Skip to content

[https://nvbugs/6160085][fix] At tensorrt_llm/tokenizer/tokenizer.py import time, re-export `bytes_to_unicod #14116

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
ziyixiong-nv merged 1 commit into from
May 28, 2026
Merged

[https://nvbugs/6160085][fix] At tensorrt_llm/tokenizer/tokenizer.py import time, re-export `bytes_to_unicod #14116

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
13 changes: 13 additions & 0 deletions tensorrt_llm/tokenizer/tokenizer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,6 +10,19 @@
from .._utils import nvtx_range_debug
from ..logger import logger

# Transformers 5.x moved ``bytes_to_unicode`` out of
# ``transformers.models.gpt2.tokenization_gpt2`` into
# ``transformers.convert_slow_tokenizer``. Some ``trust_remote_code=True``
# checkpoints (e.g. Kimi-K2's ``tokenization_kimi.py``) still import it from
# the legacy location; re-export the symbol so those tokenizers keep loading.
try:
from transformers.models.gpt2 import tokenization_gpt2 as _gpt2_mod
if not hasattr(_gpt2_mod, "bytes_to_unicode"):
from transformers.convert_slow_tokenizer import bytes_to_unicode
_gpt2_mod.bytes_to_unicode = bytes_to_unicode
except ImportError:
pass
Comment thread
ziyixiong-nv marked this conversation as resolved.

# Aliases for built-in custom tokenizers.
TOKENIZER_ALIASES = {
"deepseek_v32": "tensorrt_llm.tokenizer.deepseek_v32.DeepseekV32Tokenizer",
Expand Down
2 changes: 0 additions & 2 deletions tests/integration/test_lists/waives.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -95,8 +95,6 @@ accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v2_kv_cache-ep4-cutl
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v2_kv_cache-tp4-cutlass-auto] SKIP (https://nvbugs/5596343)
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_chunked_prefill[cutlass-auto] SKIP (https://nvbugs/5596343)
accuracy/test_llm_api_pytorch.py::TestKanana_Instruct::test_auto_dtype SKIP (https://nvbugs/6209806)
accuracy/test_llm_api_pytorch.py::TestKimiK2::test_nvfp4_longseq_trtllm_moe_async_cancel SKIP (https://nvbugs/6160085)
accuracy/test_llm_api_pytorch.py::TestKimiK2::test_nvfp4_longseq_trtllm_moe_stress SKIP (https://nvbugs/6160085)
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=False-enable_padding=False-disable_overlap_scheduler=False-sampler_async_worker=False] SKIP (https://nvbugs/6141653)
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=False-enable_padding=False-disable_overlap_scheduler=True-sampler_async_worker=False] SKIP (https://nvbugs/6141653)
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=False-disable_overlap_scheduler=False-sampler_async_worker=False] SKIP (https://nvbugs/6141653)
Expand Down
Loading