Open NGC2237-wei opened 2 years ago
tensor([[0., -0., -0., ..., -0., -0., 0.], [0., -0., -0., ..., -0., -0., 0.], [0., -0., -0., ..., -0., -0., 0.], ..., [-0., -0., 0., ..., -0., -0., 0.], [0., -0., 0., ..., -0., -0., 0.], [0., -0., 0., ..., -0., -0., 0.]]) predictions.shape torch.Size([416, 63]) gold_labels.shape torch.Size([42412])
42412 is my train set size
The prediction shape is not equal the gold_labels.shape
Traceback (most recent call last):
File "/home/anaconda3/envs/scispacy/lib/python3.6/runpy.py", line 193, in _run_module_as_main
"main", mod_spec)
File "/home/anaconda3/envs/scispacy/lib/python3.6/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/home/home/scib/src/allennlp/allennlp/run.py", line 21, in
I'm trying to predient with my own train set using ner model, i create a new train set like provided,but i meet the problem like this.