[Model] Add VaultGemma via Transformers modeling backend (#49803)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
@@ -55,7 +55,7 @@ from vllm.model_executor.layers.vocab_parallel_embedding import (
|
||||
VocabParallelEmbedding,
|
||||
)
|
||||
from vllm.sequence import IntermediateTensors
|
||||
from vllm.transformers_utils.config import is_interleaved, set_default_rope_theta
|
||||
from vllm.transformers_utils.config import set_default_rope_theta
|
||||
from vllm.v1.attention.backend import AttentionType
|
||||
|
||||
from .interfaces import (
|
||||
@@ -345,7 +345,7 @@ class Qwen2Model(nn.Module, EagleModelMixin):
|
||||
quant_config = vllm_config.quant_config
|
||||
|
||||
# TODO (@robertgshaw2): see if this can be moved out
|
||||
if is_interleaved(vllm_config.model_config.hf_text_config):
|
||||
if len(set(getattr(config, "layer_types", []))) > 1:
|
||||
assert config.max_window_layers == config.num_hidden_layers, (
|
||||
"Sliding window for some but all layers is not supported. "
|
||||
"This model uses sliding window but `max_window_layers` = {} "
|
||||
|
||||
@@ -659,6 +659,7 @@ _TRANSFORMERS_SUPPORTED_MODELS = {
|
||||
"Olmo2ForCausalLM": ("transformers", "TransformersForCausalLM"),
|
||||
"SmolLM3ForCausalLM": ("transformers", "TransformersForCausalLM"),
|
||||
"Starcoder2ForCausalLM": ("transformers", "TransformersForCausalLM"),
|
||||
"VaultGemmaForCausalLM": ("transformers", "TransformersForCausalLM"),
|
||||
# Multimodal models
|
||||
"Emu3ForConditionalGeneration": (
|
||||
"transformers",
|
||||
|
||||
Reference in New Issue
Block a user