Skip to content

[ray.data.llm] Update release test: enforce pp_backend #52923

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 3 commits into from
Closed

[ray.data.llm] Update release test: enforce pp_backend #52923

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
29096cf
enforce pp_backend and mp_method
lk-chen May 9, 2025
61e2c6e
revert mp_method
lk-chen May 9, 2025
6738d95
enforce ray backend for all
lk-chen May 18, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions release/llm_tests/batch/test_batch_vllm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,7 @@ def test_chat_template_with_vllm():
processor_config = vLLMEngineProcessorConfig(
model_source="unsloth/Llama-3.2-1B-Instruct",
engine_kwargs=dict(
distributed_executor_backend="ray",
max_model_len=16384,
enable_chunked_prefill=True,
max_num_batched_tokens=2048,
Expand Down Expand Up @@ -78,6 +79,7 @@ def test_vllm_llama_parallel(tp_size, pp_size, concurrency, vllm_use_v1):
engine_kwargs=dict(
tensor_parallel_size=tp_size,
pipeline_parallel_size=pp_size,
distributed_executor_backend="ray",
max_model_len=16384,
enable_chunked_prefill=True,
max_num_batched_tokens=2048,
Expand Down Expand Up @@ -130,6 +132,7 @@ def test_vllm_llama_lora():
model_source=model_source,
dynamic_lora_loading_path=lora_path,
engine_kwargs=dict(
distributed_executor_backend="ray",
max_model_len=4096,
enable_chunked_prefill=True,
enable_lora=True,
Expand All @@ -141,6 +144,7 @@ def test_vllm_llama_lora():
concurrency=1,
runtime_env={
"env_vars": {
# TODO(lk-chen): Remove this once ray.data.llm enables LoRA in v1.
"VLLM_USE_V1": "0",
}
},
Expand Down Expand Up @@ -212,6 +216,7 @@ def test_vllm_vision_language_models(
engine_kwargs=dict(
tensor_parallel_size=tp_size,
pipeline_parallel_size=pp_size,
distributed_executor_backend="ray",
max_model_len=4096,
enable_chunked_prefill=True,
),
Expand Down