a: Tuple=["<s>NOTUSED", "</s>NOTUSED"] , **__a: str , )-> Optional[int]:
# Mask token behave like a normal word, i.e. include the space before it
lowerCamelCase : Dict = AddedToken(_a , lstrip=_a , rstrip=_a ) if isinstance(_a , _a ) else mask_token
super().__init__(
_a , tokenizer_file=_a , bos_token=_a , eos_token=_a , sep_token=_a , cls_token=_a , unk_token=_a , pad_token=_a , mask_token=_a , additional_special_tokens=_a , **_a , )
lowerCamelCase : int = vocab_file
lowerCamelCase : int = False if not self.vocab_file else True
def a__ ( self: Optional[int] , __a: Tuple , __a: Tuple = None )-> int:
if token_ids_a is None:
return [self.cls_token_id] + token_ids_a + [self.sep_token_id]
lowerCamelCase : str = [self.cls_token_id]
lowerCamelCase : Any = [self.sep_token_id]
return cls + token_ids_a + sep + sep + token_ids_a + sep
def a__ ( self: Tuple , __a: int , __a: str = None )-> int:
lowerCamelCase : Tuple = [self.sep_token_id]
lowerCamelCase : Tuple = [self.cls_token_id]
if token_ids_a is None:
return len(cls + token_ids_a + sep ) * [0]
return len(cls + token_ids_a + sep + sep + token_ids_a + sep ) * [0]
def a__ ( self: Union[str, Any] , __a: str , __a: List[str] = None )-> Union[str, Any]:
if not self.can_save_slow_tokenizer:
raise ValueError(
"""Your fast tokenizer does not have the necessary information to save the vocabulary for a slow """
"""tokenizer.""" )
if not os.path.isdir(_a ):
logger.error(f'Vocabulary path ({save_directory}) should be a directory' )
return
lowerCamelCase : Tuple = os.path.join(
_a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""vocab_file"""] )
if os.path.abspath(self.vocab_file ) != os.path.abspath(_a ):
copyfile(self.vocab_file , _a )
return (out_vocab_file,)
| 720
|
"""simple docstring"""
import gc
import random
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTextModelWithProjection, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DiffusionPipeline,
EulerDiscreteScheduler,
StableDiffusionXLImgaImgPipeline,
UNetaDConditionModel,
)
from diffusers.utils import floats_tensor, slow, torch_device
from diffusers.utils.testing_utils import enable_full_determinism, require_torch_gpu
from ..pipeline_params import (
IMAGE_TO_IMAGE_IMAGE_PARAMS,
TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS,
TEXT_GUIDED_IMAGE_VARIATION_PARAMS,
)
from ..test_pipelines_common import PipelineLatentTesterMixin, PipelineTesterMixin
enable_full_determinism()
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : str =StableDiffusionXLImgaImgPipeline
snake_case__ : Any =TEXT_GUIDED_IMAGE_VARIATION_PARAMS - {'''height''', '''width'''}
snake_case__ : Optional[int] =PipelineTesterMixin.required_optional_params - {'''latents'''}
snake_case__ : Dict =TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS
snake_case__ : Tuple =IMAGE_TO_IMAGE_IMAGE_PARAMS
snake_case__ : List[str] =IMAGE_TO_IMAGE_IMAGE_PARAMS
def a__ ( self: List[str] )-> int:
torch.manual_seed(0 )
lowerCamelCase : Any = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=2 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , attention_head_dim=(2, 4) , use_linear_projection=__a , addition_embed_type="""text_time""" , addition_time_embed_dim=8 , transformer_layers_per_block=(1, 2) , projection_class_embeddings_input_dim=80 , cross_attention_dim=64 , )
lowerCamelCase : Any = EulerDiscreteScheduler(
beta_start=0.0_00_85 , beta_end=0.0_12 , steps_offset=1 , beta_schedule="""scaled_linear""" , timestep_spacing="""leading""" , )
torch.manual_seed(0 )
lowerCamelCase : Any = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
lowerCamelCase : Optional[Any] = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-0_5 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1_000 , hidden_act="""gelu""" , projection_dim=32 , )
lowerCamelCase : Dict = CLIPTextModel(__a )
lowerCamelCase : Union[str, Any] = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" , local_files_only=__a )
lowerCamelCase : Dict = CLIPTextModelWithProjection(__a )
lowerCamelCase : Optional[int] = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" , local_files_only=__a )
lowerCamelCase : str = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""text_encoder_2""": text_encoder_a,
"""tokenizer_2""": tokenizer_a,
# "safety_checker": None,
# "feature_extractor": None,
}
return components
def a__ ( self: Any , __a: str , __a: Tuple=0 )-> Union[str, Any]:
lowerCamelCase : List[Any] = floats_tensor((1, 3, 32, 32) , rng=random.Random(__a ) ).to(__a )
lowerCamelCase : Any = image / 2 + 0.5
if str(__a ).startswith("""mps""" ):
lowerCamelCase : Dict = torch.manual_seed(__a )
else:
lowerCamelCase : Tuple = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : Tuple = {
"""prompt""": """A painting of a squirrel eating a burger""",
"""image""": image,
"""generator""": generator,
"""num_inference_steps""": 2,
"""guidance_scale""": 5.0,
"""output_type""": """numpy""",
"""strength""": 0.75,
}
return inputs
def a__ ( self: Dict )-> Optional[Any]:
lowerCamelCase : Any = """cpu""" # ensure determinism for the device-dependent torch.Generator
lowerCamelCase : Union[str, Any] = self.get_dummy_components()
lowerCamelCase : Optional[int] = StableDiffusionXLImgaImgPipeline(**__a )
lowerCamelCase : int = sd_pipe.to(__a )
sd_pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Optional[Any] = self.get_dummy_inputs(__a )
lowerCamelCase : Optional[int] = sd_pipe(**__a ).images
lowerCamelCase : Tuple = image[0, -3:, -3:, -1]
assert image.shape == (1, 32, 32, 3)
lowerCamelCase : Any = np.array([0.46_56, 0.48_40, 0.44_39, 0.66_98, 0.55_74, 0.45_24, 0.57_99, 0.59_43, 0.51_65] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-2
def a__ ( self: Optional[int] )-> Union[str, Any]:
super().test_attention_slicing_forward_pass(expected_max_diff=3e-3 )
def a__ ( self: Optional[Any] )-> str:
super().test_inference_batch_single_identical(expected_max_diff=3e-3 )
def a__ ( self: List[str] )-> Optional[Any]:
pass
def a__ ( self: List[Any] )-> Union[str, Any]:
lowerCamelCase : Tuple = self.get_dummy_components()
lowerCamelCase : Union[str, Any] = StableDiffusionXLImgaImgPipeline(**__a )
lowerCamelCase : str = sd_pipe.to(__a )
lowerCamelCase : Any = sd_pipe.to(__a )
sd_pipe.set_progress_bar_config(disable=__a )
# forward without prompt embeds
lowerCamelCase : Dict = self.get_dummy_inputs(__a )
lowerCamelCase : Any = 3 * ["""this is a negative prompt"""]
lowerCamelCase : Optional[int] = negative_prompt
lowerCamelCase : Tuple = 3 * [inputs["""prompt"""]]
lowerCamelCase : List[Any] = sd_pipe(**__a )
lowerCamelCase : Optional[int] = output.images[0, -3:, -3:, -1]
# forward with prompt embeds
lowerCamelCase : Tuple = self.get_dummy_inputs(__a )
lowerCamelCase : List[Any] = 3 * ["""this is a negative prompt"""]
lowerCamelCase : Tuple = 3 * [inputs.pop("""prompt""" )]
(
(
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) ,
) : Union[str, Any] = sd_pipe.encode_prompt(__a , negative_prompt=__a )
lowerCamelCase : int = sd_pipe(
**__a , prompt_embeds=__a , negative_prompt_embeds=__a , pooled_prompt_embeds=__a , negative_pooled_prompt_embeds=__a , )
lowerCamelCase : Union[str, Any] = output.images[0, -3:, -3:, -1]
# make sure that it's equal
assert np.abs(image_slice_a.flatten() - image_slice_a.flatten() ).max() < 1e-4
@slow
@require_torch_gpu
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: Dict )-> str:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: Union[str, Any] , __a: Any , __a: Any="cpu" , __a: str=torch.floataa , __a: Any=0 )-> Optional[Any]:
lowerCamelCase : Optional[Any] = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : List[Any] = np.random.RandomState(__a ).standard_normal((1, 4, 64, 64) )
lowerCamelCase : List[str] = torch.from_numpy(__a ).to(device=__a , dtype=__a )
lowerCamelCase : int = {
"""prompt""": """a photograph of an astronaut riding a horse""",
"""latents""": latents,
"""generator""": generator,
"""num_inference_steps""": 3,
"""guidance_scale""": 7.5,
"""output_type""": """numpy""",
}
return inputs
def a__ ( self: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = DiffusionPipeline.from_pretrained("""stabilityai/stable-diffusion-2-base""" )
pipe.to(__a )
pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Optional[int] = self.get_inputs(__a )
lowerCamelCase : Optional[Any] = pipe(**__a ).images
lowerCamelCase : Dict = image[0, -3:, -3:, -1].flatten()
assert image.shape == (1, 512, 512, 3)
lowerCamelCase : List[str] = np.array([0.4_94_93, 0.4_78_96, 0.4_07_98, 0.5_42_14, 0.5_32_12, 0.4_82_02, 0.4_76_56, 0.4_63_29, 0.4_85_06] )
assert np.abs(image_slice - expected_slice ).max() < 7e-3
| 42
| 0
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_flax_available,
is_tf_available,
is_torch_available,
is_vision_available,
)
__lowerCamelCase :Union[str, Any] = {"""configuration_vit""": ["""VIT_PRETRAINED_CONFIG_ARCHIVE_MAP""", """ViTConfig""", """ViTOnnxConfig"""]}
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :str = ["""ViTFeatureExtractor"""]
__lowerCamelCase :Dict = ["""ViTImageProcessor"""]
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :str = [
"""VIT_PRETRAINED_MODEL_ARCHIVE_LIST""",
"""ViTForImageClassification""",
"""ViTForMaskedImageModeling""",
"""ViTModel""",
"""ViTPreTrainedModel""",
]
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Dict = [
"""TFViTForImageClassification""",
"""TFViTModel""",
"""TFViTPreTrainedModel""",
]
try:
if not is_flax_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :List[str] = [
"""FlaxViTForImageClassification""",
"""FlaxViTModel""",
"""FlaxViTPreTrainedModel""",
]
if TYPE_CHECKING:
from .configuration_vit import VIT_PRETRAINED_CONFIG_ARCHIVE_MAP, ViTConfig, ViTOnnxConfig
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .feature_extraction_vit import ViTFeatureExtractor
from .image_processing_vit import ViTImageProcessor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_vit import (
VIT_PRETRAINED_MODEL_ARCHIVE_LIST,
ViTForImageClassification,
ViTForMaskedImageModeling,
ViTModel,
ViTPreTrainedModel,
)
try:
if not is_tf_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_tf_vit import TFViTForImageClassification, TFViTModel, TFViTPreTrainedModel
try:
if not is_flax_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_flax_vit import FlaxViTForImageClassification, FlaxViTModel, FlaxViTPreTrainedModel
else:
import sys
__lowerCamelCase :Optional[int] = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 721
|
"""simple docstring"""
import unittest
from pathlib import Path
from tempfile import NamedTemporaryFile, TemporaryDirectory
from transformers import BertConfig, BertTokenizerFast, FeatureExtractionPipeline
from transformers.convert_graph_to_onnx import (
convert,
ensure_valid_input,
generate_identified_filename,
infer_shapes,
quantize,
)
from transformers.testing_utils import require_tf, require_tokenizers, require_torch, slow
class A__ :
"""simple docstring"""
def a__ ( self: Optional[int] , __a: Optional[int] , __a: Tuple , __a: Optional[int] )-> List[str]:
return None
class A__ :
"""simple docstring"""
def a__ ( self: Optional[int] , __a: Tuple , __a: str , __a: str , __a: str )-> Tuple:
return None
class A__ ( unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[Any] =[
# (model_name, model_kwargs)
('''bert-base-cased''', {}),
('''gpt2''', {'''use_cache''': False}), # We don't support exporting GPT2 past keys anymore
]
@require_tf
@slow
def a__ ( self: Optional[Any] )-> int:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
self._test_export(__a , """tf""" , 12 , **__a )
@require_torch
@slow
def a__ ( self: str )-> int:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
self._test_export(__a , """pt""" , 12 , **__a )
@require_torch
@slow
def a__ ( self: Union[str, Any] )-> Dict:
from transformers import BertModel
lowerCamelCase : int = ["""[UNK]""", """[SEP]""", """[CLS]""", """[PAD]""", """[MASK]""", """some""", """other""", """words"""]
with NamedTemporaryFile(mode="""w+t""" ) as vocab_file:
vocab_file.write("""\n""".join(__a ) )
vocab_file.flush()
lowerCamelCase : Dict = BertTokenizerFast(vocab_file.name )
with TemporaryDirectory() as bert_save_dir:
lowerCamelCase : List[str] = BertModel(BertConfig(vocab_size=len(__a ) ) )
model.save_pretrained(__a )
self._test_export(__a , """pt""" , 12 , __a )
@require_tf
@slow
def a__ ( self: Optional[Any] )-> Optional[int]:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
lowerCamelCase : Optional[int] = self._test_export(__a , """tf""" , 12 , **__a )
lowerCamelCase : Tuple = quantize(Path(__a ) )
# Ensure the actual quantized model is not bigger than the original one
if quantized_path.stat().st_size >= Path(__a ).stat().st_size:
self.fail("""Quantized model is bigger than initial ONNX model""" )
@require_torch
@slow
def a__ ( self: Any )-> Optional[int]:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
lowerCamelCase : Any = self._test_export(__a , """pt""" , 12 , **__a )
lowerCamelCase : Dict = quantize(__a )
# Ensure the actual quantized model is not bigger than the original one
if quantized_path.stat().st_size >= Path(__a ).stat().st_size:
self.fail("""Quantized model is bigger than initial ONNX model""" )
def a__ ( self: List[Any] , __a: Optional[Any] , __a: List[Any] , __a: Union[str, Any] , __a: Optional[Any]=None , **__a: Optional[int] )-> Any:
try:
# Compute path
with TemporaryDirectory() as tempdir:
lowerCamelCase : Optional[Any] = Path(__a ).joinpath("""model.onnx""" )
# Remove folder if exists
if path.parent.exists():
path.parent.rmdir()
# Export
convert(__a , __a , __a , __a , __a , **__a )
return path
except Exception as e:
self.fail(__a )
@require_torch
@require_tokenizers
@slow
def a__ ( self: Tuple )-> Dict:
from transformers import BertModel
lowerCamelCase : int = BertModel(BertConfig.from_pretrained("""lysandre/tiny-bert-random""" ) )
lowerCamelCase : List[Any] = BertTokenizerFast.from_pretrained("""lysandre/tiny-bert-random""" )
self._test_infer_dynamic_axis(__a , __a , """pt""" )
@require_tf
@require_tokenizers
@slow
def a__ ( self: Optional[Any] )-> List[Any]:
from transformers import TFBertModel
lowerCamelCase : Union[str, Any] = TFBertModel(BertConfig.from_pretrained("""lysandre/tiny-bert-random""" ) )
lowerCamelCase : str = BertTokenizerFast.from_pretrained("""lysandre/tiny-bert-random""" )
self._test_infer_dynamic_axis(__a , __a , """tf""" )
def a__ ( self: List[str] , __a: str , __a: Optional[Any] , __a: str )-> List[Any]:
lowerCamelCase : List[str] = FeatureExtractionPipeline(__a , __a )
lowerCamelCase : List[str] = ["""input_ids""", """token_type_ids""", """attention_mask""", """output_0""", """output_1"""]
lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase : Tuple = infer_shapes(__a , __a )
# Assert all variables are present
self.assertEqual(len(__a ) , len(__a ) )
self.assertTrue(all(var_name in shapes for var_name in variable_names ) )
self.assertSequenceEqual(variable_names[:3] , __a )
self.assertSequenceEqual(variable_names[3:] , __a )
# Assert inputs are {0: batch, 1: sequence}
for var_name in ["input_ids", "token_type_ids", "attention_mask"]:
self.assertDictEqual(shapes[var_name] , {0: """batch""", 1: """sequence"""} )
# Assert outputs are {0: batch, 1: sequence} and {0: batch}
self.assertDictEqual(shapes["""output_0"""] , {0: """batch""", 1: """sequence"""} )
self.assertDictEqual(shapes["""output_1"""] , {0: """batch"""} )
def a__ ( self: List[Any] )-> int:
lowerCamelCase : List[str] = ["""input_ids""", """attention_mask""", """token_type_ids"""]
lowerCamelCase : str = {"""input_ids""": [1, 2, 3, 4], """attention_mask""": [0, 0, 0, 0], """token_type_ids""": [1, 1, 1, 1]}
lowerCamelCase , lowerCamelCase : List[Any] = ensure_valid_input(FuncContiguousArgs() , __a , __a )
# Should have exactly the same number of args (all are valid)
self.assertEqual(len(__a ) , 3 )
# Should have exactly the same input names
self.assertEqual(set(__a ) , set(__a ) )
# Parameter should be reordered according to their respective place in the function:
# (input_ids, token_type_ids, attention_mask)
self.assertEqual(__a , (tokens["""input_ids"""], tokens["""token_type_ids"""], tokens["""attention_mask"""]) )
# Generated args are interleaved with another args (for instance parameter "past" in GPT2)
lowerCamelCase , lowerCamelCase : List[Any] = ensure_valid_input(FuncNonContiguousArgs() , __a , __a )
# Should have exactly the one arg (all before the one not provided "some_other_args")
self.assertEqual(len(__a ) , 1 )
self.assertEqual(len(__a ) , 1 )
# Should have only "input_ids"
self.assertEqual(inputs_args[0] , tokens["""input_ids"""] )
self.assertEqual(ordered_input_names[0] , """input_ids""" )
def a__ ( self: Tuple )-> Tuple:
lowerCamelCase : Optional[int] = generate_identified_filename(Path("""/home/something/my_fake_model.onnx""" ) , """-test""" )
self.assertEqual("""/home/something/my_fake_model-test.onnx""" , generated.as_posix() )
| 42
| 0
|
"""simple docstring"""
from urllib.parse import quote
import pytest
from datasets.utils.hub import hf_hub_url
@pytest.mark.parametrize("""repo_id""" , ["""canonical_dataset_name""", """org-name/dataset-name"""] )
@pytest.mark.parametrize("""path""" , ["""filename.csv""", """filename with blanks.csv"""] )
@pytest.mark.parametrize("""revision""" , [None, """v2"""] )
def snake_case ( UpperCamelCase__ : List[Any] , UpperCamelCase__ : Optional[int] , UpperCamelCase__ : str ) -> Union[str, Any]:
lowerCamelCase : Optional[int] = hf_hub_url(repo_id=UpperCamelCase__ , path=UpperCamelCase__ , revision=UpperCamelCase__ )
assert url == F'https://huggingface.co/datasets/{repo_id}/resolve/{revision or "main"}/{quote(UpperCamelCase__ )}'
| 700
|
"""simple docstring"""
import unittest
from knapsack import greedy_knapsack as kp
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: Optional[int] )-> Union[str, Any]:
lowerCamelCase : Tuple = [10, 20, 30, 40, 50, 60]
lowerCamelCase : Union[str, Any] = [2, 4, 6, 8, 10, 12]
lowerCamelCase : Union[str, Any] = 100
self.assertEqual(kp.calc_profit(__a , __a , __a ) , 210 )
def a__ ( self: str )-> str:
self.assertRaisesRegex(__a , """max_weight must greater than zero.""" )
def a__ ( self: str )-> List[Any]:
self.assertRaisesRegex(__a , """Weight can not be negative.""" )
def a__ ( self: Any )-> Dict:
self.assertRaisesRegex(__a , """Profit can not be negative.""" )
def a__ ( self: Optional[Any] )-> List[Any]:
self.assertRaisesRegex(__a , """max_weight must greater than zero.""" )
def a__ ( self: Optional[Any] )-> Tuple:
self.assertRaisesRegex(
__a , """The length of profit and weight must be same.""" )
if __name__ == "__main__":
unittest.main()
| 42
| 0
|
"""simple docstring"""
def snake_case ( UpperCamelCase__ : List[Any] , UpperCamelCase__ : Optional[int] , UpperCamelCase__ : int ) -> List[Any]:
if n == 0:
return 1
elif n % 2 == 1:
return (binary_exponentiation(SCREAMING_SNAKE_CASE_ , n - 1 , SCREAMING_SNAKE_CASE_ ) * a) % mod
else:
lowerCamelCase : List[str] = binary_exponentiation(SCREAMING_SNAKE_CASE_ , n / 2 , SCREAMING_SNAKE_CASE_ )
return (b * b) % mod
# a prime number
__lowerCamelCase :Tuple = 701
__lowerCamelCase :str = 1_000_000_000
__lowerCamelCase :Any = 10
# using binary exponentiation function, O(log(p)):
print((a / b) % p == (a * binary_exponentiation(b, p - 2, p)) % p)
print((a / b) % p == (a * b ** (p - 2)) % p)
| 701
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_flax_available,
is_tf_available,
is_tokenizers_available,
is_torch_available,
is_vision_available,
)
__lowerCamelCase :List[str] = {
'configuration_owlvit': [
'OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP',
'OwlViTConfig',
'OwlViTOnnxConfig',
'OwlViTTextConfig',
'OwlViTVisionConfig',
],
'processing_owlvit': ['OwlViTProcessor'],
}
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[int] = ['OwlViTFeatureExtractor']
__lowerCamelCase :List[str] = ['OwlViTImageProcessor']
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[Any] = [
'OWLVIT_PRETRAINED_MODEL_ARCHIVE_LIST',
'OwlViTModel',
'OwlViTPreTrainedModel',
'OwlViTTextModel',
'OwlViTVisionModel',
'OwlViTForObjectDetection',
]
if TYPE_CHECKING:
from .configuration_owlvit import (
OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP,
OwlViTConfig,
OwlViTOnnxConfig,
OwlViTTextConfig,
OwlViTVisionConfig,
)
from .processing_owlvit import OwlViTProcessor
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .feature_extraction_owlvit import OwlViTFeatureExtractor
from .image_processing_owlvit import OwlViTImageProcessor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_owlvit import (
OWLVIT_PRETRAINED_MODEL_ARCHIVE_LIST,
OwlViTForObjectDetection,
OwlViTModel,
OwlViTPreTrainedModel,
OwlViTTextModel,
OwlViTVisionModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :List[str] = logging.get_logger(__name__)
__lowerCamelCase :str = {'openai-gpt': 'https://huggingface.co/openai-gpt/resolve/main/config.json'}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Dict ='''openai-gpt'''
snake_case__ : Optional[Any] ={
'''max_position_embeddings''': '''n_positions''',
'''hidden_size''': '''n_embd''',
'''num_attention_heads''': '''n_head''',
'''num_hidden_layers''': '''n_layer''',
}
def __init__( self: str , __a: List[str]=40_478 , __a: Union[str, Any]=512 , __a: List[str]=768 , __a: str=12 , __a: Dict=12 , __a: str="gelu" , __a: Dict=0.1 , __a: Any=0.1 , __a: Optional[int]=0.1 , __a: str=1e-5 , __a: Dict=0.02 , __a: Any="cls_index" , __a: Tuple=True , __a: Union[str, Any]=None , __a: Union[str, Any]=True , __a: Any=0.1 , **__a: int , )-> Tuple:
lowerCamelCase : Union[str, Any] = vocab_size
lowerCamelCase : str = n_positions
lowerCamelCase : str = n_embd
lowerCamelCase : Dict = n_layer
lowerCamelCase : Any = n_head
lowerCamelCase : int = afn
lowerCamelCase : Dict = resid_pdrop
lowerCamelCase : Union[str, Any] = embd_pdrop
lowerCamelCase : int = attn_pdrop
lowerCamelCase : List[Any] = layer_norm_epsilon
lowerCamelCase : Union[str, Any] = initializer_range
lowerCamelCase : Optional[Any] = summary_type
lowerCamelCase : int = summary_use_proj
lowerCamelCase : Tuple = summary_activation
lowerCamelCase : Union[str, Any] = summary_first_dropout
lowerCamelCase : str = summary_proj_to_labels
super().__init__(**__a )
| 702
|
"""simple docstring"""
import collections
import inspect
import unittest
from transformers import FocalNetConfig
from transformers.testing_utils import require_torch, require_vision, slow, torch_device
from transformers.utils import cached_property, is_torch_available, is_vision_available
from ...test_backbone_common import BackboneTesterMixin
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, _config_zero_init, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from torch import nn
from transformers import (
FocalNetBackbone,
FocalNetForImageClassification,
FocalNetForMaskedImageModeling,
FocalNetModel,
)
from transformers.models.focalnet.modeling_focalnet import FOCALNET_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import AutoImageProcessor
class A__ :
"""simple docstring"""
def __init__( self: List[Any] , __a: List[str] , __a: Optional[int]=13 , __a: List[str]=32 , __a: int=2 , __a: List[str]=3 , __a: Union[str, Any]=16 , __a: int=[32, 64, 128] , __a: Optional[Any]=[1, 2, 1] , __a: Optional[int]=[2, 2, 4] , __a: Tuple=2 , __a: Dict=2.0 , __a: List[str]=True , __a: Optional[Any]=0.0 , __a: Any=0.0 , __a: List[Any]=0.1 , __a: List[str]="gelu" , __a: Tuple=False , __a: Union[str, Any]=True , __a: Optional[int]=0.02 , __a: Tuple=1e-5 , __a: int=True , __a: List[Any]=None , __a: Optional[int]=True , __a: Dict=10 , __a: List[str]=8 , __a: Any=["stage1", "stage2"] , __a: Union[str, Any]=[1, 2] , )-> Dict:
lowerCamelCase : Dict = parent
lowerCamelCase : Optional[Any] = batch_size
lowerCamelCase : Union[str, Any] = image_size
lowerCamelCase : Optional[int] = patch_size
lowerCamelCase : Any = num_channels
lowerCamelCase : Any = embed_dim
lowerCamelCase : Dict = hidden_sizes
lowerCamelCase : List[Any] = depths
lowerCamelCase : Tuple = num_heads
lowerCamelCase : List[Any] = window_size
lowerCamelCase : str = mlp_ratio
lowerCamelCase : str = qkv_bias
lowerCamelCase : str = hidden_dropout_prob
lowerCamelCase : Dict = attention_probs_dropout_prob
lowerCamelCase : Tuple = drop_path_rate
lowerCamelCase : Dict = hidden_act
lowerCamelCase : Tuple = use_absolute_embeddings
lowerCamelCase : List[str] = patch_norm
lowerCamelCase : List[str] = layer_norm_eps
lowerCamelCase : str = initializer_range
lowerCamelCase : Tuple = is_training
lowerCamelCase : int = scope
lowerCamelCase : Union[str, Any] = use_labels
lowerCamelCase : List[str] = type_sequence_label_size
lowerCamelCase : str = encoder_stride
lowerCamelCase : List[str] = out_features
lowerCamelCase : Optional[int] = out_indices
def a__ ( self: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : str = None
if self.use_labels:
lowerCamelCase : List[str] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : str = self.get_config()
return config, pixel_values, labels
def a__ ( self: List[Any] )-> Optional[int]:
return FocalNetConfig(
image_size=self.image_size , patch_size=self.patch_size , num_channels=self.num_channels , embed_dim=self.embed_dim , hidden_sizes=self.hidden_sizes , depths=self.depths , num_heads=self.num_heads , window_size=self.window_size , mlp_ratio=self.mlp_ratio , qkv_bias=self.qkv_bias , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , drop_path_rate=self.drop_path_rate , hidden_act=self.hidden_act , use_absolute_embeddings=self.use_absolute_embeddings , path_norm=self.patch_norm , layer_norm_eps=self.layer_norm_eps , initializer_range=self.initializer_range , encoder_stride=self.encoder_stride , out_features=self.out_features , out_indices=self.out_indices , )
def a__ ( self: Tuple , __a: Optional[int] , __a: Optional[int] , __a: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = FocalNetModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Tuple = model(__a )
lowerCamelCase : Any = ((config.image_size // config.patch_size) ** 2) // (4 ** (len(config.depths ) - 1))
lowerCamelCase : List[Any] = int(config.embed_dim * 2 ** (len(config.depths ) - 1) )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, expected_seq_len, expected_dim) )
def a__ ( self: Optional[int] , __a: Dict , __a: Tuple , __a: List[Any] )-> int:
lowerCamelCase : List[Any] = FocalNetBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Optional[Any] = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , len(config.out_features ) )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.image_size, 8, 8] )
# verify channels
self.parent.assertEqual(len(model.channels ) , len(config.out_features ) )
self.parent.assertListEqual(model.channels , config.hidden_sizes[:-1] )
# verify backbone works with out_features=None
lowerCamelCase : Dict = None
lowerCamelCase : Dict = FocalNetBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , 1 )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.image_size * 2, 4, 4] )
# verify channels
self.parent.assertEqual(len(model.channels ) , 1 )
self.parent.assertListEqual(model.channels , [config.hidden_sizes[-1]] )
def a__ ( self: Optional[int] , __a: Optional[int] , __a: Optional[int] , __a: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = FocalNetForMaskedImageModeling(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[str] = model(__a )
self.parent.assertEqual(
result.reconstruction.shape , (self.batch_size, self.num_channels, self.image_size, self.image_size) )
# test greyscale images
lowerCamelCase : List[str] = 1
lowerCamelCase : Any = FocalNetForMaskedImageModeling(__a )
model.to(__a )
model.eval()
lowerCamelCase : str = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Tuple = model(__a )
self.parent.assertEqual(result.reconstruction.shape , (self.batch_size, 1, self.image_size, self.image_size) )
def a__ ( self: str , __a: Optional[Any] , __a: Optional[Any] , __a: Tuple )-> str:
lowerCamelCase : Optional[Any] = self.type_sequence_label_size
lowerCamelCase : Optional[Any] = FocalNetForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : List[str] = model(__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
# test greyscale images
lowerCamelCase : int = 1
lowerCamelCase : List[Any] = FocalNetForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Optional[Any] = model(__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
def a__ ( self: int )-> Optional[int]:
lowerCamelCase : str = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = config_and_inputs
lowerCamelCase : List[Any] = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =(
(
FocalNetModel,
FocalNetForImageClassification,
FocalNetForMaskedImageModeling,
FocalNetBackbone,
)
if is_torch_available()
else ()
)
snake_case__ : Optional[int] =(
{'''feature-extraction''': FocalNetModel, '''image-classification''': FocalNetForImageClassification}
if is_torch_available()
else {}
)
snake_case__ : Tuple =False
snake_case__ : Dict =False
snake_case__ : Dict =False
snake_case__ : Tuple =False
snake_case__ : Optional[int] =False
def a__ ( self: Union[str, Any] )-> Optional[int]:
lowerCamelCase : List[str] = FocalNetModelTester(self )
lowerCamelCase : Optional[Any] = ConfigTester(self , config_class=__a , embed_dim=37 , has_text_modality=__a )
def a__ ( self: List[str] )-> List[str]:
self.create_and_test_config_common_properties()
self.config_tester.create_and_test_config_to_json_string()
self.config_tester.create_and_test_config_to_json_file()
self.config_tester.create_and_test_config_from_and_save_pretrained()
self.config_tester.create_and_test_config_with_num_labels()
self.config_tester.check_config_can_be_init_without_params()
self.config_tester.check_config_arguments_init()
def a__ ( self: List[str] )-> Union[str, Any]:
return
def a__ ( self: Tuple )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: List[Any] )-> Dict:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_backbone(*__a )
def a__ ( self: List[Any] )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_masked_image_modeling(*__a )
def a__ ( self: List[str] )-> Dict:
lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_classification(*__a )
@unittest.skip(reason="""FocalNet does not use inputs_embeds""" )
def a__ ( self: Optional[Any] )-> str:
pass
@unittest.skip(reason="""FocalNet does not use feedforward chunking""" )
def a__ ( self: Optional[Any] )-> Dict:
pass
def a__ ( self: Optional[Any] )-> Dict:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : Any = model_class(__a )
self.assertIsInstance(model.get_input_embeddings() , (nn.Module) )
lowerCamelCase : Dict = model.get_output_embeddings()
self.assertTrue(x is None or isinstance(__a , nn.Linear ) )
def a__ ( self: Tuple )-> Optional[int]:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : int = model_class(__a )
lowerCamelCase : int = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Any = [*signature.parameters.keys()]
lowerCamelCase : List[Any] = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: str , __a: Union[str, Any] , __a: int , __a: Tuple , __a: List[str] )-> Union[str, Any]:
lowerCamelCase : List[Any] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : List[str] = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : List[str] = outputs.hidden_states
lowerCamelCase : Tuple = getattr(
self.model_tester , """expected_num_hidden_layers""" , len(self.model_tester.depths ) + 1 )
self.assertEqual(len(__a ) , __a )
# FocalNet has a different seq_length
lowerCamelCase : Tuple = (
config.patch_size
if isinstance(config.patch_size , collections.abc.Iterable )
else (config.patch_size, config.patch_size)
)
lowerCamelCase : Dict = (image_size[1] // patch_size[1]) * (image_size[0] // patch_size[0])
self.assertListEqual(
list(hidden_states[0].shape[-2:] ) , [num_patches, self.model_tester.embed_dim] , )
lowerCamelCase : Optional[Any] = outputs.reshaped_hidden_states
self.assertEqual(len(__a ) , __a )
lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = reshaped_hidden_states[0].shape
lowerCamelCase : Tuple = (
reshaped_hidden_states[0].view(__a , __a , height * width ).permute(0 , 2 , 1 )
)
self.assertListEqual(
list(reshaped_hidden_states.shape[-2:] ) , [num_patches, self.model_tester.embed_dim] , )
def a__ ( self: Any )-> Any:
lowerCamelCase , lowerCamelCase : str = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Union[str, Any] = (
self.model_tester.image_size
if isinstance(self.model_tester.image_size , collections.abc.Iterable )
else (self.model_tester.image_size, self.model_tester.image_size)
)
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : List[str] = True
self.check_hidden_states_output(__a , __a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : List[Any] = True
self.check_hidden_states_output(__a , __a , __a , __a )
def a__ ( self: str )-> Union[str, Any]:
lowerCamelCase , lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : List[str] = 3
lowerCamelCase : Any = (
self.model_tester.image_size
if isinstance(self.model_tester.image_size , collections.abc.Iterable )
else (self.model_tester.image_size, self.model_tester.image_size)
)
lowerCamelCase : Optional[int] = (
config.patch_size
if isinstance(config.patch_size , collections.abc.Iterable )
else (config.patch_size, config.patch_size)
)
lowerCamelCase : Optional[Any] = image_size[0] + patch_size[0] - (image_size[0] % patch_size[0])
lowerCamelCase : List[str] = image_size[1] + patch_size[1] - (image_size[1] % patch_size[1])
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : str = True
self.check_hidden_states_output(__a , __a , __a , (padded_height, padded_width) )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : Union[str, Any] = True
self.check_hidden_states_output(__a , __a , __a , (padded_height, padded_width) )
@slow
def a__ ( self: Optional[int] )-> List[Any]:
for model_name in FOCALNET_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : List[str] = FocalNetModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def a__ ( self: str )-> Any:
lowerCamelCase , lowerCamelCase : Optional[Any] = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : int = _config_zero_init(__a )
for model_class in self.all_model_classes:
lowerCamelCase : int = model_class(config=__a )
for name, param in model.named_parameters():
if "embeddings" not in name and param.requires_grad:
self.assertIn(
((param.data.mean() * 1e9).round() / 1e9).item() , [0.0, 1.0] , msg=f'Parameter {name} of model {model_class} seems not properly initialized' , )
@require_vision
@require_torch
class A__ ( unittest.TestCase):
"""simple docstring"""
@cached_property
def a__ ( self: Optional[int] )-> Optional[Any]:
# TODO update organization
return AutoImageProcessor.from_pretrained("""microsoft/focalnet-tiny""" ) if is_vision_available() else None
@slow
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : Tuple = FocalNetForImageClassification.from_pretrained("""microsoft/focalnet-tiny""" ).to(__a )
lowerCamelCase : Any = self.default_image_processor
lowerCamelCase : Union[str, Any] = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
lowerCamelCase : int = image_processor(images=__a , return_tensors="""pt""" ).to(__a )
# forward pass
with torch.no_grad():
lowerCamelCase : Any = model(**__a )
# verify the logits
lowerCamelCase : Tuple = torch.Size((1, 1_000) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : List[str] = torch.tensor([0.21_66, -0.43_68, 0.21_91] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3] , __a , atol=1e-4 ) )
self.assertTrue(outputs.logits.argmax(dim=-1 ).item() , 281 )
@require_torch
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : str =(FocalNetBackbone,) if is_torch_available() else ()
snake_case__ : Optional[int] =FocalNetConfig
snake_case__ : str =False
def a__ ( self: Union[str, Any] )-> Tuple:
lowerCamelCase : str = FocalNetModelTester(self )
| 42
| 0
|
"""simple docstring"""
import os
import re
import warnings
from shutil import copyfile
from typing import List, Optional, Tuple
from ...tokenization_utils_fast import PreTrainedTokenizerFast
from ...utils import is_sentencepiece_available, logging
if is_sentencepiece_available():
from .tokenization_ta import TaTokenizer
else:
__lowerCamelCase :List[str] = None
__lowerCamelCase :Tuple = logging.get_logger(__name__)
__lowerCamelCase :List[str] = {'vocab_file': 'spiece.model', 'tokenizer_file': 'tokenizer.json'}
__lowerCamelCase :List[str] = {
'vocab_file': {
't5-small': 'https://huggingface.co/t5-small/resolve/main/spiece.model',
't5-base': 'https://huggingface.co/t5-base/resolve/main/spiece.model',
't5-large': 'https://huggingface.co/t5-large/resolve/main/spiece.model',
't5-3b': 'https://huggingface.co/t5-3b/resolve/main/spiece.model',
't5-11b': 'https://huggingface.co/t5-11b/resolve/main/spiece.model',
},
'tokenizer_file': {
't5-small': 'https://huggingface.co/t5-small/resolve/main/tokenizer.json',
't5-base': 'https://huggingface.co/t5-base/resolve/main/tokenizer.json',
't5-large': 'https://huggingface.co/t5-large/resolve/main/tokenizer.json',
't5-3b': 'https://huggingface.co/t5-3b/resolve/main/tokenizer.json',
't5-11b': 'https://huggingface.co/t5-11b/resolve/main/tokenizer.json',
},
}
# TODO(PVP) - this should be removed in Transformers v5
__lowerCamelCase :Any = {
't5-small': 512,
't5-base': 512,
't5-large': 512,
't5-3b': 512,
't5-11b': 512,
}
class A__ ( __A):
"""simple docstring"""
snake_case__ : Optional[Any] =VOCAB_FILES_NAMES
snake_case__ : Optional[int] =PRETRAINED_VOCAB_FILES_MAP
snake_case__ : Any =PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
snake_case__ : List[str] =['''input_ids''', '''attention_mask''']
snake_case__ : str =TaTokenizer
snake_case__ : Tuple =[]
def __init__( self: Tuple , __a: str=None , __a: Union[str, Any]=None , __a: Optional[int]="</s>" , __a: Dict="<unk>" , __a: Dict="<pad>" , __a: Union[str, Any]=100 , __a: List[str]=None , **__a: int , )-> Any:
if extra_ids > 0 and additional_special_tokens is None:
lowerCamelCase : Tuple = [f'<extra_id_{i}>' for i in range(__a )]
elif extra_ids > 0 and additional_special_tokens is not None:
# Check that we have the right number of extra special tokens
lowerCamelCase : Dict = len(set(filter(lambda __a : bool("""extra_id_""" in str(__a ) ) , __a ) ) )
if extra_tokens != extra_ids:
raise ValueError(
f'Both extra_ids ({extra_ids}) and additional_special_tokens ({additional_special_tokens}) are'
""" provided to T5Tokenizer. In this case the additional_special_tokens must include the extra_ids"""
""" tokens""" )
super().__init__(
__a , tokenizer_file=__a , eos_token=__a , unk_token=__a , pad_token=__a , extra_ids=__a , additional_special_tokens=__a , **__a , )
lowerCamelCase : List[str] = vocab_file
lowerCamelCase : List[Any] = False if not self.vocab_file else True
lowerCamelCase : Optional[int] = extra_ids
@staticmethod
def a__ ( __a: str , __a: Tuple , __a: Dict )-> Union[str, Any]:
if pretrained_model_name_or_path in TaTokenizerFast.max_model_input_sizes:
lowerCamelCase : Optional[Any] = TaTokenizerFast.max_model_input_sizes[pretrained_model_name_or_path]
if init_max_model_length is not None and init_max_model_length != max_model_length:
return init_max_model_length
elif init_max_model_length is None:
warnings.warn(
"""This tokenizer was incorrectly instantiated with a model max length of"""
f' {deprecated_max_model_length} which will be corrected in Transformers v5.\nFor now, this'
""" behavior is kept to avoid breaking backwards compatibility when padding/encoding with"""
""" `truncation is True`.\n- Be aware that you SHOULD NOT rely on"""
f' {pretrained_model_name_or_path} automatically truncating your input to'
f' {deprecated_max_model_length} when padding/encoding.\n- If you want to encode/pad to sequences'
f' longer than {deprecated_max_model_length} you can either instantiate this tokenizer with'
""" `model_max_length` or pass `max_length` when encoding/padding.\n- To avoid this warning, please"""
""" instantiate this tokenizer with `model_max_length` set to your preferred value.""" , __a , )
return max_model_length
def a__ ( self: Optional[Any] , __a: List[str] , __a: Union[str, Any] = None )-> List[str]:
if not self.can_save_slow_tokenizer:
raise ValueError(
"""Your fast tokenizer does not have the necessary information to save the vocabulary for a slow """
"""tokenizer.""" )
if not os.path.isdir(__a ):
logger.error(f'Vocabulary path ({save_directory}) should be a directory' )
return
lowerCamelCase : Union[str, Any] = os.path.join(
__a , (filename_prefix + """-""" if filename_prefix else """""") + VOCAB_FILES_NAMES["""vocab_file"""] )
if os.path.abspath(self.vocab_file ) != os.path.abspath(__a ):
copyfile(self.vocab_file , __a )
logger.info(f'Copy vocab file to {out_vocab_file}' )
return (out_vocab_file,)
def a__ ( self: int , __a: Optional[int] , __a: Optional[int] = None )-> Dict:
lowerCamelCase : int = token_ids_a + [self.eos_token_id]
if token_ids_a is None:
return self.prefix_tokens + token_ids_a
else:
lowerCamelCase : Tuple = token_ids_a + [self.eos_token_id]
return self.prefix_tokens + token_ids_a + token_ids_a
def a__ ( self: Optional[Any] , __a: Optional[int] , __a: Dict = None )-> List[Any]:
lowerCamelCase : Dict = [self.eos_token_id]
if token_ids_a is None:
return len(token_ids_a + eos ) * [0]
return len(token_ids_a + eos + token_ids_a + eos ) * [0]
def a__ ( self: Optional[Any] )-> List[Any]:
return list(
set(filter(lambda __a : bool(re.search(r"""<extra_id_\d+>""" , __a ) ) is not None , self.additional_special_tokens ) ) )
def a__ ( self: List[Any] )-> str:
return [self.convert_tokens_to_ids(__a ) for token in self.get_sentinel_tokens()]
| 703
|
"""simple docstring"""
import os
def snake_case ( ) -> Optional[Any]:
with open(os.path.dirname(UpperCamelCase__ ) + """/grid.txt""" ) as f:
lowerCamelCase : int = [] # noqa: E741
for _ in range(20 ):
l.append([int(UpperCamelCase__ ) for x in f.readline().split()] )
lowerCamelCase : Union[str, Any] = 0
# right
for i in range(20 ):
for j in range(17 ):
lowerCamelCase : Dict = l[i][j] * l[i][j + 1] * l[i][j + 2] * l[i][j + 3]
if temp > maximum:
lowerCamelCase : Tuple = temp
# down
for i in range(17 ):
for j in range(20 ):
lowerCamelCase : Any = l[i][j] * l[i + 1][j] * l[i + 2][j] * l[i + 3][j]
if temp > maximum:
lowerCamelCase : Optional[Any] = temp
# diagonal 1
for i in range(17 ):
for j in range(17 ):
lowerCamelCase : List[Any] = l[i][j] * l[i + 1][j + 1] * l[i + 2][j + 2] * l[i + 3][j + 3]
if temp > maximum:
lowerCamelCase : List[str] = temp
# diagonal 2
for i in range(17 ):
for j in range(3 , 20 ):
lowerCamelCase : List[str] = l[i][j] * l[i + 1][j - 1] * l[i + 2][j - 2] * l[i + 3][j - 3]
if temp > maximum:
lowerCamelCase : List[Any] = temp
return maximum
if __name__ == "__main__":
print(solution())
| 42
| 0
|
"""simple docstring"""
import unittest
from transformers import DebertaVaTokenizer, DebertaVaTokenizerFast
from transformers.testing_utils import get_tests_dir, require_sentencepiece, require_tokenizers, slow
from ...test_tokenization_common import TokenizerTesterMixin
__lowerCamelCase :Any = get_tests_dir('fixtures/spiece.model')
@require_sentencepiece
@require_tokenizers
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Union[str, Any] =DebertaVaTokenizer
snake_case__ : Union[str, Any] =DebertaVaTokenizerFast
snake_case__ : int =True
snake_case__ : Optional[Any] =True
def a__ ( self: Optional[int] )-> Optional[Any]:
super().setUp()
# We have a SentencePiece fixture for testing
lowerCamelCase : Tuple = DebertaVaTokenizer(_A , unk_token="""<unk>""" )
tokenizer.save_pretrained(self.tmpdirname )
def a__ ( self: List[Any] , __a: Dict )-> Optional[Any]:
lowerCamelCase : Any = """this is a test"""
lowerCamelCase : int = """this is a test"""
return input_text, output_text
def a__ ( self: Optional[Any] )-> str:
lowerCamelCase : Tuple = """<pad>"""
lowerCamelCase : List[str] = 0
self.assertEqual(self.get_tokenizer()._convert_token_to_id(_A ) , _A )
self.assertEqual(self.get_tokenizer()._convert_id_to_token(_A ) , _A )
def a__ ( self: Tuple )-> int:
lowerCamelCase : Tuple = list(self.get_tokenizer().get_vocab().keys() )
self.assertEqual(vocab_keys[0] , """<pad>""" )
self.assertEqual(vocab_keys[1] , """<unk>""" )
self.assertEqual(vocab_keys[-1] , """[PAD]""" )
self.assertEqual(len(_A ) , 30_001 )
def a__ ( self: Dict )-> Dict:
self.assertEqual(self.get_tokenizer().vocab_size , 30_000 )
def a__ ( self: Optional[Any] )-> Optional[Any]:
# fmt: off
lowerCamelCase : Tuple = """ \tHeLLo!how \n Are yoU? """
lowerCamelCase : Any = ["""▁hello""", """!""", """how""", """▁are""", """▁you""", """?"""]
# fmt: on
lowerCamelCase : Tuple = DebertaVaTokenizer(_A , do_lower_case=_A )
lowerCamelCase : Any = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
lowerCamelCase : List[str] = DebertaVaTokenizerFast(_A , do_lower_case=_A )
lowerCamelCase : Dict = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
@unittest.skip("""There is an inconsistency between slow and fast tokenizer due to a bug in the fast one.""" )
def a__ ( self: Any )-> Any:
pass
@unittest.skip("""There is an inconsistency between slow and fast tokenizer due to a bug in the fast one.""" )
def a__ ( self: Any )-> List[str]:
pass
def a__ ( self: Tuple )-> str:
# fmt: off
lowerCamelCase : List[str] = """I was born in 92000, and this is falsé."""
lowerCamelCase : List[Any] = ["""▁""", """<unk>""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """▁""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """<unk>""", """▁""", """.""", ]
# fmt: on
lowerCamelCase : List[str] = DebertaVaTokenizer(_A , split_by_punct=_A )
lowerCamelCase : List[str] = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
lowerCamelCase : List[str] = DebertaVaTokenizerFast(_A , split_by_punct=_A )
lowerCamelCase : str = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
def a__ ( self: List[str] )-> List[Any]:
# fmt: off
lowerCamelCase : Any = """I was born in 92000, and this is falsé."""
lowerCamelCase : Any = ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """▁""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """<unk>""", """▁""", """.""", ]
# fmt: on
lowerCamelCase : Optional[Any] = DebertaVaTokenizer(_A , do_lower_case=_A , split_by_punct=_A )
lowerCamelCase : List[Any] = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
lowerCamelCase : Tuple = DebertaVaTokenizerFast(_A , do_lower_case=_A , split_by_punct=_A )
lowerCamelCase : int = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
def a__ ( self: Optional[Any] )-> str:
# fmt: off
lowerCamelCase : List[Any] = """I was born in 92000, and this is falsé."""
lowerCamelCase : Dict = ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """<unk>""", """.""", ]
# fmt: on
lowerCamelCase : List[str] = DebertaVaTokenizer(_A , do_lower_case=_A , split_by_punct=_A )
lowerCamelCase : Any = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
lowerCamelCase : Optional[int] = DebertaVaTokenizerFast(_A , do_lower_case=_A , split_by_punct=_A )
lowerCamelCase : str = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
def a__ ( self: int )-> int:
# fmt: off
lowerCamelCase : Any = """I was born in 92000, and this is falsé."""
lowerCamelCase : Any = ["""▁""", """<unk>""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """▁""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """<unk>""", """▁""", """.""", ]
# fmt: on
lowerCamelCase : int = DebertaVaTokenizer(_A , do_lower_case=_A , split_by_punct=_A )
lowerCamelCase : Optional[Any] = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
lowerCamelCase : Union[str, Any] = DebertaVaTokenizerFast(_A , do_lower_case=_A , split_by_punct=_A )
lowerCamelCase : Any = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
def a__ ( self: str )-> int:
# fmt: off
lowerCamelCase : Union[str, Any] = """ \tHeLLo!how \n Are yoU? """
lowerCamelCase : Optional[int] = ["""▁""", """<unk>""", """e""", """<unk>""", """o""", """!""", """how""", """▁""", """<unk>""", """re""", """▁yo""", """<unk>""", """?"""]
# fmt: on
lowerCamelCase : str = DebertaVaTokenizer(_A , do_lower_case=_A , split_by_punct=_A )
lowerCamelCase : str = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
lowerCamelCase : str = DebertaVaTokenizerFast(_A , do_lower_case=_A , split_by_punct=_A )
lowerCamelCase : Dict = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
def a__ ( self: Optional[Any] )-> Tuple:
lowerCamelCase : Union[str, Any] = self.get_tokenizer()
lowerCamelCase : List[Any] = self.get_rust_tokenizer()
lowerCamelCase : int = """I was born in 92000, and this is falsé."""
lowerCamelCase : Union[str, Any] = tokenizer.convert_ids_to_tokens(tokenizer.encode(_A , add_special_tokens=_A ) )
lowerCamelCase : Tuple = rust_tokenizer.convert_ids_to_tokens(rust_tokenizer.encode(_A , add_special_tokens=_A ) )
self.assertListEqual(_A , _A )
lowerCamelCase : str = tokenizer.encode(_A , add_special_tokens=_A )
lowerCamelCase : Optional[int] = rust_tokenizer.encode(_A , add_special_tokens=_A )
self.assertListEqual(_A , _A )
lowerCamelCase : Dict = self.get_rust_tokenizer()
lowerCamelCase : Any = tokenizer.encode(_A )
lowerCamelCase : int = rust_tokenizer.encode(_A )
self.assertListEqual(_A , _A )
def a__ ( self: str )-> List[Any]:
lowerCamelCase : Optional[int] = """This is a test"""
lowerCamelCase : Tuple = [13, 1, 4_398, 25, 21, 1_289]
lowerCamelCase : Dict = ["""▁""", """T""", """his""", """▁is""", """▁a""", """▁test"""]
lowerCamelCase : Tuple = ["""▁""", """<unk>""", """his""", """▁is""", """▁a""", """▁test"""]
lowerCamelCase : Optional[Any] = DebertaVaTokenizer(_A , keep_accents=_A )
lowerCamelCase : List[Any] = DebertaVaTokenizerFast(_A , keep_accents=_A )
lowerCamelCase : str = tokenizer.encode(_A , add_special_tokens=_A )
self.assertListEqual(_A , _A )
lowerCamelCase : List[str] = tokenizer.tokenize(_A )
self.assertListEqual(_A , _A )
lowerCamelCase : int = tokenizer.convert_ids_to_tokens(_A )
self.assertListEqual(_A , _A )
lowerCamelCase : Optional[int] = rust_tokenizer.encode(_A , add_special_tokens=_A )
self.assertListEqual(_A , _A )
lowerCamelCase : str = rust_tokenizer.tokenize(_A )
self.assertListEqual(_A , _A )
lowerCamelCase : List[Any] = rust_tokenizer.convert_ids_to_tokens(_A )
self.assertListEqual(_A , _A )
# fmt: off
lowerCamelCase : Optional[Any] = """I was born in 92000, and this is falsé."""
lowerCamelCase : Optional[Any] = [13, 1, 23, 386, 19, 561, 3_050, 15, 17, 48, 25, 8_256, 18, 1, 9]
lowerCamelCase : Optional[Any] = ["""▁""", """I""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """é""", """.""", ]
lowerCamelCase : Optional[int] = ["""▁""", """<unk>""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """<unk>""", """.""", ]
# fmt: on
lowerCamelCase : Tuple = tokenizer.encode(_A , add_special_tokens=_A )
self.assertListEqual(_A , _A )
lowerCamelCase : List[Any] = tokenizer.tokenize(_A )
self.assertListEqual(_A , _A )
lowerCamelCase : int = tokenizer.convert_ids_to_tokens(_A )
self.assertListEqual(_A , _A )
lowerCamelCase : List[Any] = rust_tokenizer.encode(_A , add_special_tokens=_A )
self.assertListEqual(_A , _A )
lowerCamelCase : Union[str, Any] = rust_tokenizer.tokenize(_A )
self.assertListEqual(_A , _A )
lowerCamelCase : Optional[Any] = rust_tokenizer.convert_ids_to_tokens(_A )
self.assertListEqual(_A , _A )
def a__ ( self: List[str] )-> Any:
lowerCamelCase : str = DebertaVaTokenizer(_A )
lowerCamelCase : Any = tokenizer.encode("""sequence builders""" )
lowerCamelCase : str = tokenizer.encode("""multi-sequence build""" )
lowerCamelCase : str = tokenizer.build_inputs_with_special_tokens(_A )
lowerCamelCase : Optional[int] = tokenizer.build_inputs_with_special_tokens(_A , _A )
self.assertEqual([tokenizer.cls_token_id] + text + [tokenizer.sep_token_id] , _A )
self.assertEqual(
[tokenizer.cls_token_id] + text + [tokenizer.sep_token_id] + text_a + [tokenizer.sep_token_id] , _A , )
@slow
def a__ ( self: List[str] )-> int:
# fmt: off
lowerCamelCase : Any = {"""input_ids""": [[1, 39_867, 36, 19_390, 486, 27, 35_052, 81_436, 18, 60_685, 1_225, 7, 35_052, 81_436, 18, 9_367, 16_899, 18, 15_937, 53, 594, 773, 18, 16_287, 30_465, 36, 15_937, 6, 41_139, 38, 36_979, 60_763, 191, 6, 34_132, 99, 6, 50_538, 390, 43_230, 6, 34_132, 2_779, 20_850, 14, 699, 1_072, 1_194, 36, 382, 10_901, 53, 7, 699, 1_072, 2_084, 36, 20_422, 630, 53, 19, 105, 3_049, 1_896, 1_053, 16_899, 1_506, 11, 37_978, 4_243, 7, 1_237, 31_869, 200, 16_566, 654, 6, 35_052, 81_436, 7, 55_630, 13_593, 4, 2], [1, 26, 15_011, 13, 667, 8, 1_053, 18, 23_611, 1_237, 72_356, 12_820, 34, 104_134, 1_209, 35, 13_313, 6_627, 21, 202, 347, 7, 164, 2_399, 11, 46, 4_485, 4, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 5, 1_232, 2_864, 15_785, 14_951, 105, 5, 8_581, 1_250, 4, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """token_type_ids""": [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """attention_mask""": [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]} # noqa: E501
# fmt: on
self.tokenizer_integration_test_util(
expected_encoding=_A , model_name="""microsoft/deberta-v2-xlarge""" , revision="""ad6e42c1532ddf3a15c39246b63f5559d558b670""" , )
| 704
|
"""simple docstring"""
import gc
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DDIMScheduler,
StableDiffusionAttendAndExcitePipeline,
UNetaDConditionModel,
)
from diffusers.utils import load_numpy, skip_mps, slow
from diffusers.utils.testing_utils import require_torch_gpu
from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
from ..test_pipelines_common import PipelineKarrasSchedulerTesterMixin, PipelineLatentTesterMixin, PipelineTesterMixin
__lowerCamelCase :Any = False
@skip_mps
class A__ ( __lowercase , __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[Any] =StableDiffusionAttendAndExcitePipeline
snake_case__ : Any =False
snake_case__ : Dict =TEXT_TO_IMAGE_PARAMS
snake_case__ : Any =TEXT_TO_IMAGE_BATCH_PARAMS.union({'''token_indices'''})
snake_case__ : Dict =TEXT_TO_IMAGE_IMAGE_PARAMS
snake_case__ : str =TEXT_TO_IMAGE_IMAGE_PARAMS
@classmethod
def a__ ( cls: Dict )-> Tuple:
super().setUpClass()
torch.use_deterministic_algorithms(__a )
@classmethod
def a__ ( cls: Union[str, Any] )-> Any:
super().tearDownClass()
torch.use_deterministic_algorithms(__a )
def a__ ( self: Tuple )-> Union[str, Any]:
torch.manual_seed(0 )
lowerCamelCase : str = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=1 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , cross_attention_dim=32 , attention_head_dim=(2, 4) , use_linear_projection=__a , )
lowerCamelCase : Union[str, Any] = DDIMScheduler(
beta_start=0.0_00_85 , beta_end=0.0_12 , beta_schedule="""scaled_linear""" , clip_sample=__a , set_alpha_to_one=__a , )
torch.manual_seed(0 )
lowerCamelCase : Union[str, Any] = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
lowerCamelCase : str = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-0_5 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1_000 , hidden_act="""gelu""" , projection_dim=512 , )
lowerCamelCase : Optional[int] = CLIPTextModel(__a )
lowerCamelCase : str = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" )
lowerCamelCase : List[str] = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""safety_checker""": None,
"""feature_extractor""": None,
}
return components
def a__ ( self: Tuple , __a: int , __a: Union[str, Any]=0 )-> Optional[Any]:
if str(__a ).startswith("""mps""" ):
lowerCamelCase : Tuple = torch.manual_seed(__a )
else:
lowerCamelCase : str = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : Dict = {
"""prompt""": """a cat and a frog""",
"""token_indices""": [2, 5],
"""generator""": generator,
"""num_inference_steps""": 1,
"""guidance_scale""": 6.0,
"""output_type""": """numpy""",
"""max_iter_to_alter""": 2,
"""thresholds""": {0: 0.7},
}
return inputs
def a__ ( self: Dict )-> str:
lowerCamelCase : Tuple = """cpu"""
lowerCamelCase : List[str] = self.get_dummy_components()
lowerCamelCase : List[Any] = self.pipeline_class(**__a )
pipe.to(__a )
pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Any = self.get_dummy_inputs(__a )
lowerCamelCase : Union[str, Any] = pipe(**__a ).images
lowerCamelCase : Tuple = image[0, -3:, -3:, -1]
self.assertEqual(image.shape , (1, 64, 64, 3) )
lowerCamelCase : Optional[Any] = np.array(
[0.63_90_53_64, 0.62_89_73_07, 0.48_59_90_17, 0.5_13_36_24, 0.5_55_00_48, 0.45_76_95_16, 0.50_32_69_73, 0.5_02_31_39, 0.45_38_44_96] )
lowerCamelCase : Optional[Any] = np.abs(image_slice.flatten() - expected_slice ).max()
self.assertLessEqual(__a , 1e-3 )
def a__ ( self: int )-> Optional[Any]:
super().test_cpu_offload_forward_pass(expected_max_diff=5e-4 )
def a__ ( self: Union[str, Any] )-> Optional[int]:
# NOTE: Larger batch sizes cause this test to timeout, only test on smaller batches
self._test_inference_batch_consistent(batch_sizes=[1, 2] )
def a__ ( self: Tuple )-> int:
self._test_inference_batch_single_identical(batch_size=2 , expected_max_diff=7e-4 )
def a__ ( self: Dict )-> List[Any]:
super().test_dict_tuple_outputs_equivalent(expected_max_difference=3e-3 )
def a__ ( self: Optional[int] )-> Dict:
super().test_pt_np_pil_outputs_equivalent(expected_max_diff=5e-4 )
def a__ ( self: Any )-> Tuple:
super().test_save_load_local(expected_max_difference=5e-4 )
def a__ ( self: str )-> str:
super().test_save_load_optional_components(expected_max_difference=4e-4 )
@require_torch_gpu
@slow
class A__ ( unittest.TestCase):
"""simple docstring"""
@classmethod
def a__ ( cls: Any )-> Tuple:
super().setUpClass()
torch.use_deterministic_algorithms(__a )
@classmethod
def a__ ( cls: Dict )-> Optional[int]:
super().tearDownClass()
torch.use_deterministic_algorithms(__a )
def a__ ( self: int )-> Optional[int]:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : List[Any] = torch.manual_seed(51 )
lowerCamelCase : List[str] = StableDiffusionAttendAndExcitePipeline.from_pretrained(
"""CompVis/stable-diffusion-v1-4""" , safety_checker=__a , torch_dtype=torch.floataa )
pipe.to("""cuda""" )
lowerCamelCase : Dict = """a painting of an elephant with glasses"""
lowerCamelCase : Any = [5, 7]
lowerCamelCase : Tuple = pipe(
prompt=__a , token_indices=__a , guidance_scale=7.5 , generator=__a , num_inference_steps=5 , max_iter_to_alter=5 , output_type="""numpy""" , ).images[0]
lowerCamelCase : Union[str, Any] = load_numpy(
"""https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/attend-and-excite/elephant_glasses.npy""" )
assert np.abs((expected_image - image).max() ) < 5e-1
| 42
| 0
|
"""simple docstring"""
from math import loga
def snake_case ( UpperCamelCase__ : Optional[int] ) -> List[str]:
if a < 0:
raise ValueError("""Input value must be a positive integer""" )
elif isinstance(lowerCAmelCase__ , lowerCAmelCase__ ):
raise TypeError("""Input value must be a 'int' type""" )
return 0 if (a == 0) else int(loga(a & -a ) )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 705
|
"""simple docstring"""
import unittest
from transformers import EsmConfig, is_torch_available
from transformers.testing_utils import TestCasePlus, require_torch, slow, torch_device
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, ids_tensor, random_attention_mask
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import EsmForMaskedLM, EsmForSequenceClassification, EsmForTokenClassification, EsmModel
from transformers.models.esm.modeling_esm import (
ESM_PRETRAINED_MODEL_ARCHIVE_LIST,
EsmEmbeddings,
create_position_ids_from_input_ids,
)
class A__ :
"""simple docstring"""
def __init__( self: List[str] , __a: List[str] , __a: Dict=13 , __a: Tuple=7 , __a: Dict=False , __a: str=True , __a: List[Any]=False , __a: Dict=True , __a: Any=33 , __a: Optional[Any]=32 , __a: List[Any]=5 , __a: Any=4 , __a: Dict=37 , __a: str="gelu" , __a: str=0.1 , __a: int=0.1 , __a: Optional[int]=512 , __a: List[Any]=16 , __a: int=2 , __a: int=0.02 , __a: Optional[int]=3 , __a: str=4 , __a: Tuple=None , )-> Tuple:
lowerCamelCase : Union[str, Any] = parent
lowerCamelCase : Tuple = batch_size
lowerCamelCase : Any = seq_length
lowerCamelCase : Any = is_training
lowerCamelCase : Tuple = use_input_mask
lowerCamelCase : int = use_token_type_ids
lowerCamelCase : List[str] = use_labels
lowerCamelCase : Optional[int] = vocab_size
lowerCamelCase : Tuple = hidden_size
lowerCamelCase : List[str] = num_hidden_layers
lowerCamelCase : Optional[int] = num_attention_heads
lowerCamelCase : Optional[Any] = intermediate_size
lowerCamelCase : Optional[Any] = hidden_act
lowerCamelCase : Union[str, Any] = hidden_dropout_prob
lowerCamelCase : Optional[Any] = attention_probs_dropout_prob
lowerCamelCase : Any = max_position_embeddings
lowerCamelCase : str = type_vocab_size
lowerCamelCase : List[Any] = type_sequence_label_size
lowerCamelCase : Optional[Any] = initializer_range
lowerCamelCase : Union[str, Any] = num_labels
lowerCamelCase : Optional[Any] = num_choices
lowerCamelCase : Any = scope
def a__ ( self: Optional[int] )-> List[Any]:
lowerCamelCase : List[Any] = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : Dict = None
if self.use_input_mask:
lowerCamelCase : Tuple = random_attention_mask([self.batch_size, self.seq_length] )
lowerCamelCase : Any = None
lowerCamelCase : int = None
lowerCamelCase : Union[str, Any] = None
if self.use_labels:
lowerCamelCase : Optional[Any] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : Optional[int] = ids_tensor([self.batch_size, self.seq_length] , self.num_labels )
lowerCamelCase : Union[str, Any] = ids_tensor([self.batch_size] , self.num_choices )
lowerCamelCase : List[str] = self.get_config()
return config, input_ids, input_mask, sequence_labels, token_labels, choice_labels
def a__ ( self: Tuple )-> Union[str, Any]:
return EsmConfig(
vocab_size=self.vocab_size , hidden_size=self.hidden_size , pad_token_id=1 , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , initializer_range=self.initializer_range , )
def a__ ( self: List[Any] , __a: List[str] , __a: str , __a: Tuple , __a: List[str] , __a: List[str] , __a: str )-> int:
lowerCamelCase : Optional[int] = EsmModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a , attention_mask=__a )
lowerCamelCase : str = model(__a )
lowerCamelCase : Optional[Any] = model(__a )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) )
self.parent.assertEqual(result.pooler_output.shape , (self.batch_size, self.hidden_size) )
def a__ ( self: int , __a: Union[str, Any] , __a: Optional[int] , __a: List[str] , __a: str , __a: List[str] , __a: Tuple )-> int:
lowerCamelCase : str = EsmForMaskedLM(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[Any] = model(__a , attention_mask=__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) )
def a__ ( self: List[str] , __a: List[Any] , __a: List[str] , __a: int , __a: Union[str, Any] , __a: List[Any] , __a: Tuple )-> List[str]:
lowerCamelCase : Tuple = self.num_labels
lowerCamelCase : Dict = EsmForTokenClassification(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a , attention_mask=__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) )
def a__ ( self: Optional[int] )-> Optional[int]:
lowerCamelCase : Any = self.prepare_config_and_inputs()
(
(
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) ,
) : Tuple = config_and_inputs
lowerCamelCase : List[Any] = {"""input_ids""": input_ids, """attention_mask""": input_mask}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Any =False
snake_case__ : Dict =(
(
EsmForMaskedLM,
EsmModel,
EsmForSequenceClassification,
EsmForTokenClassification,
)
if is_torch_available()
else ()
)
snake_case__ : Dict =()
snake_case__ : Optional[int] =(
{
'''feature-extraction''': EsmModel,
'''fill-mask''': EsmForMaskedLM,
'''text-classification''': EsmForSequenceClassification,
'''token-classification''': EsmForTokenClassification,
'''zero-shot''': EsmForSequenceClassification,
}
if is_torch_available()
else {}
)
snake_case__ : Any =True
def a__ ( self: Optional[int] )-> Optional[int]:
lowerCamelCase : Optional[Any] = EsmModelTester(self )
lowerCamelCase : Any = ConfigTester(self , config_class=__a , hidden_size=37 )
def a__ ( self: List[Any] )-> Optional[Any]:
self.config_tester.run_common_tests()
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: Tuple )-> Any:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
for type in ["absolute", "relative_key", "relative_key_query"]:
lowerCamelCase : Tuple = type
self.model_tester.create_and_check_model(*__a )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Optional[int] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_masked_lm(*__a )
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_token_classification(*__a )
@slow
def a__ ( self: Any )-> List[Any]:
for model_name in ESM_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : int = EsmModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def a__ ( self: str )-> List[str]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()[0]
lowerCamelCase : Union[str, Any] = EsmEmbeddings(config=__a )
lowerCamelCase : List[str] = torch.as_tensor([[12, 31, 13, model.padding_idx]] )
lowerCamelCase : Union[str, Any] = torch.as_tensor(
[
[
0 + model.padding_idx + 1,
1 + model.padding_idx + 1,
2 + model.padding_idx + 1,
model.padding_idx,
]
] )
lowerCamelCase : Optional[Any] = create_position_ids_from_input_ids(__a , model.padding_idx )
self.assertEqual(position_ids.shape , expected_positions.shape )
self.assertTrue(torch.all(torch.eq(__a , __a ) ) )
def a__ ( self: Optional[int] )-> int:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()[0]
lowerCamelCase : Any = EsmEmbeddings(config=__a )
lowerCamelCase : Dict = torch.empty(2 , 4 , 30 )
lowerCamelCase : List[Any] = [
0 + embeddings.padding_idx + 1,
1 + embeddings.padding_idx + 1,
2 + embeddings.padding_idx + 1,
3 + embeddings.padding_idx + 1,
]
lowerCamelCase : Any = torch.as_tensor([expected_single_positions, expected_single_positions] )
lowerCamelCase : List[str] = embeddings.create_position_ids_from_inputs_embeds(__a )
self.assertEqual(position_ids.shape , expected_positions.shape )
self.assertTrue(torch.all(torch.eq(__a , __a ) ) )
@unittest.skip("""Esm does not support embedding resizing""" )
def a__ ( self: Any )-> Optional[Any]:
pass
@unittest.skip("""Esm does not support embedding resizing""" )
def a__ ( self: Dict )-> Dict:
pass
@unittest.skip("""Will be fixed soon by reducing the size of the model used for common tests.""" )
def a__ ( self: List[str] )-> Dict:
pass
@require_torch
class A__ ( __lowercase):
"""simple docstring"""
@slow
def a__ ( self: Any )-> Union[str, Any]:
with torch.no_grad():
lowerCamelCase : Union[str, Any] = EsmForMaskedLM.from_pretrained("""facebook/esm2_t6_8M_UR50D""" )
model.eval()
lowerCamelCase : List[str] = torch.tensor([[0, 1, 2, 3, 4, 5]] )
lowerCamelCase : Tuple = model(__a )[0]
lowerCamelCase : Dict = 33
lowerCamelCase : List[str] = torch.Size((1, 6, vocab_size) )
self.assertEqual(output.shape , __a )
lowerCamelCase : Tuple = torch.tensor(
[[[8.92_15, -10.58_98, -6.46_71], [-6.39_67, -13.91_14, -1.12_12], [-7.78_12, -13.95_16, -3.74_06]]] )
self.assertTrue(torch.allclose(output[:, :3, :3] , __a , atol=1e-4 ) )
@slow
def a__ ( self: Dict )-> str:
with torch.no_grad():
lowerCamelCase : Any = EsmModel.from_pretrained("""facebook/esm2_t6_8M_UR50D""" )
model.eval()
lowerCamelCase : Optional[Any] = torch.tensor([[0, 6, 4, 13, 5, 4, 16, 12, 11, 7, 2]] )
lowerCamelCase : Any = model(__a )[0]
# compare the actual values for a slice.
lowerCamelCase : Tuple = torch.tensor(
[[[0.14_44, 0.54_13, 0.32_48], [0.30_34, 0.00_53, 0.31_08], [0.32_28, -0.24_99, 0.34_15]]] )
self.assertTrue(torch.allclose(output[:, :3, :3] , __a , atol=1e-4 ) )
| 42
| 0
|
"""simple docstring"""
import numpy as np
def snake_case ( UpperCamelCase__ : Any ) -> int:
return 1 / (1 + np.exp(-vector ))
def snake_case ( UpperCamelCase__ : Optional[int] ) -> Union[str, Any]:
return vector * sigmoid(lowerCAmelCase_ )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 706
|
"""simple docstring"""
import unittest
from transformers import AlbertTokenizer, AlbertTokenizerFast
from transformers.testing_utils import get_tests_dir, require_sentencepiece, require_tokenizers, slow
from ...test_tokenization_common import TokenizerTesterMixin
__lowerCamelCase :str = get_tests_dir('fixtures/spiece.model')
@require_sentencepiece
@require_tokenizers
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =AlbertTokenizer
snake_case__ : Optional[Any] =AlbertTokenizerFast
snake_case__ : Optional[int] =True
snake_case__ : Any =True
snake_case__ : Optional[int] =True
def a__ ( self: Dict )-> Optional[Any]:
super().setUp()
# We have a SentencePiece fixture for testing
lowerCamelCase : int = AlbertTokenizer(__a )
tokenizer.save_pretrained(self.tmpdirname )
def a__ ( self: Tuple , __a: Tuple )-> Union[str, Any]:
lowerCamelCase : List[str] = """this is a test"""
lowerCamelCase : int = """this is a test"""
return input_text, output_text
def a__ ( self: Any )-> List[Any]:
lowerCamelCase : int = """<pad>"""
lowerCamelCase : Optional[int] = 0
self.assertEqual(self.get_tokenizer()._convert_token_to_id(__a ) , __a )
self.assertEqual(self.get_tokenizer()._convert_id_to_token(__a ) , __a )
def a__ ( self: Tuple )-> str:
lowerCamelCase : Optional[int] = list(self.get_tokenizer().get_vocab().keys() )
self.assertEqual(vocab_keys[0] , """<pad>""" )
self.assertEqual(vocab_keys[1] , """<unk>""" )
self.assertEqual(vocab_keys[-1] , """▁eloquent""" )
self.assertEqual(len(__a ) , 30_000 )
def a__ ( self: List[str] )-> Any:
self.assertEqual(self.get_tokenizer().vocab_size , 30_000 )
def a__ ( self: Optional[Any] )-> Union[str, Any]:
if not self.test_rust_tokenizer:
return
lowerCamelCase : str = self.get_tokenizer()
lowerCamelCase : Tuple = self.get_rust_tokenizer()
lowerCamelCase : Union[str, Any] = """I was born in 92000, and this is falsé."""
lowerCamelCase : List[str] = tokenizer.tokenize(__a )
lowerCamelCase : Tuple = rust_tokenizer.tokenize(__a )
self.assertListEqual(__a , __a )
lowerCamelCase : Dict = tokenizer.encode(__a , add_special_tokens=__a )
lowerCamelCase : List[str] = rust_tokenizer.encode(__a , add_special_tokens=__a )
self.assertListEqual(__a , __a )
lowerCamelCase : Any = self.get_rust_tokenizer()
lowerCamelCase : List[str] = tokenizer.encode(__a )
lowerCamelCase : str = rust_tokenizer.encode(__a )
self.assertListEqual(__a , __a )
def a__ ( self: Tuple )-> List[Any]:
lowerCamelCase : List[str] = AlbertTokenizer(__a , keep_accents=__a )
lowerCamelCase : int = tokenizer.tokenize("""This is a test""" )
self.assertListEqual(__a , ["""▁this""", """▁is""", """▁a""", """▁test"""] )
self.assertListEqual(tokenizer.convert_tokens_to_ids(__a ) , [48, 25, 21, 1_289] )
lowerCamelCase : List[str] = tokenizer.tokenize("""I was born in 92000, and this is falsé.""" )
self.assertListEqual(
__a , ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """é""", """."""] )
lowerCamelCase : List[str] = tokenizer.convert_tokens_to_ids(__a )
self.assertListEqual(__a , [31, 23, 386, 19, 561, 3_050, 15, 17, 48, 25, 8_256, 18, 1, 9] )
lowerCamelCase : Tuple = tokenizer.convert_ids_to_tokens(__a )
self.assertListEqual(
__a , ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """<unk>""", """."""] , )
def a__ ( self: Tuple )-> str:
lowerCamelCase : str = AlbertTokenizer(__a )
lowerCamelCase : Union[str, Any] = tokenizer.encode("""sequence builders""" )
lowerCamelCase : List[Any] = tokenizer.encode("""multi-sequence build""" )
lowerCamelCase : Any = tokenizer.build_inputs_with_special_tokens(__a )
lowerCamelCase : str = tokenizer.build_inputs_with_special_tokens(__a , __a )
assert encoded_sentence == [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id]
assert encoded_pair == [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id] + text_a + [
tokenizer.sep_token_id
]
@slow
def a__ ( self: Any )-> Dict:
# fmt: off
lowerCamelCase : Optional[Any] = {"""attention_mask""": [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """input_ids""": [[2, 21_970, 13, 5, 6_092, 167, 28, 7_103, 2_153, 673, 8, 7_028, 12_051, 18, 17, 7_103, 2_153, 673, 8, 3_515, 18_684, 8, 4_461, 6, 1_927, 297, 8, 12_060, 2_607, 18, 13, 5, 4_461, 15, 10_538, 38, 8, 135, 15, 822, 58, 15, 993, 10_363, 15, 1_460, 8_005, 4_461, 15, 993, 255, 2_328, 9, 9, 9, 6, 26, 1_112, 816, 3_260, 13, 5, 103, 2_377, 6, 17, 1_112, 816, 2_782, 13, 5, 103, 10_641, 6, 29, 84, 2_512, 2_430, 782, 18_684, 2_761, 19, 808, 2_430, 2_556, 17, 855, 1_480, 9_477, 4_091, 128, 11_712, 15, 7_103, 2_153, 673, 17, 24_883, 9_990, 9, 3], [2, 11_502, 25, 1_006, 20, 782, 8, 11_809, 855, 1_732, 19_393, 18_667, 37, 367, 21_018, 69, 1_854, 34, 11_860, 19_124, 27, 156, 225, 17, 193, 4_141, 19, 65, 9_124, 9, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [2, 14, 2_231, 886, 2_385, 17_659, 84, 14, 16_792, 1_952, 9, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """token_type_ids""": [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]} # noqa: E501
# fmt: on
self.tokenizer_integration_test_util(
expected_encoding=__a , model_name="""albert-base-v2""" , revision="""6b6560eaf5ff2e250b00c50f380c5389a9c2d82e""" , )
| 42
| 0
|
import argparse
from collections import OrderedDict
from pathlib import Path
import requests
import torch
from PIL import Image
from transformers import GLPNConfig, GLPNForDepthEstimation, GLPNImageProcessor
from transformers.utils import logging
logging.set_verbosity_info()
__lowerCamelCase :Union[str, Any] = logging.get_logger(__name__)
def snake_case ( UpperCamelCase__ : Optional[int] ) -> Any:
lowerCamelCase : Any = OrderedDict()
for key, value in state_dict.items():
if key.startswith("""module.encoder""" ):
lowerCamelCase : List[Any] = key.replace("""module.encoder""" , """glpn.encoder""" )
if key.startswith("""module.decoder""" ):
lowerCamelCase : str = key.replace("""module.decoder""" , """decoder.stages""" )
if "patch_embed" in key:
# replace for example patch_embed1 by patch_embeddings.0
lowerCamelCase : Union[str, Any] = key[key.find("""patch_embed""" ) + len("""patch_embed""" )]
lowerCamelCase : Optional[Any] = key.replace(F'patch_embed{idx}' , F'patch_embeddings.{int(_snake_case )-1}' )
if "norm" in key:
lowerCamelCase : str = key.replace("""norm""" , """layer_norm""" )
if "glpn.encoder.layer_norm" in key:
# replace for example layer_norm1 by layer_norm.0
lowerCamelCase : Any = key[key.find("""glpn.encoder.layer_norm""" ) + len("""glpn.encoder.layer_norm""" )]
lowerCamelCase : List[Any] = key.replace(F'layer_norm{idx}' , F'layer_norm.{int(_snake_case )-1}' )
if "layer_norm1" in key:
lowerCamelCase : Tuple = key.replace("""layer_norm1""" , """layer_norm_1""" )
if "layer_norm2" in key:
lowerCamelCase : List[Any] = key.replace("""layer_norm2""" , """layer_norm_2""" )
if "block" in key:
# replace for example block1 by block.0
lowerCamelCase : Union[str, Any] = key[key.find("""block""" ) + len("""block""" )]
lowerCamelCase : Dict = key.replace(F'block{idx}' , F'block.{int(_snake_case )-1}' )
if "attn.q" in key:
lowerCamelCase : Any = key.replace("""attn.q""" , """attention.self.query""" )
if "attn.proj" in key:
lowerCamelCase : Tuple = key.replace("""attn.proj""" , """attention.output.dense""" )
if "attn" in key:
lowerCamelCase : List[str] = key.replace("""attn""" , """attention.self""" )
if "fc1" in key:
lowerCamelCase : Dict = key.replace("""fc1""" , """dense1""" )
if "fc2" in key:
lowerCamelCase : List[Any] = key.replace("""fc2""" , """dense2""" )
if "linear_pred" in key:
lowerCamelCase : Any = key.replace("""linear_pred""" , """classifier""" )
if "linear_fuse" in key:
lowerCamelCase : Optional[Any] = key.replace("""linear_fuse.conv""" , """linear_fuse""" )
lowerCamelCase : List[str] = key.replace("""linear_fuse.bn""" , """batch_norm""" )
if "linear_c" in key:
# replace for example linear_c4 by linear_c.3
lowerCamelCase : Optional[Any] = key[key.find("""linear_c""" ) + len("""linear_c""" )]
lowerCamelCase : Optional[int] = key.replace(F'linear_c{idx}' , F'linear_c.{int(_snake_case )-1}' )
if "bot_conv" in key:
lowerCamelCase : Dict = key.replace("""bot_conv""" , """0.convolution""" )
if "skip_conv1" in key:
lowerCamelCase : List[str] = key.replace("""skip_conv1""" , """1.convolution""" )
if "skip_conv2" in key:
lowerCamelCase : Optional[int] = key.replace("""skip_conv2""" , """2.convolution""" )
if "fusion1" in key:
lowerCamelCase : Any = key.replace("""fusion1""" , """1.fusion""" )
if "fusion2" in key:
lowerCamelCase : Optional[Any] = key.replace("""fusion2""" , """2.fusion""" )
if "fusion3" in key:
lowerCamelCase : Dict = key.replace("""fusion3""" , """3.fusion""" )
if "fusion" in key and "conv" in key:
lowerCamelCase : List[Any] = key.replace("""conv""" , """convolutional_layer""" )
if key.startswith("""module.last_layer_depth""" ):
lowerCamelCase : Optional[Any] = key.replace("""module.last_layer_depth""" , """head.head""" )
lowerCamelCase : Union[str, Any] = value
return new_state_dict
def snake_case ( UpperCamelCase__ : Tuple , UpperCamelCase__ : Tuple ) -> Tuple:
for i in range(config.num_encoder_blocks ):
for j in range(config.depths[i] ):
# read in weights + bias of keys and values (which is a single matrix in the original implementation)
lowerCamelCase : Optional[Any] = state_dict.pop(F'glpn.encoder.block.{i}.{j}.attention.self.kv.weight' )
lowerCamelCase : Dict = state_dict.pop(F'glpn.encoder.block.{i}.{j}.attention.self.kv.bias' )
# next, add keys and values (in that order) to the state dict
lowerCamelCase : Optional[int] = kv_weight[
: config.hidden_sizes[i], :
]
lowerCamelCase : Optional[Any] = kv_bias[: config.hidden_sizes[i]]
lowerCamelCase : str = kv_weight[
config.hidden_sizes[i] :, :
]
lowerCamelCase : Optional[int] = kv_bias[config.hidden_sizes[i] :]
def snake_case ( ) -> List[Any]:
lowerCamelCase : Optional[Any] = """http://images.cocodataset.org/val2017/000000039769.jpg"""
lowerCamelCase : Tuple = Image.open(requests.get(_snake_case , stream=_snake_case ).raw )
return image
@torch.no_grad()
def snake_case ( UpperCamelCase__ : str , UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Tuple=False , UpperCamelCase__ : str=None ) -> Any:
lowerCamelCase : Dict = GLPNConfig(hidden_sizes=[64, 128, 320, 512] , decoder_hidden_size=64 , depths=[3, 8, 27, 3] )
# load image processor (only resize + rescale)
lowerCamelCase : str = GLPNImageProcessor()
# prepare image
lowerCamelCase : Optional[Any] = prepare_img()
lowerCamelCase : int = image_processor(images=_snake_case , return_tensors="""pt""" ).pixel_values
logger.info("""Converting model...""" )
# load original state dict
lowerCamelCase : Optional[Any] = torch.load(_snake_case , map_location=torch.device("""cpu""" ) )
# rename keys
lowerCamelCase : List[str] = rename_keys(_snake_case )
# key and value matrices need special treatment
read_in_k_v(_snake_case , _snake_case )
# create HuggingFace model and load state dict
lowerCamelCase : Optional[int] = GLPNForDepthEstimation(_snake_case )
model.load_state_dict(_snake_case )
model.eval()
# forward pass
lowerCamelCase : Tuple = model(_snake_case )
lowerCamelCase : Tuple = outputs.predicted_depth
# verify output
if model_name is not None:
if "nyu" in model_name:
lowerCamelCase : int = torch.tensor(
[[4.4_1_4_7, 4.0_8_7_3, 4.0_6_7_3], [3.7_8_9_0, 3.2_8_8_1, 3.1_5_2_5], [3.7_6_7_4, 3.5_4_2_3, 3.4_9_1_3]] )
elif "kitti" in model_name:
lowerCamelCase : List[Any] = torch.tensor(
[[3.4_2_9_1, 2.7_8_6_5, 2.5_1_5_1], [3.2_8_4_1, 2.7_0_2_1, 2.3_5_0_2], [3.1_1_4_7, 2.4_6_2_5, 2.2_4_8_1]] )
else:
raise ValueError(F'Unknown model name: {model_name}' )
lowerCamelCase : List[Any] = torch.Size([1, 480, 640] )
assert predicted_depth.shape == expected_shape
assert torch.allclose(predicted_depth[0, :3, :3] , _snake_case , atol=1E-4 )
print("""Looks ok!""" )
# finally, push to hub if required
if push_to_hub:
logger.info("""Pushing model and image processor to the hub...""" )
model.push_to_hub(
repo_path_or_name=Path(_snake_case , _snake_case ) , organization="""nielsr""" , commit_message="""Add model""" , use_temp_dir=_snake_case , )
image_processor.push_to_hub(
repo_path_or_name=Path(_snake_case , _snake_case ) , organization="""nielsr""" , commit_message="""Add image processor""" , use_temp_dir=_snake_case , )
if __name__ == "__main__":
__lowerCamelCase :Any = argparse.ArgumentParser()
parser.add_argument(
'--checkpoint_path',
default=None,
type=str,
help='Path to the original PyTorch checkpoint (.pth file).',
)
parser.add_argument(
'--pytorch_dump_folder_path', default=None, type=str, help='Path to the folder to output PyTorch model.'
)
parser.add_argument(
'--push_to_hub', action='store_true', help='Whether to upload the model to the HuggingFace hub.'
)
parser.add_argument(
'--model_name',
default='glpn-kitti',
type=str,
help='Name of the model in case you\'re pushing to the hub.',
)
__lowerCamelCase :List[Any] = parser.parse_args()
convert_glpn_checkpoint(args.checkpoint_path, args.pytorch_dump_folder_path, args.push_to_hub, args.model_name)
| 707
|
"""simple docstring"""
__lowerCamelCase :List[Any] = {0: [2, 3], 1: [0], 2: [1], 3: [4], 4: []}
__lowerCamelCase :Union[str, Any] = {0: [1, 2, 3], 1: [2], 2: [0], 3: [4], 4: [5], 5: [3]}
def snake_case ( UpperCamelCase__ : dict[int, list[int]] , UpperCamelCase__ : int , UpperCamelCase__ : list[bool] ) -> list[int]:
lowerCamelCase : Tuple = True
lowerCamelCase : Any = []
for neighbour in graph[vert]:
if not visited[neighbour]:
order += topology_sort(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
order.append(UpperCamelCase__ )
return order
def snake_case ( UpperCamelCase__ : dict[int, list[int]] , UpperCamelCase__ : int , UpperCamelCase__ : list[bool] ) -> list[int]:
lowerCamelCase : List[Any] = True
lowerCamelCase : int = [vert]
for neighbour in reversed_graph[vert]:
if not visited[neighbour]:
component += find_components(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return component
def snake_case ( UpperCamelCase__ : dict[int, list[int]] ) -> list[list[int]]:
lowerCamelCase : int = len(UpperCamelCase__ ) * [False]
lowerCamelCase : dict[int, list[int]] = {vert: [] for vert in range(len(UpperCamelCase__ ) )}
for vert, neighbours in graph.items():
for neighbour in neighbours:
reversed_graph[neighbour].append(UpperCamelCase__ )
lowerCamelCase : int = []
for i, was_visited in enumerate(UpperCamelCase__ ):
if not was_visited:
order += topology_sort(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
lowerCamelCase : Tuple = []
lowerCamelCase : str = len(UpperCamelCase__ ) * [False]
for i in range(len(UpperCamelCase__ ) ):
lowerCamelCase : Any = order[len(UpperCamelCase__ ) - i - 1]
if not visited[vert]:
lowerCamelCase : List[str] = find_components(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
components_list.append(UpperCamelCase__ )
return components_list
| 42
| 0
|
"""simple docstring"""
import pytest
import datasets.config
from datasets.utils.info_utils import is_small_dataset
@pytest.mark.parametrize("""dataset_size""" , [None, 400 * 2**20, 600 * 2**20] )
@pytest.mark.parametrize("""input_in_memory_max_size""" , ["""default""", 0, 100 * 2**20, 900 * 2**20] )
def lowerCAmelCase ( UpperCamelCase__ : int , UpperCamelCase__ : Dict , UpperCamelCase__ : Tuple ) -> Optional[int]:
if input_in_memory_max_size != "default":
monkeypatch.setattr(datasets.config , """IN_MEMORY_MAX_SIZE""" , _A )
lowerCamelCase : Any = datasets.config.IN_MEMORY_MAX_SIZE
if input_in_memory_max_size == "default":
assert in_memory_max_size == 0
else:
assert in_memory_max_size == input_in_memory_max_size
if dataset_size and in_memory_max_size:
lowerCamelCase : List[str] = dataset_size < in_memory_max_size
else:
lowerCamelCase : Tuple = False
lowerCamelCase : Optional[int] = is_small_dataset(_A )
assert result == expected
| 708
|
"""simple docstring"""
from typing import List, Optional, Union
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :str = logging.get_logger(__name__)
__lowerCamelCase :Any = {
'huggingface/time-series-transformer-tourism-monthly': (
'https://huggingface.co/huggingface/time-series-transformer-tourism-monthly/resolve/main/config.json'
),
# See all TimeSeriesTransformer models at https://huggingface.co/models?filter=time_series_transformer
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : List[Any] ='''time_series_transformer'''
snake_case__ : List[Any] ={
'''hidden_size''': '''d_model''',
'''num_attention_heads''': '''encoder_attention_heads''',
'''num_hidden_layers''': '''encoder_layers''',
}
def __init__( self: List[str] , __a: Optional[int] = None , __a: Optional[int] = None , __a: str = "student_t" , __a: str = "nll" , __a: int = 1 , __a: List[int] = [1, 2, 3, 4, 5, 6, 7] , __a: Optional[Union[str, bool]] = "mean" , __a: int = 0 , __a: int = 0 , __a: int = 0 , __a: int = 0 , __a: Optional[List[int]] = None , __a: Optional[List[int]] = None , __a: int = 32 , __a: int = 32 , __a: int = 2 , __a: int = 2 , __a: int = 2 , __a: int = 2 , __a: bool = True , __a: str = "gelu" , __a: int = 64 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: int = 100 , __a: float = 0.02 , __a: Tuple=True , **__a: str , )-> Any:
# time series specific configuration
lowerCamelCase : str = prediction_length
lowerCamelCase : Optional[Any] = context_length or prediction_length
lowerCamelCase : Tuple = distribution_output
lowerCamelCase : Any = loss
lowerCamelCase : List[Any] = input_size
lowerCamelCase : int = num_time_features
lowerCamelCase : Dict = lags_sequence
lowerCamelCase : Optional[int] = scaling
lowerCamelCase : int = num_dynamic_real_features
lowerCamelCase : Tuple = num_static_real_features
lowerCamelCase : Any = num_static_categorical_features
if cardinality and num_static_categorical_features > 0:
if len(__a ) != num_static_categorical_features:
raise ValueError(
"""The cardinality should be a list of the same length as `num_static_categorical_features`""" )
lowerCamelCase : int = cardinality
else:
lowerCamelCase : Dict = [0]
if embedding_dimension and num_static_categorical_features > 0:
if len(__a ) != num_static_categorical_features:
raise ValueError(
"""The embedding dimension should be a list of the same length as `num_static_categorical_features`""" )
lowerCamelCase : str = embedding_dimension
else:
lowerCamelCase : str = [min(50 , (cat + 1) // 2 ) for cat in self.cardinality]
lowerCamelCase : Any = num_parallel_samples
# Transformer architecture configuration
lowerCamelCase : Any = input_size * len(__a ) + self._number_of_features
lowerCamelCase : List[str] = d_model
lowerCamelCase : Tuple = encoder_attention_heads
lowerCamelCase : Optional[int] = decoder_attention_heads
lowerCamelCase : Union[str, Any] = encoder_ffn_dim
lowerCamelCase : str = decoder_ffn_dim
lowerCamelCase : str = encoder_layers
lowerCamelCase : Any = decoder_layers
lowerCamelCase : Optional[int] = dropout
lowerCamelCase : List[str] = attention_dropout
lowerCamelCase : Tuple = activation_dropout
lowerCamelCase : Optional[int] = encoder_layerdrop
lowerCamelCase : int = decoder_layerdrop
lowerCamelCase : Optional[int] = activation_function
lowerCamelCase : Optional[Any] = init_std
lowerCamelCase : Optional[Any] = use_cache
super().__init__(is_encoder_decoder=__a , **__a )
@property
def a__ ( self: int )-> int:
return (
sum(self.embedding_dimension )
+ self.num_dynamic_real_features
+ self.num_time_features
+ self.num_static_real_features
+ self.input_size * 2 # the log1p(abs(loc)) and log(scale) features
)
| 42
| 0
|
"""simple docstring"""
import tempfile
import unittest
from transformers import AutoModelForSeqaSeqLM, AutoTokenizer
from transformers.testing_utils import (
is_torch_available,
require_optimum,
require_torch,
slow,
)
if is_torch_available():
import torch
@require_torch
@require_optimum
@slow
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: List[Any] )-> Optional[int]:
lowerCamelCase : Union[str, Any] = """hf-internal-testing/tiny-random-t5"""
lowerCamelCase : int = AutoTokenizer.from_pretrained(UpperCamelCase__ )
lowerCamelCase : Optional[int] = AutoModelForSeqaSeqLM.from_pretrained(UpperCamelCase__ )
lowerCamelCase : Tuple = tokenizer("""This is me""" , return_tensors="""pt""" )
lowerCamelCase : Union[str, Any] = model.to_bettertransformer()
self.assertTrue(any("""BetterTransformer""" in mod.__class__.__name__ for _, mod in model.named_modules() ) )
lowerCamelCase : Dict = model.generate(**UpperCamelCase__ )
lowerCamelCase : Dict = model.reverse_bettertransformer()
self.assertFalse(any("""BetterTransformer""" in mod.__class__.__name__ for _, mod in model.named_modules() ) )
with tempfile.TemporaryDirectory() as tmpdirname:
model.save_pretrained(UpperCamelCase__ )
lowerCamelCase : Union[str, Any] = AutoModelForSeqaSeqLM.from_pretrained(UpperCamelCase__ )
self.assertFalse(
any("""BetterTransformer""" in mod.__class__.__name__ for _, mod in model_reloaded.named_modules() ) )
lowerCamelCase : str = model_reloaded.generate(**UpperCamelCase__ )
self.assertTrue(torch.allclose(UpperCamelCase__ , UpperCamelCase__ ) )
def a__ ( self: int )-> int:
lowerCamelCase : Union[str, Any] = """hf-internal-testing/tiny-random-t5"""
lowerCamelCase : Union[str, Any] = AutoModelForSeqaSeqLM.from_pretrained(UpperCamelCase__ )
lowerCamelCase : Optional[Any] = model.to_bettertransformer()
with tempfile.TemporaryDirectory() as tmpdirname:
with self.assertRaises(UpperCamelCase__ ):
model.save_pretrained(UpperCamelCase__ )
lowerCamelCase : Optional[Any] = model.reverse_bettertransformer()
model.save_pretrained(UpperCamelCase__ )
| 709
|
"""simple docstring"""
from __future__ import annotations
__lowerCamelCase :int = 10
def snake_case ( UpperCamelCase__ : list[int] ) -> list[int]:
lowerCamelCase : int = 1
lowerCamelCase : Union[str, Any] = max(UpperCamelCase__ )
while placement <= max_digit:
# declare and initialize empty buckets
lowerCamelCase : list[list] = [[] for _ in range(UpperCamelCase__ )]
# split list_of_ints between the buckets
for i in list_of_ints:
lowerCamelCase : Any = int((i / placement) % RADIX )
buckets[tmp].append(UpperCamelCase__ )
# put each buckets' contents into list_of_ints
lowerCamelCase : Dict = 0
for b in range(UpperCamelCase__ ):
for i in buckets[b]:
lowerCamelCase : List[str] = i
a += 1
# move to next
placement *= RADIX
return list_of_ints
if __name__ == "__main__":
import doctest
doctest.testmod()
| 42
| 0
|
"""simple docstring"""
from typing import List, Optional, Tuple, Union
import PIL
import torch
from torchvision import transforms
from diffusers.pipeline_utils import DiffusionPipeline, ImagePipelineOutput
from diffusers.schedulers import DDIMScheduler
from diffusers.utils import randn_tensor
__lowerCamelCase :Optional[Any] = transforms.Compose(
[
transforms.Resize((256, 256)),
transforms.ToTensor(),
transforms.Normalize([0.5], [0.5]),
]
)
def snake_case ( UpperCamelCase__ : Optional[int] ) -> Union[str, Any]:
if isinstance(UpperCamelCase__ , torch.Tensor ):
return image
elif isinstance(UpperCamelCase__ , PIL.Image.Image ):
lowerCamelCase : int = [image]
lowerCamelCase : int = [trans(img.convert("""RGB""" ) ) for img in image]
lowerCamelCase : Dict = torch.stack(UpperCamelCase__ )
return image
class A__ ( snake_case__):
"""simple docstring"""
def __init__( self: str , __a: Any , __a: int )-> str:
super().__init__()
# make sure scheduler can always be converted to DDIM
lowerCamelCase : Tuple = DDIMScheduler.from_config(scheduler.config )
self.register_modules(unet=_A , scheduler=_A )
def a__ ( self: Optional[int] , __a: List[Any] )-> Optional[Any]:
if strength < 0 or strength > 1:
raise ValueError(f'The value of strength should in [0.0, 1.0] but is {strength}' )
def a__ ( self: Union[str, Any] , __a: str , __a: Union[str, Any] , __a: Optional[Any] )-> str:
# get the original timestep using init_timestep
lowerCamelCase : Optional[int] = min(int(num_inference_steps * strength ) , _A )
lowerCamelCase : str = max(num_inference_steps - init_timestep , 0 )
lowerCamelCase : Optional[int] = self.scheduler.timesteps[t_start:]
return timesteps, num_inference_steps - t_start
def a__ ( self: Optional[Any] , __a: str , __a: Tuple , __a: Union[str, Any] , __a: Any , __a: Dict , __a: Union[str, Any]=None )-> Optional[int]:
if not isinstance(_A , (torch.Tensor, PIL.Image.Image, list) ):
raise ValueError(
f'`image` has to be of type `torch.Tensor`, `PIL.Image.Image` or list but is {type(_A )}' )
lowerCamelCase : Union[str, Any] = image.to(device=_A , dtype=_A )
if isinstance(_A , _A ) and len(_A ) != batch_size:
raise ValueError(
f'You have passed a list of generators of length {len(_A )}, but requested an effective batch'
f' size of {batch_size}. Make sure the batch size matches the length of the generators.' )
lowerCamelCase : Dict = init_latents.shape
lowerCamelCase : Optional[int] = randn_tensor(_A , generator=_A , device=_A , dtype=_A )
# get latents
print("""add noise to latents at timestep""" , _A )
lowerCamelCase : List[Any] = self.scheduler.add_noise(_A , _A , _A )
lowerCamelCase : Optional[int] = init_latents
return latents
@torch.no_grad()
def __call__( self: Union[str, Any] , __a: Tuple = None , __a: int = 0.8 , __a: Dict = 1 , __a: Optional[int] = None , __a: Optional[Any] = 0.0 , __a: Any = 50 , __a: List[str] = None , __a: Optional[int] = "pil" , __a: str = True , )-> str:
self.check_inputs(_A )
# 2. Preprocess image
lowerCamelCase : int = preprocess(_A )
# 3. set timesteps
self.scheduler.set_timesteps(_A , device=self.device )
lowerCamelCase : Tuple = self.get_timesteps(_A , _A , self.device )
lowerCamelCase : int = timesteps[:1].repeat(_A )
# 4. Prepare latent variables
lowerCamelCase : Dict = self.prepare_latents(_A , _A , _A , self.unet.dtype , self.device , _A )
lowerCamelCase : List[Any] = latents
# 5. Denoising loop
for t in self.progress_bar(_A ):
# 1. predict noise model_output
lowerCamelCase : Optional[Any] = self.unet(_A , _A ).sample
# 2. predict previous mean of image x_t-1 and add variance depending on eta
# eta corresponds to η in paper and should be between [0, 1]
# do x_t -> x_t-1
lowerCamelCase : List[Any] = self.scheduler.step(
_A , _A , _A , eta=_A , use_clipped_model_output=_A , generator=_A , ).prev_sample
lowerCamelCase : List[Any] = (image / 2 + 0.5).clamp(0 , 1 )
lowerCamelCase : List[str] = image.cpu().permute(0 , 2 , 3 , 1 ).numpy()
if output_type == "pil":
lowerCamelCase : Optional[int] = self.numpy_to_pil(_A )
if not return_dict:
return (image, latent_timestep.item())
return ImagePipelineOutput(images=_A )
| 710
|
"""simple docstring"""
import argparse
import pickle
import numpy as np
import torch
from torch import nn
from transformers import ReformerConfig, ReformerModelWithLMHead
from transformers.utils import logging
logging.set_verbosity_info()
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Tuple , UpperCamelCase__ : Optional[Any]=None ) -> Tuple:
# set parameter of one layer
assert torch_layer.weight.shape == weight.shape, F'{torch_layer} layer.weight does not match'
lowerCamelCase : Dict = nn.Parameter(UpperCamelCase__ )
if bias is not None:
assert torch_layer.bias.shape == bias.shape, F'{torch_layer} layer.bias does not match'
lowerCamelCase : Any = nn.Parameter(UpperCamelCase__ )
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : Any , UpperCamelCase__ : Tuple ) -> Union[str, Any]:
# set torch weights for 1-to-1 comparison
lowerCamelCase : Dict = np.asarray(weights[0] )
lowerCamelCase : List[Any] = np.asarray(weights[1] )
lowerCamelCase : List[str] = np.asarray(weights[2] )
set_param(
torch_layer.self_attention.query_key , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.value , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.output.dense , torch.tensor(UpperCamelCase__ ).view(-1 , UpperCamelCase__ ).contiguous().transpose(0 , 1 ) , )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Any ) -> List[Any]:
# set torch weights for 1-to-1 comparison
lowerCamelCase : Tuple = np.asarray(weights[0] )
lowerCamelCase : Any = np.asarray(weights[1] )
lowerCamelCase : List[Any] = np.asarray(weights[2] )
lowerCamelCase : List[str] = np.asarray(weights[3] )
set_param(
torch_layer.self_attention.query , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.key , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.value , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.output.dense , torch.tensor(UpperCamelCase__ ).view(-1 , UpperCamelCase__ ).contiguous().transpose(0 , 1 ) , )
def snake_case ( UpperCamelCase__ : Dict , UpperCamelCase__ : Optional[int] , UpperCamelCase__ : Any ) -> Optional[Any]:
# layernorm 1
lowerCamelCase : str = weights[0][0][0]
lowerCamelCase : Optional[int] = np.asarray(layer_norm_a[0] )
lowerCamelCase : Tuple = np.asarray(layer_norm_a[1] )
set_param(
torch_block.attention.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# lsh weights + output
lowerCamelCase : List[Any] = weights[0][1]
if len(UpperCamelCase__ ) < 4:
set_layer_weights_in_torch_lsh(UpperCamelCase__ , torch_block.attention , UpperCamelCase__ )
else:
set_layer_weights_in_torch_local(UpperCamelCase__ , torch_block.attention , UpperCamelCase__ )
# intermediate weighs
lowerCamelCase : int = weights[2][0][1][2]
# Chunked Feed Forward
if len(UpperCamelCase__ ) == 4:
lowerCamelCase : Dict = intermediate_weights[2]
# layernorm 2
lowerCamelCase : Optional[int] = np.asarray(intermediate_weights[0][0] )
lowerCamelCase : Tuple = np.asarray(intermediate_weights[0][1] )
set_param(
torch_block.feed_forward.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# intermediate dense
lowerCamelCase : Optional[Any] = np.asarray(intermediate_weights[1][0] )
lowerCamelCase : Union[str, Any] = np.asarray(intermediate_weights[1][1] )
set_param(
torch_block.feed_forward.dense.dense , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
# intermediate out
lowerCamelCase : Optional[int] = np.asarray(intermediate_weights[4][0] )
lowerCamelCase : List[Any] = np.asarray(intermediate_weights[4][1] )
set_param(
torch_block.feed_forward.output.dense , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : List[str] , UpperCamelCase__ : int ) -> List[Any]:
# reformer model
lowerCamelCase : List[Any] = torch_model.reformer
# word embeds
lowerCamelCase : Union[str, Any] = np.asarray(weights[1] )
set_param(
torch_model_reformer.embeddings.word_embeddings , torch.tensor(UpperCamelCase__ ) , )
if isinstance(weights[3] , UpperCamelCase__ ):
lowerCamelCase : Union[str, Any] = torch_model_reformer.embeddings.position_embeddings
for emb_idx in range(len(position_embeddings.weights ) ):
lowerCamelCase : str = np.asarray(weights[3][emb_idx][0] )
assert (
position_embeddings.weights[emb_idx].shape == emb_weights.shape
), F'{position_embeddings[emb_idx]} emb does not match'
lowerCamelCase : Dict = nn.Parameter(torch.tensor(UpperCamelCase__ ) )
lowerCamelCase : int = weights[5]
assert len(torch_model_reformer.encoder.layers ) * 4 == len(
UpperCamelCase__ ), "HF and trax model do not have the same number of layers"
for layer_idx, layer in enumerate(torch_model_reformer.encoder.layers ):
lowerCamelCase : Dict = trax_layer_weights[4 * layer_idx : 4 * (layer_idx + 1)]
set_block_weights_in_torch(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
# output layer norm
lowerCamelCase : Any = np.asarray(weights[7][0] )
lowerCamelCase : List[str] = np.asarray(weights[7][1] )
set_param(
torch_model_reformer.encoder.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# output embeddings
lowerCamelCase : List[Any] = np.asarray(weights[9][0] )
lowerCamelCase : Optional[int] = np.asarray(weights[9][1] )
set_param(
torch_model.lm_head.decoder , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
def snake_case ( UpperCamelCase__ : str , UpperCamelCase__ : int , UpperCamelCase__ : str ) -> Optional[int]:
# Initialise PyTorch model
lowerCamelCase : Union[str, Any] = ReformerConfig.from_json_file(UpperCamelCase__ )
print(F'Building PyTorch model from configuration: {config}' )
lowerCamelCase : str = ReformerModelWithLMHead(UpperCamelCase__ )
with open(UpperCamelCase__ , """rb""" ) as f:
lowerCamelCase : str = pickle.load(UpperCamelCase__ )["""weights"""]
set_model_weights_in_torch(UpperCamelCase__ , UpperCamelCase__ , config.hidden_size )
# Save pytorch-model
print(F'Save PyTorch model to {pytorch_dump_path}' )
torch.save(model.state_dict() , UpperCamelCase__ )
if __name__ == "__main__":
__lowerCamelCase :Dict = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--trax_model_pkl_path', default=None, type=str, required=True, help='Path to the TensorFlow checkpoint path.'
)
parser.add_argument(
'--config_file',
default=None,
type=str,
required=True,
help=(
'The config json file corresponding to the pre-trained Reformer model. \n'
'This specifies the model architecture.'
),
)
parser.add_argument(
'--pytorch_dump_path', default=None, type=str, required=True, help='Path to the output PyTorch model.'
)
__lowerCamelCase :Optional[int] = parser.parse_args()
convert_trax_checkpoint_to_pytorch(args.trax_model_pkl_path, args.config_file, args.pytorch_dump_path)
| 42
| 0
|
"""simple docstring"""
def snake_case ( ) -> list[list[int]]:
return [list(range(1000 - i , -1000 - i , -1 ) ) for i in range(1000 )]
__lowerCamelCase :List[str] = generate_large_matrix()
__lowerCamelCase :List[str] = (
[[4, 3, 2, -1], [3, 2, 1, -1], [1, 1, -1, -2], [-1, -1, -2, -3]],
[[3, 2], [1, 0]],
[[7, 7, 6]],
[[7, 7, 6], [-1, -2, -3]],
grid,
)
def snake_case ( UpperCamelCase__ : Tuple ) -> None:
assert all(row == sorted(__snake_case , reverse=__snake_case ) for row in grid )
assert all(list(__snake_case ) == sorted(__snake_case , reverse=__snake_case ) for col in zip(*__snake_case ) )
def snake_case ( UpperCamelCase__ : Tuple ) -> int:
lowerCamelCase : Tuple = 0
lowerCamelCase : Optional[int] = len(__snake_case ) - 1
# Edge cases such as no values or all numbers are negative.
if not array or array[0] < 0:
return 0
while right + 1 > left:
lowerCamelCase : Any = (left + right) // 2
lowerCamelCase : Optional[int] = array[mid]
# Num must be negative and the index must be greater than or equal to 0.
if num < 0 and array[mid - 1] >= 0:
return mid
if num >= 0:
lowerCamelCase : Optional[Any] = mid + 1
else:
lowerCamelCase : Optional[Any] = mid - 1
# No negative numbers so return the last index of the array + 1 which is the length.
return len(__snake_case )
def snake_case ( UpperCamelCase__ : List[Any] ) -> int:
lowerCamelCase : int = 0
lowerCamelCase : Dict = len(grid[0] )
for i in range(len(__snake_case ) ):
lowerCamelCase : Optional[int] = find_negative_index(grid[i][:bound] )
total += bound
return (len(__snake_case ) * len(grid[0] )) - total
def snake_case ( UpperCamelCase__ : List[Any] ) -> int:
return len([number for row in grid for number in row if number < 0] )
def snake_case ( UpperCamelCase__ : Union[str, Any] ) -> int:
lowerCamelCase : Optional[int] = 0
for row in grid:
for i, number in enumerate(__snake_case ):
if number < 0:
total += len(__snake_case ) - i
break
return total
def snake_case ( ) -> None:
from timeit import timeit
print("""Running benchmarks""" )
lowerCamelCase : int = (
"""from __main__ import count_negatives_binary_search, """
"""count_negatives_brute_force, count_negatives_brute_force_with_break, grid"""
)
for func in (
"count_negatives_binary_search", # took 0.7727 seconds
"count_negatives_brute_force_with_break", # took 4.6505 seconds
"count_negatives_brute_force", # took 12.8160 seconds
):
lowerCamelCase : List[Any] = timeit(F'{func}(grid=grid)' , setup=__snake_case , number=500 )
print(F'{func}() took {time:0.4f} seconds' )
if __name__ == "__main__":
import doctest
doctest.testmod()
benchmark()
| 711
|
"""simple docstring"""
import inspect
import unittest
import torch
import torch.nn as nn
from accelerate.hooks import (
AlignDevicesHook,
ModelHook,
SequentialHook,
add_hook_to_module,
attach_align_device_hook,
remove_hook_from_module,
remove_hook_from_submodules,
)
from accelerate.test_utils import require_multi_gpu
class A__ ( nn.Module):
"""simple docstring"""
def __init__( self: Dict )-> Dict:
super().__init__()
lowerCamelCase : Tuple = nn.Linear(3 , 4 )
lowerCamelCase : Optional[Any] = nn.BatchNormad(4 )
lowerCamelCase : Optional[Any] = nn.Linear(4 , 5 )
def a__ ( self: List[str] , __a: List[Any] )-> Optional[Any]:
return self.lineara(self.batchnorm(self.lineara(__a ) ) )
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Tuple , __a: int , *__a: Any , **__a: Tuple )-> Tuple:
return (args[0] + 1,) + args[1:], kwargs
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Optional[int] , __a: List[str] , __a: List[Any] )-> List[str]:
return output + 1
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: int )-> str:
lowerCamelCase : List[str] = ModelForTest()
lowerCamelCase : Dict = ModelHook()
add_hook_to_module(__a , __a )
self.assertEqual(test_model._hf_hook , __a )
self.assertTrue(hasattr(__a , """_old_forward""" ) )
# Check adding the hook did not change the name or the signature
self.assertEqual(test_model.forward.__name__ , """forward""" )
self.assertListEqual(list(inspect.signature(test_model.forward ).parameters ) , ["""x"""] )
remove_hook_from_module(__a )
self.assertFalse(hasattr(__a , """_hf_hook""" ) )
self.assertFalse(hasattr(__a , """_old_forward""" ) )
def a__ ( self: int )-> str:
lowerCamelCase : List[str] = ModelForTest()
lowerCamelCase : Union[str, Any] = ModelHook()
add_hook_to_module(__a , __a )
add_hook_to_module(__a , __a , append=__a )
self.assertEqual(isinstance(test_model._hf_hook , __a ) , __a )
self.assertEqual(len(test_model._hf_hook.hooks ) , 2 )
self.assertTrue(hasattr(__a , """_old_forward""" ) )
# Check adding the hook did not change the name or the signature
self.assertEqual(test_model.forward.__name__ , """forward""" )
self.assertListEqual(list(inspect.signature(test_model.forward ).parameters ) , ["""x"""] )
remove_hook_from_module(__a )
self.assertFalse(hasattr(__a , """_hf_hook""" ) )
self.assertFalse(hasattr(__a , """_old_forward""" ) )
def a__ ( self: List[Any] )-> List[str]:
lowerCamelCase : str = ModelForTest()
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : Union[str, Any] = test_model(x + 1 )
lowerCamelCase : Optional[int] = test_model(x + 2 )
lowerCamelCase : List[Any] = PreForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[int] = test_model(__a )
self.assertTrue(torch.allclose(__a , __a , atol=1e-5 ) )
# Attaching a hook to a model when it already has one replaces, does not chain
lowerCamelCase : Dict = PreForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Tuple = test_model(__a )
self.assertTrue(torch.allclose(__a , __a , atol=1e-5 ) )
# You need to use the sequential hook to chain two or more hooks
lowerCamelCase : Any = SequentialHook(PreForwardHook() , PreForwardHook() )
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[Any] = test_model(__a )
assert torch.allclose(__a , __a , atol=1e-5 )
def a__ ( self: Any )-> Optional[int]:
lowerCamelCase : str = ModelForTest()
lowerCamelCase : List[str] = torch.randn(2 , 3 )
lowerCamelCase : int = test_model(__a )
lowerCamelCase : Dict = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Tuple = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 , atol=1e-5 ) )
# Attaching a hook to a model when it already has one replaces, does not chain
lowerCamelCase : str = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[Any] = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 , atol=1e-5 ) )
# You need to use the sequential hook to chain two or more hooks
lowerCamelCase : Union[str, Any] = SequentialHook(PostForwardHook() , PostForwardHook() )
add_hook_to_module(__a , __a )
lowerCamelCase : str = test_model(__a )
assert torch.allclose(__a , output + 2 , atol=1e-5 )
def a__ ( self: int )-> Dict:
lowerCamelCase : List[Any] = ModelForTest()
lowerCamelCase : Optional[int] = torch.randn(2 , 3 )
lowerCamelCase : List[str] = test_model(__a )
lowerCamelCase : Any = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : str = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 ) )
self.assertTrue(outputa.requires_grad )
lowerCamelCase : Optional[int] = True
lowerCamelCase : Optional[int] = test_model(__a )
self.assertFalse(outputa.requires_grad )
@require_multi_gpu
def a__ ( self: List[str] )-> Union[str, Any]:
lowerCamelCase : int = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
add_hook_to_module(model.lineara , AlignDevicesHook(execution_device=0 ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(execution_device=0 ) )
add_hook_to_module(model.lineara , AlignDevicesHook(execution_device=1 ) )
self.assertEqual(model.lineara.weight.device , torch.device(0 ) )
self.assertEqual(model.batchnorm.weight.device , torch.device(0 ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device(0 ) )
self.assertEqual(model.lineara.weight.device , torch.device(1 ) )
# We can still make a forward pass. The input does not need to be on any particular device
lowerCamelCase : str = torch.randn(2 , 3 )
lowerCamelCase : Dict = model(__a )
self.assertEqual(output.device , torch.device(1 ) )
# We can add a general hook to put back output on same device as input.
add_hook_to_module(__a , AlignDevicesHook(io_same_device=__a ) )
lowerCamelCase : Optional[int] = torch.randn(2 , 3 ).to(0 )
lowerCamelCase : str = model(__a )
self.assertEqual(output.device , torch.device(0 ) )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Union[str, Any] = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : Tuple = {"""execution_device""": 0 if torch.cuda.is_available() else """cpu""", """offload""": True}
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(**__a ) )
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(hook_kwargs["""execution_device"""] )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Optional[Any] = torch.randn(2 , 3 )
lowerCamelCase : Optional[Any] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_module(model.lineara )
remove_hook_from_module(model.batchnorm )
remove_hook_from_module(model.lineara )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
lowerCamelCase : Any = {
"""execution_device""": 0 if torch.cuda.is_available() else """cpu""",
"""offload""": True,
"""offload_buffers""": True,
}
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(**__a ) )
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : int = torch.randn(2 , 3 )
lowerCamelCase : Optional[int] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_module(model.lineara )
remove_hook_from_module(model.batchnorm )
remove_hook_from_module(model.lineara )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
def a__ ( self: Any )-> List[str]:
lowerCamelCase : int = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : int = 0 if torch.cuda.is_available() else """cpu"""
attach_align_device_hook(__a , execution_device=__a , offload=__a )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(__a )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : Optional[Any] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
attach_align_device_hook(__a , execution_device=__a , offload=__a , offload_buffers=__a )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : Optional[int] = torch.randn(2 , 3 )
lowerCamelCase : int = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
def a__ ( self: Optional[Any] )-> List[Any]:
lowerCamelCase : List[Any] = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : Any = 0 if torch.cuda.is_available() else """cpu"""
attach_align_device_hook(
__a , execution_device=__a , offload=__a , weights_map=model.state_dict() )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(__a )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : int = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
attach_align_device_hook(
__a , execution_device=__a , offload=__a , weights_map=model.state_dict() , offload_buffers=__a , )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : Tuple = torch.randn(2 , 3 )
lowerCamelCase : Any = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
| 42
| 0
|
"""simple docstring"""
import gc
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DDIMScheduler,
StableDiffusionAttendAndExcitePipeline,
UNetaDConditionModel,
)
from diffusers.utils import load_numpy, skip_mps, slow
from diffusers.utils.testing_utils import require_torch_gpu
from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
from ..test_pipelines_common import PipelineKarrasSchedulerTesterMixin, PipelineLatentTesterMixin, PipelineTesterMixin
__lowerCamelCase :Dict = False
@skip_mps
class A__ ( lowercase_ , lowercase_ , lowercase_ , unittest.TestCase):
"""simple docstring"""
snake_case__ : Dict =StableDiffusionAttendAndExcitePipeline
snake_case__ : Union[str, Any] =False
snake_case__ : List[Any] =TEXT_TO_IMAGE_PARAMS
snake_case__ : Any =TEXT_TO_IMAGE_BATCH_PARAMS.union({'''token_indices'''})
snake_case__ : Optional[int] =TEXT_TO_IMAGE_IMAGE_PARAMS
snake_case__ : str =TEXT_TO_IMAGE_IMAGE_PARAMS
@classmethod
def a__ ( cls: Optional[int] )-> List[Any]:
super().setUpClass()
torch.use_deterministic_algorithms(lowerCamelCase_ )
@classmethod
def a__ ( cls: List[str] )-> Union[str, Any]:
super().tearDownClass()
torch.use_deterministic_algorithms(lowerCamelCase_ )
def a__ ( self: Dict )-> Union[str, Any]:
torch.manual_seed(0 )
lowerCamelCase : str = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=1 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , cross_attention_dim=32 , attention_head_dim=(2, 4) , use_linear_projection=lowerCamelCase_ , )
lowerCamelCase : Dict = DDIMScheduler(
beta_start=0.0_00_85 , beta_end=0.0_12 , beta_schedule="""scaled_linear""" , clip_sample=lowerCamelCase_ , set_alpha_to_one=lowerCamelCase_ , )
torch.manual_seed(0 )
lowerCamelCase : int = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
lowerCamelCase : List[str] = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-0_5 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1_000 , hidden_act="""gelu""" , projection_dim=512 , )
lowerCamelCase : List[Any] = CLIPTextModel(lowerCamelCase_ )
lowerCamelCase : Optional[int] = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" )
lowerCamelCase : List[Any] = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""safety_checker""": None,
"""feature_extractor""": None,
}
return components
def a__ ( self: Dict , __a: List[str] , __a: Dict=0 )-> Optional[Any]:
if str(lowerCamelCase_ ).startswith("""mps""" ):
lowerCamelCase : Optional[Any] = torch.manual_seed(lowerCamelCase_ )
else:
lowerCamelCase : Optional[int] = torch.Generator(device=lowerCamelCase_ ).manual_seed(lowerCamelCase_ )
lowerCamelCase : List[Any] = {
"""prompt""": """a cat and a frog""",
"""token_indices""": [2, 5],
"""generator""": generator,
"""num_inference_steps""": 1,
"""guidance_scale""": 6.0,
"""output_type""": """numpy""",
"""max_iter_to_alter""": 2,
"""thresholds""": {0: 0.7},
}
return inputs
def a__ ( self: Optional[int] )-> Union[str, Any]:
lowerCamelCase : Optional[Any] = """cpu"""
lowerCamelCase : Any = self.get_dummy_components()
lowerCamelCase : Tuple = self.pipeline_class(**lowerCamelCase_ )
pipe.to(lowerCamelCase_ )
pipe.set_progress_bar_config(disable=lowerCamelCase_ )
lowerCamelCase : List[str] = self.get_dummy_inputs(lowerCamelCase_ )
lowerCamelCase : Tuple = pipe(**lowerCamelCase_ ).images
lowerCamelCase : Any = image[0, -3:, -3:, -1]
self.assertEqual(image.shape , (1, 64, 64, 3) )
lowerCamelCase : Optional[int] = np.array(
[0.63_90_53_64, 0.62_89_73_07, 0.48_59_90_17, 0.5_13_36_24, 0.5_55_00_48, 0.45_76_95_16, 0.50_32_69_73, 0.5_02_31_39, 0.45_38_44_96] )
lowerCamelCase : Any = np.abs(image_slice.flatten() - expected_slice ).max()
self.assertLessEqual(lowerCamelCase_ , 1e-3 )
def a__ ( self: List[str] )-> str:
super().test_cpu_offload_forward_pass(expected_max_diff=5e-4 )
def a__ ( self: Any )-> Optional[Any]:
self._test_inference_batch_consistent(batch_sizes=[1, 2] )
def a__ ( self: Tuple )-> Any:
self._test_inference_batch_single_identical(batch_size=2 , expected_max_diff=7e-4 )
def a__ ( self: List[str] )-> Dict:
super().test_dict_tuple_outputs_equivalent(expected_max_difference=3e-3 )
def a__ ( self: str )-> Union[str, Any]:
super().test_pt_np_pil_outputs_equivalent(expected_max_diff=5e-4 )
def a__ ( self: Any )-> Optional[int]:
super().test_save_load_local(expected_max_difference=5e-4 )
def a__ ( self: int )-> Optional[int]:
super().test_save_load_optional_components(expected_max_difference=4e-4 )
@require_torch_gpu
@slow
class A__ ( unittest.TestCase):
"""simple docstring"""
@classmethod
def a__ ( cls: List[Any] )-> List[Any]:
super().setUpClass()
torch.use_deterministic_algorithms(lowerCamelCase_ )
@classmethod
def a__ ( cls: int )-> Optional[int]:
super().tearDownClass()
torch.use_deterministic_algorithms(lowerCamelCase_ )
def a__ ( self: Tuple )-> int:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: int )-> Any:
lowerCamelCase : Dict = torch.manual_seed(51 )
lowerCamelCase : List[Any] = StableDiffusionAttendAndExcitePipeline.from_pretrained(
"""CompVis/stable-diffusion-v1-4""" , safety_checker=lowerCamelCase_ , torch_dtype=torch.floataa )
pipe.to("""cuda""" )
lowerCamelCase : List[Any] = """a painting of an elephant with glasses"""
lowerCamelCase : Optional[Any] = [5, 7]
lowerCamelCase : Optional[Any] = pipe(
prompt=lowerCamelCase_ , token_indices=lowerCamelCase_ , guidance_scale=7.5 , generator=lowerCamelCase_ , num_inference_steps=5 , max_iter_to_alter=5 , output_type="""numpy""" , ).images[0]
lowerCamelCase : Dict = load_numpy(
"""https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/attend-and-excite/elephant_glasses.npy""" )
assert np.abs((expected_image - image).max() ) < 5e-1
| 712
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_torch_available,
)
__lowerCamelCase :Optional[Any] = {
'configuration_encodec': [
'ENCODEC_PRETRAINED_CONFIG_ARCHIVE_MAP',
'EncodecConfig',
],
'feature_extraction_encodec': ['EncodecFeatureExtractor'],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Union[str, Any] = [
'ENCODEC_PRETRAINED_MODEL_ARCHIVE_LIST',
'EncodecModel',
'EncodecPreTrainedModel',
]
if TYPE_CHECKING:
from .configuration_encodec import (
ENCODEC_PRETRAINED_CONFIG_ARCHIVE_MAP,
EncodecConfig,
)
from .feature_extraction_encodec import EncodecFeatureExtractor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_encodec import (
ENCODEC_PRETRAINED_MODEL_ARCHIVE_LIST,
EncodecModel,
EncodecPreTrainedModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
def snake_case ( UpperCamelCase__ : list , UpperCamelCase__ : int , UpperCamelCase__ : int = 0 , UpperCamelCase__ : int = 0 ) -> Union[str, Any]:
lowerCamelCase : int = right or len(A__ ) - 1
if left > right:
return -1
elif list_data[left] == key:
return left
elif list_data[right] == key:
return right
else:
return search(A__ , A__ , left + 1 , right - 1 )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 713
|
"""simple docstring"""
import inspect
import unittest
from transformers import ConvNextConfig
from transformers.testing_utils import require_torch, require_vision, slow, torch_device
from transformers.utils import cached_property, is_torch_available, is_vision_available
from ...test_backbone_common import BackboneTesterMixin
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import ConvNextBackbone, ConvNextForImageClassification, ConvNextModel
from transformers.models.convnext.modeling_convnext import CONVNEXT_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import AutoImageProcessor
class A__ :
"""simple docstring"""
def __init__( self: Optional[Any] , __a: Union[str, Any] , __a: Optional[Any]=13 , __a: Optional[Any]=32 , __a: Dict=3 , __a: int=4 , __a: Dict=[10, 20, 30, 40] , __a: int=[2, 2, 3, 2] , __a: Any=True , __a: List[Any]=True , __a: Any=37 , __a: Optional[int]="gelu" , __a: List[str]=10 , __a: Optional[int]=0.02 , __a: Dict=["stage2", "stage3", "stage4"] , __a: List[str]=[2, 3, 4] , __a: List[str]=None , )-> Union[str, Any]:
lowerCamelCase : Optional[int] = parent
lowerCamelCase : Optional[int] = batch_size
lowerCamelCase : Any = image_size
lowerCamelCase : Tuple = num_channels
lowerCamelCase : str = num_stages
lowerCamelCase : List[str] = hidden_sizes
lowerCamelCase : str = depths
lowerCamelCase : Dict = is_training
lowerCamelCase : Optional[Any] = use_labels
lowerCamelCase : List[str] = intermediate_size
lowerCamelCase : List[str] = hidden_act
lowerCamelCase : List[str] = num_labels
lowerCamelCase : Union[str, Any] = initializer_range
lowerCamelCase : List[Any] = out_features
lowerCamelCase : Optional[Any] = out_indices
lowerCamelCase : int = scope
def a__ ( self: str )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : Dict = None
if self.use_labels:
lowerCamelCase : Dict = ids_tensor([self.batch_size] , self.num_labels )
lowerCamelCase : Any = self.get_config()
return config, pixel_values, labels
def a__ ( self: Dict )-> Union[str, Any]:
return ConvNextConfig(
num_channels=self.num_channels , hidden_sizes=self.hidden_sizes , depths=self.depths , num_stages=self.num_stages , hidden_act=self.hidden_act , is_decoder=__a , initializer_range=self.initializer_range , out_features=self.out_features , out_indices=self.out_indices , num_labels=self.num_labels , )
def a__ ( self: Optional[Any] , __a: List[Any] , __a: Any , __a: int )-> List[Any]:
lowerCamelCase : Optional[int] = ConvNextModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a )
# expected last hidden states: B, C, H // 32, W // 32
self.parent.assertEqual(
result.last_hidden_state.shape , (self.batch_size, self.hidden_sizes[-1], self.image_size // 32, self.image_size // 32) , )
def a__ ( self: int , __a: Union[str, Any] , __a: List[Any] , __a: Tuple )-> Optional[int]:
lowerCamelCase : str = ConvNextForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_labels) )
def a__ ( self: List[Any] , __a: Any , __a: Optional[int] , __a: Tuple )-> List[str]:
lowerCamelCase : List[str] = ConvNextBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a )
# verify hidden states
self.parent.assertEqual(len(result.feature_maps ) , len(config.out_features ) )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.hidden_sizes[1], 4, 4] )
# verify channels
self.parent.assertEqual(len(model.channels ) , len(config.out_features ) )
self.parent.assertListEqual(model.channels , config.hidden_sizes[1:] )
# verify backbone works with out_features=None
lowerCamelCase : Tuple = None
lowerCamelCase : List[str] = ConvNextBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[Any] = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , 1 )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.hidden_sizes[-1], 1, 1] )
# verify channels
self.parent.assertEqual(len(model.channels ) , 1 )
self.parent.assertListEqual(model.channels , [config.hidden_sizes[-1]] )
def a__ ( self: Optional[Any] )-> Any:
lowerCamelCase : List[Any] = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : Tuple = config_and_inputs
lowerCamelCase : int = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : int =(
(
ConvNextModel,
ConvNextForImageClassification,
ConvNextBackbone,
)
if is_torch_available()
else ()
)
snake_case__ : str =(
{'''feature-extraction''': ConvNextModel, '''image-classification''': ConvNextForImageClassification}
if is_torch_available()
else {}
)
snake_case__ : Union[str, Any] =True
snake_case__ : Optional[int] =False
snake_case__ : Tuple =False
snake_case__ : Union[str, Any] =False
snake_case__ : Tuple =False
def a__ ( self: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Tuple = ConvNextModelTester(self )
lowerCamelCase : List[Any] = ConfigTester(self , config_class=__a , has_text_modality=__a , hidden_size=37 )
def a__ ( self: Optional[int] )-> Dict:
self.create_and_test_config_common_properties()
self.config_tester.create_and_test_config_to_json_string()
self.config_tester.create_and_test_config_to_json_file()
self.config_tester.create_and_test_config_from_and_save_pretrained()
self.config_tester.create_and_test_config_with_num_labels()
self.config_tester.check_config_can_be_init_without_params()
self.config_tester.check_config_arguments_init()
def a__ ( self: Optional[int] )-> Optional[Any]:
return
@unittest.skip(reason="""ConvNext does not use inputs_embeds""" )
def a__ ( self: int )-> Dict:
pass
@unittest.skip(reason="""ConvNext does not support input and output embeddings""" )
def a__ ( self: Dict )-> Optional[Any]:
pass
@unittest.skip(reason="""ConvNext does not use feedforward chunking""" )
def a__ ( self: int )-> List[Any]:
pass
def a__ ( self: Union[str, Any] )-> int:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : Any = model_class(__a )
lowerCamelCase : str = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Optional[Any] = [*signature.parameters.keys()]
lowerCamelCase : List[str] = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: Optional[int] )-> str:
lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: str )-> int:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_backbone(*__a )
def a__ ( self: int )-> Optional[int]:
def check_hidden_states_output(__a: Tuple , __a: int , __a: Tuple ):
lowerCamelCase : str = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : Tuple = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : int = outputs.encoder_hidden_states if config.is_encoder_decoder else outputs.hidden_states
lowerCamelCase : Optional[int] = self.model_tester.num_stages
self.assertEqual(len(__a ) , expected_num_stages + 1 )
# ConvNext's feature maps are of shape (batch_size, num_channels, height, width)
self.assertListEqual(
list(hidden_states[0].shape[-2:] ) , [self.model_tester.image_size // 4, self.model_tester.image_size // 4] , )
lowerCamelCase , lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : List[Any] = True
check_hidden_states_output(__a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : Tuple = True
check_hidden_states_output(__a , __a , __a )
def a__ ( self: Dict )-> Optional[Any]:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_classification(*__a )
@slow
def a__ ( self: Optional[Any] )-> Tuple:
for model_name in CONVNEXT_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : str = ConvNextModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def snake_case ( ) -> Optional[int]:
lowerCamelCase : int = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
return image
@require_torch
@require_vision
class A__ ( unittest.TestCase):
"""simple docstring"""
@cached_property
def a__ ( self: Dict )-> Union[str, Any]:
return AutoImageProcessor.from_pretrained("""facebook/convnext-tiny-224""" ) if is_vision_available() else None
@slow
def a__ ( self: List[str] )-> Dict:
lowerCamelCase : Tuple = ConvNextForImageClassification.from_pretrained("""facebook/convnext-tiny-224""" ).to(__a )
lowerCamelCase : Dict = self.default_image_processor
lowerCamelCase : Union[str, Any] = prepare_img()
lowerCamelCase : Optional[Any] = image_processor(images=__a , return_tensors="""pt""" ).to(__a )
# forward pass
with torch.no_grad():
lowerCamelCase : Any = model(**__a )
# verify the logits
lowerCamelCase : Union[str, Any] = torch.Size((1, 1_000) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : Tuple = torch.tensor([-0.02_60, -0.47_39, 0.19_11] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3] , __a , atol=1e-4 ) )
@require_torch
class A__ ( unittest.TestCase , __lowercase):
"""simple docstring"""
snake_case__ : Union[str, Any] =(ConvNextBackbone,) if is_torch_available() else ()
snake_case__ : Optional[Any] =ConvNextConfig
snake_case__ : Optional[Any] =False
def a__ ( self: List[str] )-> int:
lowerCamelCase : Dict = ConvNextModelTester(self )
| 42
| 0
|
"""simple docstring"""
import inspect
import unittest
from transformers import SegformerConfig, is_torch_available, is_vision_available
from transformers.models.auto import get_values
from transformers.testing_utils import require_torch, slow, torch_device
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import (
MODEL_MAPPING,
SegformerForImageClassification,
SegformerForSemanticSegmentation,
SegformerModel,
)
from transformers.models.segformer.modeling_segformer import SEGFORMER_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import SegformerImageProcessor
class A__ ( _UpperCamelCase):
"""simple docstring"""
def a__ ( self: Tuple )-> List[Any]:
lowerCamelCase : int = self.config_class(**self.inputs_dict )
self.parent.assertTrue(hasattr(__a , """hidden_sizes""" ) )
self.parent.assertTrue(hasattr(__a , """num_attention_heads""" ) )
self.parent.assertTrue(hasattr(__a , """num_encoder_blocks""" ) )
class A__ :
"""simple docstring"""
def __init__( self: List[str] , __a: List[Any] , __a: int=13 , __a: Dict=64 , __a: Union[str, Any]=3 , __a: int=4 , __a: Union[str, Any]=[2, 2, 2, 2] , __a: List[Any]=[8, 4, 2, 1] , __a: Tuple=[16, 32, 64, 128] , __a: Any=[1, 4, 8, 16] , __a: Optional[Any]=[1, 2, 4, 8] , __a: str=True , __a: Any=True , __a: Union[str, Any]="gelu" , __a: List[str]=0.1 , __a: Tuple=0.1 , __a: List[str]=0.02 , __a: Optional[int]=3 , __a: Union[str, Any]=None , )-> List[Any]:
lowerCamelCase : Optional[int] = parent
lowerCamelCase : Optional[Any] = batch_size
lowerCamelCase : Tuple = image_size
lowerCamelCase : int = num_channels
lowerCamelCase : Optional[int] = num_encoder_blocks
lowerCamelCase : Optional[Any] = sr_ratios
lowerCamelCase : Any = depths
lowerCamelCase : Dict = hidden_sizes
lowerCamelCase : Union[str, Any] = downsampling_rates
lowerCamelCase : str = num_attention_heads
lowerCamelCase : List[Any] = is_training
lowerCamelCase : Dict = use_labels
lowerCamelCase : str = hidden_act
lowerCamelCase : Optional[Any] = hidden_dropout_prob
lowerCamelCase : Optional[int] = attention_probs_dropout_prob
lowerCamelCase : Tuple = initializer_range
lowerCamelCase : int = num_labels
lowerCamelCase : Union[str, Any] = scope
def a__ ( self: Dict )-> Tuple:
lowerCamelCase : List[str] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : List[Any] = None
if self.use_labels:
lowerCamelCase : Union[str, Any] = ids_tensor([self.batch_size, self.image_size, self.image_size] , self.num_labels )
lowerCamelCase : List[Any] = self.get_config()
return config, pixel_values, labels
def a__ ( self: List[str] )-> Any:
return SegformerConfig(
image_size=self.image_size , num_channels=self.num_channels , num_encoder_blocks=self.num_encoder_blocks , depths=self.depths , hidden_sizes=self.hidden_sizes , num_attention_heads=self.num_attention_heads , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , initializer_range=self.initializer_range , )
def a__ ( self: str , __a: Union[str, Any] , __a: Tuple , __a: Any )-> Optional[int]:
lowerCamelCase : Tuple = SegformerModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Optional[Any] = model(__a )
lowerCamelCase : Optional[Any] = self.image_size // (self.downsampling_rates[-1] * 2)
self.parent.assertEqual(
result.last_hidden_state.shape , (self.batch_size, self.hidden_sizes[-1], expected_height, expected_width) )
def a__ ( self: Union[str, Any] , __a: List[Any] , __a: Any , __a: Union[str, Any] )-> int:
lowerCamelCase : List[str] = self.num_labels
lowerCamelCase : List[str] = SegformerForSemanticSegmentation(__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a )
self.parent.assertEqual(
result.logits.shape , (self.batch_size, self.num_labels, self.image_size // 4, self.image_size // 4) )
lowerCamelCase : Optional[Any] = model(__a , labels=__a )
self.parent.assertEqual(
result.logits.shape , (self.batch_size, self.num_labels, self.image_size // 4, self.image_size // 4) )
self.parent.assertGreater(result.loss , 0.0 )
def a__ ( self: Union[str, Any] , __a: Any , __a: List[Any] , __a: Tuple )-> Tuple:
lowerCamelCase : List[str] = 1
lowerCamelCase : List[Any] = SegformerForSemanticSegmentation(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = torch.randint(0 , 1 , (self.batch_size, self.image_size, self.image_size) ).to(__a )
lowerCamelCase : int = model(__a , labels=__a )
self.parent.assertGreater(result.loss , 0.0 )
def a__ ( self: Any )-> List[Any]:
lowerCamelCase : int = self.prepare_config_and_inputs()
lowerCamelCase : Any = config_and_inputs
lowerCamelCase : Optional[int] = {"pixel_values": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( _UpperCamelCase , _UpperCamelCase , unittest.TestCase):
"""simple docstring"""
snake_case__ : int =(
(
SegformerModel,
SegformerForSemanticSegmentation,
SegformerForImageClassification,
)
if is_torch_available()
else ()
)
snake_case__ : Tuple =(
{
'''feature-extraction''': SegformerModel,
'''image-classification''': SegformerForImageClassification,
'''image-segmentation''': SegformerForSemanticSegmentation,
}
if is_torch_available()
else {}
)
snake_case__ : str =True
snake_case__ : int =False
snake_case__ : Dict =False
snake_case__ : int =False
def a__ ( self: str )-> Optional[Any]:
lowerCamelCase : Dict = SegformerModelTester(self )
lowerCamelCase : Union[str, Any] = SegformerConfigTester(self , config_class=__a )
def a__ ( self: str )-> Any:
self.config_tester.run_common_tests()
def a__ ( self: Union[str, Any] )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: str )-> str:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_binary_image_segmentation(*__a )
def a__ ( self: Tuple )-> int:
lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_segmentation(*__a )
@unittest.skip("""SegFormer does not use inputs_embeds""" )
def a__ ( self: Any )-> Dict:
pass
@unittest.skip("""SegFormer does not have get_input_embeddings method and get_output_embeddings methods""" )
def a__ ( self: Optional[Any] )-> Tuple:
pass
def a__ ( self: List[str] )-> int:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : Any = model_class(__a )
lowerCamelCase : Dict = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Tuple = [*signature.parameters.keys()]
lowerCamelCase : Dict = ["pixel_values"]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: Optional[int] )-> int:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Union[str, Any] = True
for model_class in self.all_model_classes:
lowerCamelCase : Any = True
lowerCamelCase : int = False
lowerCamelCase : str = True
lowerCamelCase : Union[str, Any] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : Optional[int] = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : Dict = outputs.attentions
lowerCamelCase : Optional[int] = sum(self.model_tester.depths )
self.assertEqual(len(__a ) , __a )
# check that output_attentions also work using config
del inputs_dict["output_attentions"]
lowerCamelCase : Union[str, Any] = True
lowerCamelCase : Union[str, Any] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : List[Any] = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : Optional[int] = outputs.attentions
self.assertEqual(len(__a ) , __a )
# verify the first attentions (first block, first layer)
lowerCamelCase : Optional[int] = (self.model_tester.image_size // 4) ** 2
lowerCamelCase : Dict = (self.model_tester.image_size // (4 * self.model_tester.sr_ratios[0])) ** 2
self.assertListEqual(
list(attentions[0].shape[-3:] ) , [self.model_tester.num_attention_heads[0], expected_seq_len, expected_reduced_seq_len] , )
# verify the last attentions (last block, last layer)
lowerCamelCase : Tuple = (self.model_tester.image_size // 32) ** 2
lowerCamelCase : Optional[int] = (self.model_tester.image_size // (32 * self.model_tester.sr_ratios[-1])) ** 2
self.assertListEqual(
list(attentions[-1].shape[-3:] ) , [self.model_tester.num_attention_heads[-1], expected_seq_len, expected_reduced_seq_len] , )
lowerCamelCase : Optional[Any] = len(__a )
# Check attention is always last and order is fine
lowerCamelCase : Optional[int] = True
lowerCamelCase : Optional[int] = True
lowerCamelCase : List[str] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : Optional[Any] = model(**self._prepare_for_class(__a , __a ) )
self.assertEqual(out_len + 1 , len(__a ) )
lowerCamelCase : Any = outputs.attentions
self.assertEqual(len(__a ) , __a )
# verify the first attentions (first block, first layer)
lowerCamelCase : Dict = (self.model_tester.image_size // 4) ** 2
lowerCamelCase : int = (self.model_tester.image_size // (4 * self.model_tester.sr_ratios[0])) ** 2
self.assertListEqual(
list(self_attentions[0].shape[-3:] ) , [self.model_tester.num_attention_heads[0], expected_seq_len, expected_reduced_seq_len] , )
def a__ ( self: List[Any] )-> Dict:
def check_hidden_states_output(__a: Any , __a: Union[str, Any] , __a: str ):
lowerCamelCase : List[str] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : str = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : Dict = outputs.hidden_states
lowerCamelCase : int = self.model_tester.num_encoder_blocks
self.assertEqual(len(__a ) , __a )
# verify the first hidden states (first block)
self.assertListEqual(
list(hidden_states[0].shape[-3:] ) , [
self.model_tester.hidden_sizes[0],
self.model_tester.image_size // 4,
self.model_tester.image_size // 4,
] , )
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes:
lowerCamelCase : List[Any] = True
check_hidden_states_output(__a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : str = True
check_hidden_states_output(__a , __a , __a )
def a__ ( self: Union[str, Any] )-> List[Any]:
if not self.model_tester.is_training:
return
lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : List[Any] = True
for model_class in self.all_model_classes:
if model_class in get_values(__a ):
continue
lowerCamelCase : Optional[int] = model_class(__a )
model.to(__a )
model.train()
lowerCamelCase : List[Any] = self._prepare_for_class(__a , __a , return_labels=__a )
lowerCamelCase : Any = model(**__a ).loss
loss.backward()
@unittest.skip("""Will be fixed soon by reducing the size of the model used for common tests.""" )
def a__ ( self: Union[str, Any] )-> List[Any]:
pass
@slow
def a__ ( self: Any )-> int:
for model_name in SEGFORMER_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : List[str] = SegformerModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def snake_case ( ) -> Any:
lowerCamelCase : Dict = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
return image
@require_torch
class A__ ( unittest.TestCase):
"""simple docstring"""
@slow
def a__ ( self: Dict )-> Union[str, Any]:
# only resize + normalize
lowerCamelCase : str = SegformerImageProcessor(
image_scale=(512, 512) , keep_ratio=__a , align=__a , do_random_crop=__a )
lowerCamelCase : str = SegformerForSemanticSegmentation.from_pretrained("""nvidia/segformer-b0-finetuned-ade-512-512""" ).to(
__a )
lowerCamelCase : str = prepare_img()
lowerCamelCase : List[Any] = image_processor(images=__a , return_tensors="""pt""" )
lowerCamelCase : Optional[Any] = encoded_inputs.pixel_values.to(__a )
with torch.no_grad():
lowerCamelCase : Optional[Any] = model(__a )
lowerCamelCase : List[Any] = torch.Size((1, model.config.num_labels, 128, 128) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : int = torch.tensor(
[
[[-4.63_10, -5.52_32, -6.23_56], [-5.19_21, -6.14_44, -6.59_96], [-5.44_24, -6.27_90, -6.75_74]],
[[-12.13_91, -13.31_22, -13.95_54], [-12.87_32, -13.93_52, -14.35_63], [-12.94_38, -13.82_26, -14.25_13]],
[[-12.51_34, -13.46_86, -14.49_15], [-12.86_69, -14.43_43, -14.77_58], [-13.25_23, -14.58_19, -15.06_94]],
] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3, :3, :3] , __a , atol=1e-4 ) )
@slow
def a__ ( self: Optional[Any] )-> List[Any]:
# only resize + normalize
lowerCamelCase : List[Any] = SegformerImageProcessor(
image_scale=(512, 512) , keep_ratio=__a , align=__a , do_random_crop=__a )
lowerCamelCase : Dict = SegformerForSemanticSegmentation.from_pretrained(
"""nvidia/segformer-b1-finetuned-cityscapes-1024-1024""" ).to(__a )
lowerCamelCase : Union[str, Any] = prepare_img()
lowerCamelCase : List[str] = image_processor(images=__a , return_tensors="""pt""" )
lowerCamelCase : Any = encoded_inputs.pixel_values.to(__a )
with torch.no_grad():
lowerCamelCase : Any = model(__a )
lowerCamelCase : List[Any] = torch.Size((1, model.config.num_labels, 128, 128) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : Dict = torch.tensor(
[
[[-13.57_48, -13.91_11, -12.65_00], [-14.35_00, -15.36_83, -14.23_28], [-14.75_32, -16.04_24, -15.60_87]],
[[-17.16_51, -15.87_25, -12.96_53], [-17.25_80, -17.37_18, -14.82_23], [-16.60_58, -16.87_83, -16.74_52]],
[[-3.64_56, -3.02_09, -1.42_03], [-3.07_97, -3.19_59, -2.00_00], [-1.87_57, -1.92_17, -1.69_97]],
] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3, :3, :3] , __a , atol=1e-1 ) )
@slow
def a__ ( self: Tuple )-> Dict:
# only resize + normalize
lowerCamelCase : int = SegformerImageProcessor(
image_scale=(512, 512) , keep_ratio=__a , align=__a , do_random_crop=__a )
lowerCamelCase : Dict = SegformerForSemanticSegmentation.from_pretrained("""nvidia/segformer-b0-finetuned-ade-512-512""" ).to(
__a )
lowerCamelCase : Tuple = prepare_img()
lowerCamelCase : Dict = image_processor(images=__a , return_tensors="""pt""" )
lowerCamelCase : int = encoded_inputs.pixel_values.to(__a )
with torch.no_grad():
lowerCamelCase : List[str] = model(__a )
lowerCamelCase : Optional[Any] = outputs.logits.detach().cpu()
lowerCamelCase : Union[str, Any] = image_processor.post_process_semantic_segmentation(outputs=__a , target_sizes=[(500, 300)] )
lowerCamelCase : Union[str, Any] = torch.Size((500, 300) )
self.assertEqual(segmentation[0].shape , __a )
lowerCamelCase : Any = image_processor.post_process_semantic_segmentation(outputs=__a )
lowerCamelCase : List[str] = torch.Size((128, 128) )
self.assertEqual(segmentation[0].shape , __a )
| 714
|
"""simple docstring"""
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :Optional[int] = logging.get_logger(__name__)
__lowerCamelCase :List[str] = {
'google/realm-cc-news-pretrained-embedder': (
'https://huggingface.co/google/realm-cc-news-pretrained-embedder/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-encoder': (
'https://huggingface.co/google/realm-cc-news-pretrained-encoder/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-scorer': (
'https://huggingface.co/google/realm-cc-news-pretrained-scorer/resolve/main/config.json'
),
'google/realm-cc-news-pretrained-openqa': (
'https://huggingface.co/google/realm-cc-news-pretrained-openqa/aresolve/main/config.json'
),
'google/realm-orqa-nq-openqa': 'https://huggingface.co/google/realm-orqa-nq-openqa/resolve/main/config.json',
'google/realm-orqa-nq-reader': 'https://huggingface.co/google/realm-orqa-nq-reader/resolve/main/config.json',
'google/realm-orqa-wq-openqa': 'https://huggingface.co/google/realm-orqa-wq-openqa/resolve/main/config.json',
'google/realm-orqa-wq-reader': 'https://huggingface.co/google/realm-orqa-wq-reader/resolve/main/config.json',
# See all REALM models at https://huggingface.co/models?filter=realm
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Optional[Any] ='''realm'''
def __init__( self: Union[str, Any] , __a: List[Any]=30_522 , __a: List[Any]=768 , __a: List[Any]=128 , __a: Union[str, Any]=12 , __a: Union[str, Any]=12 , __a: Optional[Any]=8 , __a: Dict=3_072 , __a: List[Any]="gelu_new" , __a: List[Any]=0.1 , __a: Tuple=0.1 , __a: Optional[Any]=512 , __a: Optional[int]=2 , __a: str=0.02 , __a: int=1e-1_2 , __a: Optional[Any]=256 , __a: Any=10 , __a: Dict=1e-3 , __a: Optional[Any]=5 , __a: Dict=320 , __a: Tuple=13_353_718 , __a: List[Any]=5_000 , __a: Dict=1 , __a: int=0 , __a: Dict=2 , **__a: List[str] , )-> Any:
super().__init__(pad_token_id=__a , bos_token_id=__a , eos_token_id=__a , **__a )
# Common config
lowerCamelCase : Optional[Any] = vocab_size
lowerCamelCase : str = max_position_embeddings
lowerCamelCase : Dict = hidden_size
lowerCamelCase : Dict = retriever_proj_size
lowerCamelCase : Optional[Any] = num_hidden_layers
lowerCamelCase : List[str] = num_attention_heads
lowerCamelCase : Tuple = num_candidates
lowerCamelCase : int = intermediate_size
lowerCamelCase : Dict = hidden_act
lowerCamelCase : List[str] = hidden_dropout_prob
lowerCamelCase : Dict = attention_probs_dropout_prob
lowerCamelCase : Optional[int] = initializer_range
lowerCamelCase : Dict = type_vocab_size
lowerCamelCase : Optional[Any] = layer_norm_eps
# Reader config
lowerCamelCase : List[str] = span_hidden_size
lowerCamelCase : Dict = max_span_width
lowerCamelCase : Optional[Any] = reader_layer_norm_eps
lowerCamelCase : Optional[int] = reader_beam_size
lowerCamelCase : List[Any] = reader_seq_len
# Retrieval config
lowerCamelCase : int = num_block_records
lowerCamelCase : Dict = searcher_beam_size
| 42
| 0
|
"""simple docstring"""
import unittest
from transformers import is_flax_available
from transformers.testing_utils import require_flax, require_sentencepiece, require_tokenizers, require_torch, slow
if is_flax_available():
import optax
from flax.training.common_utils import onehot
from transformers import AutoTokenizer, FlaxMTaForConditionalGeneration
from transformers.models.ta.modeling_flax_ta import shift_tokens_right
@require_torch
@require_sentencepiece
@require_tokenizers
@require_flax
class A__ ( unittest.TestCase):
"""simple docstring"""
@slow
def a__ ( self: Dict )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = FlaxMTaForConditionalGeneration.from_pretrained("""google/mt5-small""" )
lowerCamelCase : int = AutoTokenizer.from_pretrained("""google/mt5-small""" )
lowerCamelCase : Any = tokenizer("""Hello there""" , return_tensors="""np""" ).input_ids
lowerCamelCase : List[str] = tokenizer("""Hi I am""" , return_tensors="""np""" ).input_ids
lowerCamelCase : Any = shift_tokens_right(lowercase_ , model.config.pad_token_id , model.config.decoder_start_token_id )
lowerCamelCase : int = model(lowercase_ , decoder_input_ids=lowercase_ ).logits
lowerCamelCase : List[Any] = optax.softmax_cross_entropy(lowercase_ , onehot(lowercase_ , logits.shape[-1] ) ).mean()
lowerCamelCase : Any = -(labels.shape[-1] * loss.item())
lowerCamelCase : str = -84.91_27
self.assertTrue(abs(mtf_score - EXPECTED_SCORE ) < 1e-4 )
| 715
|
"""simple docstring"""
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :Tuple = logging.get_logger(__name__)
__lowerCamelCase :Any = {
'vinvino02/glpn-kitti': 'https://huggingface.co/vinvino02/glpn-kitti/resolve/main/config.json',
# See all GLPN models at https://huggingface.co/models?filter=glpn
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Tuple ='''glpn'''
def __init__( self: Dict , __a: List[str]=3 , __a: Optional[int]=4 , __a: Dict=[2, 2, 2, 2] , __a: str=[8, 4, 2, 1] , __a: Optional[int]=[32, 64, 160, 256] , __a: Dict=[7, 3, 3, 3] , __a: Dict=[4, 2, 2, 2] , __a: Optional[Any]=[1, 2, 5, 8] , __a: Tuple=[4, 4, 4, 4] , __a: int="gelu" , __a: Union[str, Any]=0.0 , __a: str=0.0 , __a: Union[str, Any]=0.02 , __a: str=0.1 , __a: Union[str, Any]=1e-6 , __a: Any=64 , __a: Dict=10 , __a: Union[str, Any]=-1 , **__a: Optional[Any] , )-> Dict:
super().__init__(**__a )
lowerCamelCase : Dict = num_channels
lowerCamelCase : Any = num_encoder_blocks
lowerCamelCase : Dict = depths
lowerCamelCase : List[str] = sr_ratios
lowerCamelCase : Dict = hidden_sizes
lowerCamelCase : Tuple = patch_sizes
lowerCamelCase : Optional[int] = strides
lowerCamelCase : Optional[Any] = mlp_ratios
lowerCamelCase : Union[str, Any] = num_attention_heads
lowerCamelCase : List[str] = hidden_act
lowerCamelCase : Any = hidden_dropout_prob
lowerCamelCase : Optional[int] = attention_probs_dropout_prob
lowerCamelCase : List[Any] = initializer_range
lowerCamelCase : Dict = drop_path_rate
lowerCamelCase : Any = layer_norm_eps
lowerCamelCase : Optional[Any] = decoder_hidden_size
lowerCamelCase : Tuple = max_depth
lowerCamelCase : Optional[Any] = head_in_index
| 42
| 0
|
"""simple docstring"""
import qiskit
def snake_case ( UpperCamelCase__ : int = 2 ) -> qiskit.result.counts.Counts:
lowerCamelCase : str = qubits
# Using Aer's simulator
lowerCamelCase : Optional[int] = qiskit.Aer.get_backend("""aer_simulator""" )
# Creating a Quantum Circuit acting on the q register
lowerCamelCase : Dict = qiskit.QuantumCircuit(snake_case__ , snake_case__ )
# Adding a H gate on qubit 0 (now q0 in superposition)
circuit.h(0 )
for i in range(1 , snake_case__ ):
# Adding CX (CNOT) gate
circuit.cx(i - 1 , snake_case__ )
# Mapping the quantum measurement to the classical bits
circuit.measure(list(range(snake_case__ ) ) , list(range(snake_case__ ) ) )
# Now measuring any one qubit would affect other qubits to collapse
# their super position and have same state as the measured one.
# Executing the circuit on the simulator
lowerCamelCase : List[Any] = qiskit.execute(snake_case__ , snake_case__ , shots=1000 )
return job.result().get_counts(snake_case__ )
if __name__ == "__main__":
print(f"""Total count for various states are: {quantum_entanglement(3)}""")
| 716
|
"""simple docstring"""
from __future__ import annotations
import math
def snake_case ( UpperCamelCase__ : float , UpperCamelCase__ : int ) -> float:
lowerCamelCase : Dict = u
for i in range(1 , UpperCamelCase__ ):
lowerCamelCase : List[str] = temp * (u - i)
return temp
def snake_case ( ) -> None:
lowerCamelCase : List[Any] = int(input("""enter the numbers of values: """ ) )
lowerCamelCase : list[list[float]] = []
for _ in range(UpperCamelCase__ ):
y.append([] )
for i in range(UpperCamelCase__ ):
for j in range(UpperCamelCase__ ):
y[i].append(UpperCamelCase__ )
lowerCamelCase : Union[str, Any] = 0
print("""enter the values of parameters in a list: """ )
lowerCamelCase : Any = list(map(UpperCamelCase__ , input().split() ) )
print("""enter the values of corresponding parameters: """ )
for i in range(UpperCamelCase__ ):
lowerCamelCase : int = float(input() )
lowerCamelCase : Dict = int(input("""enter the value to interpolate: """ ) )
lowerCamelCase : List[Any] = (value - x[0]) / (x[1] - x[0])
# for calculating forward difference table
for i in range(1 , UpperCamelCase__ ):
for j in range(n - i ):
lowerCamelCase : str = y[j + 1][i - 1] - y[j][i - 1]
lowerCamelCase : Any = y[0][0]
for i in range(1 , UpperCamelCase__ ):
summ += (ucal(UpperCamelCase__ , UpperCamelCase__ ) * y[0][i]) / math.factorial(UpperCamelCase__ )
print(F'the value at {value} is {summ}' )
if __name__ == "__main__":
main()
| 42
| 0
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_tensorflow_text_available, is_torch_available
__lowerCamelCase :Dict = {
'configuration_ernie': ['ERNIE_PRETRAINED_CONFIG_ARCHIVE_MAP', 'ErnieConfig', 'ErnieOnnxConfig'],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :List[str] = [
'ERNIE_PRETRAINED_MODEL_ARCHIVE_LIST',
'ErnieForCausalLM',
'ErnieForMaskedLM',
'ErnieForMultipleChoice',
'ErnieForNextSentencePrediction',
'ErnieForPreTraining',
'ErnieForQuestionAnswering',
'ErnieForSequenceClassification',
'ErnieForTokenClassification',
'ErnieModel',
'ErniePreTrainedModel',
]
if TYPE_CHECKING:
from .configuration_ernie import ERNIE_PRETRAINED_CONFIG_ARCHIVE_MAP, ErnieConfig, ErnieOnnxConfig
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_ernie import (
ERNIE_PRETRAINED_MODEL_ARCHIVE_LIST,
ErnieForCausalLM,
ErnieForMaskedLM,
ErnieForMultipleChoice,
ErnieForNextSentencePrediction,
ErnieForPreTraining,
ErnieForQuestionAnswering,
ErnieForSequenceClassification,
ErnieForTokenClassification,
ErnieModel,
ErniePreTrainedModel,
)
else:
import sys
__lowerCamelCase :Tuple = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 717
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_sentencepiece_available
__lowerCamelCase :str = {}
try:
if not is_sentencepiece_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[Any] = ['GPTSw3Tokenizer']
if TYPE_CHECKING:
try:
if not is_sentencepiece_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .tokenization_gpt_swa import GPTSwaTokenizer
else:
import sys
__lowerCamelCase :Tuple = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
from scipy.stats import spearmanr
import datasets
__lowerCamelCase :Optional[Any] = """
The Spearman rank-order correlation coefficient is a measure of the
relationship between two datasets. Like other correlation coefficients,
this one varies between -1 and +1 with 0 implying no correlation.
Positive correlations imply that as data in dataset x increases, so
does data in dataset y. Negative correlations imply that as x increases,
y decreases. Correlations of -1 or +1 imply an exact monotonic relationship.
Unlike the Pearson correlation, the Spearman correlation does not
assume that both datasets are normally distributed.
The p-value roughly indicates the probability of an uncorrelated system
producing datasets that have a Spearman correlation at least as extreme
as the one computed from these datasets. The p-values are not entirely
reliable but are probably reasonable for datasets larger than 500 or so.
"""
__lowerCamelCase :Any = """
Args:
predictions (`List[float]`): Predicted labels, as returned by a model.
references (`List[float]`): Ground truth labels.
return_pvalue (`bool`): If `True`, returns the p-value. If `False`, returns
only the spearmanr score. Defaults to `False`.
Returns:
spearmanr (`float`): Spearman correlation coefficient.
p-value (`float`): p-value. **Note**: is only returned if `return_pvalue=True` is input.
Examples:
Example 1:
>>> spearmanr_metric = datasets.load_metric(\"spearmanr\")
>>> results = spearmanr_metric.compute(references=[1, 2, 3, 4, 5], predictions=[10, 9, 2.5, 6, 4])
>>> print(results)
{'spearmanr': -0.7}
Example 2:
>>> spearmanr_metric = datasets.load_metric(\"spearmanr\")
>>> results = spearmanr_metric.compute(references=[1, 2, 3, 4, 5],
... predictions=[10, 9, 2.5, 6, 4],
... return_pvalue=True)
>>> print(results['spearmanr'])
-0.7
>>> print(round(results['spearmanr_pvalue'], 2))
0.19
"""
__lowerCamelCase :int = r"""\
@book{kokoska2000crc,
title={CRC standard probability and statistics tables and formulae},
author={Kokoska, Stephen and Zwillinger, Daniel},
year={2000},
publisher={Crc Press}
}
@article{2020SciPy-NMeth,
author = {Virtanen, Pauli and Gommers, Ralf and Oliphant, Travis E. and
Haberland, Matt and Reddy, Tyler and Cournapeau, David and
Burovski, Evgeni and Peterson, Pearu and Weckesser, Warren and
Bright, Jonathan and {van der Walt}, St{\'e}fan J. and
Brett, Matthew and Wilson, Joshua and Millman, K. Jarrod and
Mayorov, Nikolay and Nelson, Andrew R. J. and Jones, Eric and
Kern, Robert and Larson, Eric and Carey, C J and
Polat, {\.I}lhan and Feng, Yu and Moore, Eric W. and
{VanderPlas}, Jake and Laxalde, Denis and Perktold, Josef and
Cimrman, Robert and Henriksen, Ian and Quintero, E. A. and
Harris, Charles R. and Archibald, Anne M. and
Ribeiro, Ant{\^o}nio H. and Pedregosa, Fabian and
{van Mulbregt}, Paul and {SciPy 1.0 Contributors}},
title = {{{SciPy} 1.0: Fundamental Algorithms for Scientific
Computing in Python}},
journal = {Nature Methods},
year = {2020},
volume = {17},
pages = {261--272},
adsurl = {https://rdcu.be/b08Wh},
doi = {10.1038/s41592-019-0686-2},
}
"""
@datasets.utils.file_utils.add_start_docstrings(_DESCRIPTION , _KWARGS_DESCRIPTION)
class A__ ( datasets.Metric):
"""simple docstring"""
def a__ ( self: Any )-> List[str]:
return datasets.MetricInfo(
description=_DESCRIPTION , citation=_CITATION , inputs_description=_KWARGS_DESCRIPTION , features=datasets.Features(
{
"""predictions""": datasets.Value("""float""" ),
"""references""": datasets.Value("""float""" ),
} ) , reference_urls=["""https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.spearmanr.html"""] , )
def a__ ( self: Optional[int] , __a: Any , __a: int , __a: Any=False )-> List[Any]:
lowerCamelCase : List[str] = spearmanr(snake_case__ , snake_case__ )
if return_pvalue:
return {"spearmanr": results[0], "spearmanr_pvalue": results[1]}
else:
return {"spearmanr": results[0]}
| 718
|
"""simple docstring"""
import argparse
import json
from collections import OrderedDict
from functools import partial
from pathlib import Path
import timm
import torch
from huggingface_hub import hf_hub_download
from transformers import LevitConfig, LevitForImageClassificationWithTeacher, LevitImageProcessor
from transformers.utils import logging
logging.set_verbosity_info()
__lowerCamelCase :Dict = logging.get_logger()
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : str , UpperCamelCase__ : LevitConfig , UpperCamelCase__ : Path , UpperCamelCase__ : bool = True ) -> Dict:
print(F'Converting {name}...' )
with torch.no_grad():
if hidden_sizes == 128:
if name[-1] == "S":
lowerCamelCase : Optional[Any] = timm.create_model("""levit_128s""" , pretrained=UpperCamelCase__ )
else:
lowerCamelCase : Dict = timm.create_model("""levit_128""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 192:
lowerCamelCase : Tuple = timm.create_model("""levit_192""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 256:
lowerCamelCase : Optional[int] = timm.create_model("""levit_256""" , pretrained=UpperCamelCase__ )
if hidden_sizes == 384:
lowerCamelCase : Dict = timm.create_model("""levit_384""" , pretrained=UpperCamelCase__ )
from_model.eval()
lowerCamelCase : Optional[Any] = LevitForImageClassificationWithTeacher(UpperCamelCase__ ).eval()
lowerCamelCase : Tuple = OrderedDict()
lowerCamelCase : Optional[Any] = from_model.state_dict()
lowerCamelCase : str = list(from_model.state_dict().keys() )
lowerCamelCase : List[Any] = list(our_model.state_dict().keys() )
print(len(UpperCamelCase__ ) , len(UpperCamelCase__ ) )
for i in range(len(UpperCamelCase__ ) ):
lowerCamelCase : str = weights[og_keys[i]]
our_model.load_state_dict(UpperCamelCase__ )
lowerCamelCase : int = torch.randn((2, 3, 224, 224) )
lowerCamelCase : Any = from_model(UpperCamelCase__ )
lowerCamelCase : List[Any] = our_model(UpperCamelCase__ ).logits
assert torch.allclose(UpperCamelCase__ , UpperCamelCase__ ), "The model logits don't match the original one."
lowerCamelCase : Dict = name
print(UpperCamelCase__ )
if push_to_hub:
our_model.save_pretrained(save_directory / checkpoint_name )
lowerCamelCase : Optional[int] = LevitImageProcessor()
image_processor.save_pretrained(save_directory / checkpoint_name )
print(F'Pushed {checkpoint_name}' )
def snake_case ( UpperCamelCase__ : Path , UpperCamelCase__ : str = None , UpperCamelCase__ : bool = True ) -> Optional[int]:
lowerCamelCase : Optional[Any] = """imagenet-1k-id2label.json"""
lowerCamelCase : List[Any] = 1000
lowerCamelCase : Dict = (1, num_labels)
lowerCamelCase : List[Any] = """huggingface/label-files"""
lowerCamelCase : Optional[int] = num_labels
lowerCamelCase : List[str] = json.load(open(hf_hub_download(UpperCamelCase__ , UpperCamelCase__ , repo_type="""dataset""" ) , """r""" ) )
lowerCamelCase : Any = {int(UpperCamelCase__ ): v for k, v in idalabel.items()}
lowerCamelCase : List[Any] = idalabel
lowerCamelCase : str = {v: k for k, v in idalabel.items()}
lowerCamelCase : Tuple = partial(UpperCamelCase__ , num_labels=UpperCamelCase__ , idalabel=UpperCamelCase__ , labelaid=UpperCamelCase__ )
lowerCamelCase : Optional[int] = {
"""levit-128S""": 128,
"""levit-128""": 128,
"""levit-192""": 192,
"""levit-256""": 256,
"""levit-384""": 384,
}
lowerCamelCase : List[Any] = {
"""levit-128S""": ImageNetPreTrainedConfig(
hidden_sizes=[128, 256, 384] , num_attention_heads=[4, 6, 8] , depths=[2, 3, 4] , key_dim=[16, 16, 16] , drop_path_rate=0 , ),
"""levit-128""": ImageNetPreTrainedConfig(
hidden_sizes=[128, 256, 384] , num_attention_heads=[4, 8, 12] , depths=[4, 4, 4] , key_dim=[16, 16, 16] , drop_path_rate=0 , ),
"""levit-192""": ImageNetPreTrainedConfig(
hidden_sizes=[192, 288, 384] , num_attention_heads=[3, 5, 6] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0 , ),
"""levit-256""": ImageNetPreTrainedConfig(
hidden_sizes=[256, 384, 512] , num_attention_heads=[4, 6, 8] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0 , ),
"""levit-384""": ImageNetPreTrainedConfig(
hidden_sizes=[384, 512, 768] , num_attention_heads=[6, 9, 12] , depths=[4, 4, 4] , key_dim=[32, 32, 32] , drop_path_rate=0.1 , ),
}
if model_name:
convert_weight_and_push(
names_to_hidden_sizes[model_name] , UpperCamelCase__ , names_to_config[model_name] , UpperCamelCase__ , UpperCamelCase__ )
else:
for model_name, config in names_to_config.items():
convert_weight_and_push(names_to_hidden_sizes[model_name] , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return config, expected_shape
if __name__ == "__main__":
__lowerCamelCase :Union[str, Any] = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--model_name',
default=None,
type=str,
help='The name of the model you wish to convert, it must be one of the supported Levit* architecture,',
)
parser.add_argument(
'--pytorch_dump_folder_path',
default='levit-dump-folder/',
type=Path,
required=False,
help='Path to the output PyTorch model directory.',
)
parser.add_argument('--push_to_hub', action='store_true', help='Push model and image processor to the hub')
parser.add_argument(
'--no-push_to_hub',
dest='push_to_hub',
action='store_false',
help='Do not push model and image processor to the hub',
)
__lowerCamelCase :List[Any] = parser.parse_args()
__lowerCamelCase :Path = args.pytorch_dump_folder_path
pytorch_dump_folder_path.mkdir(exist_ok=True, parents=True)
convert_weights_and_push(pytorch_dump_folder_path, args.model_name, args.push_to_hub)
| 42
| 0
|
"""simple docstring"""
# Copyright (c) 2021-, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
####################################################################################################
#
# Note: If when running this conversion script you're getting an exception:
# ModuleNotFoundError: No module named 'megatron.model.enums'
# you need to tell python where to find the clone of Megatron-LM, e.g.:
#
# cd /tmp
# git clone https://github.com/NVIDIA/Megatron-LM
# PYTHONPATH=/tmp/Megatron-LM python src/transformers/models/megatron_gpt2/convert_megatron_gpt2_checkpoint.py ...
#
# if you already have it cloned elsewhere, simply adjust the path to the existing path
#
# If the training was done using a Megatron-LM fork, e.g.,
# https://github.com/microsoft/Megatron-DeepSpeed/ then chances are that you need to have that one
# in your path, i.e., /path/to/Megatron-DeepSpeed/
#
import argparse
import os
import re
import zipfile
import torch
from transformers import AutoTokenizer, GPTaConfig
def snake_case ( UpperCamelCase__ : Optional[int] , UpperCamelCase__ : Dict , UpperCamelCase__ : List[str]=0 ) -> Optional[Any]:
if name is None:
lowerCamelCase : Any = None
else:
lowerCamelCase : Dict = """.""" * max(0 , spaces - 2 ) + """# {:""" + str(50 - spaces ) + """s}"""
lowerCamelCase : Dict = fmt.format(__a )
# Print and recurse (if needed).
if isinstance(__a , __a ):
if msg is not None:
print(__a )
for k in val.keys():
recursive_print(__a , val[k] , spaces + 2 )
elif isinstance(__a , torch.Tensor ):
print(__a , """:""" , val.size() )
else:
print(__a , """:""" , __a )
def snake_case ( UpperCamelCase__ : Any , UpperCamelCase__ : List[Any] , UpperCamelCase__ : int , UpperCamelCase__ : str , UpperCamelCase__ : Tuple ) -> Optional[int]:
lowerCamelCase : str = param.size()
if checkpoint_version == 1.0:
# version 1.0 stores [num_heads * hidden_size * num_splits, :]
lowerCamelCase : Tuple = (num_heads, hidden_size, num_splits) + input_shape[1:]
lowerCamelCase : Tuple = param.view(*__a )
lowerCamelCase : Dict = param.transpose(0 , 2 )
lowerCamelCase : Tuple = param.transpose(1 , 2 ).contiguous()
elif checkpoint_version >= 2.0:
# other versions store [num_heads * num_splits * hidden_size, :]
lowerCamelCase : int = (num_heads, num_splits, hidden_size) + input_shape[1:]
lowerCamelCase : str = param.view(*__a )
lowerCamelCase : List[Any] = param.transpose(0 , 1 ).contiguous()
lowerCamelCase : Tuple = param.view(*__a )
return param
def snake_case ( UpperCamelCase__ : Dict , UpperCamelCase__ : Any , UpperCamelCase__ : List[str] ) -> Any:
lowerCamelCase : Dict = {}
# old versions did not store training args
lowerCamelCase : List[str] = input_state_dict.get("""args""" , __a )
if ds_args is not None:
# do not make the user write a config file when the exact dimensions/sizes are already in the checkpoint
# from pprint import pprint
# pprint(vars(ds_args))
lowerCamelCase : Any = ds_args.padded_vocab_size
lowerCamelCase : List[Any] = ds_args.max_position_embeddings
lowerCamelCase : Any = ds_args.hidden_size
lowerCamelCase : Tuple = ds_args.num_layers
lowerCamelCase : Any = ds_args.num_attention_heads
lowerCamelCase : Optional[int] = ds_args.ffn_hidden_size
# pprint(config)
# The number of heads.
lowerCamelCase : Union[str, Any] = config.n_head
# The hidden_size per head.
lowerCamelCase : List[Any] = config.n_embd // config.n_head
# Megatron-LM checkpoint version
if "checkpoint_version" in input_state_dict.keys():
lowerCamelCase : str = input_state_dict["""checkpoint_version"""]
else:
lowerCamelCase : str = 0.0
# The model.
lowerCamelCase : Optional[int] = input_state_dict["""model"""]
# The language model.
lowerCamelCase : Union[str, Any] = model["""language_model"""]
# The embeddings.
lowerCamelCase : str = lm["""embedding"""]
# The word embeddings.
lowerCamelCase : List[str] = embeddings["""word_embeddings"""]["""weight"""]
# Truncate the embedding table to vocab_size rows.
lowerCamelCase : Optional[int] = word_embeddings[: config.vocab_size, :]
lowerCamelCase : Union[str, Any] = word_embeddings
# The position embeddings.
lowerCamelCase : List[Any] = embeddings["""position_embeddings"""]["""weight"""]
# Read the causal mask dimension (seqlen). [max_sequence_length, hidden_size]
lowerCamelCase : List[Any] = pos_embeddings.size(0 )
if n_positions != config.n_positions:
raise ValueError(
F'pos_embeddings.max_sequence_length={n_positions} and config.n_positions={config.n_positions} don\'t match' )
# Store the position embeddings.
lowerCamelCase : Optional[int] = pos_embeddings
# The transformer.
lowerCamelCase : int = lm["""transformer"""] if """transformer""" in lm.keys() else lm["""encoder"""]
# The regex to extract layer names.
lowerCamelCase : Any = re.compile(R"""layers\.(\d+)\.([a-z0-9_.]+)\.([a-z]+)""" )
# The simple map of names for "automated" rules.
lowerCamelCase : Optional[Any] = {
"""attention.dense""": """.attn.c_proj.""",
"""self_attention.dense""": """.attn.c_proj.""",
"""mlp.dense_h_to_4h""": """.mlp.c_fc.""",
"""mlp.dense_4h_to_h""": """.mlp.c_proj.""",
}
# Extract the layers.
for key, val in transformer.items():
# Match the name.
lowerCamelCase : Optional[Any] = layer_re.match(__a )
# Stop if that's not a layer
if m is None:
break
# The index of the layer.
lowerCamelCase : Optional[int] = int(m.group(1 ) )
# The name of the operation.
lowerCamelCase : int = m.group(2 )
# Is it a weight or a bias?
lowerCamelCase : Union[str, Any] = m.group(3 )
# The name of the layer.
lowerCamelCase : List[Any] = F'transformer.h.{layer_idx}'
# For layernorm(s), simply store the layer norm.
if op_name.endswith("""layernorm""" ):
lowerCamelCase : int = """ln_1""" if op_name.startswith("""input""" ) else """ln_2"""
lowerCamelCase : str = val
# Transpose the QKV matrix.
elif (
op_name == "attention.query_key_value" or op_name == "self_attention.query_key_value"
) and weight_or_bias == "weight":
# Insert a tensor of 1x1xDxD bias.
lowerCamelCase : Dict = torch.tril(torch.ones((n_positions, n_positions) , dtype=torch.floataa ) ).view(
1 , 1 , __a , __a )
lowerCamelCase : str = causal_mask
# Insert a "dummy" tensor for masked_bias.
lowerCamelCase : Dict = torch.tensor(-1E4 , dtype=torch.floataa )
lowerCamelCase : int = masked_bias
lowerCamelCase : Tuple = fix_query_key_value_ordering(__a , __a , 3 , __a , __a )
# Megatron stores (3*D) x D but transformers-GPT2 expects D x 3*D.
lowerCamelCase : List[str] = out_val.transpose(0 , 1 ).contiguous()
# Store.
lowerCamelCase : Union[str, Any] = out_val
# Transpose the bias.
elif (
op_name == "attention.query_key_value" or op_name == "self_attention.query_key_value"
) and weight_or_bias == "bias":
lowerCamelCase : int = fix_query_key_value_ordering(__a , __a , 3 , __a , __a )
# Store. No change of shape.
lowerCamelCase : Optional[int] = out_val
# Transpose the weights.
elif weight_or_bias == "weight":
lowerCamelCase : Optional[Any] = megatron_to_transformers[op_name]
lowerCamelCase : List[str] = val.transpose(0 , 1 )
# Copy the bias.
elif weight_or_bias == "bias":
lowerCamelCase : str = megatron_to_transformers[op_name]
lowerCamelCase : Union[str, Any] = val
# DEBUG.
assert config.n_layer == layer_idx + 1
# The final layernorm.
lowerCamelCase : List[Any] = transformer["""final_layernorm.weight"""]
lowerCamelCase : Optional[Any] = transformer["""final_layernorm.bias"""]
# For LM head, transformers' wants the matrix to weight embeddings.
lowerCamelCase : Tuple = word_embeddings
# It should be done!
return output_state_dict
def snake_case ( ) -> List[str]:
lowerCamelCase : Optional[Any] = argparse.ArgumentParser()
parser.add_argument("""--print-checkpoint-structure""" , action="""store_true""" )
parser.add_argument(
"""path_to_checkpoint""" , type=__a , help="""Path to the checkpoint file (.zip archive or direct .pt file)""" , )
parser.add_argument(
"""--config_file""" , default="""""" , type=__a , help="""An optional config json file describing the pre-trained model.""" , )
lowerCamelCase : Dict = parser.parse_args()
# Extract the basename.
lowerCamelCase : Dict = os.path.dirname(args.path_to_checkpoint )
# Load the model.
# the .zip is very optional, let's keep it for backward compatibility
print(F'Extracting PyTorch state dictionary from {args.path_to_checkpoint}' )
if args.path_to_checkpoint.endswith(""".zip""" ):
with zipfile.ZipFile(args.path_to_checkpoint , """r""" ) as checkpoint:
with checkpoint.open("""release/mp_rank_00/model_optim_rng.pt""" ) as pytorch_dict:
lowerCamelCase : List[Any] = torch.load(__a , map_location="""cpu""" )
else:
lowerCamelCase : str = torch.load(args.path_to_checkpoint , map_location="""cpu""" )
lowerCamelCase : int = input_state_dict.get("""args""" , __a )
# Read the config, or default to the model released by NVIDIA.
if args.config_file == "":
if ds_args is not None:
if ds_args.bias_gelu_fusion:
lowerCamelCase : Tuple = """gelu_fast"""
elif ds_args.openai_gelu:
lowerCamelCase : Union[str, Any] = """gelu_new"""
else:
lowerCamelCase : Tuple = """gelu"""
else:
# in the very early days this used to be "gelu_new"
lowerCamelCase : Tuple = """gelu_new"""
# Spell out all parameters in case the defaults change.
lowerCamelCase : List[Any] = GPTaConfig(
vocab_size=50257 , n_positions=1024 , n_embd=1024 , n_layer=24 , n_head=16 , n_inner=4096 , activation_function=__a , resid_pdrop=0.1 , embd_pdrop=0.1 , attn_pdrop=0.1 , layer_norm_epsilon=1E-5 , initializer_range=0.0_2 , summary_type="""cls_index""" , summary_use_proj=__a , summary_activation=__a , summary_proj_to_labels=__a , summary_first_dropout=0.1 , scale_attn_weights=__a , use_cache=__a , bos_token_id=50256 , eos_token_id=50256 , )
else:
lowerCamelCase : Any = GPTaConfig.from_json_file(args.config_file )
lowerCamelCase : List[Any] = ["""GPT2LMHeadModel"""]
# Convert.
print("""Converting""" )
lowerCamelCase : int = convert_megatron_checkpoint(__a , __a , __a )
# Print the structure of converted state dict.
if args.print_checkpoint_structure:
recursive_print(__a , __a )
# Add tokenizer class info to config
# see https://github.com/huggingface/transformers/issues/13906)
if ds_args is not None:
lowerCamelCase : Optional[Any] = ds_args.tokenizer_type
if tokenizer_type == "GPT2BPETokenizer":
lowerCamelCase : int = """gpt2"""
elif tokenizer_type == "PretrainedFromHF":
lowerCamelCase : Any = ds_args.tokenizer_name_or_path
else:
raise ValueError(F'Unrecognized tokenizer_type {tokenizer_type}' )
else:
lowerCamelCase : List[Any] = """gpt2"""
lowerCamelCase : List[Any] = AutoTokenizer.from_pretrained(__a )
lowerCamelCase : Dict = type(__a ).__name__
lowerCamelCase : Dict = tokenizer_class
# Store the config to file.
print("""Saving config""" )
config.save_pretrained(__a )
# Save tokenizer based on args
print(F'Adding {tokenizer_class} tokenizer files' )
tokenizer.save_pretrained(__a )
# Store the state_dict to file.
lowerCamelCase : Optional[int] = os.path.join(__a , """pytorch_model.bin""" )
print(F'Saving checkpoint to \"{output_checkpoint_file}\"' )
torch.save(__a , __a )
####################################################################################################
if __name__ == "__main__":
main()
####################################################################################################
| 719
|
"""simple docstring"""
import torch
from diffusers import KDPMaDiscreteScheduler
from diffusers.utils import torch_device
from .test_schedulers import SchedulerCommonTest
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : Tuple =(KDPMaDiscreteScheduler,)
snake_case__ : Tuple =10
def a__ ( self: List[Any] , **__a: Optional[int] )-> Union[str, Any]:
lowerCamelCase : int = {
"""num_train_timesteps""": 1_100,
"""beta_start""": 0.00_01,
"""beta_end""": 0.02,
"""beta_schedule""": """linear""",
}
config.update(**__a )
return config
def a__ ( self: Union[str, Any] )-> Any:
for timesteps in [10, 50, 100, 1_000]:
self.check_over_configs(num_train_timesteps=__a )
def a__ ( self: str )-> int:
for beta_start, beta_end in zip([0.0_00_01, 0.00_01, 0.0_01] , [0.00_02, 0.0_02, 0.02] ):
self.check_over_configs(beta_start=__a , beta_end=__a )
def a__ ( self: int )-> Union[str, Any]:
for schedule in ["linear", "scaled_linear"]:
self.check_over_configs(beta_schedule=__a )
def a__ ( self: List[Any] )-> List[Any]:
for prediction_type in ["epsilon", "v_prediction"]:
self.check_over_configs(prediction_type=__a )
def a__ ( self: Union[str, Any] )-> int:
lowerCamelCase : List[str] = self.scheduler_classes[0]
lowerCamelCase : Union[str, Any] = self.get_scheduler_config(prediction_type="""v_prediction""" )
lowerCamelCase : List[str] = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps )
lowerCamelCase : Dict = self.dummy_model()
lowerCamelCase : List[Any] = self.dummy_sample_deter * scheduler.init_noise_sigma
lowerCamelCase : List[Any] = sample.to(__a )
for i, t in enumerate(scheduler.timesteps ):
lowerCamelCase : Optional[Any] = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[int] = model(__a , __a )
lowerCamelCase : Tuple = scheduler.step(__a , __a , __a )
lowerCamelCase : Optional[Any] = output.prev_sample
lowerCamelCase : List[str] = torch.sum(torch.abs(__a ) )
lowerCamelCase : Tuple = torch.mean(torch.abs(__a ) )
if torch_device in ["cpu", "mps"]:
assert abs(result_sum.item() - 4.6_9_3_4e-0_7 ) < 1e-2
assert abs(result_mean.item() - 6.1_1_1_2e-1_0 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 4.6_9_3_4_2_8_6_5_0_1_7_0_9_7_2e-0_7 ) < 1e-2
assert abs(result_mean.item() - 0.00_02 ) < 1e-3
def a__ ( self: Any )-> Any:
if torch_device == "mps":
return
lowerCamelCase : Dict = self.scheduler_classes[0]
lowerCamelCase : Dict = self.get_scheduler_config()
lowerCamelCase : int = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps )
lowerCamelCase : List[Any] = self.dummy_model()
lowerCamelCase : Optional[Any] = self.dummy_sample_deter * scheduler.init_noise_sigma
lowerCamelCase : Optional[int] = sample.to(__a )
for i, t in enumerate(scheduler.timesteps ):
lowerCamelCase : Dict = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[Any] = model(__a , __a )
lowerCamelCase : Tuple = scheduler.step(__a , __a , __a )
lowerCamelCase : str = output.prev_sample
lowerCamelCase : Tuple = torch.sum(torch.abs(__a ) )
lowerCamelCase : Tuple = torch.mean(torch.abs(__a ) )
if torch_device in ["cpu", "mps"]:
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
def a__ ( self: Optional[Any] )-> List[Any]:
if torch_device == "mps":
return
lowerCamelCase : Any = self.scheduler_classes[0]
lowerCamelCase : Union[str, Any] = self.get_scheduler_config()
lowerCamelCase : Optional[Any] = scheduler_class(**__a )
scheduler.set_timesteps(self.num_inference_steps , device=__a )
lowerCamelCase : Union[str, Any] = self.dummy_model()
lowerCamelCase : List[str] = self.dummy_sample_deter.to(__a ) * scheduler.init_noise_sigma
for t in scheduler.timesteps:
lowerCamelCase : Union[str, Any] = scheduler.scale_model_input(__a , __a )
lowerCamelCase : Optional[int] = model(__a , __a )
lowerCamelCase : int = scheduler.step(__a , __a , __a )
lowerCamelCase : int = output.prev_sample
lowerCamelCase : Union[str, Any] = torch.sum(torch.abs(__a ) )
lowerCamelCase : int = torch.mean(torch.abs(__a ) )
if str(__a ).startswith("""cpu""" ):
# The following sum varies between 148 and 156 on mps. Why?
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
else:
# CUDA
assert abs(result_sum.item() - 20.41_25 ) < 1e-2
assert abs(result_mean.item() - 0.02_66 ) < 1e-3
| 42
| 0
|
"""simple docstring"""
import unittest
from transformers import XLMConfig, is_torch_available
from transformers.testing_utils import require_torch, slow, torch_device
from ...generation.test_utils import GenerationTesterMixin
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, ids_tensor, random_attention_mask
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import (
XLMForMultipleChoice,
XLMForQuestionAnswering,
XLMForQuestionAnsweringSimple,
XLMForSequenceClassification,
XLMForTokenClassification,
XLMModel,
XLMWithLMHeadModel,
)
from transformers.models.xlm.modeling_xlm import XLM_PRETRAINED_MODEL_ARCHIVE_LIST
class A__ :
"""simple docstring"""
def __init__( self: int , __a: Any , __a: List[Any]=13 , __a: Optional[Any]=7 , __a: Optional[int]=True , __a: Any=True , __a: Optional[Any]=True , __a: Any=True , __a: int=True , __a: Optional[int]=False , __a: Dict=False , __a: List[Any]=False , __a: str=2 , __a: List[str]=99 , __a: int=0 , __a: int=32 , __a: List[Any]=5 , __a: Tuple=4 , __a: Dict=0.1 , __a: Tuple=0.1 , __a: str=512 , __a: Optional[int]=2 , __a: Tuple=0.02 , __a: List[str]=2 , __a: Optional[int]=4 , __a: Union[str, Any]="last" , __a: Union[str, Any]=True , __a: Optional[Any]=None , __a: int=0 , )-> Dict:
lowerCamelCase : List[str] = parent
lowerCamelCase : Tuple = batch_size
lowerCamelCase : Union[str, Any] = seq_length
lowerCamelCase : Dict = is_training
lowerCamelCase : Dict = use_input_lengths
lowerCamelCase : List[str] = use_token_type_ids
lowerCamelCase : Dict = use_labels
lowerCamelCase : int = gelu_activation
lowerCamelCase : List[str] = sinusoidal_embeddings
lowerCamelCase : Tuple = causal
lowerCamelCase : Tuple = asm
lowerCamelCase : List[Any] = n_langs
lowerCamelCase : List[Any] = vocab_size
lowerCamelCase : List[Any] = n_special
lowerCamelCase : Dict = hidden_size
lowerCamelCase : str = num_hidden_layers
lowerCamelCase : Optional[Any] = num_attention_heads
lowerCamelCase : Dict = hidden_dropout_prob
lowerCamelCase : Optional[Any] = attention_probs_dropout_prob
lowerCamelCase : str = max_position_embeddings
lowerCamelCase : List[Any] = type_sequence_label_size
lowerCamelCase : Optional[int] = initializer_range
lowerCamelCase : Optional[Any] = num_labels
lowerCamelCase : str = num_choices
lowerCamelCase : Tuple = summary_type
lowerCamelCase : Tuple = use_proj
lowerCamelCase : List[Any] = scope
lowerCamelCase : int = bos_token_id
def a__ ( self: Optional[Any] )-> str:
lowerCamelCase : Any = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : List[str] = random_attention_mask([self.batch_size, self.seq_length] )
lowerCamelCase : Union[str, Any] = None
if self.use_input_lengths:
lowerCamelCase : int = (
ids_tensor([self.batch_size] , vocab_size=2 ) + self.seq_length - 2
) # small variation of seq_length
lowerCamelCase : Optional[int] = None
if self.use_token_type_ids:
lowerCamelCase : int = ids_tensor([self.batch_size, self.seq_length] , self.n_langs )
lowerCamelCase : Optional[int] = None
lowerCamelCase : str = None
lowerCamelCase : Optional[Any] = None
if self.use_labels:
lowerCamelCase : Optional[Any] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : Tuple = ids_tensor([self.batch_size, self.seq_length] , self.num_labels )
lowerCamelCase : Optional[Any] = ids_tensor([self.batch_size] , 2 ).float()
lowerCamelCase : Optional[int] = ids_tensor([self.batch_size] , self.num_choices )
lowerCamelCase : Optional[int] = self.get_config()
return (
config,
input_ids,
token_type_ids,
input_lengths,
sequence_labels,
token_labels,
is_impossible_labels,
choice_labels,
input_mask,
)
def a__ ( self: Tuple )-> List[Any]:
return XLMConfig(
vocab_size=self.vocab_size , n_special=self.n_special , emb_dim=self.hidden_size , n_layers=self.num_hidden_layers , n_heads=self.num_attention_heads , dropout=self.hidden_dropout_prob , attention_dropout=self.attention_probs_dropout_prob , gelu_activation=self.gelu_activation , sinusoidal_embeddings=self.sinusoidal_embeddings , asm=self.asm , causal=self.causal , n_langs=self.n_langs , max_position_embeddings=self.max_position_embeddings , initializer_range=self.initializer_range , summary_type=self.summary_type , use_proj=self.use_proj , num_labels=self.num_labels , bos_token_id=self.bos_token_id , )
def a__ ( self: int , __a: int , __a: Dict , __a: Any , __a: str , __a: Dict , __a: Optional[int] , __a: Optional[int] , __a: List[Any] , __a: Optional[Any] , )-> Any:
lowerCamelCase : Tuple = XLMModel(config=snake_case_ )
model.to(snake_case_ )
model.eval()
lowerCamelCase : List[str] = model(snake_case_ , lengths=snake_case_ , langs=snake_case_ )
lowerCamelCase : Optional[Any] = model(snake_case_ , langs=snake_case_ )
lowerCamelCase : str = model(snake_case_ )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) )
def a__ ( self: List[Any] , __a: List[Any] , __a: List[Any] , __a: Any , __a: Optional[int] , __a: int , __a: Tuple , __a: Dict , __a: Union[str, Any] , __a: str , )-> Optional[Any]:
lowerCamelCase : Optional[Any] = XLMWithLMHeadModel(snake_case_ )
model.to(snake_case_ )
model.eval()
lowerCamelCase : int = model(snake_case_ , token_type_ids=snake_case_ , labels=snake_case_ )
self.parent.assertEqual(result.loss.shape , () )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) )
def a__ ( self: Any , __a: List[Any] , __a: List[str] , __a: Optional[int] , __a: Dict , __a: Union[str, Any] , __a: Dict , __a: int , __a: Optional[int] , __a: Optional[Any] , )-> Tuple:
lowerCamelCase : List[Any] = XLMForQuestionAnsweringSimple(snake_case_ )
model.to(snake_case_ )
model.eval()
lowerCamelCase : Optional[int] = model(snake_case_ )
lowerCamelCase : Tuple = model(snake_case_ , start_positions=snake_case_ , end_positions=snake_case_ )
lowerCamelCase : Union[str, Any] = outputs
self.parent.assertEqual(result.start_logits.shape , (self.batch_size, self.seq_length) )
self.parent.assertEqual(result.end_logits.shape , (self.batch_size, self.seq_length) )
def a__ ( self: Optional[Any] , __a: Optional[Any] , __a: List[Any] , __a: int , __a: Union[str, Any] , __a: Dict , __a: Dict , __a: str , __a: int , __a: Optional[int] , )-> Dict:
lowerCamelCase : Union[str, Any] = XLMForQuestionAnswering(snake_case_ )
model.to(snake_case_ )
model.eval()
lowerCamelCase : int = model(snake_case_ )
lowerCamelCase : List[Any] = model(
snake_case_ , start_positions=snake_case_ , end_positions=snake_case_ , cls_index=snake_case_ , is_impossible=snake_case_ , p_mask=snake_case_ , )
lowerCamelCase : Optional[Any] = model(
snake_case_ , start_positions=snake_case_ , end_positions=snake_case_ , cls_index=snake_case_ , is_impossible=snake_case_ , )
(lowerCamelCase ) : Dict = result_with_labels.to_tuple()
lowerCamelCase : Optional[Any] = model(snake_case_ , start_positions=snake_case_ , end_positions=snake_case_ )
(lowerCamelCase ) : List[str] = result_with_labels.to_tuple()
self.parent.assertEqual(result_with_labels.loss.shape , () )
self.parent.assertEqual(result.start_top_log_probs.shape , (self.batch_size, model.config.start_n_top) )
self.parent.assertEqual(result.start_top_index.shape , (self.batch_size, model.config.start_n_top) )
self.parent.assertEqual(
result.end_top_log_probs.shape , (self.batch_size, model.config.start_n_top * model.config.end_n_top) )
self.parent.assertEqual(
result.end_top_index.shape , (self.batch_size, model.config.start_n_top * model.config.end_n_top) )
self.parent.assertEqual(result.cls_logits.shape , (self.batch_size,) )
def a__ ( self: Union[str, Any] , __a: List[Any] , __a: str , __a: List[Any] , __a: List[str] , __a: Dict , __a: Dict , __a: List[Any] , __a: Any , __a: str , )-> Dict:
lowerCamelCase : Union[str, Any] = XLMForSequenceClassification(snake_case_ )
model.to(snake_case_ )
model.eval()
lowerCamelCase : Optional[int] = model(snake_case_ )
lowerCamelCase : Union[str, Any] = model(snake_case_ , labels=snake_case_ )
self.parent.assertEqual(result.loss.shape , () )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
def a__ ( self: Dict , __a: Union[str, Any] , __a: List[Any] , __a: Union[str, Any] , __a: str , __a: Union[str, Any] , __a: str , __a: Tuple , __a: str , __a: int , )-> Dict:
lowerCamelCase : List[str] = self.num_labels
lowerCamelCase : Union[str, Any] = XLMForTokenClassification(snake_case_ )
model.to(snake_case_ )
model.eval()
lowerCamelCase : Optional[Any] = model(snake_case_ , attention_mask=snake_case_ , labels=snake_case_ )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) )
def a__ ( self: List[Any] , __a: Optional[int] , __a: str , __a: Dict , __a: Optional[int] , __a: Optional[Any] , __a: str , __a: Optional[Any] , __a: Any , __a: Dict , )-> List[str]:
lowerCamelCase : List[str] = self.num_choices
lowerCamelCase : Union[str, Any] = XLMForMultipleChoice(config=snake_case_ )
model.to(snake_case_ )
model.eval()
lowerCamelCase : int = input_ids.unsqueeze(1 ).expand(-1 , self.num_choices , -1 ).contiguous()
lowerCamelCase : int = token_type_ids.unsqueeze(1 ).expand(-1 , self.num_choices , -1 ).contiguous()
lowerCamelCase : List[str] = input_mask.unsqueeze(1 ).expand(-1 , self.num_choices , -1 ).contiguous()
lowerCamelCase : List[Any] = model(
snake_case_ , attention_mask=snake_case_ , token_type_ids=snake_case_ , labels=snake_case_ , )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.num_choices) )
def a__ ( self: Optional[Any] )-> Tuple:
lowerCamelCase : Dict = self.prepare_config_and_inputs()
(
lowerCamelCase
) : Any = config_and_inputs
lowerCamelCase : str = {"input_ids": input_ids, "token_type_ids": token_type_ids, "lengths": input_lengths}
return config, inputs_dict
@require_torch
class A__ ( _snake_case , _snake_case , _snake_case , unittest.TestCase):
"""simple docstring"""
snake_case__ : Dict =(
(
XLMModel,
XLMWithLMHeadModel,
XLMForQuestionAnswering,
XLMForSequenceClassification,
XLMForQuestionAnsweringSimple,
XLMForTokenClassification,
XLMForMultipleChoice,
)
if is_torch_available()
else ()
)
snake_case__ : Union[str, Any] =(
(XLMWithLMHeadModel,) if is_torch_available() else ()
) # TODO (PVP): Check other models whether language generation is also applicable
snake_case__ : Optional[int] =(
{
"""feature-extraction""": XLMModel,
"""fill-mask""": XLMWithLMHeadModel,
"""question-answering""": XLMForQuestionAnsweringSimple,
"""text-classification""": XLMForSequenceClassification,
"""text-generation""": XLMWithLMHeadModel,
"""token-classification""": XLMForTokenClassification,
"""zero-shot""": XLMForSequenceClassification,
}
if is_torch_available()
else {}
)
def a__ ( self: str , __a: Optional[int] , __a: Dict , __a: Union[str, Any] , __a: List[Any] , __a: int )-> List[str]:
if (
pipeline_test_casse_name == "QAPipelineTests"
and tokenizer_name is not None
and not tokenizer_name.endswith("""Fast""" )
):
# `QAPipelineTests` fails for a few models when the slower tokenizer are used.
# (The slower tokenizers were never used for pipeline tests before the pipeline testing rework)
# TODO: check (and possibly fix) the `QAPipelineTests` with slower tokenizer
return True
return False
def a__ ( self: Any , __a: Optional[Any] , __a: str , __a: int=False )-> Tuple:
lowerCamelCase : Optional[Any] = super()._prepare_for_class(snake_case_ , snake_case_ , return_labels=snake_case_ )
if return_labels:
if model_class.__name__ == "XLMForQuestionAnswering":
lowerCamelCase : Any = torch.zeros(
self.model_tester.batch_size , dtype=torch.long , device=snake_case_ )
lowerCamelCase : Union[str, Any] = torch.zeros(
self.model_tester.batch_size , dtype=torch.long , device=snake_case_ )
return inputs_dict
def a__ ( self: Optional[Any] )-> Any:
lowerCamelCase : int = XLMModelTester(self )
lowerCamelCase : Any = ConfigTester(self , config_class=snake_case_ , emb_dim=37 )
def a__ ( self: List[Any] )-> Tuple:
self.config_tester.run_common_tests()
def a__ ( self: Tuple )-> Optional[int]:
lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_xlm_model(*snake_case_ )
def a__ ( self: Optional[Any] )-> int:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_xlm_lm_head(*snake_case_ )
def a__ ( self: Optional[Any] )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_xlm_simple_qa(*snake_case_ )
def a__ ( self: Optional[int] )-> int:
lowerCamelCase : str = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_xlm_qa(*snake_case_ )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_xlm_sequence_classif(*snake_case_ )
def a__ ( self: Tuple )-> Union[str, Any]:
lowerCamelCase : Any = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_xlm_token_classif(*snake_case_ )
def a__ ( self: str )-> List[str]:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_xlm_for_multiple_choice(*snake_case_ )
def a__ ( self: Any , __a: Optional[int] , __a: Optional[Any] , __a: Optional[int] , __a: int , __a: Dict , __a: Optional[int]=False , __a: Union[str, Any]=1 )-> Any:
self.assertIsInstance(snake_case_ , snake_case_ )
self.assertListEqual(
[isinstance(snake_case_ , snake_case_ ) for iter_attentions in attentions] , [True] * len(snake_case_ ) )
self.assertEqual(len(snake_case_ ) , (max_length - min_length) * num_beam_groups )
for idx, iter_attentions in enumerate(snake_case_ ):
# adds PAD dummy token
lowerCamelCase : Optional[Any] = min_length + idx + 1
lowerCamelCase : Optional[Any] = min_length + idx + 1
lowerCamelCase : Any = (
batch_size * num_beam_groups,
config.num_attention_heads,
tgt_len,
src_len,
)
# check attn size
self.assertListEqual(
[layer_attention.shape for layer_attention in iter_attentions] , [expected_shape] * len(snake_case_ ) )
def a__ ( self: Dict , __a: Any , __a: List[str] , __a: Any , __a: Dict , __a: int , __a: Dict=False , __a: Optional[int]=1 )-> Union[str, Any]:
self.assertIsInstance(snake_case_ , snake_case_ )
self.assertListEqual(
[isinstance(snake_case_ , snake_case_ ) for iter_hidden_states in hidden_states] , [True] * len(snake_case_ ) , )
self.assertEqual(len(snake_case_ ) , (max_length - min_length) * num_beam_groups )
for idx, iter_hidden_states in enumerate(snake_case_ ):
# adds PAD dummy token
lowerCamelCase : Dict = min_length + idx + 1
lowerCamelCase : Optional[Any] = (batch_size * num_beam_groups, seq_len, config.hidden_size)
# check hidden size
self.assertListEqual(
[layer_hidden_states.shape for layer_hidden_states in iter_hidden_states] , [expected_shape] * len(snake_case_ ) , )
pass
@slow
def a__ ( self: Dict )-> str:
for model_name in XLM_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : Any = XLMModel.from_pretrained(snake_case_ )
self.assertIsNotNone(snake_case_ )
@require_torch
class A__ ( unittest.TestCase):
"""simple docstring"""
@slow
def a__ ( self: str )-> Any:
lowerCamelCase : Optional[Any] = XLMWithLMHeadModel.from_pretrained("""xlm-mlm-en-2048""" )
model.to(snake_case_ )
lowerCamelCase : List[Any] = torch.tensor([[14, 447]] , dtype=torch.long , device=snake_case_ ) # the president
lowerCamelCase : List[Any] = [
14,
447,
14,
447,
14,
447,
14,
447,
14,
447,
14,
447,
14,
447,
14,
447,
14,
447,
14,
447,
] # the president the president the president the president the president the president the president the president the president the president
# TODO(PVP): this and other input_ids I tried for generation give pretty bad results. Not sure why. Model might just not be made for auto-regressive inference
lowerCamelCase : Any = model.generate(snake_case_ , do_sample=snake_case_ )
self.assertListEqual(output_ids[0].cpu().numpy().tolist() , snake_case_ )
| 720
|
"""simple docstring"""
import gc
import random
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTextModelWithProjection, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DiffusionPipeline,
EulerDiscreteScheduler,
StableDiffusionXLImgaImgPipeline,
UNetaDConditionModel,
)
from diffusers.utils import floats_tensor, slow, torch_device
from diffusers.utils.testing_utils import enable_full_determinism, require_torch_gpu
from ..pipeline_params import (
IMAGE_TO_IMAGE_IMAGE_PARAMS,
TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS,
TEXT_GUIDED_IMAGE_VARIATION_PARAMS,
)
from ..test_pipelines_common import PipelineLatentTesterMixin, PipelineTesterMixin
enable_full_determinism()
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : str =StableDiffusionXLImgaImgPipeline
snake_case__ : Any =TEXT_GUIDED_IMAGE_VARIATION_PARAMS - {'''height''', '''width'''}
snake_case__ : Optional[int] =PipelineTesterMixin.required_optional_params - {'''latents'''}
snake_case__ : Dict =TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS
snake_case__ : Tuple =IMAGE_TO_IMAGE_IMAGE_PARAMS
snake_case__ : List[str] =IMAGE_TO_IMAGE_IMAGE_PARAMS
def a__ ( self: List[str] )-> int:
torch.manual_seed(0 )
lowerCamelCase : Any = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=2 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , attention_head_dim=(2, 4) , use_linear_projection=__a , addition_embed_type="""text_time""" , addition_time_embed_dim=8 , transformer_layers_per_block=(1, 2) , projection_class_embeddings_input_dim=80 , cross_attention_dim=64 , )
lowerCamelCase : Any = EulerDiscreteScheduler(
beta_start=0.0_00_85 , beta_end=0.0_12 , steps_offset=1 , beta_schedule="""scaled_linear""" , timestep_spacing="""leading""" , )
torch.manual_seed(0 )
lowerCamelCase : Any = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
lowerCamelCase : Optional[Any] = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-0_5 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1_000 , hidden_act="""gelu""" , projection_dim=32 , )
lowerCamelCase : Dict = CLIPTextModel(__a )
lowerCamelCase : Union[str, Any] = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" , local_files_only=__a )
lowerCamelCase : Dict = CLIPTextModelWithProjection(__a )
lowerCamelCase : Optional[int] = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" , local_files_only=__a )
lowerCamelCase : str = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""text_encoder_2""": text_encoder_a,
"""tokenizer_2""": tokenizer_a,
# "safety_checker": None,
# "feature_extractor": None,
}
return components
def a__ ( self: Any , __a: str , __a: Tuple=0 )-> Union[str, Any]:
lowerCamelCase : List[Any] = floats_tensor((1, 3, 32, 32) , rng=random.Random(__a ) ).to(__a )
lowerCamelCase : Any = image / 2 + 0.5
if str(__a ).startswith("""mps""" ):
lowerCamelCase : Dict = torch.manual_seed(__a )
else:
lowerCamelCase : Tuple = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : Tuple = {
"""prompt""": """A painting of a squirrel eating a burger""",
"""image""": image,
"""generator""": generator,
"""num_inference_steps""": 2,
"""guidance_scale""": 5.0,
"""output_type""": """numpy""",
"""strength""": 0.75,
}
return inputs
def a__ ( self: Dict )-> Optional[Any]:
lowerCamelCase : Any = """cpu""" # ensure determinism for the device-dependent torch.Generator
lowerCamelCase : Union[str, Any] = self.get_dummy_components()
lowerCamelCase : Optional[int] = StableDiffusionXLImgaImgPipeline(**__a )
lowerCamelCase : int = sd_pipe.to(__a )
sd_pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Optional[Any] = self.get_dummy_inputs(__a )
lowerCamelCase : Optional[int] = sd_pipe(**__a ).images
lowerCamelCase : Tuple = image[0, -3:, -3:, -1]
assert image.shape == (1, 32, 32, 3)
lowerCamelCase : Any = np.array([0.46_56, 0.48_40, 0.44_39, 0.66_98, 0.55_74, 0.45_24, 0.57_99, 0.59_43, 0.51_65] )
assert np.abs(image_slice.flatten() - expected_slice ).max() < 1e-2
def a__ ( self: Optional[int] )-> Union[str, Any]:
super().test_attention_slicing_forward_pass(expected_max_diff=3e-3 )
def a__ ( self: Optional[Any] )-> str:
super().test_inference_batch_single_identical(expected_max_diff=3e-3 )
def a__ ( self: List[str] )-> Optional[Any]:
pass
def a__ ( self: List[Any] )-> Union[str, Any]:
lowerCamelCase : Tuple = self.get_dummy_components()
lowerCamelCase : Union[str, Any] = StableDiffusionXLImgaImgPipeline(**__a )
lowerCamelCase : str = sd_pipe.to(__a )
lowerCamelCase : Any = sd_pipe.to(__a )
sd_pipe.set_progress_bar_config(disable=__a )
# forward without prompt embeds
lowerCamelCase : Dict = self.get_dummy_inputs(__a )
lowerCamelCase : Any = 3 * ["""this is a negative prompt"""]
lowerCamelCase : Optional[int] = negative_prompt
lowerCamelCase : Tuple = 3 * [inputs["""prompt"""]]
lowerCamelCase : List[Any] = sd_pipe(**__a )
lowerCamelCase : Optional[int] = output.images[0, -3:, -3:, -1]
# forward with prompt embeds
lowerCamelCase : Tuple = self.get_dummy_inputs(__a )
lowerCamelCase : List[Any] = 3 * ["""this is a negative prompt"""]
lowerCamelCase : Tuple = 3 * [inputs.pop("""prompt""" )]
(
(
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) ,
) : Union[str, Any] = sd_pipe.encode_prompt(__a , negative_prompt=__a )
lowerCamelCase : int = sd_pipe(
**__a , prompt_embeds=__a , negative_prompt_embeds=__a , pooled_prompt_embeds=__a , negative_pooled_prompt_embeds=__a , )
lowerCamelCase : Union[str, Any] = output.images[0, -3:, -3:, -1]
# make sure that it's equal
assert np.abs(image_slice_a.flatten() - image_slice_a.flatten() ).max() < 1e-4
@slow
@require_torch_gpu
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: Dict )-> str:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: Union[str, Any] , __a: Any , __a: Any="cpu" , __a: str=torch.floataa , __a: Any=0 )-> Optional[Any]:
lowerCamelCase : Optional[Any] = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : List[Any] = np.random.RandomState(__a ).standard_normal((1, 4, 64, 64) )
lowerCamelCase : List[str] = torch.from_numpy(__a ).to(device=__a , dtype=__a )
lowerCamelCase : int = {
"""prompt""": """a photograph of an astronaut riding a horse""",
"""latents""": latents,
"""generator""": generator,
"""num_inference_steps""": 3,
"""guidance_scale""": 7.5,
"""output_type""": """numpy""",
}
return inputs
def a__ ( self: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = DiffusionPipeline.from_pretrained("""stabilityai/stable-diffusion-2-base""" )
pipe.to(__a )
pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Optional[int] = self.get_inputs(__a )
lowerCamelCase : Optional[Any] = pipe(**__a ).images
lowerCamelCase : Dict = image[0, -3:, -3:, -1].flatten()
assert image.shape == (1, 512, 512, 3)
lowerCamelCase : List[str] = np.array([0.4_94_93, 0.4_78_96, 0.4_07_98, 0.5_42_14, 0.5_32_12, 0.4_82_02, 0.4_76_56, 0.4_63_29, 0.4_85_06] )
assert np.abs(image_slice - expected_slice ).max() < 7e-3
| 42
| 0
|
"""simple docstring"""
import functools
import operator
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :Tuple = logging.get_logger(__name__)
__lowerCamelCase :List[str] = {
'facebook/wav2vec2-base-960h': 'https://huggingface.co/facebook/wav2vec2-base-960h/resolve/main/config.json',
# See all Wav2Vec2 models at https://huggingface.co/models?filter=wav2vec2
}
class A__ ( lowercase__):
"""simple docstring"""
snake_case__ : Any ='''wav2vec2'''
def __init__( self: List[Any] , __a: Union[str, Any]=32 , __a: int=768 , __a: Optional[Any]=12 , __a: Union[str, Any]=12 , __a: Union[str, Any]=3_072 , __a: Tuple="gelu" , __a: List[Any]=0.1 , __a: int=0.1 , __a: Any=0.1 , __a: Optional[Any]=0.0 , __a: Tuple=0.0 , __a: Optional[int]=0.1 , __a: Tuple=0.1 , __a: List[str]=0.02 , __a: Optional[Any]=1e-5 , __a: List[Any]="group" , __a: Dict="gelu" , __a: Union[str, Any]=(512, 512, 512, 512, 512, 512, 512) , __a: Optional[Any]=(5, 2, 2, 2, 2, 2, 2) , __a: Any=(10, 3, 3, 3, 3, 2, 2) , __a: Dict=False , __a: List[str]=128 , __a: Union[str, Any]=16 , __a: List[str]=False , __a: int=True , __a: Any=0.05 , __a: Union[str, Any]=10 , __a: List[str]=2 , __a: List[Any]=0.0 , __a: List[str]=10 , __a: Union[str, Any]=0 , __a: List[str]=320 , __a: List[str]=2 , __a: List[str]=0.1 , __a: Tuple=100 , __a: Optional[Any]=256 , __a: List[Any]=256 , __a: str=0.1 , __a: Dict="sum" , __a: List[str]=False , __a: Optional[int]=False , __a: Union[str, Any]=256 , __a: Dict=(512, 512, 512, 512, 1_500) , __a: Optional[int]=(5, 3, 3, 1, 1) , __a: Optional[int]=(1, 2, 3, 1, 1) , __a: List[str]=512 , __a: Optional[Any]=0 , __a: List[str]=1 , __a: List[Any]=2 , __a: Dict=False , __a: str=3 , __a: int=2 , __a: Optional[Any]=3 , __a: str=None , __a: Tuple=None , **__a: Optional[Any] , )-> Optional[int]:
super().__init__(**UpperCAmelCase__ , pad_token_id=UpperCAmelCase__ , bos_token_id=UpperCAmelCase__ , eos_token_id=UpperCAmelCase__ )
lowerCamelCase : Union[str, Any] = hidden_size
lowerCamelCase : Optional[int] = feat_extract_norm
lowerCamelCase : Optional[int] = feat_extract_activation
lowerCamelCase : Dict = list(UpperCAmelCase__ )
lowerCamelCase : str = list(UpperCAmelCase__ )
lowerCamelCase : str = list(UpperCAmelCase__ )
lowerCamelCase : List[Any] = conv_bias
lowerCamelCase : List[Any] = num_conv_pos_embeddings
lowerCamelCase : Dict = num_conv_pos_embedding_groups
lowerCamelCase : Optional[int] = len(self.conv_dim )
lowerCamelCase : Any = num_hidden_layers
lowerCamelCase : int = intermediate_size
lowerCamelCase : List[str] = hidden_act
lowerCamelCase : Dict = num_attention_heads
lowerCamelCase : Tuple = hidden_dropout
lowerCamelCase : Optional[int] = attention_dropout
lowerCamelCase : Tuple = activation_dropout
lowerCamelCase : str = feat_proj_dropout
lowerCamelCase : int = final_dropout
lowerCamelCase : Optional[Any] = layerdrop
lowerCamelCase : Union[str, Any] = layer_norm_eps
lowerCamelCase : List[str] = initializer_range
lowerCamelCase : str = vocab_size
lowerCamelCase : Optional[Any] = do_stable_layer_norm
lowerCamelCase : List[Any] = use_weighted_layer_sum
if (
(len(self.conv_stride ) != self.num_feat_extract_layers)
or (len(self.conv_kernel ) != self.num_feat_extract_layers)
or (len(self.conv_dim ) != self.num_feat_extract_layers)
):
raise ValueError(
"""Configuration for convolutional layers is incorrect. It is required that `len(config.conv_dim)` =="""
""" `len(config.conv_stride)` == `len(config.conv_kernel)`, but is `len(config.conv_dim) ="""
f' {len(self.conv_dim )}`, `len(config.conv_stride) = {len(self.conv_stride )}`,'
f' `len(config.conv_kernel) = {len(self.conv_kernel )}`.' )
# fine-tuning config parameters for SpecAugment: https://arxiv.org/abs/1904.08779
lowerCamelCase : List[Any] = apply_spec_augment
lowerCamelCase : Optional[Any] = mask_time_prob
lowerCamelCase : Union[str, Any] = mask_time_length
lowerCamelCase : Optional[int] = mask_time_min_masks
lowerCamelCase : Union[str, Any] = mask_feature_prob
lowerCamelCase : Any = mask_feature_length
lowerCamelCase : int = mask_feature_min_masks
# parameters for pretraining with codevector quantized representations
lowerCamelCase : Optional[int] = num_codevectors_per_group
lowerCamelCase : int = num_codevector_groups
lowerCamelCase : Optional[Any] = contrastive_logits_temperature
lowerCamelCase : str = feat_quantizer_dropout
lowerCamelCase : List[str] = num_negatives
lowerCamelCase : int = codevector_dim
lowerCamelCase : Any = proj_codevector_dim
lowerCamelCase : List[Any] = diversity_loss_weight
# ctc loss
lowerCamelCase : Tuple = ctc_loss_reduction
lowerCamelCase : Optional[Any] = ctc_zero_infinity
# adapter
lowerCamelCase : Dict = add_adapter
lowerCamelCase : str = adapter_kernel_size
lowerCamelCase : Optional[Any] = adapter_stride
lowerCamelCase : Union[str, Any] = num_adapter_layers
lowerCamelCase : Union[str, Any] = output_hidden_size or hidden_size
lowerCamelCase : Dict = adapter_attn_dim
# SequenceClassification-specific parameter. Feel free to ignore for other classes.
lowerCamelCase : Union[str, Any] = classifier_proj_size
# XVector-specific parameters. Feel free to ignore for other classes.
lowerCamelCase : List[str] = list(UpperCAmelCase__ )
lowerCamelCase : List[str] = list(UpperCAmelCase__ )
lowerCamelCase : Dict = list(UpperCAmelCase__ )
lowerCamelCase : Optional[int] = xvector_output_dim
@property
def a__ ( self: Any )-> Any:
return functools.reduce(operator.mul , self.conv_stride , 1 )
| 721
|
"""simple docstring"""
import unittest
from pathlib import Path
from tempfile import NamedTemporaryFile, TemporaryDirectory
from transformers import BertConfig, BertTokenizerFast, FeatureExtractionPipeline
from transformers.convert_graph_to_onnx import (
convert,
ensure_valid_input,
generate_identified_filename,
infer_shapes,
quantize,
)
from transformers.testing_utils import require_tf, require_tokenizers, require_torch, slow
class A__ :
"""simple docstring"""
def a__ ( self: Optional[int] , __a: Optional[int] , __a: Tuple , __a: Optional[int] )-> List[str]:
return None
class A__ :
"""simple docstring"""
def a__ ( self: Optional[int] , __a: Tuple , __a: str , __a: str , __a: str )-> Tuple:
return None
class A__ ( unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[Any] =[
# (model_name, model_kwargs)
('''bert-base-cased''', {}),
('''gpt2''', {'''use_cache''': False}), # We don't support exporting GPT2 past keys anymore
]
@require_tf
@slow
def a__ ( self: Optional[Any] )-> int:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
self._test_export(__a , """tf""" , 12 , **__a )
@require_torch
@slow
def a__ ( self: str )-> int:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
self._test_export(__a , """pt""" , 12 , **__a )
@require_torch
@slow
def a__ ( self: Union[str, Any] )-> Dict:
from transformers import BertModel
lowerCamelCase : int = ["""[UNK]""", """[SEP]""", """[CLS]""", """[PAD]""", """[MASK]""", """some""", """other""", """words"""]
with NamedTemporaryFile(mode="""w+t""" ) as vocab_file:
vocab_file.write("""\n""".join(__a ) )
vocab_file.flush()
lowerCamelCase : Dict = BertTokenizerFast(vocab_file.name )
with TemporaryDirectory() as bert_save_dir:
lowerCamelCase : List[str] = BertModel(BertConfig(vocab_size=len(__a ) ) )
model.save_pretrained(__a )
self._test_export(__a , """pt""" , 12 , __a )
@require_tf
@slow
def a__ ( self: Optional[Any] )-> Optional[int]:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
lowerCamelCase : Optional[int] = self._test_export(__a , """tf""" , 12 , **__a )
lowerCamelCase : Tuple = quantize(Path(__a ) )
# Ensure the actual quantized model is not bigger than the original one
if quantized_path.stat().st_size >= Path(__a ).stat().st_size:
self.fail("""Quantized model is bigger than initial ONNX model""" )
@require_torch
@slow
def a__ ( self: Any )-> Optional[int]:
for model, model_kwargs in OnnxExportTestCase.MODEL_TO_TEST:
lowerCamelCase : Any = self._test_export(__a , """pt""" , 12 , **__a )
lowerCamelCase : Dict = quantize(__a )
# Ensure the actual quantized model is not bigger than the original one
if quantized_path.stat().st_size >= Path(__a ).stat().st_size:
self.fail("""Quantized model is bigger than initial ONNX model""" )
def a__ ( self: List[Any] , __a: Optional[Any] , __a: List[Any] , __a: Union[str, Any] , __a: Optional[Any]=None , **__a: Optional[int] )-> Any:
try:
# Compute path
with TemporaryDirectory() as tempdir:
lowerCamelCase : Optional[Any] = Path(__a ).joinpath("""model.onnx""" )
# Remove folder if exists
if path.parent.exists():
path.parent.rmdir()
# Export
convert(__a , __a , __a , __a , __a , **__a )
return path
except Exception as e:
self.fail(__a )
@require_torch
@require_tokenizers
@slow
def a__ ( self: Tuple )-> Dict:
from transformers import BertModel
lowerCamelCase : int = BertModel(BertConfig.from_pretrained("""lysandre/tiny-bert-random""" ) )
lowerCamelCase : List[Any] = BertTokenizerFast.from_pretrained("""lysandre/tiny-bert-random""" )
self._test_infer_dynamic_axis(__a , __a , """pt""" )
@require_tf
@require_tokenizers
@slow
def a__ ( self: Optional[Any] )-> List[Any]:
from transformers import TFBertModel
lowerCamelCase : Union[str, Any] = TFBertModel(BertConfig.from_pretrained("""lysandre/tiny-bert-random""" ) )
lowerCamelCase : str = BertTokenizerFast.from_pretrained("""lysandre/tiny-bert-random""" )
self._test_infer_dynamic_axis(__a , __a , """tf""" )
def a__ ( self: List[str] , __a: str , __a: Optional[Any] , __a: str )-> List[Any]:
lowerCamelCase : List[str] = FeatureExtractionPipeline(__a , __a )
lowerCamelCase : List[str] = ["""input_ids""", """token_type_ids""", """attention_mask""", """output_0""", """output_1"""]
lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase : Tuple = infer_shapes(__a , __a )
# Assert all variables are present
self.assertEqual(len(__a ) , len(__a ) )
self.assertTrue(all(var_name in shapes for var_name in variable_names ) )
self.assertSequenceEqual(variable_names[:3] , __a )
self.assertSequenceEqual(variable_names[3:] , __a )
# Assert inputs are {0: batch, 1: sequence}
for var_name in ["input_ids", "token_type_ids", "attention_mask"]:
self.assertDictEqual(shapes[var_name] , {0: """batch""", 1: """sequence"""} )
# Assert outputs are {0: batch, 1: sequence} and {0: batch}
self.assertDictEqual(shapes["""output_0"""] , {0: """batch""", 1: """sequence"""} )
self.assertDictEqual(shapes["""output_1"""] , {0: """batch"""} )
def a__ ( self: List[Any] )-> int:
lowerCamelCase : List[str] = ["""input_ids""", """attention_mask""", """token_type_ids"""]
lowerCamelCase : str = {"""input_ids""": [1, 2, 3, 4], """attention_mask""": [0, 0, 0, 0], """token_type_ids""": [1, 1, 1, 1]}
lowerCamelCase , lowerCamelCase : List[Any] = ensure_valid_input(FuncContiguousArgs() , __a , __a )
# Should have exactly the same number of args (all are valid)
self.assertEqual(len(__a ) , 3 )
# Should have exactly the same input names
self.assertEqual(set(__a ) , set(__a ) )
# Parameter should be reordered according to their respective place in the function:
# (input_ids, token_type_ids, attention_mask)
self.assertEqual(__a , (tokens["""input_ids"""], tokens["""token_type_ids"""], tokens["""attention_mask"""]) )
# Generated args are interleaved with another args (for instance parameter "past" in GPT2)
lowerCamelCase , lowerCamelCase : List[Any] = ensure_valid_input(FuncNonContiguousArgs() , __a , __a )
# Should have exactly the one arg (all before the one not provided "some_other_args")
self.assertEqual(len(__a ) , 1 )
self.assertEqual(len(__a ) , 1 )
# Should have only "input_ids"
self.assertEqual(inputs_args[0] , tokens["""input_ids"""] )
self.assertEqual(ordered_input_names[0] , """input_ids""" )
def a__ ( self: Tuple )-> Tuple:
lowerCamelCase : Optional[int] = generate_identified_filename(Path("""/home/something/my_fake_model.onnx""" ) , """-test""" )
self.assertEqual("""/home/something/my_fake_model-test.onnx""" , generated.as_posix() )
| 42
| 0
|
"""simple docstring"""
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :List[Any] = logging.get_logger(__name__)
__lowerCamelCase :Dict = {
'''caidas/swin2sr-classicalsr-x2-64''': (
'''https://huggingface.co/caidas/swin2sr-classicalsr-x2-64/resolve/main/config.json'''
),
}
class A__ ( __UpperCAmelCase):
"""simple docstring"""
snake_case__ : Optional[Any] ="swin2sr"
snake_case__ : str ={
"hidden_size": "embed_dim",
"num_attention_heads": "num_heads",
"num_hidden_layers": "num_layers",
}
def __init__( self: Optional[int] , __a: Dict=64 , __a: str=1 , __a: Optional[int]=3 , __a: List[Any]=180 , __a: Optional[int]=[6, 6, 6, 6, 6, 6] , __a: Union[str, Any]=[6, 6, 6, 6, 6, 6] , __a: Tuple=8 , __a: List[Any]=2.0 , __a: Dict=True , __a: Dict=0.0 , __a: int=0.0 , __a: Optional[int]=0.1 , __a: Any="gelu" , __a: int=False , __a: Tuple=0.02 , __a: List[str]=1e-5 , __a: List[Any]=2 , __a: str=1.0 , __a: List[Any]="1conv" , __a: str="pixelshuffle" , **__a: List[str] , )-> str:
super().__init__(**_lowerCamelCase )
lowerCamelCase : int = image_size
lowerCamelCase : List[str] = patch_size
lowerCamelCase : Dict = num_channels
lowerCamelCase : Optional[int] = embed_dim
lowerCamelCase : List[Any] = depths
lowerCamelCase : Optional[Any] = len(_lowerCamelCase )
lowerCamelCase : Any = num_heads
lowerCamelCase : Any = window_size
lowerCamelCase : int = mlp_ratio
lowerCamelCase : Optional[int] = qkv_bias
lowerCamelCase : int = hidden_dropout_prob
lowerCamelCase : Union[str, Any] = attention_probs_dropout_prob
lowerCamelCase : Optional[int] = drop_path_rate
lowerCamelCase : Dict = hidden_act
lowerCamelCase : int = use_absolute_embeddings
lowerCamelCase : Any = layer_norm_eps
lowerCamelCase : List[Any] = initializer_range
lowerCamelCase : Optional[int] = upscale
lowerCamelCase : Tuple = img_range
lowerCamelCase : Optional[int] = resi_connection
lowerCamelCase : Optional[Any] = upsampler
| 700
|
"""simple docstring"""
import unittest
from knapsack import greedy_knapsack as kp
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: Optional[int] )-> Union[str, Any]:
lowerCamelCase : Tuple = [10, 20, 30, 40, 50, 60]
lowerCamelCase : Union[str, Any] = [2, 4, 6, 8, 10, 12]
lowerCamelCase : Union[str, Any] = 100
self.assertEqual(kp.calc_profit(__a , __a , __a ) , 210 )
def a__ ( self: str )-> str:
self.assertRaisesRegex(__a , """max_weight must greater than zero.""" )
def a__ ( self: str )-> List[Any]:
self.assertRaisesRegex(__a , """Weight can not be negative.""" )
def a__ ( self: Any )-> Dict:
self.assertRaisesRegex(__a , """Profit can not be negative.""" )
def a__ ( self: Optional[Any] )-> List[Any]:
self.assertRaisesRegex(__a , """max_weight must greater than zero.""" )
def a__ ( self: Optional[Any] )-> Tuple:
self.assertRaisesRegex(
__a , """The length of profit and weight must be same.""" )
if __name__ == "__main__":
unittest.main()
| 42
| 0
|
"""simple docstring"""
import unittest
import numpy as np
from transformers import BertConfig, is_flax_available
from transformers.testing_utils import require_flax, slow
from ...test_modeling_flax_common import FlaxModelTesterMixin, floats_tensor, ids_tensor, random_attention_mask
if is_flax_available():
from transformers.models.bert.modeling_flax_bert import (
FlaxBertForMaskedLM,
FlaxBertForMultipleChoice,
FlaxBertForNextSentencePrediction,
FlaxBertForPreTraining,
FlaxBertForQuestionAnswering,
FlaxBertForSequenceClassification,
FlaxBertForTokenClassification,
FlaxBertModel,
)
class A__ ( unittest.TestCase):
"""simple docstring"""
def __init__( self: int , __a: Any , __a: int=13 , __a: Any=7 , __a: Union[str, Any]=True , __a: List[Any]=True , __a: List[str]=True , __a: List[Any]=True , __a: Tuple=99 , __a: str=32 , __a: Union[str, Any]=5 , __a: List[str]=4 , __a: List[Any]=37 , __a: Optional[Any]="gelu" , __a: Dict=0.1 , __a: List[str]=0.1 , __a: Tuple=512 , __a: Union[str, Any]=16 , __a: int=2 , __a: List[str]=0.02 , __a: List[str]=4 , )-> Optional[int]:
lowerCamelCase : Optional[Any] = parent
lowerCamelCase : Dict = batch_size
lowerCamelCase : Dict = seq_length
lowerCamelCase : Any = is_training
lowerCamelCase : List[str] = use_attention_mask
lowerCamelCase : int = use_token_type_ids
lowerCamelCase : Union[str, Any] = use_labels
lowerCamelCase : Optional[int] = vocab_size
lowerCamelCase : Tuple = hidden_size
lowerCamelCase : Tuple = num_hidden_layers
lowerCamelCase : Optional[int] = num_attention_heads
lowerCamelCase : Tuple = intermediate_size
lowerCamelCase : Any = hidden_act
lowerCamelCase : Optional[Any] = hidden_dropout_prob
lowerCamelCase : int = attention_probs_dropout_prob
lowerCamelCase : Dict = max_position_embeddings
lowerCamelCase : List[str] = type_vocab_size
lowerCamelCase : Union[str, Any] = type_sequence_label_size
lowerCamelCase : Any = initializer_range
lowerCamelCase : List[Any] = num_choices
def a__ ( self: str )-> int:
lowerCamelCase : Union[str, Any] = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : List[Any] = None
if self.use_attention_mask:
lowerCamelCase : Any = random_attention_mask([self.batch_size, self.seq_length] )
lowerCamelCase : Union[str, Any] = None
if self.use_token_type_ids:
lowerCamelCase : Optional[Any] = ids_tensor([self.batch_size, self.seq_length] , self.type_vocab_size )
lowerCamelCase : Union[str, Any] = BertConfig(
vocab_size=self.vocab_size , hidden_size=self.hidden_size , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , is_decoder=__lowerCamelCase , initializer_range=self.initializer_range , )
return config, input_ids, token_type_ids, attention_mask
def a__ ( self: List[str] )-> List[Any]:
lowerCamelCase : Optional[int] = self.prepare_config_and_inputs()
lowerCamelCase : List[str] = config_and_inputs
lowerCamelCase : Tuple = {'''input_ids''': input_ids, '''token_type_ids''': token_type_ids, '''attention_mask''': attention_mask}
return config, inputs_dict
def a__ ( self: Union[str, Any] )-> Optional[Any]:
lowerCamelCase : str = self.prepare_config_and_inputs()
lowerCamelCase : Tuple = config_and_inputs
lowerCamelCase : Optional[Any] = True
lowerCamelCase : Dict = floats_tensor([self.batch_size, self.seq_length, self.hidden_size] )
lowerCamelCase : Dict = ids_tensor([self.batch_size, self.seq_length] , vocab_size=2 )
return (
config,
input_ids,
attention_mask,
encoder_hidden_states,
encoder_attention_mask,
)
@require_flax
class A__ ( lowerCamelCase__ , unittest.TestCase):
"""simple docstring"""
snake_case__ : str =True
snake_case__ : Dict =(
(
FlaxBertModel,
FlaxBertForPreTraining,
FlaxBertForMaskedLM,
FlaxBertForMultipleChoice,
FlaxBertForQuestionAnswering,
FlaxBertForNextSentencePrediction,
FlaxBertForSequenceClassification,
FlaxBertForTokenClassification,
FlaxBertForQuestionAnswering,
)
if is_flax_available()
else ()
)
def a__ ( self: Dict )-> Dict:
lowerCamelCase : Optional[int] = FlaxBertModelTester(self )
@slow
def a__ ( self: List[Any] )-> str:
lowerCamelCase : Tuple = FlaxBertModel.from_pretrained("""bert-base-cased""" )
lowerCamelCase : int = model(np.ones((1, 1) ) )
self.assertIsNotNone(__lowerCamelCase )
| 701
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_flax_available,
is_tf_available,
is_tokenizers_available,
is_torch_available,
is_vision_available,
)
__lowerCamelCase :List[str] = {
'configuration_owlvit': [
'OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP',
'OwlViTConfig',
'OwlViTOnnxConfig',
'OwlViTTextConfig',
'OwlViTVisionConfig',
],
'processing_owlvit': ['OwlViTProcessor'],
}
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[int] = ['OwlViTFeatureExtractor']
__lowerCamelCase :List[str] = ['OwlViTImageProcessor']
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Optional[Any] = [
'OWLVIT_PRETRAINED_MODEL_ARCHIVE_LIST',
'OwlViTModel',
'OwlViTPreTrainedModel',
'OwlViTTextModel',
'OwlViTVisionModel',
'OwlViTForObjectDetection',
]
if TYPE_CHECKING:
from .configuration_owlvit import (
OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP,
OwlViTConfig,
OwlViTOnnxConfig,
OwlViTTextConfig,
OwlViTVisionConfig,
)
from .processing_owlvit import OwlViTProcessor
try:
if not is_vision_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .feature_extraction_owlvit import OwlViTFeatureExtractor
from .image_processing_owlvit import OwlViTImageProcessor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_owlvit import (
OWLVIT_PRETRAINED_MODEL_ARCHIVE_LIST,
OwlViTForObjectDetection,
OwlViTModel,
OwlViTPreTrainedModel,
OwlViTTextModel,
OwlViTVisionModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
"""simple docstring"""
import gc
import unittest
from parameterized import parameterized
from diffusers import FlaxUNetaDConditionModel
from diffusers.utils import is_flax_available
from diffusers.utils.testing_utils import load_hf_numpy, require_flax, slow
if is_flax_available():
import jax
import jax.numpy as jnp
@slow
@require_flax
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: Tuple , __a: Tuple , __a: str )-> int:
return f'gaussian_noise_s={seed}_shape={"_".join([str(_lowercase ) for s in shape] )}.npy'
def a__ ( self: Dict )-> Any:
# clean up the VRAM after each test
super().tearDown()
gc.collect()
def a__ ( self: Dict , __a: Any=0 , __a: int=(4, 4, 64, 64) , __a: Dict=False )-> Optional[Any]:
lowerCamelCase : Optional[int] = jnp.bfloataa if fpaa else jnp.floataa
lowerCamelCase : List[Any] = jnp.array(load_hf_numpy(self.get_file_format(_lowercase , _lowercase ) ) , dtype=_lowercase )
return image
def a__ ( self: str , __a: List[str]=False , __a: Optional[int]="CompVis/stable-diffusion-v1-4" )-> str:
lowerCamelCase : str = jnp.bfloataa if fpaa else jnp.floataa
lowerCamelCase : int = """bf16""" if fpaa else None
lowerCamelCase , lowerCamelCase : int = FlaxUNetaDConditionModel.from_pretrained(
_lowercase , subfolder="""unet""" , dtype=_lowercase , revision=_lowercase )
return model, params
def a__ ( self: Tuple , __a: Any=0 , __a: Dict=(4, 77, 768) , __a: List[str]=False )-> Any:
lowerCamelCase : Optional[int] = jnp.bfloataa if fpaa else jnp.floataa
lowerCamelCase : Optional[Any] = jnp.array(load_hf_numpy(self.get_file_format(_lowercase , _lowercase ) ) , dtype=_lowercase )
return hidden_states
@parameterized.expand(
[
# fmt: off
[83, 4, [-0.23_23, -0.13_04, 0.08_13, -0.30_93, -0.09_19, -0.15_71, -0.11_25, -0.58_06]],
[17, 0.55, [-0.08_31, -0.24_43, 0.09_01, -0.09_19, 0.33_96, 0.01_03, -0.37_43, 0.07_01]],
[8, 0.89, [-0.48_63, 0.08_59, 0.08_75, -0.16_58, 0.91_99, -0.01_14, 0.48_39, 0.46_39]],
[3, 1_000, [-0.56_49, 0.24_02, -0.55_18, 0.12_48, 1.13_28, -0.24_43, -0.03_25, -1.00_78]],
# fmt: on
] )
def a__ ( self: List[Any] , __a: str , __a: Tuple , __a: Optional[Any] )-> int:
lowerCamelCase , lowerCamelCase : Dict = self.get_unet_model(model_id="""CompVis/stable-diffusion-v1-4""" , fpaa=_lowercase )
lowerCamelCase : Dict = self.get_latents(_lowercase , fpaa=_lowercase )
lowerCamelCase : str = self.get_encoder_hidden_states(_lowercase , fpaa=_lowercase )
lowerCamelCase : Union[str, Any] = model.apply(
{"""params""": params} , _lowercase , jnp.array(_lowercase , dtype=jnp.intaa ) , encoder_hidden_states=_lowercase , ).sample
assert sample.shape == latents.shape
lowerCamelCase : Optional[Any] = jnp.asarray(jax.device_get((sample[-1, -2:, -2:, :2].flatten()) ) , dtype=jnp.floataa )
lowerCamelCase : Union[str, Any] = jnp.array(_lowercase , dtype=jnp.floataa )
# Found torch (float16) and flax (bfloat16) outputs to be within this tolerance, in the same hardware
assert jnp.allclose(_lowercase , _lowercase , atol=1e-2 )
@parameterized.expand(
[
# fmt: off
[83, 4, [0.15_14, 0.08_07, 0.16_24, 0.10_16, -0.18_96, 0.02_63, 0.06_77, 0.23_10]],
[17, 0.55, [0.11_64, -0.02_16, 0.01_70, 0.15_89, -0.31_20, 0.10_05, -0.05_81, -0.14_58]],
[8, 0.89, [-0.17_58, -0.01_69, 0.10_04, -0.14_11, 0.13_12, 0.11_03, -0.19_96, 0.21_39]],
[3, 1_000, [0.12_14, 0.03_52, -0.07_31, -0.15_62, -0.09_94, -0.09_06, -0.23_40, -0.05_39]],
# fmt: on
] )
def a__ ( self: int , __a: Optional[Any] , __a: str , __a: List[Any] )-> Any:
lowerCamelCase , lowerCamelCase : Optional[Any] = self.get_unet_model(model_id="""stabilityai/stable-diffusion-2""" , fpaa=_lowercase )
lowerCamelCase : Dict = self.get_latents(_lowercase , shape=(4, 4, 96, 96) , fpaa=_lowercase )
lowerCamelCase : Union[str, Any] = self.get_encoder_hidden_states(_lowercase , shape=(4, 77, 1_024) , fpaa=_lowercase )
lowerCamelCase : Any = model.apply(
{"""params""": params} , _lowercase , jnp.array(_lowercase , dtype=jnp.intaa ) , encoder_hidden_states=_lowercase , ).sample
assert sample.shape == latents.shape
lowerCamelCase : int = jnp.asarray(jax.device_get((sample[-1, -2:, -2:, :2].flatten()) ) , dtype=jnp.floataa )
lowerCamelCase : str = jnp.array(_lowercase , dtype=jnp.floataa )
# Found torch (float16) and flax (bfloat16) outputs to be within this tolerance, on the same hardware
assert jnp.allclose(_lowercase , _lowercase , atol=1e-2 )
| 702
|
"""simple docstring"""
import collections
import inspect
import unittest
from transformers import FocalNetConfig
from transformers.testing_utils import require_torch, require_vision, slow, torch_device
from transformers.utils import cached_property, is_torch_available, is_vision_available
from ...test_backbone_common import BackboneTesterMixin
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, _config_zero_init, floats_tensor, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from torch import nn
from transformers import (
FocalNetBackbone,
FocalNetForImageClassification,
FocalNetForMaskedImageModeling,
FocalNetModel,
)
from transformers.models.focalnet.modeling_focalnet import FOCALNET_PRETRAINED_MODEL_ARCHIVE_LIST
if is_vision_available():
from PIL import Image
from transformers import AutoImageProcessor
class A__ :
"""simple docstring"""
def __init__( self: List[Any] , __a: List[str] , __a: Optional[int]=13 , __a: List[str]=32 , __a: int=2 , __a: List[str]=3 , __a: Union[str, Any]=16 , __a: int=[32, 64, 128] , __a: Optional[Any]=[1, 2, 1] , __a: Optional[int]=[2, 2, 4] , __a: Tuple=2 , __a: Dict=2.0 , __a: List[str]=True , __a: Optional[Any]=0.0 , __a: Any=0.0 , __a: List[Any]=0.1 , __a: List[str]="gelu" , __a: Tuple=False , __a: Union[str, Any]=True , __a: Optional[int]=0.02 , __a: Tuple=1e-5 , __a: int=True , __a: List[Any]=None , __a: Optional[int]=True , __a: Dict=10 , __a: List[str]=8 , __a: Any=["stage1", "stage2"] , __a: Union[str, Any]=[1, 2] , )-> Dict:
lowerCamelCase : Dict = parent
lowerCamelCase : Optional[Any] = batch_size
lowerCamelCase : Union[str, Any] = image_size
lowerCamelCase : Optional[int] = patch_size
lowerCamelCase : Any = num_channels
lowerCamelCase : Any = embed_dim
lowerCamelCase : Dict = hidden_sizes
lowerCamelCase : List[Any] = depths
lowerCamelCase : Tuple = num_heads
lowerCamelCase : List[Any] = window_size
lowerCamelCase : str = mlp_ratio
lowerCamelCase : str = qkv_bias
lowerCamelCase : str = hidden_dropout_prob
lowerCamelCase : Dict = attention_probs_dropout_prob
lowerCamelCase : Tuple = drop_path_rate
lowerCamelCase : Dict = hidden_act
lowerCamelCase : Tuple = use_absolute_embeddings
lowerCamelCase : List[str] = patch_norm
lowerCamelCase : List[str] = layer_norm_eps
lowerCamelCase : str = initializer_range
lowerCamelCase : Tuple = is_training
lowerCamelCase : int = scope
lowerCamelCase : Union[str, Any] = use_labels
lowerCamelCase : List[str] = type_sequence_label_size
lowerCamelCase : str = encoder_stride
lowerCamelCase : List[str] = out_features
lowerCamelCase : Optional[int] = out_indices
def a__ ( self: Optional[Any] )-> Union[str, Any]:
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size] )
lowerCamelCase : str = None
if self.use_labels:
lowerCamelCase : List[str] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : str = self.get_config()
return config, pixel_values, labels
def a__ ( self: List[Any] )-> Optional[int]:
return FocalNetConfig(
image_size=self.image_size , patch_size=self.patch_size , num_channels=self.num_channels , embed_dim=self.embed_dim , hidden_sizes=self.hidden_sizes , depths=self.depths , num_heads=self.num_heads , window_size=self.window_size , mlp_ratio=self.mlp_ratio , qkv_bias=self.qkv_bias , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , drop_path_rate=self.drop_path_rate , hidden_act=self.hidden_act , use_absolute_embeddings=self.use_absolute_embeddings , path_norm=self.patch_norm , layer_norm_eps=self.layer_norm_eps , initializer_range=self.initializer_range , encoder_stride=self.encoder_stride , out_features=self.out_features , out_indices=self.out_indices , )
def a__ ( self: Tuple , __a: Optional[int] , __a: Optional[int] , __a: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = FocalNetModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Tuple = model(__a )
lowerCamelCase : Any = ((config.image_size // config.patch_size) ** 2) // (4 ** (len(config.depths ) - 1))
lowerCamelCase : List[Any] = int(config.embed_dim * 2 ** (len(config.depths ) - 1) )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, expected_seq_len, expected_dim) )
def a__ ( self: Optional[int] , __a: Dict , __a: Tuple , __a: List[Any] )-> int:
lowerCamelCase : List[Any] = FocalNetBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Optional[Any] = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , len(config.out_features ) )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.image_size, 8, 8] )
# verify channels
self.parent.assertEqual(len(model.channels ) , len(config.out_features ) )
self.parent.assertListEqual(model.channels , config.hidden_sizes[:-1] )
# verify backbone works with out_features=None
lowerCamelCase : Dict = None
lowerCamelCase : Dict = FocalNetBackbone(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : Any = model(__a )
# verify feature maps
self.parent.assertEqual(len(result.feature_maps ) , 1 )
self.parent.assertListEqual(list(result.feature_maps[0].shape ) , [self.batch_size, self.image_size * 2, 4, 4] )
# verify channels
self.parent.assertEqual(len(model.channels ) , 1 )
self.parent.assertListEqual(model.channels , [config.hidden_sizes[-1]] )
def a__ ( self: Optional[int] , __a: Optional[int] , __a: Optional[int] , __a: Optional[int] )-> List[str]:
lowerCamelCase : Tuple = FocalNetForMaskedImageModeling(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[str] = model(__a )
self.parent.assertEqual(
result.reconstruction.shape , (self.batch_size, self.num_channels, self.image_size, self.image_size) )
# test greyscale images
lowerCamelCase : List[str] = 1
lowerCamelCase : Any = FocalNetForMaskedImageModeling(__a )
model.to(__a )
model.eval()
lowerCamelCase : str = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Tuple = model(__a )
self.parent.assertEqual(result.reconstruction.shape , (self.batch_size, 1, self.image_size, self.image_size) )
def a__ ( self: str , __a: Optional[Any] , __a: Optional[Any] , __a: Tuple )-> str:
lowerCamelCase : Optional[Any] = self.type_sequence_label_size
lowerCamelCase : Optional[Any] = FocalNetForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : List[str] = model(__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
# test greyscale images
lowerCamelCase : int = 1
lowerCamelCase : List[Any] = FocalNetForImageClassification(__a )
model.to(__a )
model.eval()
lowerCamelCase : Union[str, Any] = floats_tensor([self.batch_size, 1, self.image_size, self.image_size] )
lowerCamelCase : Optional[Any] = model(__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.type_sequence_label_size) )
def a__ ( self: int )-> Optional[int]:
lowerCamelCase : str = self.prepare_config_and_inputs()
lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = config_and_inputs
lowerCamelCase : List[Any] = {"""pixel_values""": pixel_values}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =(
(
FocalNetModel,
FocalNetForImageClassification,
FocalNetForMaskedImageModeling,
FocalNetBackbone,
)
if is_torch_available()
else ()
)
snake_case__ : Optional[int] =(
{'''feature-extraction''': FocalNetModel, '''image-classification''': FocalNetForImageClassification}
if is_torch_available()
else {}
)
snake_case__ : Tuple =False
snake_case__ : Dict =False
snake_case__ : Dict =False
snake_case__ : Tuple =False
snake_case__ : Optional[int] =False
def a__ ( self: Union[str, Any] )-> Optional[int]:
lowerCamelCase : List[str] = FocalNetModelTester(self )
lowerCamelCase : Optional[Any] = ConfigTester(self , config_class=__a , embed_dim=37 , has_text_modality=__a )
def a__ ( self: List[str] )-> List[str]:
self.create_and_test_config_common_properties()
self.config_tester.create_and_test_config_to_json_string()
self.config_tester.create_and_test_config_to_json_file()
self.config_tester.create_and_test_config_from_and_save_pretrained()
self.config_tester.create_and_test_config_with_num_labels()
self.config_tester.check_config_can_be_init_without_params()
self.config_tester.check_config_arguments_init()
def a__ ( self: List[str] )-> Union[str, Any]:
return
def a__ ( self: Tuple )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: List[Any] )-> Dict:
lowerCamelCase : int = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_backbone(*__a )
def a__ ( self: List[Any] )-> Tuple:
lowerCamelCase : Dict = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_masked_image_modeling(*__a )
def a__ ( self: List[str] )-> Dict:
lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_image_classification(*__a )
@unittest.skip(reason="""FocalNet does not use inputs_embeds""" )
def a__ ( self: Optional[Any] )-> str:
pass
@unittest.skip(reason="""FocalNet does not use feedforward chunking""" )
def a__ ( self: Optional[Any] )-> Dict:
pass
def a__ ( self: Optional[Any] )-> Dict:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : Any = model_class(__a )
self.assertIsInstance(model.get_input_embeddings() , (nn.Module) )
lowerCamelCase : Dict = model.get_output_embeddings()
self.assertTrue(x is None or isinstance(__a , nn.Linear ) )
def a__ ( self: Tuple )-> Optional[int]:
lowerCamelCase , lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs_for_common()
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : int = model_class(__a )
lowerCamelCase : int = inspect.signature(model.forward )
# signature.parameters is an OrderedDict => so arg_names order is deterministic
lowerCamelCase : Any = [*signature.parameters.keys()]
lowerCamelCase : List[Any] = ["""pixel_values"""]
self.assertListEqual(arg_names[:1] , __a )
def a__ ( self: str , __a: Union[str, Any] , __a: int , __a: Tuple , __a: List[str] )-> Union[str, Any]:
lowerCamelCase : List[Any] = model_class(__a )
model.to(__a )
model.eval()
with torch.no_grad():
lowerCamelCase : List[str] = model(**self._prepare_for_class(__a , __a ) )
lowerCamelCase : List[str] = outputs.hidden_states
lowerCamelCase : Tuple = getattr(
self.model_tester , """expected_num_hidden_layers""" , len(self.model_tester.depths ) + 1 )
self.assertEqual(len(__a ) , __a )
# FocalNet has a different seq_length
lowerCamelCase : Tuple = (
config.patch_size
if isinstance(config.patch_size , collections.abc.Iterable )
else (config.patch_size, config.patch_size)
)
lowerCamelCase : Dict = (image_size[1] // patch_size[1]) * (image_size[0] // patch_size[0])
self.assertListEqual(
list(hidden_states[0].shape[-2:] ) , [num_patches, self.model_tester.embed_dim] , )
lowerCamelCase : Optional[Any] = outputs.reshaped_hidden_states
self.assertEqual(len(__a ) , __a )
lowerCamelCase , lowerCamelCase , lowerCamelCase , lowerCamelCase : Optional[int] = reshaped_hidden_states[0].shape
lowerCamelCase : Tuple = (
reshaped_hidden_states[0].view(__a , __a , height * width ).permute(0 , 2 , 1 )
)
self.assertListEqual(
list(reshaped_hidden_states.shape[-2:] ) , [num_patches, self.model_tester.embed_dim] , )
def a__ ( self: Any )-> Any:
lowerCamelCase , lowerCamelCase : str = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : Union[str, Any] = (
self.model_tester.image_size
if isinstance(self.model_tester.image_size , collections.abc.Iterable )
else (self.model_tester.image_size, self.model_tester.image_size)
)
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : List[str] = True
self.check_hidden_states_output(__a , __a , __a , __a )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : List[Any] = True
self.check_hidden_states_output(__a , __a , __a , __a )
def a__ ( self: str )-> Union[str, Any]:
lowerCamelCase , lowerCamelCase : Tuple = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : List[str] = 3
lowerCamelCase : Any = (
self.model_tester.image_size
if isinstance(self.model_tester.image_size , collections.abc.Iterable )
else (self.model_tester.image_size, self.model_tester.image_size)
)
lowerCamelCase : Optional[int] = (
config.patch_size
if isinstance(config.patch_size , collections.abc.Iterable )
else (config.patch_size, config.patch_size)
)
lowerCamelCase : Optional[Any] = image_size[0] + patch_size[0] - (image_size[0] % patch_size[0])
lowerCamelCase : List[str] = image_size[1] + patch_size[1] - (image_size[1] % patch_size[1])
for model_class in self.all_model_classes[:-1]:
lowerCamelCase : str = True
self.check_hidden_states_output(__a , __a , __a , (padded_height, padded_width) )
# check that output_hidden_states also work using config
del inputs_dict["output_hidden_states"]
lowerCamelCase : Union[str, Any] = True
self.check_hidden_states_output(__a , __a , __a , (padded_height, padded_width) )
@slow
def a__ ( self: Optional[int] )-> List[Any]:
for model_name in FOCALNET_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : List[str] = FocalNetModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def a__ ( self: str )-> Any:
lowerCamelCase , lowerCamelCase : Optional[Any] = self.model_tester.prepare_config_and_inputs_for_common()
lowerCamelCase : int = _config_zero_init(__a )
for model_class in self.all_model_classes:
lowerCamelCase : int = model_class(config=__a )
for name, param in model.named_parameters():
if "embeddings" not in name and param.requires_grad:
self.assertIn(
((param.data.mean() * 1e9).round() / 1e9).item() , [0.0, 1.0] , msg=f'Parameter {name} of model {model_class} seems not properly initialized' , )
@require_vision
@require_torch
class A__ ( unittest.TestCase):
"""simple docstring"""
@cached_property
def a__ ( self: Optional[int] )-> Optional[Any]:
# TODO update organization
return AutoImageProcessor.from_pretrained("""microsoft/focalnet-tiny""" ) if is_vision_available() else None
@slow
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : Tuple = FocalNetForImageClassification.from_pretrained("""microsoft/focalnet-tiny""" ).to(__a )
lowerCamelCase : Any = self.default_image_processor
lowerCamelCase : Union[str, Any] = Image.open("""./tests/fixtures/tests_samples/COCO/000000039769.png""" )
lowerCamelCase : int = image_processor(images=__a , return_tensors="""pt""" ).to(__a )
# forward pass
with torch.no_grad():
lowerCamelCase : Any = model(**__a )
# verify the logits
lowerCamelCase : Tuple = torch.Size((1, 1_000) )
self.assertEqual(outputs.logits.shape , __a )
lowerCamelCase : List[str] = torch.tensor([0.21_66, -0.43_68, 0.21_91] ).to(__a )
self.assertTrue(torch.allclose(outputs.logits[0, :3] , __a , atol=1e-4 ) )
self.assertTrue(outputs.logits.argmax(dim=-1 ).item() , 281 )
@require_torch
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : str =(FocalNetBackbone,) if is_torch_available() else ()
snake_case__ : Optional[int] =FocalNetConfig
snake_case__ : str =False
def a__ ( self: Union[str, Any] )-> Tuple:
lowerCamelCase : str = FocalNetModelTester(self )
| 42
| 0
|
"""simple docstring"""
import numpy as np
import torch
import torch.nn as nn
from transformers import CLIPConfig, CLIPVisionModelWithProjection, PreTrainedModel
from ...utils import logging
__lowerCamelCase :List[str] = logging.get_logger(__name__)
class A__ ( snake_case__):
"""simple docstring"""
snake_case__ : int =CLIPConfig
snake_case__ : Dict =['''CLIPEncoderLayer''']
def __init__( self: Any , __a: int )-> Optional[int]:
super().__init__(lowercase_ )
lowerCamelCase : Dict = CLIPVisionModelWithProjection(config.vision_config )
lowerCamelCase : List[str] = nn.Linear(config.vision_config.projection_dim , 1 )
lowerCamelCase : Dict = nn.Linear(config.vision_config.projection_dim , 1 )
@torch.no_grad()
def a__ ( self: Tuple , __a: List[str] , __a: List[Any] , __a: Optional[int]=0.5 , __a: int=0.5 )-> Optional[int]:
lowerCamelCase : Any = self.vision_model(lowercase_ )[0]
lowerCamelCase : Optional[int] = self.p_head(lowercase_ )
lowerCamelCase : Any = nsfw_detected.flatten()
lowerCamelCase : Dict = nsfw_detected > p_threshold
lowerCamelCase : Optional[int] = nsfw_detected.tolist()
if any(lowercase_ ):
logger.warning(
"""Potential NSFW content was detected in one or more images. A black image will be returned instead."""
""" Try again with a different prompt and/or seed.""" )
for idx, nsfw_detected_ in enumerate(lowercase_ ):
if nsfw_detected_:
lowerCamelCase : Any = np.zeros(images[idx].shape )
lowerCamelCase : List[str] = self.w_head(lowercase_ )
lowerCamelCase : Union[str, Any] = watermark_detected.flatten()
lowerCamelCase : str = watermark_detected > w_threshold
lowerCamelCase : str = watermark_detected.tolist()
if any(lowercase_ ):
logger.warning(
"""Potential watermarked content was detected in one or more images. A black image will be returned instead."""
""" Try again with a different prompt and/or seed.""" )
for idx, watermark_detected_ in enumerate(lowercase_ ):
if watermark_detected_:
lowerCamelCase : List[Any] = np.zeros(images[idx].shape )
return images, nsfw_detected, watermark_detected
| 703
|
"""simple docstring"""
import os
def snake_case ( ) -> Optional[Any]:
with open(os.path.dirname(UpperCamelCase__ ) + """/grid.txt""" ) as f:
lowerCamelCase : int = [] # noqa: E741
for _ in range(20 ):
l.append([int(UpperCamelCase__ ) for x in f.readline().split()] )
lowerCamelCase : Union[str, Any] = 0
# right
for i in range(20 ):
for j in range(17 ):
lowerCamelCase : Dict = l[i][j] * l[i][j + 1] * l[i][j + 2] * l[i][j + 3]
if temp > maximum:
lowerCamelCase : Tuple = temp
# down
for i in range(17 ):
for j in range(20 ):
lowerCamelCase : Any = l[i][j] * l[i + 1][j] * l[i + 2][j] * l[i + 3][j]
if temp > maximum:
lowerCamelCase : Optional[Any] = temp
# diagonal 1
for i in range(17 ):
for j in range(17 ):
lowerCamelCase : List[Any] = l[i][j] * l[i + 1][j + 1] * l[i + 2][j + 2] * l[i + 3][j + 3]
if temp > maximum:
lowerCamelCase : List[str] = temp
# diagonal 2
for i in range(17 ):
for j in range(3 , 20 ):
lowerCamelCase : List[str] = l[i][j] * l[i + 1][j - 1] * l[i + 2][j - 2] * l[i + 3][j - 3]
if temp > maximum:
lowerCamelCase : List[Any] = temp
return maximum
if __name__ == "__main__":
print(solution())
| 42
| 0
|
"""simple docstring"""
from __future__ import annotations
import unittest
from transformers import BlenderbotConfig, BlenderbotTokenizer, is_tf_available
from transformers.testing_utils import require_tf, require_tokenizers, slow
from transformers.utils import cached_property
from ...test_configuration_common import ConfigTester
from ...test_modeling_tf_common import TFModelTesterMixin, ids_tensor
from ...test_pipeline_mixin import PipelineTesterMixin
if is_tf_available():
import tensorflow as tf
from transformers import TFAutoModelForSeqaSeqLM, TFBlenderbotForConditionalGeneration, TFBlenderbotModel
@require_tf
class A__ :
"""simple docstring"""
snake_case__ : Optional[Any] =BlenderbotConfig
snake_case__ : List[Any] ={}
snake_case__ : int ='''gelu'''
def __init__( self: Any , __a: str , __a: Any=13 , __a: List[str]=7 , __a: Any=True , __a: Optional[Any]=False , __a: Tuple=99 , __a: Optional[int]=32 , __a: str=2 , __a: Dict=4 , __a: Dict=37 , __a: Tuple=0.1 , __a: Optional[int]=0.1 , __a: Optional[Any]=20 , __a: Optional[int]=2 , __a: Union[str, Any]=1 , __a: int=0 , )-> Tuple:
lowerCamelCase : Union[str, Any] = parent
lowerCamelCase : str = batch_size
lowerCamelCase : Optional[Any] = seq_length
lowerCamelCase : str = is_training
lowerCamelCase : Any = use_labels
lowerCamelCase : List[str] = vocab_size
lowerCamelCase : Any = hidden_size
lowerCamelCase : Union[str, Any] = num_hidden_layers
lowerCamelCase : Optional[Any] = num_attention_heads
lowerCamelCase : Union[str, Any] = intermediate_size
lowerCamelCase : int = hidden_dropout_prob
lowerCamelCase : List[Any] = attention_probs_dropout_prob
lowerCamelCase : Dict = max_position_embeddings
lowerCamelCase : Any = eos_token_id
lowerCamelCase : str = pad_token_id
lowerCamelCase : Optional[Any] = bos_token_id
def a__ ( self: Any )-> Optional[Any]:
lowerCamelCase : str = ids_tensor([self.batch_size, self.seq_length - 1] , self.vocab_size )
lowerCamelCase : Optional[Any] = tf.expand_dims(tf.constant([self.eos_token_id] * self.batch_size ) , 1 )
lowerCamelCase : str = tf.concat([input_ids, eos_tensor] , axis=1 )
lowerCamelCase : Optional[Any] = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : List[str] = self.config_cls(
vocab_size=self.vocab_size , d_model=self.hidden_size , encoder_layers=self.num_hidden_layers , decoder_layers=self.num_hidden_layers , encoder_attention_heads=self.num_attention_heads , decoder_attention_heads=self.num_attention_heads , encoder_ffn_dim=self.intermediate_size , decoder_ffn_dim=self.intermediate_size , dropout=self.hidden_dropout_prob , attention_dropout=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , eos_token_ids=[2] , bos_token_id=self.bos_token_id , pad_token_id=self.pad_token_id , decoder_start_token_id=self.pad_token_id , **self.config_updates , )
lowerCamelCase : str = prepare_blenderbot_inputs_dict(A_ , A_ , A_ )
return config, inputs_dict
def a__ ( self: Any , __a: int , __a: Tuple )-> List[Any]:
lowerCamelCase : Dict = TFBlenderbotModel(config=A_ ).get_decoder()
lowerCamelCase : List[str] = inputs_dict["""input_ids"""]
lowerCamelCase : Optional[int] = input_ids[:1, :]
lowerCamelCase : Any = inputs_dict["""attention_mask"""][:1, :]
lowerCamelCase : Union[str, Any] = inputs_dict["""head_mask"""]
lowerCamelCase : List[Any] = 1
# first forward pass
lowerCamelCase : int = model(A_ , attention_mask=A_ , head_mask=A_ , use_cache=A_ )
lowerCamelCase , lowerCamelCase : Union[str, Any] = outputs.to_tuple()
# create hypothetical next token and extent to next_input_ids
lowerCamelCase : Any = ids_tensor((self.batch_size, 3) , config.vocab_size )
lowerCamelCase : List[Any] = tf.cast(ids_tensor((self.batch_size, 3) , 2 ) , tf.inta )
# append to next input_ids and
lowerCamelCase : int = tf.concat([input_ids, next_tokens] , axis=-1 )
lowerCamelCase : Optional[Any] = tf.concat([attention_mask, next_attn_mask] , axis=-1 )
lowerCamelCase : Optional[int] = model(A_ , attention_mask=A_ )[0]
lowerCamelCase : Dict = model(A_ , attention_mask=A_ , past_key_values=A_ )[0]
self.parent.assertEqual(next_tokens.shape[1] , output_from_past.shape[1] )
# select random slice
lowerCamelCase : Union[str, Any] = int(ids_tensor((1,) , output_from_past.shape[-1] ) )
lowerCamelCase : Tuple = output_from_no_past[:, -3:, random_slice_idx]
lowerCamelCase : List[Any] = output_from_past[:, :, random_slice_idx]
# test that outputs are equal for slice
tf.debugging.assert_near(A_ , A_ , rtol=1e-3 )
def snake_case ( UpperCamelCase__ : List[Any] , UpperCamelCase__ : List[str] , UpperCamelCase__ : List[str] , UpperCamelCase__ : Optional[Any]=None , UpperCamelCase__ : Union[str, Any]=None , UpperCamelCase__ : int=None , UpperCamelCase__ : Tuple=None , UpperCamelCase__ : List[Any]=None , ) -> List[Any]:
if attention_mask is None:
lowerCamelCase : Any = tf.cast(tf.math.not_equal(UpperCamelCase__ , config.pad_token_id ) , tf.inta )
if decoder_attention_mask is None:
lowerCamelCase : Any = tf.concat(
[
tf.ones(decoder_input_ids[:, :1].shape , dtype=tf.inta ),
tf.cast(tf.math.not_equal(decoder_input_ids[:, 1:] , config.pad_token_id ) , tf.inta ),
] , axis=-1 , )
if head_mask is None:
lowerCamelCase : Dict = tf.ones((config.encoder_layers, config.encoder_attention_heads) )
if decoder_head_mask is None:
lowerCamelCase : Tuple = tf.ones((config.decoder_layers, config.decoder_attention_heads) )
if cross_attn_head_mask is None:
lowerCamelCase : Union[str, Any] = tf.ones((config.decoder_layers, config.decoder_attention_heads) )
return {
"input_ids": input_ids,
"decoder_input_ids": decoder_input_ids,
"attention_mask": attention_mask,
"decoder_attention_mask": decoder_attention_mask,
"head_mask": head_mask,
"decoder_head_mask": decoder_head_mask,
"cross_attn_head_mask": cross_attn_head_mask,
}
@require_tf
class A__ ( _UpperCAmelCase , _UpperCAmelCase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =(TFBlenderbotForConditionalGeneration, TFBlenderbotModel) if is_tf_available() else ()
snake_case__ : Optional[int] =(TFBlenderbotForConditionalGeneration,) if is_tf_available() else ()
snake_case__ : Optional[Any] =(
{
'''conversational''': TFBlenderbotForConditionalGeneration,
'''feature-extraction''': TFBlenderbotModel,
'''summarization''': TFBlenderbotForConditionalGeneration,
'''text2text-generation''': TFBlenderbotForConditionalGeneration,
'''translation''': TFBlenderbotForConditionalGeneration,
}
if is_tf_available()
else {}
)
snake_case__ : Optional[Any] =True
snake_case__ : Any =False
snake_case__ : Optional[Any] =False
def a__ ( self: str )-> List[str]:
lowerCamelCase : Tuple = TFBlenderbotModelTester(self )
lowerCamelCase : Union[str, Any] = ConfigTester(self , config_class=A_ )
def a__ ( self: List[Any] )-> str:
self.config_tester.run_common_tests()
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs_for_common()
self.model_tester.check_decoder_model_past_large_inputs(*A_ )
@require_tokenizers
@require_tf
class A__ ( unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[Any] =['''My friends are cool but they eat too many carbs.''']
snake_case__ : Any ='''facebook/blenderbot-400M-distill'''
@cached_property
def a__ ( self: str )-> Optional[int]:
return BlenderbotTokenizer.from_pretrained(self.model_name )
@cached_property
def a__ ( self: Tuple )-> List[str]:
lowerCamelCase : Any = TFAutoModelForSeqaSeqLM.from_pretrained(self.model_name )
return model
@slow
def a__ ( self: List[Any] )-> str:
lowerCamelCase : List[Any] = self.tokenizer(self.src_text , return_tensors="""tf""" )
lowerCamelCase : int = self.model.generate(
model_inputs.input_ids , )
lowerCamelCase : List[Any] = self.tokenizer.batch_decode(generated_ids.numpy() , skip_special_tokens=A_ )[0]
assert (
generated_words
== " That's unfortunate. Are they trying to lose weight or are they just trying to be healthier?"
)
| 704
|
"""simple docstring"""
import gc
import unittest
import numpy as np
import torch
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer
from diffusers import (
AutoencoderKL,
DDIMScheduler,
StableDiffusionAttendAndExcitePipeline,
UNetaDConditionModel,
)
from diffusers.utils import load_numpy, skip_mps, slow
from diffusers.utils.testing_utils import require_torch_gpu
from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
from ..test_pipelines_common import PipelineKarrasSchedulerTesterMixin, PipelineLatentTesterMixin, PipelineTesterMixin
__lowerCamelCase :Any = False
@skip_mps
class A__ ( __lowercase , __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Optional[Any] =StableDiffusionAttendAndExcitePipeline
snake_case__ : Any =False
snake_case__ : Dict =TEXT_TO_IMAGE_PARAMS
snake_case__ : Any =TEXT_TO_IMAGE_BATCH_PARAMS.union({'''token_indices'''})
snake_case__ : Dict =TEXT_TO_IMAGE_IMAGE_PARAMS
snake_case__ : str =TEXT_TO_IMAGE_IMAGE_PARAMS
@classmethod
def a__ ( cls: Dict )-> Tuple:
super().setUpClass()
torch.use_deterministic_algorithms(__a )
@classmethod
def a__ ( cls: Union[str, Any] )-> Any:
super().tearDownClass()
torch.use_deterministic_algorithms(__a )
def a__ ( self: Tuple )-> Union[str, Any]:
torch.manual_seed(0 )
lowerCamelCase : str = UNetaDConditionModel(
block_out_channels=(32, 64) , layers_per_block=1 , sample_size=32 , in_channels=4 , out_channels=4 , down_block_types=("""DownBlock2D""", """CrossAttnDownBlock2D""") , up_block_types=("""CrossAttnUpBlock2D""", """UpBlock2D""") , cross_attention_dim=32 , attention_head_dim=(2, 4) , use_linear_projection=__a , )
lowerCamelCase : Union[str, Any] = DDIMScheduler(
beta_start=0.0_00_85 , beta_end=0.0_12 , beta_schedule="""scaled_linear""" , clip_sample=__a , set_alpha_to_one=__a , )
torch.manual_seed(0 )
lowerCamelCase : Union[str, Any] = AutoencoderKL(
block_out_channels=[32, 64] , in_channels=3 , out_channels=3 , down_block_types=["""DownEncoderBlock2D""", """DownEncoderBlock2D"""] , up_block_types=["""UpDecoderBlock2D""", """UpDecoderBlock2D"""] , latent_channels=4 , sample_size=128 , )
torch.manual_seed(0 )
lowerCamelCase : str = CLIPTextConfig(
bos_token_id=0 , eos_token_id=2 , hidden_size=32 , intermediate_size=37 , layer_norm_eps=1e-0_5 , num_attention_heads=4 , num_hidden_layers=5 , pad_token_id=1 , vocab_size=1_000 , hidden_act="""gelu""" , projection_dim=512 , )
lowerCamelCase : Optional[int] = CLIPTextModel(__a )
lowerCamelCase : str = CLIPTokenizer.from_pretrained("""hf-internal-testing/tiny-random-clip""" )
lowerCamelCase : List[str] = {
"""unet""": unet,
"""scheduler""": scheduler,
"""vae""": vae,
"""text_encoder""": text_encoder,
"""tokenizer""": tokenizer,
"""safety_checker""": None,
"""feature_extractor""": None,
}
return components
def a__ ( self: Tuple , __a: int , __a: Union[str, Any]=0 )-> Optional[Any]:
if str(__a ).startswith("""mps""" ):
lowerCamelCase : Tuple = torch.manual_seed(__a )
else:
lowerCamelCase : str = torch.Generator(device=__a ).manual_seed(__a )
lowerCamelCase : Dict = {
"""prompt""": """a cat and a frog""",
"""token_indices""": [2, 5],
"""generator""": generator,
"""num_inference_steps""": 1,
"""guidance_scale""": 6.0,
"""output_type""": """numpy""",
"""max_iter_to_alter""": 2,
"""thresholds""": {0: 0.7},
}
return inputs
def a__ ( self: Dict )-> str:
lowerCamelCase : Tuple = """cpu"""
lowerCamelCase : List[str] = self.get_dummy_components()
lowerCamelCase : List[Any] = self.pipeline_class(**__a )
pipe.to(__a )
pipe.set_progress_bar_config(disable=__a )
lowerCamelCase : Any = self.get_dummy_inputs(__a )
lowerCamelCase : Union[str, Any] = pipe(**__a ).images
lowerCamelCase : Tuple = image[0, -3:, -3:, -1]
self.assertEqual(image.shape , (1, 64, 64, 3) )
lowerCamelCase : Optional[Any] = np.array(
[0.63_90_53_64, 0.62_89_73_07, 0.48_59_90_17, 0.5_13_36_24, 0.5_55_00_48, 0.45_76_95_16, 0.50_32_69_73, 0.5_02_31_39, 0.45_38_44_96] )
lowerCamelCase : Optional[Any] = np.abs(image_slice.flatten() - expected_slice ).max()
self.assertLessEqual(__a , 1e-3 )
def a__ ( self: int )-> Optional[Any]:
super().test_cpu_offload_forward_pass(expected_max_diff=5e-4 )
def a__ ( self: Union[str, Any] )-> Optional[int]:
# NOTE: Larger batch sizes cause this test to timeout, only test on smaller batches
self._test_inference_batch_consistent(batch_sizes=[1, 2] )
def a__ ( self: Tuple )-> int:
self._test_inference_batch_single_identical(batch_size=2 , expected_max_diff=7e-4 )
def a__ ( self: Dict )-> List[Any]:
super().test_dict_tuple_outputs_equivalent(expected_max_difference=3e-3 )
def a__ ( self: Optional[int] )-> Dict:
super().test_pt_np_pil_outputs_equivalent(expected_max_diff=5e-4 )
def a__ ( self: Any )-> Tuple:
super().test_save_load_local(expected_max_difference=5e-4 )
def a__ ( self: str )-> str:
super().test_save_load_optional_components(expected_max_difference=4e-4 )
@require_torch_gpu
@slow
class A__ ( unittest.TestCase):
"""simple docstring"""
@classmethod
def a__ ( cls: Any )-> Tuple:
super().setUpClass()
torch.use_deterministic_algorithms(__a )
@classmethod
def a__ ( cls: Dict )-> Optional[int]:
super().tearDownClass()
torch.use_deterministic_algorithms(__a )
def a__ ( self: int )-> Optional[int]:
super().tearDown()
gc.collect()
torch.cuda.empty_cache()
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : List[Any] = torch.manual_seed(51 )
lowerCamelCase : List[str] = StableDiffusionAttendAndExcitePipeline.from_pretrained(
"""CompVis/stable-diffusion-v1-4""" , safety_checker=__a , torch_dtype=torch.floataa )
pipe.to("""cuda""" )
lowerCamelCase : Dict = """a painting of an elephant with glasses"""
lowerCamelCase : Any = [5, 7]
lowerCamelCase : Tuple = pipe(
prompt=__a , token_indices=__a , guidance_scale=7.5 , generator=__a , num_inference_steps=5 , max_iter_to_alter=5 , output_type="""numpy""" , ).images[0]
lowerCamelCase : Union[str, Any] = load_numpy(
"""https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/attend-and-excite/elephant_glasses.npy""" )
assert np.abs((expected_image - image).max() ) < 5e-1
| 42
| 0
|
"""simple docstring"""
from collections import UserDict
from typing import List, Union
from ..utils import (
add_end_docstrings,
is_tf_available,
is_torch_available,
is_vision_available,
logging,
requires_backends,
)
from .base import PIPELINE_INIT_ARGS, Pipeline
if is_vision_available():
from PIL import Image
from ..image_utils import load_image
if is_torch_available():
from ..models.auto.modeling_auto import MODEL_FOR_ZERO_SHOT_IMAGE_CLASSIFICATION_MAPPING
if is_tf_available():
from ..models.auto.modeling_tf_auto import TF_MODEL_FOR_ZERO_SHOT_IMAGE_CLASSIFICATION_MAPPING
from ..tf_utils import stable_softmax
__lowerCamelCase :int = logging.get_logger(__name__)
@add_end_docstrings(__lowercase)
class A__ ( __lowercase):
"""simple docstring"""
def __init__( self: Any , **__a: Any )-> str:
super().__init__(**__a )
requires_backends(self , """vision""" )
self.check_model_type(
TF_MODEL_FOR_ZERO_SHOT_IMAGE_CLASSIFICATION_MAPPING
if self.framework == """tf"""
else MODEL_FOR_ZERO_SHOT_IMAGE_CLASSIFICATION_MAPPING )
def __call__( self: List[str] , __a: Union[str, List[str], "Image", List["Image"]] , **__a: Dict )-> Union[str, Any]:
return super().__call__(__a , **__a )
def a__ ( self: int , **__a: Any )-> Dict:
lowerCamelCase : Dict = {}
if "candidate_labels" in kwargs:
lowerCamelCase : List[str] = kwargs["candidate_labels"]
if "hypothesis_template" in kwargs:
lowerCamelCase : List[Any] = kwargs["hypothesis_template"]
return preprocess_params, {}, {}
def a__ ( self: Tuple , __a: Tuple , __a: Tuple=None , __a: Optional[Any]="This is a photo of {}." )-> List[str]:
lowerCamelCase : Optional[int] = load_image(__a )
lowerCamelCase : Union[str, Any] = self.image_processor(images=[image] , return_tensors=self.framework )
lowerCamelCase : Optional[Any] = candidate_labels
lowerCamelCase : str = [hypothesis_template.format(__a ) for x in candidate_labels]
lowerCamelCase : List[str] = self.tokenizer(__a , return_tensors=self.framework , padding=__a )
lowerCamelCase : Tuple = [text_inputs]
return inputs
def a__ ( self: int , __a: str )-> List[Any]:
lowerCamelCase : int = model_inputs.pop("""candidate_labels""" )
lowerCamelCase : Optional[int] = model_inputs.pop("""text_inputs""" )
if isinstance(text_inputs[0] , __a ):
lowerCamelCase : int = text_inputs[0]
else:
# Batching case.
lowerCamelCase : Optional[int] = text_inputs[0][0]
lowerCamelCase : Any = self.model(**__a , **__a )
lowerCamelCase : Union[str, Any] = {
"candidate_labels": candidate_labels,
"logits": outputs.logits_per_image,
}
return model_outputs
def a__ ( self: Optional[int] , __a: Union[str, Any] )-> str:
lowerCamelCase : List[str] = model_outputs.pop("""candidate_labels""" )
lowerCamelCase : Optional[Any] = model_outputs["logits"][0]
if self.framework == "pt":
lowerCamelCase : str = logits.softmax(dim=-1 ).squeeze(-1 )
lowerCamelCase : Any = probs.tolist()
if not isinstance(__a , __a ):
lowerCamelCase : Optional[int] = [scores]
elif self.framework == "tf":
lowerCamelCase : Union[str, Any] = stable_softmax(__a , axis=-1 )
lowerCamelCase : str = probs.numpy().tolist()
else:
raise ValueError(f'Unsupported framework: {self.framework}' )
lowerCamelCase : Any = [
{"score": score, "label": candidate_label}
for score, candidate_label in sorted(zip(__a , __a ) , key=lambda __a : -x[0] )
]
return result
| 705
|
"""simple docstring"""
import unittest
from transformers import EsmConfig, is_torch_available
from transformers.testing_utils import TestCasePlus, require_torch, slow, torch_device
from ...test_configuration_common import ConfigTester
from ...test_modeling_common import ModelTesterMixin, ids_tensor, random_attention_mask
from ...test_pipeline_mixin import PipelineTesterMixin
if is_torch_available():
import torch
from transformers import EsmForMaskedLM, EsmForSequenceClassification, EsmForTokenClassification, EsmModel
from transformers.models.esm.modeling_esm import (
ESM_PRETRAINED_MODEL_ARCHIVE_LIST,
EsmEmbeddings,
create_position_ids_from_input_ids,
)
class A__ :
"""simple docstring"""
def __init__( self: List[str] , __a: List[str] , __a: Dict=13 , __a: Tuple=7 , __a: Dict=False , __a: str=True , __a: List[Any]=False , __a: Dict=True , __a: Any=33 , __a: Optional[Any]=32 , __a: List[Any]=5 , __a: Any=4 , __a: Dict=37 , __a: str="gelu" , __a: str=0.1 , __a: int=0.1 , __a: Optional[int]=512 , __a: List[Any]=16 , __a: int=2 , __a: int=0.02 , __a: Optional[int]=3 , __a: str=4 , __a: Tuple=None , )-> Tuple:
lowerCamelCase : Union[str, Any] = parent
lowerCamelCase : Tuple = batch_size
lowerCamelCase : Any = seq_length
lowerCamelCase : Any = is_training
lowerCamelCase : Tuple = use_input_mask
lowerCamelCase : int = use_token_type_ids
lowerCamelCase : List[str] = use_labels
lowerCamelCase : Optional[int] = vocab_size
lowerCamelCase : Tuple = hidden_size
lowerCamelCase : List[str] = num_hidden_layers
lowerCamelCase : Optional[int] = num_attention_heads
lowerCamelCase : Optional[Any] = intermediate_size
lowerCamelCase : Optional[Any] = hidden_act
lowerCamelCase : Union[str, Any] = hidden_dropout_prob
lowerCamelCase : Optional[Any] = attention_probs_dropout_prob
lowerCamelCase : Any = max_position_embeddings
lowerCamelCase : str = type_vocab_size
lowerCamelCase : List[Any] = type_sequence_label_size
lowerCamelCase : Optional[Any] = initializer_range
lowerCamelCase : Union[str, Any] = num_labels
lowerCamelCase : Optional[Any] = num_choices
lowerCamelCase : Any = scope
def a__ ( self: Optional[int] )-> List[Any]:
lowerCamelCase : List[Any] = ids_tensor([self.batch_size, self.seq_length] , self.vocab_size )
lowerCamelCase : Dict = None
if self.use_input_mask:
lowerCamelCase : Tuple = random_attention_mask([self.batch_size, self.seq_length] )
lowerCamelCase : Any = None
lowerCamelCase : int = None
lowerCamelCase : Union[str, Any] = None
if self.use_labels:
lowerCamelCase : Optional[Any] = ids_tensor([self.batch_size] , self.type_sequence_label_size )
lowerCamelCase : Optional[int] = ids_tensor([self.batch_size, self.seq_length] , self.num_labels )
lowerCamelCase : Union[str, Any] = ids_tensor([self.batch_size] , self.num_choices )
lowerCamelCase : List[str] = self.get_config()
return config, input_ids, input_mask, sequence_labels, token_labels, choice_labels
def a__ ( self: Tuple )-> Union[str, Any]:
return EsmConfig(
vocab_size=self.vocab_size , hidden_size=self.hidden_size , pad_token_id=1 , num_hidden_layers=self.num_hidden_layers , num_attention_heads=self.num_attention_heads , intermediate_size=self.intermediate_size , hidden_act=self.hidden_act , hidden_dropout_prob=self.hidden_dropout_prob , attention_probs_dropout_prob=self.attention_probs_dropout_prob , max_position_embeddings=self.max_position_embeddings , type_vocab_size=self.type_vocab_size , initializer_range=self.initializer_range , )
def a__ ( self: List[Any] , __a: List[str] , __a: str , __a: Tuple , __a: List[str] , __a: List[str] , __a: str )-> int:
lowerCamelCase : Optional[int] = EsmModel(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a , attention_mask=__a )
lowerCamelCase : str = model(__a )
lowerCamelCase : Optional[Any] = model(__a )
self.parent.assertEqual(result.last_hidden_state.shape , (self.batch_size, self.seq_length, self.hidden_size) )
self.parent.assertEqual(result.pooler_output.shape , (self.batch_size, self.hidden_size) )
def a__ ( self: int , __a: Union[str, Any] , __a: Optional[int] , __a: List[str] , __a: str , __a: List[str] , __a: Tuple )-> int:
lowerCamelCase : str = EsmForMaskedLM(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : List[Any] = model(__a , attention_mask=__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.vocab_size) )
def a__ ( self: List[str] , __a: List[Any] , __a: List[str] , __a: int , __a: Union[str, Any] , __a: List[Any] , __a: Tuple )-> List[str]:
lowerCamelCase : Tuple = self.num_labels
lowerCamelCase : Dict = EsmForTokenClassification(config=__a )
model.to(__a )
model.eval()
lowerCamelCase : int = model(__a , attention_mask=__a , labels=__a )
self.parent.assertEqual(result.logits.shape , (self.batch_size, self.seq_length, self.num_labels) )
def a__ ( self: Optional[int] )-> Optional[int]:
lowerCamelCase : Any = self.prepare_config_and_inputs()
(
(
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) , (
lowerCamelCase
) ,
) : Tuple = config_and_inputs
lowerCamelCase : List[Any] = {"""input_ids""": input_ids, """attention_mask""": input_mask}
return config, inputs_dict
@require_torch
class A__ ( __lowercase , __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : Any =False
snake_case__ : Dict =(
(
EsmForMaskedLM,
EsmModel,
EsmForSequenceClassification,
EsmForTokenClassification,
)
if is_torch_available()
else ()
)
snake_case__ : Dict =()
snake_case__ : Optional[int] =(
{
'''feature-extraction''': EsmModel,
'''fill-mask''': EsmForMaskedLM,
'''text-classification''': EsmForSequenceClassification,
'''token-classification''': EsmForTokenClassification,
'''zero-shot''': EsmForSequenceClassification,
}
if is_torch_available()
else {}
)
snake_case__ : Any =True
def a__ ( self: Optional[int] )-> Optional[int]:
lowerCamelCase : Optional[Any] = EsmModelTester(self )
lowerCamelCase : Any = ConfigTester(self , config_class=__a , hidden_size=37 )
def a__ ( self: List[Any] )-> Optional[Any]:
self.config_tester.run_common_tests()
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_model(*__a )
def a__ ( self: Tuple )-> Any:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()
for type in ["absolute", "relative_key", "relative_key_query"]:
lowerCamelCase : Tuple = type
self.model_tester.create_and_check_model(*__a )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Optional[int] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_masked_lm(*__a )
def a__ ( self: int )-> Optional[Any]:
lowerCamelCase : List[Any] = self.model_tester.prepare_config_and_inputs()
self.model_tester.create_and_check_for_token_classification(*__a )
@slow
def a__ ( self: Any )-> List[Any]:
for model_name in ESM_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
lowerCamelCase : int = EsmModel.from_pretrained(__a )
self.assertIsNotNone(__a )
def a__ ( self: str )-> List[str]:
lowerCamelCase : Union[str, Any] = self.model_tester.prepare_config_and_inputs()[0]
lowerCamelCase : Union[str, Any] = EsmEmbeddings(config=__a )
lowerCamelCase : List[str] = torch.as_tensor([[12, 31, 13, model.padding_idx]] )
lowerCamelCase : Union[str, Any] = torch.as_tensor(
[
[
0 + model.padding_idx + 1,
1 + model.padding_idx + 1,
2 + model.padding_idx + 1,
model.padding_idx,
]
] )
lowerCamelCase : Optional[Any] = create_position_ids_from_input_ids(__a , model.padding_idx )
self.assertEqual(position_ids.shape , expected_positions.shape )
self.assertTrue(torch.all(torch.eq(__a , __a ) ) )
def a__ ( self: Optional[int] )-> int:
lowerCamelCase : List[str] = self.model_tester.prepare_config_and_inputs()[0]
lowerCamelCase : Any = EsmEmbeddings(config=__a )
lowerCamelCase : Dict = torch.empty(2 , 4 , 30 )
lowerCamelCase : List[Any] = [
0 + embeddings.padding_idx + 1,
1 + embeddings.padding_idx + 1,
2 + embeddings.padding_idx + 1,
3 + embeddings.padding_idx + 1,
]
lowerCamelCase : Any = torch.as_tensor([expected_single_positions, expected_single_positions] )
lowerCamelCase : List[str] = embeddings.create_position_ids_from_inputs_embeds(__a )
self.assertEqual(position_ids.shape , expected_positions.shape )
self.assertTrue(torch.all(torch.eq(__a , __a ) ) )
@unittest.skip("""Esm does not support embedding resizing""" )
def a__ ( self: Any )-> Optional[Any]:
pass
@unittest.skip("""Esm does not support embedding resizing""" )
def a__ ( self: Dict )-> Dict:
pass
@unittest.skip("""Will be fixed soon by reducing the size of the model used for common tests.""" )
def a__ ( self: List[str] )-> Dict:
pass
@require_torch
class A__ ( __lowercase):
"""simple docstring"""
@slow
def a__ ( self: Any )-> Union[str, Any]:
with torch.no_grad():
lowerCamelCase : Union[str, Any] = EsmForMaskedLM.from_pretrained("""facebook/esm2_t6_8M_UR50D""" )
model.eval()
lowerCamelCase : List[str] = torch.tensor([[0, 1, 2, 3, 4, 5]] )
lowerCamelCase : Tuple = model(__a )[0]
lowerCamelCase : Dict = 33
lowerCamelCase : List[str] = torch.Size((1, 6, vocab_size) )
self.assertEqual(output.shape , __a )
lowerCamelCase : Tuple = torch.tensor(
[[[8.92_15, -10.58_98, -6.46_71], [-6.39_67, -13.91_14, -1.12_12], [-7.78_12, -13.95_16, -3.74_06]]] )
self.assertTrue(torch.allclose(output[:, :3, :3] , __a , atol=1e-4 ) )
@slow
def a__ ( self: Dict )-> str:
with torch.no_grad():
lowerCamelCase : Any = EsmModel.from_pretrained("""facebook/esm2_t6_8M_UR50D""" )
model.eval()
lowerCamelCase : Optional[Any] = torch.tensor([[0, 6, 4, 13, 5, 4, 16, 12, 11, 7, 2]] )
lowerCamelCase : Any = model(__a )[0]
# compare the actual values for a slice.
lowerCamelCase : Tuple = torch.tensor(
[[[0.14_44, 0.54_13, 0.32_48], [0.30_34, 0.00_53, 0.31_08], [0.32_28, -0.24_99, 0.34_15]]] )
self.assertTrue(torch.allclose(output[:, :3, :3] , __a , atol=1e-4 ) )
| 42
| 0
|
"""simple docstring"""
import numpy as np
def snake_case ( UpperCamelCase__ : Tuple , UpperCamelCase__ : int , UpperCamelCase__ : str , UpperCamelCase__ : Optional[int] , UpperCamelCase__ : Any ) -> Tuple:
lowerCamelCase : str = int(np.ceil((x_end - xa) / h ) )
lowerCamelCase : List[str] = np.zeros((n + 1,) )
lowerCamelCase : List[str] = ya
lowerCamelCase : Optional[int] = xa
for k in range(_SCREAMING_SNAKE_CASE ):
lowerCamelCase : int = f(_SCREAMING_SNAKE_CASE , y[k] )
lowerCamelCase : Optional[int] = f(x + 0.5 * h , y[k] + 0.5 * h * ka )
lowerCamelCase : List[Any] = f(x + 0.5 * h , y[k] + 0.5 * h * ka )
lowerCamelCase : Optional[Any] = f(x + h , y[k] + h * ka )
lowerCamelCase : Dict = y[k] + (1 / 6) * h * (ka + 2 * ka + 2 * ka + ka)
x += h
return y
if __name__ == "__main__":
import doctest
doctest.testmod()
| 706
|
"""simple docstring"""
import unittest
from transformers import AlbertTokenizer, AlbertTokenizerFast
from transformers.testing_utils import get_tests_dir, require_sentencepiece, require_tokenizers, slow
from ...test_tokenization_common import TokenizerTesterMixin
__lowerCamelCase :str = get_tests_dir('fixtures/spiece.model')
@require_sentencepiece
@require_tokenizers
class A__ ( __lowercase , unittest.TestCase):
"""simple docstring"""
snake_case__ : List[str] =AlbertTokenizer
snake_case__ : Optional[Any] =AlbertTokenizerFast
snake_case__ : Optional[int] =True
snake_case__ : Any =True
snake_case__ : Optional[int] =True
def a__ ( self: Dict )-> Optional[Any]:
super().setUp()
# We have a SentencePiece fixture for testing
lowerCamelCase : int = AlbertTokenizer(__a )
tokenizer.save_pretrained(self.tmpdirname )
def a__ ( self: Tuple , __a: Tuple )-> Union[str, Any]:
lowerCamelCase : List[str] = """this is a test"""
lowerCamelCase : int = """this is a test"""
return input_text, output_text
def a__ ( self: Any )-> List[Any]:
lowerCamelCase : int = """<pad>"""
lowerCamelCase : Optional[int] = 0
self.assertEqual(self.get_tokenizer()._convert_token_to_id(__a ) , __a )
self.assertEqual(self.get_tokenizer()._convert_id_to_token(__a ) , __a )
def a__ ( self: Tuple )-> str:
lowerCamelCase : Optional[int] = list(self.get_tokenizer().get_vocab().keys() )
self.assertEqual(vocab_keys[0] , """<pad>""" )
self.assertEqual(vocab_keys[1] , """<unk>""" )
self.assertEqual(vocab_keys[-1] , """▁eloquent""" )
self.assertEqual(len(__a ) , 30_000 )
def a__ ( self: List[str] )-> Any:
self.assertEqual(self.get_tokenizer().vocab_size , 30_000 )
def a__ ( self: Optional[Any] )-> Union[str, Any]:
if not self.test_rust_tokenizer:
return
lowerCamelCase : str = self.get_tokenizer()
lowerCamelCase : Tuple = self.get_rust_tokenizer()
lowerCamelCase : Union[str, Any] = """I was born in 92000, and this is falsé."""
lowerCamelCase : List[str] = tokenizer.tokenize(__a )
lowerCamelCase : Tuple = rust_tokenizer.tokenize(__a )
self.assertListEqual(__a , __a )
lowerCamelCase : Dict = tokenizer.encode(__a , add_special_tokens=__a )
lowerCamelCase : List[str] = rust_tokenizer.encode(__a , add_special_tokens=__a )
self.assertListEqual(__a , __a )
lowerCamelCase : Any = self.get_rust_tokenizer()
lowerCamelCase : List[str] = tokenizer.encode(__a )
lowerCamelCase : str = rust_tokenizer.encode(__a )
self.assertListEqual(__a , __a )
def a__ ( self: Tuple )-> List[Any]:
lowerCamelCase : List[str] = AlbertTokenizer(__a , keep_accents=__a )
lowerCamelCase : int = tokenizer.tokenize("""This is a test""" )
self.assertListEqual(__a , ["""▁this""", """▁is""", """▁a""", """▁test"""] )
self.assertListEqual(tokenizer.convert_tokens_to_ids(__a ) , [48, 25, 21, 1_289] )
lowerCamelCase : List[str] = tokenizer.tokenize("""I was born in 92000, and this is falsé.""" )
self.assertListEqual(
__a , ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """é""", """."""] )
lowerCamelCase : List[str] = tokenizer.convert_tokens_to_ids(__a )
self.assertListEqual(__a , [31, 23, 386, 19, 561, 3_050, 15, 17, 48, 25, 8_256, 18, 1, 9] )
lowerCamelCase : Tuple = tokenizer.convert_ids_to_tokens(__a )
self.assertListEqual(
__a , ["""▁i""", """▁was""", """▁born""", """▁in""", """▁9""", """2000""", """,""", """▁and""", """▁this""", """▁is""", """▁fal""", """s""", """<unk>""", """."""] , )
def a__ ( self: Tuple )-> str:
lowerCamelCase : str = AlbertTokenizer(__a )
lowerCamelCase : Union[str, Any] = tokenizer.encode("""sequence builders""" )
lowerCamelCase : List[Any] = tokenizer.encode("""multi-sequence build""" )
lowerCamelCase : Any = tokenizer.build_inputs_with_special_tokens(__a )
lowerCamelCase : str = tokenizer.build_inputs_with_special_tokens(__a , __a )
assert encoded_sentence == [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id]
assert encoded_pair == [tokenizer.cls_token_id] + text + [tokenizer.sep_token_id] + text_a + [
tokenizer.sep_token_id
]
@slow
def a__ ( self: Any )-> Dict:
# fmt: off
lowerCamelCase : Optional[Any] = {"""attention_mask""": [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """input_ids""": [[2, 21_970, 13, 5, 6_092, 167, 28, 7_103, 2_153, 673, 8, 7_028, 12_051, 18, 17, 7_103, 2_153, 673, 8, 3_515, 18_684, 8, 4_461, 6, 1_927, 297, 8, 12_060, 2_607, 18, 13, 5, 4_461, 15, 10_538, 38, 8, 135, 15, 822, 58, 15, 993, 10_363, 15, 1_460, 8_005, 4_461, 15, 993, 255, 2_328, 9, 9, 9, 6, 26, 1_112, 816, 3_260, 13, 5, 103, 2_377, 6, 17, 1_112, 816, 2_782, 13, 5, 103, 10_641, 6, 29, 84, 2_512, 2_430, 782, 18_684, 2_761, 19, 808, 2_430, 2_556, 17, 855, 1_480, 9_477, 4_091, 128, 11_712, 15, 7_103, 2_153, 673, 17, 24_883, 9_990, 9, 3], [2, 11_502, 25, 1_006, 20, 782, 8, 11_809, 855, 1_732, 19_393, 18_667, 37, 367, 21_018, 69, 1_854, 34, 11_860, 19_124, 27, 156, 225, 17, 193, 4_141, 19, 65, 9_124, 9, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [2, 14, 2_231, 886, 2_385, 17_659, 84, 14, 16_792, 1_952, 9, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], """token_type_ids""": [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]} # noqa: E501
# fmt: on
self.tokenizer_integration_test_util(
expected_encoding=__a , model_name="""albert-base-v2""" , revision="""6b6560eaf5ff2e250b00c50f380c5389a9c2d82e""" , )
| 42
| 0
|
import unittest
from transformers import (
MODEL_FOR_SEQUENCE_CLASSIFICATION_MAPPING,
TF_MODEL_FOR_SEQUENCE_CLASSIFICATION_MAPPING,
TextClassificationPipeline,
pipeline,
)
from transformers.testing_utils import is_pipeline_test, nested_simplify, require_tf, require_torch, slow
from .test_pipelines_common import ANY
# These 2 model types require different inputs than those of the usual text models.
__lowerCamelCase :Tuple = {'LayoutLMv2Config', 'LayoutLMv3Config'}
@is_pipeline_test
class A__ ( unittest.TestCase):
"""simple docstring"""
snake_case__ : List[Any] =MODEL_FOR_SEQUENCE_CLASSIFICATION_MAPPING
snake_case__ : Union[str, Any] =TF_MODEL_FOR_SEQUENCE_CLASSIFICATION_MAPPING
if model_mapping is not None:
snake_case__ : Dict ={config: model for config, model in model_mapping.items() if config.__name__ not in _TO_SKIP}
if tf_model_mapping is not None:
snake_case__ : List[str] ={
config: model for config, model in tf_model_mapping.items() if config.__name__ not in _TO_SKIP
}
@require_torch
def a__ ( self: List[str] )-> Optional[Any]:
lowerCamelCase : Optional[Any] = pipeline(
task="""text-classification""" , model="""hf-internal-testing/tiny-random-distilbert""" , framework="""pt""" )
lowerCamelCase : Tuple = text_classifier("""This is great !""" )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """LABEL_0""", """score""": 0.5_04}] )
lowerCamelCase : int = text_classifier("""This is great !""" , top_k=2 )
self.assertEqual(
nested_simplify(__a ) , [{"""label""": """LABEL_0""", """score""": 0.5_04}, {"""label""": """LABEL_1""", """score""": 0.4_96}] )
lowerCamelCase : List[str] = text_classifier(["""This is great !""", """This is bad"""] , top_k=2 )
self.assertEqual(
nested_simplify(__a ) , [
[{"""label""": """LABEL_0""", """score""": 0.5_04}, {"""label""": """LABEL_1""", """score""": 0.4_96}],
[{"""label""": """LABEL_0""", """score""": 0.5_04}, {"""label""": """LABEL_1""", """score""": 0.4_96}],
] , )
lowerCamelCase : int = text_classifier("""This is great !""" , top_k=1 )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """LABEL_0""", """score""": 0.5_04}] )
# Legacy behavior
lowerCamelCase : Optional[Any] = text_classifier("""This is great !""" , return_all_scores=__a )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """LABEL_0""", """score""": 0.5_04}] )
lowerCamelCase : List[str] = text_classifier("""This is great !""" , return_all_scores=__a )
self.assertEqual(
nested_simplify(__a ) , [[{"""label""": """LABEL_0""", """score""": 0.5_04}, {"""label""": """LABEL_1""", """score""": 0.4_96}]] )
lowerCamelCase : Union[str, Any] = text_classifier(["""This is great !""", """Something else"""] , return_all_scores=__a )
self.assertEqual(
nested_simplify(__a ) , [
[{"""label""": """LABEL_0""", """score""": 0.5_04}, {"""label""": """LABEL_1""", """score""": 0.4_96}],
[{"""label""": """LABEL_0""", """score""": 0.5_04}, {"""label""": """LABEL_1""", """score""": 0.4_96}],
] , )
lowerCamelCase : List[str] = text_classifier(["""This is great !""", """Something else"""] , return_all_scores=__a )
self.assertEqual(
nested_simplify(__a ) , [
{"""label""": """LABEL_0""", """score""": 0.5_04},
{"""label""": """LABEL_0""", """score""": 0.5_04},
] , )
@require_torch
def a__ ( self: Dict )-> List[str]:
import torch
lowerCamelCase : Tuple = pipeline(
task="""text-classification""" , model="""hf-internal-testing/tiny-random-distilbert""" , framework="""pt""" , device=torch.device("""cpu""" ) , )
lowerCamelCase : str = text_classifier("""This is great !""" )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """LABEL_0""", """score""": 0.5_04}] )
@require_tf
def a__ ( self: Optional[Any] )-> int:
lowerCamelCase : Any = pipeline(
task="""text-classification""" , model="""hf-internal-testing/tiny-random-distilbert""" , framework="""tf""" )
lowerCamelCase : Tuple = text_classifier("""This is great !""" )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """LABEL_0""", """score""": 0.5_04}] )
@slow
@require_torch
def a__ ( self: Dict )-> Dict:
lowerCamelCase : str = pipeline("""text-classification""" )
lowerCamelCase : str = text_classifier("""This is great !""" )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """POSITIVE""", """score""": 1.0}] )
lowerCamelCase : Tuple = text_classifier("""This is bad !""" )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """NEGATIVE""", """score""": 1.0}] )
lowerCamelCase : Optional[int] = text_classifier("""Birds are a type of animal""" )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """POSITIVE""", """score""": 0.9_88}] )
@slow
@require_tf
def a__ ( self: Dict )-> List[str]:
lowerCamelCase : Optional[Any] = pipeline("""text-classification""" , framework="""tf""" )
lowerCamelCase : str = text_classifier("""This is great !""" )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """POSITIVE""", """score""": 1.0}] )
lowerCamelCase : Dict = text_classifier("""This is bad !""" )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """NEGATIVE""", """score""": 1.0}] )
lowerCamelCase : str = text_classifier("""Birds are a type of animal""" )
self.assertEqual(nested_simplify(__a ) , [{"""label""": """POSITIVE""", """score""": 0.9_88}] )
def a__ ( self: str , __a: str , __a: int , __a: Dict )-> Tuple:
lowerCamelCase : List[str] = TextClassificationPipeline(model=__a , tokenizer=__a )
return text_classifier, ["HuggingFace is in", "This is another test"]
def a__ ( self: Optional[int] , __a: List[Any] , __a: List[str] )-> Optional[int]:
lowerCamelCase : Dict = text_classifier.model
# Small inputs because BartTokenizer tiny has maximum position embeddings = 22
lowerCamelCase : int = """HuggingFace is in"""
lowerCamelCase : Optional[Any] = text_classifier(__a )
self.assertEqual(nested_simplify(__a ) , [{"""label""": ANY(__a ), """score""": ANY(__a )}] )
self.assertTrue(outputs[0]["""label"""] in model.config.idalabel.values() )
lowerCamelCase : List[Any] = ["""HuggingFace is in """, """Paris is in France"""]
lowerCamelCase : Dict = text_classifier(__a )
self.assertEqual(
nested_simplify(__a ) , [{"""label""": ANY(__a ), """score""": ANY(__a )}, {"""label""": ANY(__a ), """score""": ANY(__a )}] , )
self.assertTrue(outputs[0]["""label"""] in model.config.idalabel.values() )
self.assertTrue(outputs[1]["""label"""] in model.config.idalabel.values() )
# Forcing to get all results with `top_k=None`
# This is NOT the legacy format
lowerCamelCase : Any = text_classifier(__a , top_k=__a )
lowerCamelCase : Dict = len(model.config.idalabel.values() )
self.assertEqual(
nested_simplify(__a ) , [[{"""label""": ANY(__a ), """score""": ANY(__a )}] * N, [{"""label""": ANY(__a ), """score""": ANY(__a )}] * N] , )
lowerCamelCase : Tuple = {"""text""": """HuggingFace is in """, """text_pair""": """Paris is in France"""}
lowerCamelCase : Dict = text_classifier(__a )
self.assertEqual(
nested_simplify(__a ) , {"""label""": ANY(__a ), """score""": ANY(__a )} , )
self.assertTrue(outputs["""label"""] in model.config.idalabel.values() )
# This might be used a text pair, but tokenizer + pipe interaction
# makes it hard to understand that it's not using the pair properly
# https://github.com/huggingface/transformers/issues/17305
# We disabled this usage instead as it was outputting wrong outputs.
lowerCamelCase : Union[str, Any] = [["""HuggingFace is in """, """Paris is in France"""]]
with self.assertRaises(__a ):
text_classifier(__a )
# This used to be valid for doing text pairs
# We're keeping it working because of backward compatibility
lowerCamelCase : Any = text_classifier([[["""HuggingFace is in """, """Paris is in France"""]]] )
self.assertEqual(
nested_simplify(__a ) , [{"""label""": ANY(__a ), """score""": ANY(__a )}] , )
self.assertTrue(outputs[0]["""label"""] in model.config.idalabel.values() )
| 707
|
"""simple docstring"""
__lowerCamelCase :List[Any] = {0: [2, 3], 1: [0], 2: [1], 3: [4], 4: []}
__lowerCamelCase :Union[str, Any] = {0: [1, 2, 3], 1: [2], 2: [0], 3: [4], 4: [5], 5: [3]}
def snake_case ( UpperCamelCase__ : dict[int, list[int]] , UpperCamelCase__ : int , UpperCamelCase__ : list[bool] ) -> list[int]:
lowerCamelCase : Tuple = True
lowerCamelCase : Any = []
for neighbour in graph[vert]:
if not visited[neighbour]:
order += topology_sort(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
order.append(UpperCamelCase__ )
return order
def snake_case ( UpperCamelCase__ : dict[int, list[int]] , UpperCamelCase__ : int , UpperCamelCase__ : list[bool] ) -> list[int]:
lowerCamelCase : List[Any] = True
lowerCamelCase : int = [vert]
for neighbour in reversed_graph[vert]:
if not visited[neighbour]:
component += find_components(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
return component
def snake_case ( UpperCamelCase__ : dict[int, list[int]] ) -> list[list[int]]:
lowerCamelCase : int = len(UpperCamelCase__ ) * [False]
lowerCamelCase : dict[int, list[int]] = {vert: [] for vert in range(len(UpperCamelCase__ ) )}
for vert, neighbours in graph.items():
for neighbour in neighbours:
reversed_graph[neighbour].append(UpperCamelCase__ )
lowerCamelCase : int = []
for i, was_visited in enumerate(UpperCamelCase__ ):
if not was_visited:
order += topology_sort(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
lowerCamelCase : Tuple = []
lowerCamelCase : str = len(UpperCamelCase__ ) * [False]
for i in range(len(UpperCamelCase__ ) ):
lowerCamelCase : Any = order[len(UpperCamelCase__ ) - i - 1]
if not visited[vert]:
lowerCamelCase : List[str] = find_components(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
components_list.append(UpperCamelCase__ )
return components_list
| 42
| 0
|
"""simple docstring"""
def lowerCAmelCase ( UpperCamelCase__ : float , UpperCamelCase__ : float , UpperCamelCase__ : int ) -> List[str]:
if principal <= 0:
raise Exception("""Principal borrowed must be > 0""" )
if rate_per_annum < 0:
raise Exception("""Rate of interest must be >= 0""" )
if years_to_repay <= 0 or not isinstance(lowerCAmelCase__ , lowerCAmelCase__ ):
raise Exception("""Years to repay must be an integer > 0""" )
# Yearly rate is divided by 12 to get monthly rate
lowerCamelCase : int = rate_per_annum / 12
# Years to repay is multiplied by 12 to get number of payments as payment is monthly
lowerCamelCase : Optional[int] = years_to_repay * 12
return (
principal
* rate_per_month
* (1 + rate_per_month) ** number_of_payments
/ ((1 + rate_per_month) ** number_of_payments - 1)
)
if __name__ == "__main__":
import doctest
doctest.testmod()
| 708
|
"""simple docstring"""
from typing import List, Optional, Union
from ...configuration_utils import PretrainedConfig
from ...utils import logging
__lowerCamelCase :str = logging.get_logger(__name__)
__lowerCamelCase :Any = {
'huggingface/time-series-transformer-tourism-monthly': (
'https://huggingface.co/huggingface/time-series-transformer-tourism-monthly/resolve/main/config.json'
),
# See all TimeSeriesTransformer models at https://huggingface.co/models?filter=time_series_transformer
}
class A__ ( __lowercase):
"""simple docstring"""
snake_case__ : List[Any] ='''time_series_transformer'''
snake_case__ : List[Any] ={
'''hidden_size''': '''d_model''',
'''num_attention_heads''': '''encoder_attention_heads''',
'''num_hidden_layers''': '''encoder_layers''',
}
def __init__( self: List[str] , __a: Optional[int] = None , __a: Optional[int] = None , __a: str = "student_t" , __a: str = "nll" , __a: int = 1 , __a: List[int] = [1, 2, 3, 4, 5, 6, 7] , __a: Optional[Union[str, bool]] = "mean" , __a: int = 0 , __a: int = 0 , __a: int = 0 , __a: int = 0 , __a: Optional[List[int]] = None , __a: Optional[List[int]] = None , __a: int = 32 , __a: int = 32 , __a: int = 2 , __a: int = 2 , __a: int = 2 , __a: int = 2 , __a: bool = True , __a: str = "gelu" , __a: int = 64 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: float = 0.1 , __a: int = 100 , __a: float = 0.02 , __a: Tuple=True , **__a: str , )-> Any:
# time series specific configuration
lowerCamelCase : str = prediction_length
lowerCamelCase : Optional[Any] = context_length or prediction_length
lowerCamelCase : Tuple = distribution_output
lowerCamelCase : Any = loss
lowerCamelCase : List[Any] = input_size
lowerCamelCase : int = num_time_features
lowerCamelCase : Dict = lags_sequence
lowerCamelCase : Optional[int] = scaling
lowerCamelCase : int = num_dynamic_real_features
lowerCamelCase : Tuple = num_static_real_features
lowerCamelCase : Any = num_static_categorical_features
if cardinality and num_static_categorical_features > 0:
if len(__a ) != num_static_categorical_features:
raise ValueError(
"""The cardinality should be a list of the same length as `num_static_categorical_features`""" )
lowerCamelCase : int = cardinality
else:
lowerCamelCase : Dict = [0]
if embedding_dimension and num_static_categorical_features > 0:
if len(__a ) != num_static_categorical_features:
raise ValueError(
"""The embedding dimension should be a list of the same length as `num_static_categorical_features`""" )
lowerCamelCase : str = embedding_dimension
else:
lowerCamelCase : str = [min(50 , (cat + 1) // 2 ) for cat in self.cardinality]
lowerCamelCase : Any = num_parallel_samples
# Transformer architecture configuration
lowerCamelCase : Any = input_size * len(__a ) + self._number_of_features
lowerCamelCase : List[str] = d_model
lowerCamelCase : Tuple = encoder_attention_heads
lowerCamelCase : Optional[int] = decoder_attention_heads
lowerCamelCase : Union[str, Any] = encoder_ffn_dim
lowerCamelCase : str = decoder_ffn_dim
lowerCamelCase : str = encoder_layers
lowerCamelCase : Any = decoder_layers
lowerCamelCase : Optional[int] = dropout
lowerCamelCase : List[str] = attention_dropout
lowerCamelCase : Tuple = activation_dropout
lowerCamelCase : Optional[int] = encoder_layerdrop
lowerCamelCase : int = decoder_layerdrop
lowerCamelCase : Optional[int] = activation_function
lowerCamelCase : Optional[Any] = init_std
lowerCamelCase : Optional[Any] = use_cache
super().__init__(is_encoder_decoder=__a , **__a )
@property
def a__ ( self: int )-> int:
return (
sum(self.embedding_dimension )
+ self.num_dynamic_real_features
+ self.num_time_features
+ self.num_static_real_features
+ self.input_size * 2 # the log1p(abs(loc)) and log(scale) features
)
| 42
| 0
|
"""simple docstring"""
from typing import TYPE_CHECKING
# rely on isort to merge the imports
from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available
__lowerCamelCase :Any = {
'configuration_autoformer': [
'AUTOFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP',
'AutoformerConfig',
],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Tuple = [
'AUTOFORMER_PRETRAINED_MODEL_ARCHIVE_LIST',
'AutoformerForPrediction',
'AutoformerModel',
'AutoformerPreTrainedModel',
]
if TYPE_CHECKING:
from .configuration_autoformer import (
AUTOFORMER_PRETRAINED_CONFIG_ARCHIVE_MAP,
AutoformerConfig,
)
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_autoformer import (
AUTOFORMER_PRETRAINED_MODEL_ARCHIVE_LIST,
AutoformerForPrediction,
AutoformerModel,
AutoformerPreTrainedModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 709
|
"""simple docstring"""
from __future__ import annotations
__lowerCamelCase :int = 10
def snake_case ( UpperCamelCase__ : list[int] ) -> list[int]:
lowerCamelCase : int = 1
lowerCamelCase : Union[str, Any] = max(UpperCamelCase__ )
while placement <= max_digit:
# declare and initialize empty buckets
lowerCamelCase : list[list] = [[] for _ in range(UpperCamelCase__ )]
# split list_of_ints between the buckets
for i in list_of_ints:
lowerCamelCase : Any = int((i / placement) % RADIX )
buckets[tmp].append(UpperCamelCase__ )
# put each buckets' contents into list_of_ints
lowerCamelCase : Dict = 0
for b in range(UpperCamelCase__ ):
for i in buckets[b]:
lowerCamelCase : List[str] = i
a += 1
# move to next
placement *= RADIX
return list_of_ints
if __name__ == "__main__":
import doctest
doctest.testmod()
| 42
| 0
|
"""simple docstring"""
def snake_case ( UpperCamelCase__ : str ) -> Union[str, Any]:
if not all(char in """01""" for char in bin_string ):
raise ValueError("""Non-binary value was passed to the function""" )
if not bin_string:
raise ValueError("""Empty string was passed to the function""" )
lowerCamelCase : List[str] = """"""
while len(UpperCamelCase__ ) % 3 != 0:
lowerCamelCase : Union[str, Any] = """0""" + bin_string
lowerCamelCase : Tuple = [
bin_string[index : index + 3]
for index in range(len(UpperCamelCase__ ) )
if index % 3 == 0
]
for bin_group in bin_string_in_3_list:
lowerCamelCase : int = 0
for index, val in enumerate(UpperCamelCase__ ):
oct_val += int(2 ** (2 - index) * int(UpperCamelCase__ ) )
oct_string += str(UpperCamelCase__ )
return oct_string
if __name__ == "__main__":
from doctest import testmod
testmod()
| 710
|
"""simple docstring"""
import argparse
import pickle
import numpy as np
import torch
from torch import nn
from transformers import ReformerConfig, ReformerModelWithLMHead
from transformers.utils import logging
logging.set_verbosity_info()
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Tuple , UpperCamelCase__ : Optional[Any]=None ) -> Tuple:
# set parameter of one layer
assert torch_layer.weight.shape == weight.shape, F'{torch_layer} layer.weight does not match'
lowerCamelCase : Dict = nn.Parameter(UpperCamelCase__ )
if bias is not None:
assert torch_layer.bias.shape == bias.shape, F'{torch_layer} layer.bias does not match'
lowerCamelCase : Any = nn.Parameter(UpperCamelCase__ )
def snake_case ( UpperCamelCase__ : int , UpperCamelCase__ : Any , UpperCamelCase__ : Tuple ) -> Union[str, Any]:
# set torch weights for 1-to-1 comparison
lowerCamelCase : Dict = np.asarray(weights[0] )
lowerCamelCase : List[Any] = np.asarray(weights[1] )
lowerCamelCase : List[str] = np.asarray(weights[2] )
set_param(
torch_layer.self_attention.query_key , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.value , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.output.dense , torch.tensor(UpperCamelCase__ ).view(-1 , UpperCamelCase__ ).contiguous().transpose(0 , 1 ) , )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : Any ) -> List[Any]:
# set torch weights for 1-to-1 comparison
lowerCamelCase : Tuple = np.asarray(weights[0] )
lowerCamelCase : Any = np.asarray(weights[1] )
lowerCamelCase : List[Any] = np.asarray(weights[2] )
lowerCamelCase : List[str] = np.asarray(weights[3] )
set_param(
torch_layer.self_attention.query , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.key , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.self_attention.value , torch.tensor(UpperCamelCase__ ).transpose(1 , 2 ).contiguous().view(-1 , UpperCamelCase__ ) , )
set_param(
torch_layer.output.dense , torch.tensor(UpperCamelCase__ ).view(-1 , UpperCamelCase__ ).contiguous().transpose(0 , 1 ) , )
def snake_case ( UpperCamelCase__ : Dict , UpperCamelCase__ : Optional[int] , UpperCamelCase__ : Any ) -> Optional[Any]:
# layernorm 1
lowerCamelCase : str = weights[0][0][0]
lowerCamelCase : Optional[int] = np.asarray(layer_norm_a[0] )
lowerCamelCase : Tuple = np.asarray(layer_norm_a[1] )
set_param(
torch_block.attention.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# lsh weights + output
lowerCamelCase : List[Any] = weights[0][1]
if len(UpperCamelCase__ ) < 4:
set_layer_weights_in_torch_lsh(UpperCamelCase__ , torch_block.attention , UpperCamelCase__ )
else:
set_layer_weights_in_torch_local(UpperCamelCase__ , torch_block.attention , UpperCamelCase__ )
# intermediate weighs
lowerCamelCase : int = weights[2][0][1][2]
# Chunked Feed Forward
if len(UpperCamelCase__ ) == 4:
lowerCamelCase : Dict = intermediate_weights[2]
# layernorm 2
lowerCamelCase : Optional[int] = np.asarray(intermediate_weights[0][0] )
lowerCamelCase : Tuple = np.asarray(intermediate_weights[0][1] )
set_param(
torch_block.feed_forward.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# intermediate dense
lowerCamelCase : Optional[Any] = np.asarray(intermediate_weights[1][0] )
lowerCamelCase : Union[str, Any] = np.asarray(intermediate_weights[1][1] )
set_param(
torch_block.feed_forward.dense.dense , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
# intermediate out
lowerCamelCase : Optional[int] = np.asarray(intermediate_weights[4][0] )
lowerCamelCase : List[Any] = np.asarray(intermediate_weights[4][1] )
set_param(
torch_block.feed_forward.output.dense , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
def snake_case ( UpperCamelCase__ : Optional[Any] , UpperCamelCase__ : List[str] , UpperCamelCase__ : int ) -> List[Any]:
# reformer model
lowerCamelCase : List[Any] = torch_model.reformer
# word embeds
lowerCamelCase : Union[str, Any] = np.asarray(weights[1] )
set_param(
torch_model_reformer.embeddings.word_embeddings , torch.tensor(UpperCamelCase__ ) , )
if isinstance(weights[3] , UpperCamelCase__ ):
lowerCamelCase : Union[str, Any] = torch_model_reformer.embeddings.position_embeddings
for emb_idx in range(len(position_embeddings.weights ) ):
lowerCamelCase : str = np.asarray(weights[3][emb_idx][0] )
assert (
position_embeddings.weights[emb_idx].shape == emb_weights.shape
), F'{position_embeddings[emb_idx]} emb does not match'
lowerCamelCase : Dict = nn.Parameter(torch.tensor(UpperCamelCase__ ) )
lowerCamelCase : int = weights[5]
assert len(torch_model_reformer.encoder.layers ) * 4 == len(
UpperCamelCase__ ), "HF and trax model do not have the same number of layers"
for layer_idx, layer in enumerate(torch_model_reformer.encoder.layers ):
lowerCamelCase : Dict = trax_layer_weights[4 * layer_idx : 4 * (layer_idx + 1)]
set_block_weights_in_torch(UpperCamelCase__ , UpperCamelCase__ , UpperCamelCase__ )
# output layer norm
lowerCamelCase : Any = np.asarray(weights[7][0] )
lowerCamelCase : List[str] = np.asarray(weights[7][1] )
set_param(
torch_model_reformer.encoder.layer_norm , torch.tensor(UpperCamelCase__ ) , torch.tensor(UpperCamelCase__ ) , )
# output embeddings
lowerCamelCase : List[Any] = np.asarray(weights[9][0] )
lowerCamelCase : Optional[int] = np.asarray(weights[9][1] )
set_param(
torch_model.lm_head.decoder , torch.tensor(UpperCamelCase__ ).transpose(0 , 1 ).contiguous() , torch.tensor(UpperCamelCase__ ) , )
def snake_case ( UpperCamelCase__ : str , UpperCamelCase__ : int , UpperCamelCase__ : str ) -> Optional[int]:
# Initialise PyTorch model
lowerCamelCase : Union[str, Any] = ReformerConfig.from_json_file(UpperCamelCase__ )
print(F'Building PyTorch model from configuration: {config}' )
lowerCamelCase : str = ReformerModelWithLMHead(UpperCamelCase__ )
with open(UpperCamelCase__ , """rb""" ) as f:
lowerCamelCase : str = pickle.load(UpperCamelCase__ )["""weights"""]
set_model_weights_in_torch(UpperCamelCase__ , UpperCamelCase__ , config.hidden_size )
# Save pytorch-model
print(F'Save PyTorch model to {pytorch_dump_path}' )
torch.save(model.state_dict() , UpperCamelCase__ )
if __name__ == "__main__":
__lowerCamelCase :Dict = argparse.ArgumentParser()
# Required parameters
parser.add_argument(
'--trax_model_pkl_path', default=None, type=str, required=True, help='Path to the TensorFlow checkpoint path.'
)
parser.add_argument(
'--config_file',
default=None,
type=str,
required=True,
help=(
'The config json file corresponding to the pre-trained Reformer model. \n'
'This specifies the model architecture.'
),
)
parser.add_argument(
'--pytorch_dump_path', default=None, type=str, required=True, help='Path to the output PyTorch model.'
)
__lowerCamelCase :Optional[int] = parser.parse_args()
convert_trax_checkpoint_to_pytorch(args.trax_model_pkl_path, args.config_file, args.pytorch_dump_path)
| 42
| 0
|
"""simple docstring"""
import functools
def snake_case ( UpperCamelCase__ : Dict , UpperCamelCase__ : Dict ) -> int:
if not isinstance(__lowerCAmelCase , __lowerCAmelCase ) or not all(isinstance(__lowerCAmelCase , __lowerCAmelCase ) for day in days ):
raise ValueError("""The parameter days should be a list of integers""" )
if len(__lowerCAmelCase ) != 3 or not all(isinstance(__lowerCAmelCase , __lowerCAmelCase ) for cost in costs ):
raise ValueError("""The parameter costs should be a list of three integers""" )
if len(__lowerCAmelCase ) == 0:
return 0
if min(__lowerCAmelCase ) <= 0:
raise ValueError("""All days elements should be greater than 0""" )
if max(__lowerCAmelCase ) >= 366:
raise ValueError("""All days elements should be less than 366""" )
lowerCamelCase : Tuple = set(__lowerCAmelCase )
@functools.cache
def dynamic_programming(UpperCamelCase__ : int ) -> int:
if index > 365:
return 0
if index not in days_set:
return dynamic_programming(index + 1 )
return min(
costs[0] + dynamic_programming(index + 1 ) , costs[1] + dynamic_programming(index + 7 ) , costs[2] + dynamic_programming(index + 30 ) , )
return dynamic_programming(1 )
if __name__ == "__main__":
import doctest
doctest.testmod()
| 711
|
"""simple docstring"""
import inspect
import unittest
import torch
import torch.nn as nn
from accelerate.hooks import (
AlignDevicesHook,
ModelHook,
SequentialHook,
add_hook_to_module,
attach_align_device_hook,
remove_hook_from_module,
remove_hook_from_submodules,
)
from accelerate.test_utils import require_multi_gpu
class A__ ( nn.Module):
"""simple docstring"""
def __init__( self: Dict )-> Dict:
super().__init__()
lowerCamelCase : Tuple = nn.Linear(3 , 4 )
lowerCamelCase : Optional[Any] = nn.BatchNormad(4 )
lowerCamelCase : Optional[Any] = nn.Linear(4 , 5 )
def a__ ( self: List[str] , __a: List[Any] )-> Optional[Any]:
return self.lineara(self.batchnorm(self.lineara(__a ) ) )
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Tuple , __a: int , *__a: Any , **__a: Tuple )-> Tuple:
return (args[0] + 1,) + args[1:], kwargs
class A__ ( __lowercase):
"""simple docstring"""
def a__ ( self: Optional[int] , __a: List[str] , __a: List[Any] )-> List[str]:
return output + 1
class A__ ( unittest.TestCase):
"""simple docstring"""
def a__ ( self: int )-> str:
lowerCamelCase : List[str] = ModelForTest()
lowerCamelCase : Dict = ModelHook()
add_hook_to_module(__a , __a )
self.assertEqual(test_model._hf_hook , __a )
self.assertTrue(hasattr(__a , """_old_forward""" ) )
# Check adding the hook did not change the name or the signature
self.assertEqual(test_model.forward.__name__ , """forward""" )
self.assertListEqual(list(inspect.signature(test_model.forward ).parameters ) , ["""x"""] )
remove_hook_from_module(__a )
self.assertFalse(hasattr(__a , """_hf_hook""" ) )
self.assertFalse(hasattr(__a , """_old_forward""" ) )
def a__ ( self: int )-> str:
lowerCamelCase : List[str] = ModelForTest()
lowerCamelCase : Union[str, Any] = ModelHook()
add_hook_to_module(__a , __a )
add_hook_to_module(__a , __a , append=__a )
self.assertEqual(isinstance(test_model._hf_hook , __a ) , __a )
self.assertEqual(len(test_model._hf_hook.hooks ) , 2 )
self.assertTrue(hasattr(__a , """_old_forward""" ) )
# Check adding the hook did not change the name or the signature
self.assertEqual(test_model.forward.__name__ , """forward""" )
self.assertListEqual(list(inspect.signature(test_model.forward ).parameters ) , ["""x"""] )
remove_hook_from_module(__a )
self.assertFalse(hasattr(__a , """_hf_hook""" ) )
self.assertFalse(hasattr(__a , """_old_forward""" ) )
def a__ ( self: List[Any] )-> List[str]:
lowerCamelCase : str = ModelForTest()
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : Union[str, Any] = test_model(x + 1 )
lowerCamelCase : Optional[int] = test_model(x + 2 )
lowerCamelCase : List[Any] = PreForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[int] = test_model(__a )
self.assertTrue(torch.allclose(__a , __a , atol=1e-5 ) )
# Attaching a hook to a model when it already has one replaces, does not chain
lowerCamelCase : Dict = PreForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Tuple = test_model(__a )
self.assertTrue(torch.allclose(__a , __a , atol=1e-5 ) )
# You need to use the sequential hook to chain two or more hooks
lowerCamelCase : Any = SequentialHook(PreForwardHook() , PreForwardHook() )
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[Any] = test_model(__a )
assert torch.allclose(__a , __a , atol=1e-5 )
def a__ ( self: Any )-> Optional[int]:
lowerCamelCase : str = ModelForTest()
lowerCamelCase : List[str] = torch.randn(2 , 3 )
lowerCamelCase : int = test_model(__a )
lowerCamelCase : Dict = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Tuple = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 , atol=1e-5 ) )
# Attaching a hook to a model when it already has one replaces, does not chain
lowerCamelCase : str = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : Optional[Any] = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 , atol=1e-5 ) )
# You need to use the sequential hook to chain two or more hooks
lowerCamelCase : Union[str, Any] = SequentialHook(PostForwardHook() , PostForwardHook() )
add_hook_to_module(__a , __a )
lowerCamelCase : str = test_model(__a )
assert torch.allclose(__a , output + 2 , atol=1e-5 )
def a__ ( self: int )-> Dict:
lowerCamelCase : List[Any] = ModelForTest()
lowerCamelCase : Optional[int] = torch.randn(2 , 3 )
lowerCamelCase : List[str] = test_model(__a )
lowerCamelCase : Any = PostForwardHook()
add_hook_to_module(__a , __a )
lowerCamelCase : str = test_model(__a )
self.assertTrue(torch.allclose(__a , output + 1 ) )
self.assertTrue(outputa.requires_grad )
lowerCamelCase : Optional[int] = True
lowerCamelCase : Optional[int] = test_model(__a )
self.assertFalse(outputa.requires_grad )
@require_multi_gpu
def a__ ( self: List[str] )-> Union[str, Any]:
lowerCamelCase : int = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
add_hook_to_module(model.lineara , AlignDevicesHook(execution_device=0 ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(execution_device=0 ) )
add_hook_to_module(model.lineara , AlignDevicesHook(execution_device=1 ) )
self.assertEqual(model.lineara.weight.device , torch.device(0 ) )
self.assertEqual(model.batchnorm.weight.device , torch.device(0 ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device(0 ) )
self.assertEqual(model.lineara.weight.device , torch.device(1 ) )
# We can still make a forward pass. The input does not need to be on any particular device
lowerCamelCase : str = torch.randn(2 , 3 )
lowerCamelCase : Dict = model(__a )
self.assertEqual(output.device , torch.device(1 ) )
# We can add a general hook to put back output on same device as input.
add_hook_to_module(__a , AlignDevicesHook(io_same_device=__a ) )
lowerCamelCase : Optional[int] = torch.randn(2 , 3 ).to(0 )
lowerCamelCase : str = model(__a )
self.assertEqual(output.device , torch.device(0 ) )
def a__ ( self: List[str] )-> Tuple:
lowerCamelCase : Union[str, Any] = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : Tuple = {"""execution_device""": 0 if torch.cuda.is_available() else """cpu""", """offload""": True}
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(**__a ) )
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(hook_kwargs["""execution_device"""] )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Optional[Any] = torch.randn(2 , 3 )
lowerCamelCase : Optional[Any] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_module(model.lineara )
remove_hook_from_module(model.batchnorm )
remove_hook_from_module(model.lineara )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
lowerCamelCase : Any = {
"""execution_device""": 0 if torch.cuda.is_available() else """cpu""",
"""offload""": True,
"""offload_buffers""": True,
}
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
add_hook_to_module(model.batchnorm , AlignDevicesHook(**__a ) )
add_hook_to_module(model.lineara , AlignDevicesHook(**__a ) )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : int = torch.randn(2 , 3 )
lowerCamelCase : Optional[int] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_module(model.lineara )
remove_hook_from_module(model.batchnorm )
remove_hook_from_module(model.lineara )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
def a__ ( self: Any )-> List[str]:
lowerCamelCase : int = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : int = 0 if torch.cuda.is_available() else """cpu"""
attach_align_device_hook(__a , execution_device=__a , offload=__a )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(__a )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : Optional[Any] = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
attach_align_device_hook(__a , execution_device=__a , offload=__a , offload_buffers=__a )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : Optional[int] = torch.randn(2 , 3 )
lowerCamelCase : int = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
def a__ ( self: Optional[Any] )-> List[Any]:
lowerCamelCase : List[Any] = ModelForTest()
# Everything is on CPU
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# This will move each submodule on different devices
lowerCamelCase : Any = 0 if torch.cuda.is_available() else """cpu"""
attach_align_device_hook(
__a , execution_device=__a , offload=__a , weights_map=model.state_dict() )
# Parameters have been offloaded, so on the meta device
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
# Buffers are not included in the offload by default, so are on the execution device
lowerCamelCase : List[Any] = torch.device(__a )
self.assertEqual(model.batchnorm.running_mean.device , __a )
lowerCamelCase : Dict = torch.randn(2 , 3 )
lowerCamelCase : int = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
# Now test with buffers included in the offload
attach_align_device_hook(
__a , execution_device=__a , offload=__a , weights_map=model.state_dict() , offload_buffers=__a , )
# Parameters have been offloaded, so on the meta device, buffers included
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""meta""" ) )
self.assertEqual(model.batchnorm.running_mean.device , torch.device("""meta""" ) )
lowerCamelCase : Tuple = torch.randn(2 , 3 )
lowerCamelCase : Any = model(__a )
self.assertEqual(output.device , __a )
# Removing hooks loads back the weights in the model.
remove_hook_from_submodules(__a )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.batchnorm.weight.device , torch.device("""cpu""" ) )
self.assertEqual(model.lineara.weight.device , torch.device("""cpu""" ) )
| 42
| 0
|
"""simple docstring"""
from __future__ import annotations
import os
from collections.abc import Mapping
__lowerCamelCase :Optional[int] = tuple[int, int]
class A__ :
"""simple docstring"""
def __init__( self: Union[str, Any] , __a: List[str] , __a: Union[str, Any] )-> Tuple:
lowerCamelCase : set[int] = vertices
lowerCamelCase : dict[EdgeT, int] = {
(min(_UpperCAmelCase ), max(_UpperCAmelCase )): weight for edge, weight in edges.items()
}
def a__ ( self: Any , __a: Optional[Any] , __a: str )-> Optional[Any]:
self.vertices.add(edge[0] )
self.vertices.add(edge[1] )
lowerCamelCase : Any = weight
def a__ ( self: Optional[Any] )-> Tuple:
lowerCamelCase : Graph = Graph({min(self.vertices )} , {} )
lowerCamelCase : EdgeT
lowerCamelCase : int
lowerCamelCase : EdgeT
lowerCamelCase : int
while len(subgraph.vertices ) < len(self.vertices ):
lowerCamelCase : List[Any] = max(self.edges.values() ) + 1
for edge, weight in self.edges.items():
if (edge[0] in subgraph.vertices) ^ (edge[1] in subgraph.vertices):
if weight < min_weight:
lowerCamelCase : str = edge
lowerCamelCase : int = weight
subgraph.add_edge(_UpperCAmelCase , _UpperCAmelCase )
return subgraph
def snake_case ( UpperCamelCase__ : str = "p107_network.txt" ) -> int:
lowerCamelCase : str = os.path.abspath(os.path.dirname(a_ ) )
lowerCamelCase : str = os.path.join(a_ , a_ )
lowerCamelCase : dict[EdgeT, int] = {}
lowerCamelCase : list[str]
lowerCamelCase : int
lowerCamelCase : int
with open(a_ ) as f:
lowerCamelCase : Any = f.read().strip().split("""\n""" )
lowerCamelCase : Union[str, Any] = [line.split(""",""" ) for line in data]
for edgea in range(1 , len(a_ ) ):
for edgea in range(a_ ):
if adjaceny_matrix[edgea][edgea] != "-":
lowerCamelCase : int = int(adjaceny_matrix[edgea][edgea] )
lowerCamelCase : Graph = Graph(set(range(len(a_ ) ) ) , a_ )
lowerCamelCase : Graph = graph.prims_algorithm()
lowerCamelCase : int = sum(graph.edges.values() )
lowerCamelCase : int = sum(subgraph.edges.values() )
return initial_total - optimal_total
if __name__ == "__main__":
print(f"""{solution() = }""")
| 712
|
"""simple docstring"""
from typing import TYPE_CHECKING
from ...utils import (
OptionalDependencyNotAvailable,
_LazyModule,
is_torch_available,
)
__lowerCamelCase :Optional[Any] = {
'configuration_encodec': [
'ENCODEC_PRETRAINED_CONFIG_ARCHIVE_MAP',
'EncodecConfig',
],
'feature_extraction_encodec': ['EncodecFeatureExtractor'],
}
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
__lowerCamelCase :Union[str, Any] = [
'ENCODEC_PRETRAINED_MODEL_ARCHIVE_LIST',
'EncodecModel',
'EncodecPreTrainedModel',
]
if TYPE_CHECKING:
from .configuration_encodec import (
ENCODEC_PRETRAINED_CONFIG_ARCHIVE_MAP,
EncodecConfig,
)
from .feature_extraction_encodec import EncodecFeatureExtractor
try:
if not is_torch_available():
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
pass
else:
from .modeling_encodec import (
ENCODEC_PRETRAINED_MODEL_ARCHIVE_LIST,
EncodecModel,
EncodecPreTrainedModel,
)
else:
import sys
__lowerCamelCase :Dict = _LazyModule(__name__, globals()['__file__'], _import_structure, module_spec=__spec__)
| 42
| 0
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.