chenchenghao/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

[AMD] Fix nightly 1-GPU test failures and bench_serving regression (#18761)

Browse Source 
Co-authored-by: michaelzhang-ai <michaelzhang-ai@users.noreply.github.com>
This commit is contained in:
Michael
2026-02-15 04:36:47 -08:00
committed by GitHub
parent 90555a0228
commit 88010e9601
5 changed files with 68 additions and 57 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
python/sglang/bench_serving.py
View File

@@ -1922,28 +1922,32 @@ def sample_generated_shared_prefix_requests(
range_ratio=range_ratio,
num=num_groups,
)
question_lens = compute_random_lens(
full_len=question_len,
range_ratio=range_ratio,
num=num_groups * prompts_per_group * num_turns,
question_lens = np.array(
compute_random_lens(
full_len=question_len,
range_ratio=range_ratio,
num=num_groups * prompts_per_group * num_turns,
)
).reshape(num_groups, prompts_per_group, num_turns)
output_lens = compute_random_lens(
full_len=output_len,
range_ratio=range_ratio,
num=num_groups * prompts_per_group,
output_lens = np.array(
compute_random_lens(
full_len=output_len,
range_ratio=range_ratio,
num=num_groups * prompts_per_group,
)
).reshape(num_groups, prompts_per_group)
del system_prompt_len, question_len, output_len
# Generate system prompts for each group
system_prompts = [
gen_prompt(tokenizer, system_prompt_lens[i].item()) for i in range(num_groups)
gen_prompt(tokenizer, system_prompt_lens[i]) for i in range(num_groups)
]
# Generate questions: shape (num_groups, prompts_per_group, num_turns)
questions = [
[
[
gen_prompt(tokenizer, question_lens[g, p, t].item())
gen_prompt(tokenizer, int(question_lens[g, p, t]))
for t in range(num_turns)
]
for p in range(prompts_per_group)
@@ -1976,7 +1980,7 @@ def sample_generated_shared_prefix_requests(
if getattr(args, "gsp_fast_prepare", False)
else len(tokenizer.encode(turn_prompts[0]))
)
output_len_val = output_lens[group_idx, prompt_idx].item()
output_len_val = int(output_lens[group_idx, prompt_idx])
input_requests.append(
DatasetRow(