fix tokenizers in respect to newer transformers by dimapihtar · Pull Request #4608 · NVIDIA/Megatron-LM

dimapihtar · 2026-05-04T19:17:34Z

What does this PR do ?

Minor tokenizer fix to make them compatible with transformers >= 5.3.0

Issue tracking

For PRs from open-source community contributors:

New features: a linked issue is required. Please open a feature request and reference it here before submitting the PR.
Small updates (bug fixes, minor improvements): a linked issue is recommended and will accelerate the PR review process.

Linked issue:

Contribution process

Pre-checks

I have added relevant unit tests
I have added relevant functional tests
I have added proper typing to my code Typing guidelines
I have added relevant documentation
I have run the autoformatter.sh on my PR

Code review

Feel free to message or comment the @mcore-oncall to help accelerate your merge into main. The less complex your PR is, the faster it will be approved and merged!

All PRs start as draft. If you open a non-draft PR, it will be automatically converted to draft.

Step 1: Mark PR as "Ready for Review"

When your PR is ready, click Ready for Review.
An oncall reviewer is auto-assigned and expert reviewers are notified based on your changes.
- Some PRs may jump straight to step 2. This is determined by .github/CODEOWNERS.

⚠️ Only mark as ready once merge-conflicts are resolved and the CI is passing.
Final Review might get declined if these requirements are not fulfilled.

Step 2: Final Review

For PRs that change megatron/core, once all expert reviewers have approved, the Final Review label is applied automatically and final reviewers are assigned.

For PRs outside megatron/core, this step is skipped.

Step 3: Approved

Once all required reviewers have approved, the Approved label is applied automatically.

Merge

Any member of mcore-engineers will be able to merge your PR.

For MRs into `dev` branch

The proposed review process for `dev` branch is under active discussion.

MRs are mergable after one approval by either eharper@nvidia.com or zijiey@nvidia.com.

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

copy-pr-bot · 2026-05-04T19:17:37Z

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

dimapihtar · 2026-05-04T19:18:26Z

/ok to test e01a02e

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

dimapihtar · 2026-05-04T21:26:54Z

/ok to test 9b0aa69

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

This reverts commit 62e86d1.

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

dimapihtar · 2026-05-05T14:38:25Z

/ok to test a5db619

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

dimapihtar · 2026-05-06T12:36:00Z

/ok to test 734841b

jaredcasper · 2026-05-06T19:03:09Z

-                [turn], tokenize=True, chat_template=self._prompt_config.custom_chat_template
+                [turn],
+                tokenize=True,
+                return_dict=False,


How new is the return_dict argument? Are we going to be incompatible with an older but still used transformers version?

@jaredcasper should keep backward compatibility return_dict was False by default in older versions but they changed it to True in 5.3.0+

we used 4.57.3 before this moment AFAIK

dimapihtar · 2026-05-14T16:45:29Z

/ok to test 961e804

svcnvidia-nemo-ci · 2026-05-14T18:36:01Z

🔄 Merge queue validation started!

You can track the progress here: https://github.com/NVIDIA/Megatron-LM/actions/runs/25878284960

Upstream main tip: f92a207 Pulled commits: - f92a207 Update copy-pr-bot.yaml [skip ci] - b3b6719 ci: tolerate git-gc race in /home/runner chown after checkout (NVIDIA#4808) - 9b4074b Inference: Optimize Prefill Engine Steps for Nemotron (NVIDIA#4764) - a53107c chore: Update nightly tests golden values (NVIDIA#4805) - e9a0930 ci: Update workflow to use same commit for build+test (NVIDIA#4787) - 266562f Update owners (NVIDIA#4794) - 98031e1 Bump nvidia-modelopt>=0.44.0 (NVIDIA#4803) - d167123 fix tokenizers in respect to newer transformers (NVIDIA#4608) - dbfc96b Use Protocols to type-check linear_proj submodules of Attention (NVIDIA#3434) Conflict resolutions: - .github/CODEOWNERS: --theirs (upstream granular team mapping) - megatron/core/transformer/attention.py: composed -- kept ours StreamBP imports, dropped dead ModuleSpec/build_module import (replaced by Protocol in NVIDIA#3434), kept ours attn_proj_manager pattern because our fine_grained_activation_offload.py keeps group_offload API while upstream split to static group_commit, but adopted apply_module(self.linear_proj) wrapper from NVIDIA#3434 - megatron/core/transformer/multi_latent_attention.py: same apply_module+attn_proj_manager composition; took ours ChunkRange parameter for StreamBP plus upstream BaseInferenceContext type annotation - pyproject.toml: bumped modelopt >=0.44 (upstream); kept transformer-engine[pytorch,core_cu13]>=2.9.0a0,<2.12.0 pin (custom for CUDA 13 / SM100) - uv.lock: --ours; modelopt bump is non-breaking. uv lock regenerate blocked by stale /home/sjpat/wheelhouse torchcomms path (pre-existing). Gates: - git diff --check: clean - conflict markers: none - py_compile (15 changed .py files): OK - attention.py + multi_latent_attention.py import OK - indexcache: 27/28 pass (same single GPU-env failure as pre-merge base) - transformer gdn/mtp/moe suite: 53 failed / 7 passed / 55 skipped / 5 errors -- identical to base (all failures are cudaErrorDevicesUnavailable from sglang occupying all H200s) - 2-rank torchrun smoke: blocked (no free GPUs) Custom preserved: StreamBP (megatron/core/transformer/streambp.py + tests/unit_tests/transformer/test_streambp.py + tools/streambp_prod_verify.py), IndexCache config + NVFP4 indexer, HISA topk1024 backward, emerging_optimizers v0.2.0 pin, mHC/MTP/MoE composition.

* origin/main: (138 commits) Refactor CUDA graph API: decompose cuda_graph_scope into full_iteration impl, inference scope, and per-layer capture modules (NVIDIA#4292) Add high-priority A2A stream and HybridEP preprocessing SMs (NVIDIA#4694) add is_torch_min_version in fsdp src (NVIDIA#4812) [Main][feat] Support A2A Overlap for Megatron-FSDP (NVIDIA#3797) Reorder mtp_post_process after attention backward in 1F1B schedule plan (NVIDIA#4695) [fix] Use MSC for checking checkpoint existence (NVIDIA#4251) Combine GEMM + SwiGLU fused MLP PRs (3890, 4071, 4095, 4219, 4311, 4324) → main (NVIDIA#4636) Strengthen test_checkpoint to verify distributed checkpoint behavior (NVIDIA#4711) Disable MSC by default; opt in via --enable-msc (NVIDIA#4629) additional tests for nvrx (NVIDIA#4522) Update copy-pr-bot.yaml [skip ci] ci: tolerate git-gc race in /home/runner chown after checkout (NVIDIA#4808) Inference: Optimize Prefill Engine Steps for Nemotron (NVIDIA#4764) chore: Update nightly tests golden values (NVIDIA#4805) ci: Update workflow to use same commit for building docker image and running tests (NVIDIA#4787) Update owners (NVIDIA#4794) Bump nvidia-modelopt>=0.44.0 (NVIDIA#4803) fix tokenizers in respect to newer transformers (NVIDIA#4608) Use Protocols to type-check linear_proj submodules of Attention (NVIDIA#3434) Fix recompute checkpointing + training CGs (NVIDIA#3919) ... # Conflicts: # megatron/core/transformer/moe/moe_utils.py

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

Signed-off-by: dimapihtar <dpykhtar@nvidia.com> Signed-off-by: yhgalaxy <yhgalaxy@outlook.com>

Signed-off-by: dimapihtar <dpykhtar@nvidia.com> Signed-off-by: Jon Barker <jbarker@aws-cmh-slurm-1-vscode-02.cm.cluster>

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

Signed-off-by: dimapihtar <dpykhtar@nvidia.com> Signed-off-by: Dmytro Pykhtar <dpykhtar@nvidia.com>

fix tokenizers

ee5ea00

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

Merge branch 'main' into fix_transformers_bump

e01a02e

dimapihtar marked this pull request as ready for review May 4, 2026 19:17

dimapihtar requested review from a team as code owners May 4, 2026 19:17

dimapihtar added complexity: low Expert Review [deprecated] Apply this label to indicate that your PR is ready for expert review. labels May 4, 2026

svcnvidia-nemo-ci requested a review from a team May 4, 2026 19:18

dimapihtar added the Run functional tests label May 4, 2026

svcnvidia-nemo-ci added complexity: low and removed complexity: low labels May 4, 2026

copy-pr-bot Bot temporarily deployed to test May 4, 2026 19:19 Inactive

ko3n1g approved these changes May 4, 2026

View reviewed changes

minor fix

9b0aa69

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

copy-pr-bot Bot temporarily deployed to test May 4, 2026 21:28 Inactive

dimapihtar and others added 8 commits May 5, 2026 16:42

Merge branch 'main' into fix_transformers_bump

c8253fe

set return_dict=False

0da0924

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

minor fix

95b53e5

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

upgrade transformers version

62e86d1

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

Revert "upgrade transformers version"

565cecb

This reverts commit 62e86d1.

upgrade transformers

dfd187e

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

revert changes

3f28c0c

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

Merge branch 'main' into fix_transformers_bump

a5db619

copy-pr-bot Bot temporarily deployed to test May 5, 2026 14:39 Inactive

switch back to initial solution

7f4120a

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

minor change

734841b

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

copy-pr-bot Bot temporarily deployed to test May 6, 2026 12:37 Inactive

asolergi-nv approved these changes May 6, 2026

View reviewed changes

svcnvidia-nemo-ci added the Final Review PR is in the "final review" stage label May 6, 2026

dimapihtar requested a review from ericharper May 6, 2026 18:38

jaredcasper reviewed May 6, 2026

View reviewed changes

ericharper reviewed May 6, 2026

View reviewed changes

Comment thread uv.lock

ko3n1g approved these changes May 6, 2026

View reviewed changes

Comment thread uv.lock

ericharper approved these changes May 13, 2026

View reviewed changes

ericharper enabled auto-merge May 13, 2026 15:50

jaredcasper approved these changes May 13, 2026

View reviewed changes

svcnvidia-nemo-ci added Approved All necessary approvals have been made and removed Final Review PR is in the "final review" stage labels May 13, 2026

Merge branch 'main' into fix_transformers_bump

961e804

dimapihtar removed the Run functional tests label May 14, 2026

copy-pr-bot Bot temporarily deployed to public May 14, 2026 16:46 Inactive

copy-pr-bot Bot temporarily deployed to test May 14, 2026 16:46 Inactive

ericharper added this pull request to the merge queue May 14, 2026

Merged via the queue into NVIDIA:main with commit d167123 May 14, 2026
67 of 70 checks passed

janEbert pushed a commit to janEbert/Megatron-LM that referenced this pull request Jun 2, 2026

fix tokenizers in respect to newer transformers (NVIDIA#4608)

59dc7cd

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

yhgalaxy pushed a commit to yhgalaxy/Megatron-LM that referenced this pull request Jun 17, 2026

fix tokenizers in respect to newer transformers (NVIDIA#4608)

0aa2eec

Signed-off-by: dimapihtar <dpykhtar@nvidia.com> Signed-off-by: yhgalaxy <yhgalaxy@outlook.com>

jon-barker pushed a commit to jon-barker/Megatron-LM that referenced this pull request Jul 10, 2026

fix tokenizers in respect to newer transformers (NVIDIA#4608)

7e593e2

Signed-off-by: dimapihtar <dpykhtar@nvidia.com> Signed-off-by: Jon Barker <jbarker@aws-cmh-slurm-1-vscode-02.cm.cluster>

terminator123 pushed a commit to 021ai/Megatron-LM that referenced this pull request Aug 3, 2026

fix tokenizers in respect to newer transformers (NVIDIA#4608)

7bbe25b

Signed-off-by: dimapihtar <dpykhtar@nvidia.com>

svcnvidia-nemo-ci pushed a commit to dimapihtar/Megatron-LM that referenced this pull request Aug 4, 2026

fix tokenizers in respect to newer transformers (NVIDIA#4608)

0aecb4d

Signed-off-by: dimapihtar <dpykhtar@nvidia.com> Signed-off-by: Dmytro Pykhtar <dpykhtar@nvidia.com>

Navigation Menu

Uh oh!

fix tokenizers in respect to newer transformers - #4608

fix tokenizers in respect to newer transformers#4608
ericharper merged 18 commits into
NVIDIA:mainfrom
dimapihtar:fix_transformers_bump

dimapihtar commented May 4, 2026 •

edited

Loading

Uh oh!

copy-pr-bot Bot commented May 4, 2026

Uh oh!

dimapihtar commented May 4, 2026

Uh oh!

dimapihtar commented May 4, 2026

Uh oh!

dimapihtar commented May 5, 2026

Uh oh!

dimapihtar commented May 6, 2026

Uh oh!

jaredcasper May 6, 2026

Uh oh!

dimapihtar May 7, 2026

Uh oh!

dimapihtar May 7, 2026

Uh oh!

Uh oh!

Uh oh!

dimapihtar commented May 14, 2026

Uh oh!

svcnvidia-nemo-ci commented May 14, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

Uh oh!

Conversation

dimapihtar commented May 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What does this PR do ?

Issue tracking

Contribution process

Pre-checks

Code review

Step 1: Mark PR as "Ready for Review"

Step 2: Final Review

Step 3: Approved

Merge

Uh oh!

copy-pr-bot Bot commented May 4, 2026

Uh oh!

dimapihtar commented May 4, 2026

Uh oh!

dimapihtar commented May 4, 2026

Uh oh!

dimapihtar commented May 5, 2026

Uh oh!

dimapihtar commented May 6, 2026

Uh oh!

jaredcasper May 6, 2026

Choose a reason for hiding this comment

Uh oh!

dimapihtar May 7, 2026

Choose a reason for hiding this comment

Uh oh!

dimapihtar May 7, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

dimapihtar commented May 14, 2026

Uh oh!

svcnvidia-nemo-ci commented May 14, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

dimapihtar commented May 4, 2026 •

edited

Loading