TypeError: Expected int32, got list containing Tensors of type '_Message' instead. outputs_forward,- shape,(1, 2, 100) outputs_backward-shape (1, 2, 100) output = tf.concat([outputs_forward, outputs_backward],2,name='output_sequence') tf version -tensorflow-gpu==0.12.0tir...
pGit1 Also just a quick question how exactly does coord conv work in 1d case? Does it just add a timestep integer (not integer datatype but something like 1, 2, 3.0 depending on the timestep) to the input? EDIT: Looks like it returns a normalized output between -1 and positive 1....
tensor = tensor.float() tensor = tensor.long() torch.Tensor与np.ndarray转换 除了CharTensor,其他所有CPU上的张量都支持转换为numpy格式然后再转换回来。 ndarray = tensor.cpu().numpy() tensor = torch.from_numpy(ndarray).float() tensor = torch.from_numpy(ndarray.copy()).float() # If ndarray h...
def __init__(self): # QLoRa fine-tuning: quantization_config = BitsAndBytesConfig( load_in_4bit=True, bnb_4bit_quant_type="nf4", bnb_4bit_use_double_quant=True, bnb_4bit_compute_dtype=torch.float16 ) self.model = EsmModel.from_pretrained(model_name, quantization_config=quantization_...
dropout_ratio=-1, num_classes=2, align_corners=False, loss_decode=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0, class_weight=[0.02, 0.98]), norm_cfg=dict(type='BN', requires_grad=True)) ]) train_cfg = dict( ...
This is what I get "TypeError: Expected string, got ['data/train.records'] of type 'RepeatedScalarFieldContainer' instead." I have repleased all path_to_be_configured parts properly. Could you please advise? 👍 1 👎 2 ngksai93 commented Feb 26, 2018 i have also the same issue ...