vllm

vLLM: a high-throughput and memory-efficient inference engine for LLMs

Subpackages

• vllm.adapter_commons
  • vllm.adapter_commons.layers
  • vllm.adapter_commons.models
  • vllm.adapter_commons.request
  • vllm.adapter_commons.utils
  • vllm.adapter_commons.worker_manager
• vllm.assets
  • vllm.assets.audio
  • vllm.assets.base
  • vllm.assets.image
  • vllm.assets.video
• vllm.attention
  • vllm.attention.backends
    • vllm.attention.backends.mla
    • vllm.attention.backends.abstract
    • vllm.attention.backends.blocksparse_attn
    • vllm.attention.backends.cpu_mla
    • vllm.attention.backends.dual_chunk_flash_attn
    • vllm.attention.backends.flash_attn
    • vllm.attention.backends.flashinfer
    • vllm.attention.backends.flashmla
    • vllm.attention.backends.hpu_attn
    • vllm.attention.backends.ipex_attn
    • vllm.attention.backends.pallas
    • vllm.attention.backends.placeholder_attn
    • vllm.attention.backends.rocm_aiter_mla
    • vllm.attention.backends.rocm_flash_attn
    • vllm.attention.backends.torch_sdpa
    • vllm.attention.backends.triton_mla
    • vllm.attention.backends.utils
    • vllm.attention.backends.xformers
  • vllm.attention.ops
    • vllm.attention.ops.blocksparse_attention
    • vllm.attention.ops.chunked_prefill_paged_decode
    • vllm.attention.ops.flashmla
    • vllm.attention.ops.hpu_paged_attn
    • vllm.attention.ops.ipex_attn
    • vllm.attention.ops.merge_attn_states
    • vllm.attention.ops.nki_flash_attn
    • vllm.attention.ops.paged_attn
    • vllm.attention.ops.prefix_prefill
    • vllm.attention.ops.rocm_aiter_mla
    • vllm.attention.ops.rocm_aiter_paged_attn
    • vllm.attention.ops.triton_decode_attention
    • vllm.attention.ops.triton_flash_attention
    • Fused Attention
    • vllm.attention.ops.triton_merge_attn_states
    • vllm.attention.ops.triton_unified_attention
  • vllm.attention.layer
  • vllm.attention.selector
• vllm.benchmarks
  • vllm.benchmarks.datasets
  • vllm.benchmarks.endpoint_request_func
  • vllm.benchmarks.latency
  • vllm.benchmarks.serve
  • vllm.benchmarks.throughput
  • vllm.benchmarks.utils
• vllm.compilation
  • vllm.compilation.activation_quant_fusion
  • vllm.compilation.backends
  • vllm.compilation.compiler_interface
  • vllm.compilation.counter
  • vllm.compilation.decorators
  • vllm.compilation.fix_functionalization
  • vllm.compilation.fusion
  • vllm.compilation.fx_utils
  • vllm.compilation.inductor_pass
  • vllm.compilation.monitor
  • vllm.compilation.multi_output_match
  • vllm.compilation.noop_elimination
  • vllm.compilation.pass_manager
  • vllm.compilation.sequence_parallelism
  • vllm.compilation.torch25_custom_graph_pass
  • vllm.compilation.vllm_inductor_pass
  • vllm.compilation.wrapper
• vllm.core
  • vllm.core.block
    • vllm.core.block.block_table
    • vllm.core.block.common
    • vllm.core.block.cpu_gpu_block_allocator
    • vllm.core.block.interfaces
    • vllm.core.block.naive_block
    • vllm.core.block.prefix_caching_block
    • vllm.core.block.utils
  • vllm.core.block_manager
  • vllm.core.evictor
  • vllm.core.interfaces
  • vllm.core.placeholder_block_space_manager
  • vllm.core.scheduler
• vllm.device_allocator
  • vllm.device_allocator.cumem
• vllm.distributed
  • vllm.distributed.device_communicators
    • vllm.distributed.device_communicators.all2all
    • vllm.distributed.device_communicators.base_device_communicator
    • vllm.distributed.device_communicators.cpu_communicator
    • vllm.distributed.device_communicators.cuda_communicator
    • vllm.distributed.device_communicators.cuda_wrapper
    • vllm.distributed.device_communicators.custom_all_reduce
    • vllm.distributed.device_communicators.custom_all_reduce_utils
    • vllm.distributed.device_communicators.hpu_communicator
    • vllm.distributed.device_communicators.neuron_communicator
    • vllm.distributed.device_communicators.pynccl
    • vllm.distributed.device_communicators.pynccl_wrapper
    • vllm.distributed.device_communicators.shm_broadcast
    • vllm.distributed.device_communicators.tpu_communicator
    • vllm.distributed.device_communicators.xpu_communicator
  • vllm.distributed.kv_transfer
    • vllm.distributed.kv_transfer.kv_connector
    • vllm.distributed.kv_transfer.kv_lookup_buffer
    • vllm.distributed.kv_transfer.kv_pipe
    • vllm.distributed.kv_transfer.kv_connector_agent
    • vllm.distributed.kv_transfer.kv_transfer_state
  • vllm.distributed.communication_op
  • vllm.distributed.kv_events
  • vllm.distributed.parallel_state
  • vllm.distributed.utils
• vllm.engine
  • vllm.engine.multiprocessing
    • vllm.engine.multiprocessing.client
    • vllm.engine.multiprocessing.engine
  • vllm.engine.output_processor
    • vllm.engine.output_processor.interfaces
    • vllm.engine.output_processor.multi_step
    • vllm.engine.output_processor.single_step
    • vllm.engine.output_processor.stop_checker
    • vllm.engine.output_processor.util
  • vllm.engine.arg_utils
  • vllm.engine.async_llm_engine
  • vllm.engine.async_timeout
  • vllm.engine.llm_engine
  • vllm.engine.metrics
  • vllm.engine.metrics_types
  • vllm.engine.protocol
• vllm.entrypoints
  • vllm.entrypoints.cli
    • vllm.entrypoints.cli.benchmark
    • vllm.entrypoints.cli.collect_env
    • vllm.entrypoints.cli.main
    • vllm.entrypoints.cli.openai
    • vllm.entrypoints.cli.serve
    • vllm.entrypoints.cli.types
  • vllm.entrypoints.openai
    • vllm.entrypoints.openai.tool_parsers
    • vllm.entrypoints.openai.api_server
    • vllm.entrypoints.openai.cli_args
    • vllm.entrypoints.openai.logits_processors
    • vllm.entrypoints.openai.protocol
    • vllm.entrypoints.openai.run_batch
    • vllm.entrypoints.openai.serving_chat
    • vllm.entrypoints.openai.serving_classification
    • vllm.entrypoints.openai.serving_completion
    • vllm.entrypoints.openai.serving_embedding
    • vllm.entrypoints.openai.serving_engine
    • vllm.entrypoints.openai.serving_models
    • vllm.entrypoints.openai.serving_pooling
    • vllm.entrypoints.openai.serving_score
    • vllm.entrypoints.openai.serving_tokenization
    • vllm.entrypoints.openai.serving_transcription
  • vllm.entrypoints.api_server
  • vllm.entrypoints.chat_utils
  • vllm.entrypoints.launcher
  • vllm.entrypoints.llm
  • vllm.entrypoints.logger
  • vllm.entrypoints.score_utils
  • vllm.entrypoints.ssl
  • vllm.entrypoints.utils
• vllm.executor
  • vllm.executor.executor_base
  • vllm.executor.mp_distributed_executor
  • vllm.executor.msgspec_utils
  • vllm.executor.multiproc_worker_utils
  • vllm.executor.ray_distributed_executor
  • vllm.executor.ray_utils
  • vllm.executor.uniproc_executor
• vllm.inputs
  • vllm.inputs.data
  • vllm.inputs.parse
  • vllm.inputs.preprocess
  • vllm.inputs.registry
• vllm.logging_utils
  • vllm.logging_utils.dump_input
  • vllm.logging_utils.formatter
• vllm.lora
  • vllm.lora.ops
    • vllm.lora.ops.torch_ops
    • vllm.lora.ops.triton_ops
    • vllm.lora.ops.xla_ops
  • vllm.lora.punica_wrapper
    • vllm.lora.punica_wrapper.punica_base
    • vllm.lora.punica_wrapper.punica_cpu
    • vllm.lora.punica_wrapper.punica_gpu
    • vllm.lora.punica_wrapper.punica_hpu
    • vllm.lora.punica_wrapper.punica_selector
    • vllm.lora.punica_wrapper.punica_tpu
    • vllm.lora.punica_wrapper.utils
  • vllm.lora.fully_sharded_layers
  • vllm.lora.layers
  • vllm.lora.lora
  • vllm.lora.models
  • vllm.lora.peft_helper
  • vllm.lora.request
  • vllm.lora.resolver
  • vllm.lora.utils
  • vllm.lora.worker_manager
• vllm.model_executor
  • vllm.model_executor.guided_decoding
    • vllm.model_executor.guided_decoding.guidance_decoding
    • vllm.model_executor.guided_decoding.guidance_logits_processors
    • vllm.model_executor.guided_decoding.guided_fields
    • vllm.model_executor.guided_decoding.lm_format_enforcer_decoding
    • vllm.model_executor.guided_decoding.outlines_decoding
    • vllm.model_executor.guided_decoding.outlines_logits_processors
    • vllm.model_executor.guided_decoding.utils
    • vllm.model_executor.guided_decoding.xgrammar_decoding
  • vllm.model_executor.layers
    • vllm.model_executor.layers.fused_moe
    • vllm.model_executor.layers.mamba
    • vllm.model_executor.layers.quantization
    • vllm.model_executor.layers.activation
    • vllm.model_executor.layers.layernorm
    • vllm.model_executor.layers.lightning_attn
    • vllm.model_executor.layers.linear
    • vllm.model_executor.layers.logits_processor
    • vllm.model_executor.layers.pooler
    • vllm.model_executor.layers.rejection_sampler
    • vllm.model_executor.layers.resampler
    • vllm.model_executor.layers.rotary_embedding
    • vllm.model_executor.layers.sampler
    • vllm.model_executor.layers.spec_decode_base_sampler
    • vllm.model_executor.layers.typical_acceptance_sampler
    • vllm.model_executor.layers.utils
    • vllm.model_executor.layers.vocab_parallel_embedding
  • vllm.model_executor.model_loader
    • vllm.model_executor.model_loader.base_loader
    • vllm.model_executor.model_loader.bitsandbytes_loader
    • vllm.model_executor.model_loader.default_loader
    • vllm.model_executor.model_loader.dummy_loader
    • vllm.model_executor.model_loader.gguf_loader
    • vllm.model_executor.model_loader.neuron
    • vllm.model_executor.model_loader.neuronx_distributed
    • vllm.model_executor.model_loader.runai_streamer_loader
    • vllm.model_executor.model_loader.sharded_state_loader
    • vllm.model_executor.model_loader.tensorizer
    • vllm.model_executor.model_loader.tensorizer_loader
    • vllm.model_executor.model_loader.utils
    • vllm.model_executor.model_loader.weight_utils
  • vllm.model_executor.models
    • vllm.model_executor.models.adapters
    • vllm.model_executor.models.aimv2
    • vllm.model_executor.models.arctic
    • vllm.model_executor.models.aria
    • vllm.model_executor.models.aya_vision
    • vllm.model_executor.models.baichuan
    • vllm.model_executor.models.bamba
    • vllm.model_executor.models.bart
    • vllm.model_executor.models.bert
    • vllm.model_executor.models.bert_with_rope
    • vllm.model_executor.models.blip
    • vllm.model_executor.models.blip2
    • vllm.model_executor.models.bloom
    • vllm.model_executor.models.chameleon
    • vllm.model_executor.models.chatglm
    • vllm.model_executor.models.clip
    • vllm.model_executor.models.commandr
    • vllm.model_executor.models.constant_size_cache
    • vllm.model_executor.models.dbrx
    • vllm.model_executor.models.deepseek
    • vllm.model_executor.models.deepseek_mtp
    • vllm.model_executor.models.deepseek_v2
    • vllm.model_executor.models.deepseek_vl2
    • vllm.model_executor.models.eagle
    • vllm.model_executor.models.exaone
    • vllm.model_executor.models.fairseq2_llama
    • vllm.model_executor.models.falcon
    • vllm.model_executor.models.florence2
    • vllm.model_executor.models.fuyu
    • vllm.model_executor.models.gemma
    • vllm.model_executor.models.gemma2
    • vllm.model_executor.models.gemma3
    • vllm.model_executor.models.gemma3_mm
    • vllm.model_executor.models.glm
    • vllm.model_executor.models.glm4
    • vllm.model_executor.models.glm4v
    • vllm.model_executor.models.gpt2
    • vllm.model_executor.models.gpt_bigcode
    • vllm.model_executor.models.gpt_j
    • vllm.model_executor.models.gpt_neox
    • vllm.model_executor.models.granite
    • vllm.model_executor.models.granite_speech
    • vllm.model_executor.models.granitemoe
    • vllm.model_executor.models.granitemoehybrid
    • vllm.model_executor.models.granitemoeshared
    • vllm.model_executor.models.gritlm
    • vllm.model_executor.models.grok1
    • vllm.model_executor.models.h2ovl
    • vllm.model_executor.models.idefics2_vision_model
    • vllm.model_executor.models.idefics3
    • vllm.model_executor.models.interfaces
    • vllm.model_executor.models.interfaces_base
    • vllm.model_executor.models.intern_vit
    • vllm.model_executor.models.internlm2
    • vllm.model_executor.models.internlm2_ve
    • vllm.model_executor.models.internvl
    • vllm.model_executor.models.jais
    • vllm.model_executor.models.jamba
    • vllm.model_executor.models.kimi_vl
    • vllm.model_executor.models.llama
    • vllm.model_executor.models.llama4
    • vllm.model_executor.models.llama_eagle
    • vllm.model_executor.models.llama_eagle3
    • vllm.model_executor.models.llava
    • vllm.model_executor.models.llava_next
    • vllm.model_executor.models.llava_next_video
    • vllm.model_executor.models.llava_onevision
    • vllm.model_executor.models.mamba
    • vllm.model_executor.models.mamba2
    • vllm.model_executor.models.mamba_cache
    • vllm.model_executor.models.medusa
    • vllm.model_executor.models.mimo
    • vllm.model_executor.models.mimo_mtp
    • vllm.model_executor.models.minicpm
    • vllm.model_executor.models.minicpm3
    • vllm.model_executor.models.minicpmo
    • vllm.model_executor.models.minicpmv
    • vllm.model_executor.models.minimax_cache
    • vllm.model_executor.models.minimax_text_01
    • vllm.model_executor.models.minimax_vl_01
    • vllm.model_executor.models.mistral3
    • vllm.model_executor.models.mixtral
    • vllm.model_executor.models.mixtral_quant
    • vllm.model_executor.models.mllama
    • vllm.model_executor.models.mllama4
    • vllm.model_executor.models.mlp_speculator
    • vllm.model_executor.models.modernbert
    • vllm.model_executor.models.module_mapping
    • vllm.model_executor.models.molmo
    • vllm.model_executor.models.moonvit
    • vllm.model_executor.models.mpt
    • vllm.model_executor.models.nemotron
    • vllm.model_executor.models.nemotron_nas
    • vllm.model_executor.models.nvlm_d
    • vllm.model_executor.models.olmo
    • vllm.model_executor.models.olmo2
    • vllm.model_executor.models.olmoe
    • vllm.model_executor.models.opt
    • vllm.model_executor.models.orion
    • vllm.model_executor.models.ovis
    • vllm.model_executor.models.paligemma
    • vllm.model_executor.models.persimmon
    • vllm.model_executor.models.phi
    • vllm.model_executor.models.phi3
    • vllm.model_executor.models.phi3_small
    • vllm.model_executor.models.phi3v
    • vllm.model_executor.models.phi4mm
    • vllm.model_executor.models.phi4mm_audio
    • vllm.model_executor.models.phi4mm_utils
    • vllm.model_executor.models.phimoe
    • vllm.model_executor.models.pixtral
    • vllm.model_executor.models.plamo2
    • vllm.model_executor.models.prithvi_geospatial_mae
    • vllm.model_executor.models.qwen
    • vllm.model_executor.models.qwen2
    • vllm.model_executor.models.qwen2_5_omni_thinker
    • vllm.model_executor.models.qwen2_5_vl
    • vllm.model_executor.models.qwen2_audio
    • vllm.model_executor.models.qwen2_moe
    • vllm.model_executor.models.qwen2_rm
    • vllm.model_executor.models.qwen2_vl
    • vllm.model_executor.models.qwen3
    • vllm.model_executor.models.qwen3_moe
    • vllm.model_executor.models.qwen_vl
    • vllm.model_executor.models.registry
    • vllm.model_executor.models.roberta
    • vllm.model_executor.models.siglip
    • vllm.model_executor.models.skyworkr1v
    • vllm.model_executor.models.smolvlm
    • vllm.model_executor.models.solar
    • vllm.model_executor.models.stablelm
    • vllm.model_executor.models.starcoder2
    • vllm.model_executor.models.telechat2
    • vllm.model_executor.models.teleflm
    • vllm.model_executor.models.transformers
    • vllm.model_executor.models.ultravox
    • vllm.model_executor.models.utils
    • vllm.model_executor.models.vision
    • vllm.model_executor.models.whisper
    • vllm.model_executor.models.zamba2
  • vllm.model_executor.custom_op
  • vllm.model_executor.parameter
  • vllm.model_executor.pooling_metadata
  • vllm.model_executor.sampling_metadata
  • vllm.model_executor.utils
• vllm.multimodal
  • vllm.multimodal.audio
  • vllm.multimodal.base
  • vllm.multimodal.hasher
  • vllm.multimodal.image
  • vllm.multimodal.inputs
  • vllm.multimodal.parse
  • vllm.multimodal.processing
  • vllm.multimodal.profiling
  • vllm.multimodal.registry
  • vllm.multimodal.utils
  • vllm.multimodal.video
• vllm.platforms
  • vllm.platforms.cpu
  • vllm.platforms.cuda
  • vllm.platforms.hpu
  • vllm.platforms.interface
  • vllm.platforms.neuron
  • vllm.platforms.rocm
  • vllm.platforms.tpu
  • vllm.platforms.xpu
• vllm.plugins
  • vllm.plugins.lora_resolvers
    • vllm.plugins.lora_resolvers.filesystem_resolver
• vllm.profiler
  • vllm.profiler.layerwise_profile
  • vllm.profiler.utils
• vllm.prompt_adapter
  • vllm.prompt_adapter.layers
  • vllm.prompt_adapter.models
  • vllm.prompt_adapter.request
  • vllm.prompt_adapter.utils
  • vllm.prompt_adapter.worker_manager
• vllm.reasoning
  • vllm.reasoning.abs_reasoning_parsers
  • vllm.reasoning.deepseek_r1_reasoning_parser
  • vllm.reasoning.granite_reasoning_parser
  • vllm.reasoning.qwen3_reasoning_parser
• vllm.spec_decode
  • vllm.spec_decode.batch_expansion
  • vllm.spec_decode.draft_model_runner
  • vllm.spec_decode.interfaces
  • vllm.spec_decode.medusa_worker
  • vllm.spec_decode.metrics
  • vllm.spec_decode.mlp_speculator_worker
  • vllm.spec_decode.mqa_scorer
  • vllm.spec_decode.multi_step_worker
  • vllm.spec_decode.ngram_worker
  • vllm.spec_decode.proposer_worker_base
  • vllm.spec_decode.smaller_tp_proposer_worker
  • vllm.spec_decode.spec_decode_worker
  • vllm.spec_decode.target_model_runner
  • vllm.spec_decode.top1_proposer
  • vllm.spec_decode.util
• vllm.transformers_utils
  • vllm.transformers_utils.chat_templates
    • vllm.transformers_utils.chat_templates.registry
  • vllm.transformers_utils.configs
    • vllm.transformers_utils.configs.arctic
    • vllm.transformers_utils.configs.chatglm
    • vllm.transformers_utils.configs.cohere2
    • vllm.transformers_utils.configs.dbrx
    • vllm.transformers_utils.configs.deepseek_vl2
    • vllm.transformers_utils.configs.eagle
    • vllm.transformers_utils.configs.exaone
    • vllm.transformers_utils.configs.falcon
    • vllm.transformers_utils.configs.h2ovl
    • vllm.transformers_utils.configs.internvl
    • vllm.transformers_utils.configs.jais
    • vllm.transformers_utils.configs.kimi_vl
    • vllm.transformers_utils.configs.medusa
    • vllm.transformers_utils.configs.minimax_text_01
    • vllm.transformers_utils.configs.minimax_vl_01
    • vllm.transformers_utils.configs.mllama
    • vllm.transformers_utils.configs.mlp_speculator
    • vllm.transformers_utils.configs.moonvit
    • vllm.transformers_utils.configs.mpt
    • vllm.transformers_utils.configs.nemotron
    • vllm.transformers_utils.configs.nvlm_d
    • vllm.transformers_utils.configs.ovis
    • vllm.transformers_utils.configs.skyworkr1v
    • vllm.transformers_utils.configs.solar
    • vllm.transformers_utils.configs.telechat2
    • vllm.transformers_utils.configs.ultravox
  • vllm.transformers_utils.processors
    • vllm.transformers_utils.processors.deepseek_vl2
    • vllm.transformers_utils.processors.ovis
  • vllm.transformers_utils.tokenizers
    • vllm.transformers_utils.tokenizers.mistral
  • vllm.transformers_utils.config
  • vllm.transformers_utils.detokenizer
  • vllm.transformers_utils.detokenizer_utils
  • vllm.transformers_utils.processor
  • vllm.transformers_utils.s3_utils
  • vllm.transformers_utils.tokenizer
  • vllm.transformers_utils.tokenizer_base
  • vllm.transformers_utils.tokenizer_group
  • vllm.transformers_utils.utils
• vllm.triton_utils
  • vllm.triton_utils.importing
• vllm.usage
  • vllm.usage.usage_lib
• vllm.v1
  • vllm.v1.attention
    • vllm.v1.attention.backends
  • vllm.v1.core
    • vllm.v1.core.sched
    • vllm.v1.core.block_pool
    • vllm.v1.core.encoder_cache_manager
    • vllm.v1.core.kv_cache_manager
    • vllm.v1.core.kv_cache_utils
    • vllm.v1.core.single_type_kv_cache_manager
  • vllm.v1.engine
    • vllm.v1.engine.async_llm
    • vllm.v1.engine.core
    • vllm.v1.engine.core_client
    • vllm.v1.engine.detokenizer
    • vllm.v1.engine.exceptions
    • vllm.v1.engine.llm_engine
    • vllm.v1.engine.logprobs
    • vllm.v1.engine.mm_input_cache
    • vllm.v1.engine.output_processor
    • vllm.v1.engine.parallel_sampling
    • vllm.v1.engine.processor
  • vllm.v1.executor
    • vllm.v1.executor.abstract
    • vllm.v1.executor.multiproc_executor
    • vllm.v1.executor.ray_distributed_executor
  • vllm.v1.metrics
    • vllm.v1.metrics.loggers
    • vllm.v1.metrics.ray_wrappers
    • vllm.v1.metrics.stats
  • vllm.v1.sample
    • vllm.v1.sample.ops
    • vllm.v1.sample.tpu
    • vllm.v1.sample.metadata
    • vllm.v1.sample.rejection_sampler
    • vllm.v1.sample.sampler
  • vllm.v1.spec_decode
    • vllm.v1.spec_decode.eagle
    • vllm.v1.spec_decode.medusa
    • vllm.v1.spec_decode.metadata
    • vllm.v1.spec_decode.metrics
    • vllm.v1.spec_decode.ngram_proposer
    • vllm.v1.spec_decode.utils
  • vllm.v1.structured_output
    • vllm.v1.structured_output.backend_guidance
    • vllm.v1.structured_output.backend_types
    • vllm.v1.structured_output.backend_xgrammar
    • vllm.v1.structured_output.request
    • vllm.v1.structured_output.utils
  • vllm.v1.worker
    • vllm.v1.worker.block_table
    • vllm.v1.worker.gpu_input_batch
    • vllm.v1.worker.gpu_model_runner
    • vllm.v1.worker.gpu_worker
    • vllm.v1.worker.lora_model_runner_mixin
    • vllm.v1.worker.tpu_model_runner
    • vllm.v1.worker.tpu_worker
    • vllm.v1.worker.utils
    • vllm.v1.worker.worker_base
  • vllm.v1.kv_cache_interface
  • vllm.v1.outputs
  • vllm.v1.request
  • vllm.v1.serial_utils
  • vllm.v1.utils
• vllm.worker
  • vllm.worker.cache_engine
  • vllm.worker.cpu_enc_dec_model_runner
  • vllm.worker.cpu_model_runner
  • vllm.worker.cpu_pooling_model_runner
  • vllm.worker.cpu_worker
  • vllm.worker.enc_dec_model_runner
  • vllm.worker.hpu_model_runner
  • vllm.worker.hpu_worker
  • vllm.worker.model_runner
  • vllm.worker.model_runner_base
  • vllm.worker.multi_step_hpu_worker
  • vllm.worker.multi_step_model_runner
  • vllm.worker.multi_step_neuron_model_runner
  • vllm.worker.multi_step_neuronx_distributed_model_runner
  • vllm.worker.multi_step_tpu_worker
  • vllm.worker.multi_step_worker
  • vllm.worker.neuron_model_runner
  • vllm.worker.neuron_worker
  • vllm.worker.neuronx_distributed_model_runner
  • vllm.worker.pooling_model_runner
  • vllm.worker.tpu_model_runner
  • vllm.worker.tpu_worker
  • vllm.worker.utils
  • vllm.worker.worker
  • vllm.worker.worker_base
  • vllm.worker.xpu_model_runner
  • vllm.worker.xpu_worker

Submodules

• vllm.beam_search
• vllm.collect_env
• vllm.config
• vllm.connections
• vllm.env_override
• vllm.envs
• vllm.forward_context
• vllm.jsontree
• vllm.logger
• vllm.logits_process
• vllm.outputs
• vllm.pooling_params
• vllm.sampling_params
• vllm.scalar_type
• vllm.scripts
• vllm.sequence
• vllm.test_utils
• vllm.tracing
• vllm.utils
• vllm.version