Skip to content

fix: missing shared tensor in bert::BertLMPredictionHead #2879

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.

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

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

silver-ymz
Copy link

(cls.predictions.decoder.weight, bert.embeddings.word_embeddings.weight) and (cls.predictions.decoder.bias, cls.predictions.bias) are shared tensor in BertForMaskedLM model. They may not exist in model.safetensors file.

Comment on lines +573 to +585
let decoder_weight = if vb.contains_tensor("decoder.weight") {
vb.get((config.vocab_size, config.hidden_size), "decoder.weight")?
} else {
vb.root().get(
(config.vocab_size, config.hidden_size),
"bert.embeddings.word_embeddings.weight",
)?
};
let decoder_bias = if vb.contains_tensor("decoder.bias") {
vb.get(config.vocab_size, "decoder.bias")?
} else {
vb.get(config.vocab_size, "bias")?
};
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi, not sure and better to double check with a maintainer but maybe this syntax is preferred to avoid the contains_tensor call (?)

Suggested change
let decoder_weight = if vb.contains_tensor("decoder.weight") {
vb.get((config.vocab_size, config.hidden_size), "decoder.weight")?
} else {
vb.root().get(
(config.vocab_size, config.hidden_size),
"bert.embeddings.word_embeddings.weight",
)?
};
let decoder_bias = if vb.contains_tensor("decoder.bias") {
vb.get(config.vocab_size, "decoder.bias")?
} else {
vb.get(config.vocab_size, "bias")?
};
let decoder_weight = vb
.get((config.vocab_size, config.hidden_size), "decoder.weight")
.or_else(|_| {
vb.root().get(
(config.vocab_size, config.hidden_size),
"bert.embeddings.word_embeddings.weight",
)
})?;
let decoder_bias = vb
.get(config.vocab_size, "decoder.bias")
.or_else(|_| vb.get(config.vocab_size, "bias"))?;

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants