Turn @config into a dataclass_transform (#31541)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
@@ -766,8 +766,8 @@ def assert_draft_model_correctness(args: ArgsTest, enforce_eager: bool):
|
||||
"max_model_len": args.max_model_len,
|
||||
"enforce_eager": enforce_eager,
|
||||
"draft_tensor_parallel_size": args.draft_tensor_parallel_size,
|
||||
"max_num_seqs": 100, # limit cudagraph capture runtime
|
||||
},
|
||||
max_num_seqs=100, # limit cudagraph capture runtime
|
||||
max_model_len=args.max_model_len,
|
||||
gpu_memory_utilization=args.gpu_memory_utilization,
|
||||
tensor_parallel_size=args.target_tensor_parallel_size,
|
||||
|
||||
@@ -26,11 +26,8 @@ def test_backend_guidance_rollback_terminated():
|
||||
# guidance backend. In that case we are in a stopped state, but
|
||||
# it should be reverted in case EOS is not accepted by the target
|
||||
# model.
|
||||
vllm_config = VllmConfig(
|
||||
decoding_config=StructuredOutputsConfig(
|
||||
backend="guidance",
|
||||
)
|
||||
)
|
||||
structured_outputs_config = StructuredOutputsConfig(backend="guidance")
|
||||
vllm_config = VllmConfig(structured_outputs_config=structured_outputs_config)
|
||||
tokenizer = AutoTokenizer.from_pretrained(TOKENIZER)
|
||||
|
||||
backend = GuidanceBackend(
|
||||
|
||||
Reference in New Issue
Block a user