Repository navigation
gguf-py : TemplateProcessing has final word on add_special_token - #29417
Merged
Merged
Conversation
danbev
approved these changes
Sep 25, 2026
feal87
added a commit
to feal87/myllama.cpp
that referenced
this pull request
Sep 25, 2026
Merge upstream commits: - llama: add llama_prec_policy + model-driven W4A4 path (ggml-org#24364) - llama: fix tensor split for fused qkv with uneven K/V head sizes (ggml-org#29294) - metal: split fa kernels into per-dtype libraries (ggml-org#29329) - metal: FWHT kernels for block widths above 512 (ggml-org#29095) - CUDA: fuse RMS_NORM + SCALE into one kernel (ggml-org#29393) - common: extract shared unicode path/string helpers (ggml-org#29415) - common,rpc: simplify fs_create_directory_with_parents() (ggml-org#29432) - rpc: include nb in the get_alloc_size cache key (ggml-org#29283) - [SYCL] support sparse FA (ggml-org#28796) - musa: fix PH1 operator failures and build issues (ggml-org#29193) - HIP: bump HIP_VERSION required for fp8 (ggml-org#29231) - opencl: add q5_k bin kernel (ggml-org#29401) - hexagon: add q5_k quant type support (ggml-org#29123) - hexagon: use DMA for contiguous dim1 CONCAT (ggml-org#29404) - mtmd: fix mel preprocessor in LFM2 audio (ggml-org#29403) - vulkan: fix legacy GLSLC without cooperativeMatrix (ggml-org#29409) - gguf-py: ByteLevel processing defaults bos/eos to False (ggml-org#29422) - gguf-py: TemplateProcessing has final word on add_special_token (ggml-org#29417) Assisted-by: Pi
edwardyoon
pushed a commit
to edwardyoon/focus-llama
that referenced
this pull request
Oct 7, 2026
…l-org#29417) * templateprocessing must win over tokenizer config * remove obsolete override (cherry picked from commit d81aef1)
Nathanw1014
added a commit
to Nathanw1014/llama.cpp
that referenced
this pull request
Oct 25, 2026
LFM2 (and EXL3 containers of it) converted with this fork had no tokenizer.ggml.add_bos_token / add_eos_token, so llama added a BOS per chunk: LFM2.5-8B-A1B EXL3 2.10bpw scored PPL 46.5 instead of 35.9. A ByteLevel post-processor adds no special tokens; a later TemplateProcessing or tokenizer_config add_*_token still overrides it. Port of ggml-org ggml-org#29422 (b248f4a, Sigbjørn Skjæret). The fork keeps its tokenizer_config precedence (ggml-org#29417 is not taken), so only the post-processor hunks apply. Assisted-by: Claude Opus 5.5
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
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.
Overview
TemplateProcessingmust overrule potentially erroneoustokenizer_config.json.Additional information
This affected f.ex.
nvidia/NVIDIA-Nemotron-Nano-9B-v2for which an override was added, this is no longer necessary and has been removed as it conflicts with newer models.Requirements