Fix Kimi-K2.5 tokenizer regression and _patch_mistral_regex AttributeError - #45359
Merged
Conversation
…Error Fixes #45356 Remove `kimi_k25` from `MODELS_WITH_INCORRECT_HUB_TOKENIZER_CLASS` — its remote `TikTokenTokenizer` is the only correct backend (no `tokenizer.json`, non-sequential added-token IDs that `TokenizersBackend` cannot reproduce). Also fix `_patch_mistral_regex`: the method receives the raw `tokenizers.Tokenizer` object, which has `.pre_tokenizer` directly, not `.backend_tokenizer.pre_tokenizer`. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Contributor
|
[For maintainers] Suggested jobs to run (before merge) run-slow: auto |
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
ArthurZucker
added
the
for patch
github-merge-queue
Bot
removed this pull request from the merge queue due to failed status checks
Apr 13, 2026
ArthurZucker
added a commit
that referenced
this pull request
Apr 13, 2026
…Error (#45359) Fixes #45356 Remove `kimi_k25` from `MODELS_WITH_INCORRECT_HUB_TOKENIZER_CLASS` — its remote `TikTokenTokenizer` is the only correct backend (no `tokenizer.json`, non-sequential added-token IDs that `TokenizersBackend` cannot reproduce). Also fix `_patch_mistral_regex`: the method receives the raw `tokenizers.Tokenizer` object, which has `.pre_tokenizer` directly, not `.backend_tokenizer.pre_tokenizer`. Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
sirzechs66
pushed a commit
to sirzechs66/transformers
that referenced
this pull request
Apr 18, 2026
…Error (huggingface#45359) Fixes huggingface#45356 Remove `kimi_k25` from `MODELS_WITH_INCORRECT_HUB_TOKENIZER_CLASS` — its remote `TikTokenTokenizer` is the only correct backend (no `tokenizer.json`, non-sequential added-token IDs that `TokenizersBackend` cannot reproduce). Also fix `_patch_mistral_regex`: the method receives the raw `tokenizers.Tokenizer` object, which has `.pre_tokenizer` directly, not `.backend_tokenizer.pre_tokenizer`. Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2 tasks
This was referenced Apr 23, 2026
2 tasks
wangxiyuan
pushed a commit
to vllm-project/vllm-ascend
that referenced
this pull request
Jun 4, 2026
### What this PR does / why we need it? Transformers fixed a tokenizer bug for kimi k2.6/k2.5 in version 5.5.4, and this bug affects the function calling feature. huggingface/transformers#45359 ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? by test tool calls feature - vLLM version: v0.20.2 - vLLM main: vllm-project/vllm@9090368 Signed-off-by: zjks98 <zhangjiakang4@huawei.com> Co-authored-by: zjks98 <zhangjiakang4@huawei.com>
AuroraEmiya
pushed a commit
to AuroraEmiya/vllm-ascend
that referenced
this pull request
Jun 5, 2026
…t#9996) ### What this PR does / why we need it? Transformers fixed a tokenizer bug for kimi k2.6/k2.5 in version 5.5.4, and this bug affects the function calling feature. huggingface/transformers#45359 ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? by test tool calls feature - vLLM version: v0.20.2 - vLLM main: vllm-project/vllm@9090368 Signed-off-by: zjks98 <zhangjiakang4@huawei.com> Co-authored-by: zjks98 <zhangjiakang4@huawei.com>
LostFox11
pushed a commit
to LostFox11/vllm-ascend
that referenced
this pull request
Jun 15, 2026
…t#9996) ### What this PR does / why we need it? Transformers fixed a tokenizer bug for kimi k2.6/k2.5 in version 5.5.4, and this bug affects the function calling feature. huggingface/transformers#45359 ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? by test tool calls feature - vLLM version: v0.20.2 - vLLM main: vllm-project/vllm@9090368 Signed-off-by: zjks98 <zhangjiakang4@huawei.com> Co-authored-by: zjks98 <zhangjiakang4@huawei.com>
LostFox11
pushed a commit
to LostFox11/vllm-ascend
that referenced
this pull request
Jun 15, 2026
…t#9996) ### What this PR does / why we need it? Transformers fixed a tokenizer bug for kimi k2.6/k2.5 in version 5.5.4, and this bug affects the function calling feature. huggingface/transformers#45359 ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? by test tool calls feature - vLLM version: v0.20.2 - vLLM main: vllm-project/vllm@9090368 Signed-off-by: zjks98 <zhangjiakang4@huawei.com> Co-authored-by: zjks98 <zhangjiakang4@huawei.com>
ader47
pushed a commit
to ader47/vllm-ascend
that referenced
this pull request
Jun 18, 2026
…t#9996) ### What this PR does / why we need it? Transformers fixed a tokenizer bug for kimi k2.6/k2.5 in version 5.5.4, and this bug affects the function calling feature. huggingface/transformers#45359 ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? by test tool calls feature - vLLM version: v0.20.2 - vLLM main: vllm-project/vllm@9090368 Signed-off-by: zjks98 <zhangjiakang4@huawei.com> Co-authored-by: zjks98 <zhangjiakang4@huawei.com>
CXY-Katrina
pushed a commit
to CXY-Katrina/vllm-ascend
that referenced
this pull request
Jun 27, 2026
…t#9996) ### What this PR does / why we need it? Transformers fixed a tokenizer bug for kimi k2.6/k2.5 in version 5.5.4, and this bug affects the function calling feature. huggingface/transformers#45359 ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? by test tool calls feature - vLLM version: v0.20.2 - vLLM main: vllm-project/vllm@9090368 Signed-off-by: zjks98 <zhangjiakang4@huawei.com> Co-authored-by: zjks98 <zhangjiakang4@huawei.com>
HaoxinZong
pushed a commit
to HaoxinZong/vllm-ascend
that referenced
this pull request
Jun 27, 2026
…t#9996) ### What this PR does / why we need it? Transformers fixed a tokenizer bug for kimi k2.6/k2.5 in version 5.5.4, and this bug affects the function calling feature. huggingface/transformers#45359 ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? by test tool calls feature - vLLM version: v0.20.2 - vLLM main: vllm-project/vllm@9090368 Signed-off-by: zjks98 <zhangjiakang4@huawei.com> Co-authored-by: zjks98 <zhangjiakang4@huawei.com>
pisceskkk
pushed a commit
to pisceskkk/vllm-ascend
that referenced
this pull request
Jun 29, 2026
…t#9996) ### What this PR does / why we need it? Transformers fixed a tokenizer bug for kimi k2.6/k2.5 in version 5.5.4, and this bug affects the function calling feature. huggingface/transformers#45359 ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? by test tool calls feature - vLLM version: v0.20.2 - vLLM main: vllm-project/vllm@9090368 Signed-off-by: zjks98 <zhangjiakang4@huawei.com> Co-authored-by: zjks98 <zhangjiakang4@huawei.com>
xqchen7
pushed a commit
to nv-action/vllm-benchmarks
that referenced
this pull request
Jul 15, 2026
### What this PR does / why we need it? Transformers fixed a tokenizer bug for kimi k2.6/k2.5 in version 5.5.4, and this bug affects the function calling feature. huggingface/transformers#45359 ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? by test tool calls feature - vLLM version: v0.20.2 - vLLM main: vllm-project/vllm@9090368 Signed-off-by: zjks98 <zhangjiakang4@huawei.com> Co-authored-by: zjks98 <zhangjiakang4@huawei.com> Signed-off-by: xqchen7 <chenxueqing7@huawei.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes #45356
Summary
kimi_k25fromMODELS_WITH_INCORRECT_HUB_TOKENIZER_CLASS: its remoteTikTokenTokenizeris the only correct backend — the model has notokenizer.json, and itsadded_tokens_decoderhas non-sequential IDs (gaps +[UNK]/[PAD]at 163838/163839) thatTokenizersBackend.add_tokens()can't reproduce, causing every token after ID 163588 to be assigned wrong IDs._patch_mistral_regex: usetokenizer.pre_tokenizerinstead oftokenizer.backend_tokenizer.pre_tokenizer— the method receives the rawtokenizers.Tokenizer, which doesn't have.backend_tokenizer.Test plan
AutoTokenizer.from_pretrained("moonshotai/Kimi-K2.5", trust_remote_code=True).decode([163607])returns'</think>'<think>hello</think>works