Fix weight tying (#51665)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2026-08-20 10:31:55 +00:00
committed by GitHub
parent 6e85feb1b7
commit f0c14b4f77
70 changed files with 526 additions and 238 deletions
+33
View File
@@ -30,6 +30,7 @@ from vllm.platforms import current_platform
from vllm.tasks import PoolingTask, ScoreType, SupportedTask
from vllm.transformers_utils.config import (
ConfigFormat,
checkpoint_has_lm_head,
get_config,
get_hf_image_processor_config,
get_hf_text_config,
@@ -185,6 +186,10 @@ class ModelConfig:
"""The Hugging Face config of the model."""
hf_text_config: PretrainedConfig = field(init=False)
"""The Hugging Face config of the text model (same as hf_config for text models)."""
word_embeddings_untied_by_checkpoint: bool = field(default=False, init=False)
"""Whether `tie_word_embeddings` was overridden to `False` because the checkpoint
contains an `lm_head` of its own. The two may still turn out to be identical, in
which case they are re-tied once the weights have been loaded."""
hf_config_path: str | None = None
"""Name or path of the Hugging Face config to use. If unspecified, model
name or path will be used."""
@@ -1109,6 +1114,34 @@ class ModelConfig:
def _get_encoder_config(self) -> dict[str, Any] | None:
return get_sentence_transformer_tokenizer_config(self.model, self.revision)
def maybe_untie_word_embeddings(self) -> None:
"""Stop trusting `tie_word_embeddings` when the checkpoint disagrees.
A config may claim the word embeddings are tied while the checkpoint
ships an `lm_head` of its own. Tying regardless would silently discard
that tensor, so build the `lm_head` as if untied and let it load. The
two are compared once loaded, and re-tied if they turn out to match, by
[maybe_retie_word_embeddings][vllm.model_executor.model_loader.weight_tying.maybe_retie_word_embeddings].
Transformers makes the same decision in `PreTrainedModel.tie_weights`,
where it can compare the two tensors directly.
"""
if not getattr(self.hf_config, "tie_word_embeddings", False):
return
if not checkpoint_has_lm_head(self.model, revision=self.revision):
return
logger.debug(
"The config for %s says the word embeddings are tied, but the checkpoint "
"contains an lm_head. Loading it to find out whether they really are tied.",
self.model,
)
# Both levels must be updated: `VllmConfig.with_hf_config` reads the top
# level, and the text config is what the language model itself sees.
self.hf_config.tie_word_embeddings = False
self.hf_config.get_text_config().tie_word_embeddings = False
self.word_embeddings_untied_by_checkpoint = True
def _get_default_runner_type(
self,
architectures: list[str],