Upload 860kSequences.pt

E0S=22054
df['sequence']=df['sequence'].progress_apply(lambda x: x+[22054])
PAD=22055
df['sequence'] = df['sequence'].apply(lambda x: x + [22055] * (1195 - len(x)))
SEQ_LEN=1195

def tokenize_x(lst):
return [int(round(x * 1000 + 8037)) for x in lst]

def tokenize_y(lst):
return [int(round(y * 1000 + 16953)) for y in lst]

def detokenize_x(token_list):
"""Converts a list of X integer tokens back to 3-decimal floats."""
return [round((t - 8037) / 1000, 3) for t in token_list]

def detokenize_y(token_list):
"""Converts a list of Y integer tokens back to 3-decimal floats."""
return [round((t - 16953) / 1000, 3) for t in token_list]

ADD SECOND ITEM IN LIST

df['x_path_token'] = [
path[:1] + [micro] + path[1:]
for path, micro in zip(df['x_path_token'], df['x_micro'])
]
df['y_path_token'] = [
path[:1] + [micro] + path[1:]
for path, micro in zip(df['y_path_token'], df['y_micro'])
]


CREATE ALTERNATING

df['sequence'] = [
[val for pair in zip(x, y) for val in pair]
for x, y in zip(df['x_path_token'], df['y_path_token'])
]

860kSequences.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:833ac08201e83c371b3726aa75fc478a6477d7a98bc849d52c0c19420f0c4012
+size 6510600647