All versions · Version range: v0.4.2

vLLM v0.4.2 config.py

Type: source_file
Publisher: vllm-project
Relationship: first_party
Version: vLLM v0.4.2
Retrieved: May 26
Local snapshot: corpus/domains/vllm/core-v0/evidence/snapshots/vllm-code-v042-config.py
SHA-256: e0fc689330bd57ba…

Claim	Locator	Verified
In vLLM v0.4.2 `SchedulerConfig`, when `max_num_batched_tokens` is unset and `enable_chunked_prefill=True`, code sets `max_num_batched_tokens` to 512.	vllm/config.py:615-622	1 day ago
In vLLM v0.4.2 `SchedulerConfig`, when `max_num_batched_tokens` is unset and `enable_chunked_prefill=False`, code sets `max_num_batched_tokens` to max(max_model_len, 2048).	vllm/config.py:623-626	1 day ago
In vLLM v0.4.2, `SchedulerConfig` raises an error when `max_num_batched_tokens` is smaller than `max_model_len` and `enable_chunked_prefill=False`.	vllm/config.py:639-647	1 day ago
In vLLM v0.4.2, `max_num_seqs` is exposed through the `SchedulerConfig.__init__` Python parameter `max_num_seqs`.	vllm/config.py:584-609	1 day ago
In vLLM v0.4.2 `SchedulerConfig`, `max_num_seqs` is documented and stored as the maximum number of sequences to process in a single scheduler iteration.	vllm/config.py:587-591, vllm/config.py:629-630	1 day ago
In vLLM v0.4.2, `SchedulerConfig` raises an error when `max_num_batched_tokens` is smaller than `max_num_seqs`.	vllm/config.py:649-653	1 day ago

Name	Description	Locator
max_num_batched_tokens	Maximum number of tokens to process in a scheduler iteration.	vllm/config.py:584-608
max_num_seqs	Maximum number of sequences to process in a scheduler iteration.	vllm/config.py:584-609
max_model_len	Maximum sequence length, including prompt and output tokens, used by vLLM configuration and scheduler validation.	vllm/config.py:584-610