[V1] Make v1 more testable (#9888)
Signed-off-by: Joe Runde <Joseph.Runde@ibm.com>
This commit is contained in:
@@ -10,7 +10,7 @@ from vllm.config import CacheConfig
|
||||
from vllm.model_executor.layers.logits_processor import LogitsProcessor
|
||||
from vllm.model_executor.layers.quantization.base_config import (
|
||||
QuantizationConfig)
|
||||
from vllm.model_executor.layers.sampler import Sampler, SamplerOutput
|
||||
from vllm.model_executor.layers.sampler import SamplerOutput, get_sampler
|
||||
from vllm.model_executor.model_loader.weight_utils import default_weight_loader
|
||||
from vllm.model_executor.models.bart import (BartDecoder, BartEncoder,
|
||||
BartParallelLMHead,
|
||||
@@ -112,7 +112,7 @@ class Florence2LanguageForConditionalGeneration(nn.Module):
|
||||
|
||||
self.logits_processor = LogitsProcessor(self.vocab_size,
|
||||
config.vocab_size)
|
||||
self.sampler = Sampler()
|
||||
self.sampler = get_sampler()
|
||||
|
||||
def forward(
|
||||
self,
|
||||
|
||||
Reference in New Issue
Block a user