Skip to content

Skip tokenize/detokenize when it is disabled by arg --skip-tokenizer-init #367

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

Merged
maleksan85 merged 9 commits into from
Jan 22, 2025
Merged

Skip tokenize/detokenize when it is disabled by arg --skip-tokenizer-init #367

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
b3da2f1
switching detokenize flag to be False
Dec 23, 2024
c566cd9
getting branch up to date
Jan 17, 2025
87d256a
detokenize = False for benchmarks
Jan 17, 2025
f42c45e
restoring default in main vllm code for detokenize
Jan 18, 2025
629cae8
removing extra spaces
Jan 18, 2025
5e968e3
moving detokenize to flag
Jan 18, 2025
5ab6bce
adding support for token ids
Jan 21, 2025
ce87cff
Merge branch 'main' into detokenize_false_default
maleksan85 Jan 21, 2025
f188249
Merge branch 'main' into detokenize_false_default
maleksan85 Jan 22, 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
3 changes: 2 additions & 1 deletion benchmarks/profiling/benchmark_throughput.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -272,7 +272,8 @@ def main(args: argparse.Namespace):
args.tokenizer, trust_remote_code=args.trust_remote_code)
if args.dataset is None:
# Synthesize a prompt with the given input length.
prompt = "hi" * (args.input_len - 1)
prompt = { "prompt_token_ids" : [42] * (args.input_len - 1) } \
if args.skip_tokenizer_init else "hi" * (args.input_len - 1)
requests = [(prompt, args.input_len, args.output_len)
for _ in range(args.num_prompts)]
else:
Expand Down
2 changes: 1 addition & 1 deletion vllm/engine/output_processor/multi_step.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -144,7 +144,7 @@ def process_outputs(self,
def _process_decode_and_stop(self, seq: Sequence,
sampling_params: SamplingParams) -> None:
new_char_count = 0
if sampling_params.detokenize:
if sampling_params.detokenize and self.detokenizer:
new_char_count = self.detokenizer.decode_sequence_inplace(
seq, sampling_params)

Expand Down