biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Bugfix] Fix RequestOutput miss lora_request (#30636)

Browse Source 
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
This commit is contained in:
Jee Jee Li
2025-12-16 17:36:35 +08:00
committed by GitHub
parent 0d0c929f23
commit 0e391e7570
3 changed files with 19 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/lora/test_gptoss_tp.py
View File

@@ -76,6 +76,8 @@ def test_gpt_oss_lora(gptoss20b_lora_files):
enable_lora=True,
max_loras=4,
max_lora_rank=8,
max_num_seqs=2,
max_num_batched_tokens=2048,
compilation_config=vllm.config.CompilationConfig( # Avoid OOM
cudagraph_specialize_lora=False,
),
@@ -94,8 +96,10 @@ def test_gpt_oss_lora_tp2(gptoss20b_lora_files, fully_sharded_loras):
enable_lora=True,
max_loras=2,
max_lora_rank=8,
max_num_seqs=16,
max_num_seqs=2,
max_num_batched_tokens=2048,
tensor_parallel_size=2,
gpu_memory_utilization=0.8,
fully_sharded_loras=fully_sharded_loras,
compilation_config=vllm.config.CompilationConfig( # Avoid OOM
cudagraph_specialize_lora=False,

9
tests/lora/test_llama_tp.py
View File

@@ -76,11 +76,18 @@ def do_sample(
if lora_id
else None,
)
# Print the outputs.
lora_request = LoRARequest(str(lora_id), lora_id, lora_path) if lora_id else None
generated_texts: list[str] = []
for output in outputs:
prompt = output.prompt
generated_text = output.outputs[0].text
# The output should include correct lora_request info
if lora_request is not None:
assert output.lora_request.lora_name == lora_request.lora_name
assert output.lora_request.lora_int_id == lora_request.lora_int_id
assert output.lora_request.lora_path == lora_request.lora_path
else:
assert output.lora_request is None
generated_texts.append(generated_text)
print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")
return generated_texts