Skip to content

Fix np.load allow_pickle=False to work correctly with .npz format #1329

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
wannaphong merged 2 commits into from
Mar 13, 2026
Merged

Fix np.load allow_pickle=False to work correctly with .npz format #1329

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1b72565
Initial plan
Copilot Mar 13, 2026
c707141
Fix np.load allow_pickle=False to work with .npz NpzFile format
Copilot Mar 13, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 3 additions & 1 deletion pythainlp/spell/words_spelling_correction.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -84,7 +84,9 @@ def _load_embeddings(self) -> tuple[list[str], NDArray[np.float32]]:
"""Loads embeddings matrix and vocabulary list."""
import numpy as np

input_matrix = np.load(os.path.join(self.model_dir, "embeddings.npy"))
input_matrix = np.load(
os.path.join(self.model_dir, "embeddings.npy"), allow_pickle=False
)
words = []
vocab_path = os.path.join(self.model_dir, "vocabulary.txt")
with open(vocab_path, encoding="utf-8") as f:
Expand Down
50 changes: 15 additions & 35 deletions pythainlp/transliterate/w2p.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -60,7 +60,6 @@ class Thai_W2P:
p2idx: dict[str, int]
idx2p: dict[int, str]
checkpoint: Optional[str]
variables: "NDArray"
enc_emb: "NDArray"
enc_w_ih: "NDArray"
enc_w_hh: "NDArray"
Expand Down Expand Up @@ -101,52 +100,33 @@ def _load_variables(self) -> None:

if self.checkpoint is None:
raise RuntimeError("checkpoint path is not set")
self.variables: "NDArray" = np.load(self.checkpoint, allow_pickle=False)
# (29, 64). (len(graphemes), emb)
self.enc_emb: "NDArray" = self.variables.item().get(
"encoder.emb.weight"
variables: "np.lib.npyio.NpzFile" = np.load(
self.checkpoint, allow_pickle=False
)
# (29, 64). (len(graphemes), emb)
self.enc_emb: "NDArray" = variables["encoder.emb.weight"]
# (3*128, 64)
self.enc_w_ih: "NDArray" = self.variables.item().get(
"encoder.rnn.weight_ih_l0"
)
self.enc_w_ih: "NDArray" = variables["encoder.rnn.weight_ih_l0"]
# (3*128, 128)
self.enc_w_hh: "NDArray" = self.variables.item().get(
"encoder.rnn.weight_hh_l0"
)
self.enc_w_hh: "NDArray" = variables["encoder.rnn.weight_hh_l0"]
# (3*128,)
self.enc_b_ih: "NDArray" = self.variables.item().get(
"encoder.rnn.bias_ih_l0"
)
self.enc_b_ih: "NDArray" = variables["encoder.rnn.bias_ih_l0"]
# (3*128,)
self.enc_b_hh: "NDArray" = self.variables.item().get(
"encoder.rnn.bias_hh_l0"
)

self.enc_b_hh: "NDArray" = variables["encoder.rnn.bias_hh_l0"]
# (74, 64). (len(phonemes), emb)
self.dec_emb: "NDArray" = self.variables.item().get(
"decoder.emb.weight"
)
self.dec_emb: "NDArray" = variables["decoder.emb.weight"]
# (3*128, 64)
self.dec_w_ih: "NDArray" = self.variables.item().get(
"decoder.rnn.weight_ih_l0"
)
self.dec_w_ih: "NDArray" = variables["decoder.rnn.weight_ih_l0"]
# (3*128, 128)
self.dec_w_hh: "NDArray" = self.variables.item().get(
"decoder.rnn.weight_hh_l0"
)
self.dec_w_hh: "NDArray" = variables["decoder.rnn.weight_hh_l0"]
# (3*128,)
self.dec_b_ih: "NDArray" = self.variables.item().get(
"decoder.rnn.bias_ih_l0"
)
self.dec_b_ih: "NDArray" = variables["decoder.rnn.bias_ih_l0"]
# (3*128,)
self.dec_b_hh: "NDArray" = self.variables.item().get(
"decoder.rnn.bias_hh_l0"
)
self.dec_b_hh: "NDArray" = variables["decoder.rnn.bias_hh_l0"]
# (74, 128)
self.fc_w: "NDArray" = self.variables.item().get("decoder.fc.weight")
self.fc_w: "NDArray" = variables["decoder.fc.weight"]
# (74,)
self.fc_b: "NDArray" = self.variables.item().get("decoder.fc.bias")
self.fc_b: "NDArray" = variables["decoder.fc.bias"]

def _sigmoid(self, x: "np.ndarray") -> "np.ndarray":
import numpy as np
Expand Down