3D-ResNets-PyTorch
3D-ResNets-PyTorch copied to clipboard
error in generating ucf101.json
When I generated ucf101.json, the following problems occurred “pandas.errors.ParserError : Error tokenizing data. C error: Expected 9 fields in line 3, saw 13” then I changed here: def load_ labels(label_ csv_ path) data = pd.read_ csv(label_ csv_ path, delimiter=" ", header=None, error_ bad_ Lines = false)
withou "error_ bad_ Lines = false", the json cannot be generated, delimiter="\t" doesn't work
Try using the splits from here: https://github.com/feichtenhofer/twostreamfusion/tree/master/ucf101_splits I used these and then ran the code as mentioned in the readme and it worked for me @ltjian-lttj