[Chore][V0 Deprecation] Move LogProb to a separate file (#24055)

Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
This commit is contained in:
Woosuk Kwon
2025-09-01 12:07:53 -07:00
committed by GitHub
parent a0e0efd6bd
commit 5685370271
14 changed files with 49 additions and 42 deletions
+2 -2
View File
@@ -13,14 +13,14 @@ import torch
import torch.nn as nn
import vllm.envs as envs
from vllm.logprobs import Logprob, PromptLogprobs, SampleLogprobs
from vllm.model_executor.layers.utils import apply_penalties
from vllm.model_executor.sampling_metadata import (SamplingMetadata,
SamplingTensors,
SequenceGroupToSample)
from vllm.sampling_params import SamplingType
from vllm.sequence import (VLLM_INVALID_TOKEN_ID,
CompletionSequenceGroupOutput, Logprob,
PromptLogprobs, SampleLogprobs, SequenceOutput)
CompletionSequenceGroupOutput, SequenceOutput)
if envs.VLLM_USE_FLASHINFER_SAMPLER and find_spec("flashinfer"):
# yapf: disable
+2 -2
View File
@@ -14,12 +14,12 @@ from transformers import PretrainedConfig
from vllm.config import (ModelConfig, ParallelConfig, SchedulerConfig,
SpeculativeConfig)
from vllm.logprobs import Logprob
from vllm.model_executor.layers.logits_processor import LogitsProcessor
from vllm.model_executor.layers.quantization import get_quantization_config
from vllm.model_executor.layers.sampler import Sampler, SamplerOutput
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.sequence import (CompletionSequenceGroupOutput, Logprob,
SequenceOutput)
from vllm.sequence import CompletionSequenceGroupOutput, SequenceOutput
TORCH_DTYPE_TO_NEURON_AMP = {
"auto": "f32",
@@ -27,11 +27,11 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, PretrainedConfig
from vllm.config import (ModelConfig, ParallelConfig, SchedulerConfig,
SpeculativeConfig)
from vllm.logger import init_logger
from vllm.logprobs import Logprob
from vllm.model_executor.layers.logits_processor import LogitsProcessor
from vllm.model_executor.layers.sampler import Sampler, SamplerOutput
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.sequence import (CompletionSequenceGroupOutput, Logprob,
SequenceOutput)
from vllm.sequence import CompletionSequenceGroupOutput, SequenceOutput
# yapf: enable
logger = init_logger(__name__)