Skip to content
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

Jump to bottom

fix: multilingual midel convert to tflite get wrong token #32079

Merged
merged 2 commits into from
Aug 27, 2024
Merged

fix: multilingual midel convert to tflite get wrong token #32079

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
b4073dd
fix: multilingual midel convert to tflite get wrong token
Jul 19, 2024
dac3219
fix: modify test_force_tokens_logits_processor the checking value as …
Aug 25, 2024
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
2 changes: 1 addition & 1 deletion src/transformers/generation/tf_logits_process.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -581,7 +581,7 @@ def _force_token(generation_idx):
batch_size = scores.shape[0]
current_token = self.force_token_array[generation_idx]

new_scores = tf.ones_like(scores, dtype=scores.dtype) * -float("inf")
new_scores = tf.zeros_like(scores, dtype=scores.dtype) + tf.constant([scores.dtype.min])
indices = tf.stack((tf.range(batch_size), tf.tile([current_token], [batch_size])), axis=1)
updates = tf.zeros((batch_size,), dtype=scores.dtype)
new_scores = tf.tensor_scatter_nd_update(new_scores, indices, updates)
Expand Down
7 changes: 6 additions & 1 deletion tests/generation/test_tf_logits_process.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -406,7 +406,12 @@ def test_force_tokens_logits_processor(self, use_xla):

non_forced_inds = [i for i in range(vocab_size) if i != force_token_map[cur_len]]
self.assertTrue(
tf.math.reduce_all(tf.math.is_inf(tf.gather(scores, [non_forced_inds], axis=1))),
tf.math.reduce_all(
tf.experimental.numpy.isclose(
tf.gather(scores, [non_forced_inds], axis=1),
tf.constant(scores.dtype.min),
)
)
)

# check that if the cur_len is not contained in the force_token_map, the logits are not modified
Expand Down