def convert_tf_checkpoint_to_pytorch(tf_checkpoint_path, pytorch_dump_path):
# Initialize PyTorch model
config = CanineConfig()
model = CanineModel(config)
model.eval()
print(f"Building PyTorch model from configuration: {config}")
# Load weights from tf checkpoint
load_tf_weights_in_canine(model, config, tf_checkpoint_path)
# Save pytorch-model (weights and configuration)
print(f"Save PyTorch model to {pytorch_dump_path}")
model.save_pretrained(pytorch_dump_path)
# Save tokenizer files
tokenizer = CanineTokenizer()
print(f"Save tokenizer files to {pytorch_dump_path}")
tokenizer.save_pretrained(pytorch_dump_path)
def canine_tokenizer(self):
return CanineTokenizer.from_pretrained("google/canine-s")
def setUp(self):
super().setUp()
tokenizer = CanineTokenizer()
tokenizer.save_pretrained(self.tmpdirname)
def canine_tokenizer(self):
# TODO replace nielsr by google
return CanineTokenizer.from_pretrained("nielsr/canine-s")
