vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-14 10:59:41 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

vocab : JetBrains Mellum pre-tokenizer (#15045)

Browse source
This commit is contained in:
Csaba Kecskemeti 2025-08-03 12:38:18 -07:00 • committed by GitHub
parent 83bc2f288c
commit 97366dc6ab
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 6 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

3
convert_hf_to_gguf.py
View file

@ -852,6 +852,9 @@ class TextModel(ModelBase):
if chkhsh == "2085e1638f6c377a0aa4ead21b27bb4cb941bf800df86ed391011769c1758dfb":
# ref: https://huggingface.co/LGAI-EXAONE/EXAONE-4.0-32B
res = "exaone4"
if chkhsh == "a1e163ecab2e718a4c829d1148b6e86824ec36163bb71941c3dca9cd5ac25756":
# ref: https://huggingface.co/JetBrains/Mellum-4b-base
res = "mellum"
if res is None:
logger.warning("\n")

1
convert_hf_to_gguf_update.py
View file

@ -138,6 +138,7 @@ models = [
{"name": "midm-2.0", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/K-intelligence/Midm-2.0-Base-Instruct", },
{"name": "lfm2", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/LiquidAI/LFM2-Tokenizer"},
{"name": "exaone4", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/LGAI-EXAONE/EXAONE-4.0-32B", },
{"name": "mellum", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/JetBrains/Mellum-4b-base", },
]
# some models are known to be broken upstream, so we will skip them as exceptions

3
src/llama-vocab.cpp
View file

@ -1856,7 +1856,8 @@ void llama_vocab::impl::load(llama_model_loader & ml, const LLM_KV & kv) {
tokenizer_pre == "gigachat" ||
tokenizer_pre == "jina-v2-es" ||
tokenizer_pre == "jina-v2-de" ||
tokenizer_pre == "a.x-4.0") {
tokenizer_pre == "a.x-4.0" ||
tokenizer_pre == "mellum") {
pre_type = LLAMA_VOCAB_PRE_TYPE_GPT2;
} else if (
tokenizer_pre == "jina-v1-en" ||