Skip to content

Commit

Permalink
[Core] remove temporary local variables in LLMEngine.__init__ (#10577)
Browse files Browse the repository at this point in the history 
Signed-off-by: Russell Bryant <[email protected]>
Signed-off-by: Tyler Michael Smith <[email protected]>
  • Loading branch information
@russellb @tlrmchlsmth
russellb authored and tlrmchlsmth committed Nov 23, 2024
1 parent acaaa38 commit 9c77424
Showing 1 changed file with 66 additions and 77 deletions.
143 changes: 66 additions & 77 deletions vllm/engine/llm_engine.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -231,19 +231,18 @@ def __init__(
use_cached_outputs: bool = False,
) -> None:

# TODO: remove the local variables and use self.* throughout the class.
model_config = self.model_config = vllm_config.model_config
cache_config = self.cache_config = vllm_config.cache_config
lora_config = self.lora_config = vllm_config.lora_config
parallel_config = self.parallel_config = vllm_config.parallel_config
scheduler_config = self.scheduler_config = vllm_config.scheduler_config
device_config = self.device_config = vllm_config.device_config
speculative_config = self.speculative_config = vllm_config.speculative_config # noqa
load_config = self.load_config = vllm_config.load_config
decoding_config = self.decoding_config = vllm_config.decoding_config or DecodingConfig( # noqa
self.model_config = vllm_config.model_config
self.cache_config = vllm_config.cache_config
self.lora_config = vllm_config.lora_config
self.parallel_config = vllm_config.parallel_config
self.scheduler_config = vllm_config.scheduler_config
self.device_config = vllm_config.device_config
self.speculative_config = vllm_config.speculative_config # noqa
self.load_config = vllm_config.load_config
self.decoding_config = vllm_config.decoding_config or DecodingConfig( # noqa
)
prompt_adapter_config = self.prompt_adapter_config = vllm_config.prompt_adapter_config # noqa
observability_config = self.observability_config = vllm_config.observability_config or ObservabilityConfig( # noqa
self.prompt_adapter_config = vllm_config.prompt_adapter_config # noqa
self.observability_config = vllm_config.observability_config or ObservabilityConfig( # noqa
)

logger.info(
Expand All @@ -265,54 +264,43 @@ def __init__(
"mm_processor_kwargs=%s, pooler_config=%r,"
"compilation_config=%r",
VLLM_VERSION,
model_config.model,
speculative_config,
model_config.tokenizer,
model_config.skip_tokenizer_init,
model_config.tokenizer_mode,
model_config.revision,
model_config.override_neuron_config,
model_config.tokenizer_revision,
model_config.trust_remote_code,
model_config.dtype,
model_config.max_model_len,
load_config.download_dir,
load_config.load_format,
parallel_config.tensor_parallel_size,
parallel_config.pipeline_parallel_size,
parallel_config.disable_custom_all_reduce,
model_config.quantization,
model_config.enforce_eager,
cache_config.cache_dtype,
model_config.quantization_param_path,
device_config.device,
decoding_config,
observability_config,
model_config.seed,
model_config.served_model_name,
scheduler_config.num_scheduler_steps,
scheduler_config.chunked_prefill_enabled,
scheduler_config.multi_step_stream_outputs,
cache_config.enable_prefix_caching,
model_config.use_async_output_proc,
self.model_config.model,
self.speculative_config,
self.model_config.tokenizer,
self.model_config.skip_tokenizer_init,
self.model_config.tokenizer_mode,
self.model_config.revision,
self.model_config.override_neuron_config,
self.model_config.tokenizer_revision,
self.model_config.trust_remote_code,
self.model_config.dtype,
self.model_config.max_model_len,
self.load_config.download_dir,
self.load_config.load_format,
self.parallel_config.tensor_parallel_size,
self.parallel_config.pipeline_parallel_size,
self.parallel_config.disable_custom_all_reduce,
self.model_config.quantization,
self.model_config.enforce_eager,
self.cache_config.cache_dtype,
self.model_config.quantization_param_path,
self.device_config.device,
self.decoding_config,
self.observability_config,
self.model_config.seed,
self.model_config.served_model_name,
self.scheduler_config.num_scheduler_steps,
self.scheduler_config.chunked_prefill_enabled,
self.scheduler_config.multi_step_stream_outputs,
self.cache_config.enable_prefix_caching,
self.model_config.use_async_output_proc,
use_cached_outputs,
model_config.mm_processor_kwargs,
model_config.pooler_config,
self.model_config.mm_processor_kwargs,
self.model_config.pooler_config,
vllm_config.compilation_config,
)
# TODO(woosuk): Print more configs in debug mode.
self.model_config = model_config
self.cache_config = cache_config
self.lora_config = lora_config
self.parallel_config = parallel_config
self.scheduler_config = scheduler_config
self.device_config = device_config
self.speculative_config = speculative_config
self.load_config = load_config
self.decoding_config = decoding_config or DecodingConfig()
self.prompt_adapter_config = prompt_adapter_config
self.observability_config = observability_config or ObservabilityConfig(
)

self.log_stats = log_stats
self.use_cached_outputs = use_cached_outputs

Expand All @@ -334,15 +322,15 @@ def get_tokenizer_for_seq(sequence: Sequence) -> AnyTokenizer:

self.seq_counter = Counter()
self.generation_config_fields = _load_generation_config_dict(
model_config)
self.model_config)

self.input_preprocessor = InputPreprocessor(model_config,
self.input_preprocessor = InputPreprocessor(self.model_config,
self.tokenizer,
mm_registry)

self.input_registry = input_registry
self.input_processor = input_registry.create_input_processor(
model_config)
self.model_config)

self.model_executor = executor_class(vllm_config=vllm_config, )

Expand All @@ -354,36 +342,36 @@ def get_tokenizer_for_seq(sequence: Sequence) -> AnyTokenizer:
from vllm.model_executor.model_loader import (
get_architecture_class_name)
usage_message.report_usage(
get_architecture_class_name(model_config),
get_architecture_class_name(self.model_config),
usage_context,
extra_kvs={
# Common configuration
"dtype":
str(model_config.dtype),
str(self.model_config.dtype),
"tensor_parallel_size":
parallel_config.tensor_parallel_size,
self.parallel_config.tensor_parallel_size,
"block_size":
cache_config.block_size,
self.cache_config.block_size,
"gpu_memory_utilization":
cache_config.gpu_memory_utilization,
self.cache_config.gpu_memory_utilization,

# Quantization
"quantization":
model_config.quantization,
self.model_config.quantization,
"kv_cache_dtype":
str(cache_config.cache_dtype),
str(self.cache_config.cache_dtype),

# Feature flags
"enable_lora":
bool(lora_config),
bool(self.lora_config),
"enable_prompt_adapter":
bool(prompt_adapter_config),
bool(self.prompt_adapter_config),
"enable_prefix_caching":
cache_config.enable_prefix_caching,
self.cache_config.enable_prefix_caching,
"enforce_eager":
model_config.enforce_eager,
self.model_config.enforce_eager,
"disable_custom_all_reduce":
parallel_config.disable_custom_all_reduce,
self.parallel_config.disable_custom_all_reduce,
})

if self.tokenizer:
Expand All @@ -402,7 +390,7 @@ def get_tokenizer_for_seq(sequence: Sequence) -> AnyTokenizer:
for _ in range(self.parallel_config.pipeline_parallel_size)
]

if model_config.use_async_output_proc:
if self.model_config.use_async_output_proc:
process_model_outputs = weak_bind(self._process_model_outputs)

self.async_callbacks = [
Expand All @@ -422,11 +410,11 @@ def get_tokenizer_for_seq(sequence: Sequence) -> AnyTokenizer:
# GPU and CPU blocks, which are profiled in the distributed executor.
self.scheduler = [
Scheduler(
scheduler_config, cache_config, lora_config,
parallel_config.pipeline_parallel_size,
self.scheduler_config, self.cache_config, self.lora_config,
self.parallel_config.pipeline_parallel_size,
self.async_callbacks[v_id]
if model_config.use_async_output_proc else None)
for v_id in range(parallel_config.pipeline_parallel_size)
if self.model_config.use_async_output_proc else None)
for v_id in range(self.parallel_config.pipeline_parallel_size)
]

# Metric Logging.
Expand All @@ -448,7 +436,8 @@ def get_tokenizer_for_seq(sequence: Sequence) -> AnyTokenizer:
"prometheus":
PrometheusStatLogger(
local_interval=_LOCAL_LOGGING_INTERVAL_SEC,
labels=dict(model_name=model_config.served_model_name),
labels=dict(
model_name=self.model_config.served_model_name),
max_model_len=self.model_config.max_model_len),
}
self.stat_loggers["prometheus"].info("cache_config",
Expand Down

0 comments on commit 9c77424

Please sign in to comment.