vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

llama : Add support for DeepSeek V3 (#11049)

Browse source 
* convert : extend DEEPSEEK2 model architecture to support DeepseekV3ForCausalLM by adding EXPERT_WEIGHTS_NORM and EXPERT_GATING_FUNC model parameters and FFN_EXP_PROBS_B tensor type

* vocab : add DeepSeek V3 pre-tokenizer regexes

* unicode : handle ACCENT_MARK and SYMBOL categories in regex

* llama : add DeepSeek V3 chat template, handle new model parameters and tensor types

---------

Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com>
This commit is contained in:
fairydreaming 2025-01-04 21:06:11 +01:00 • committed by GitHub
parent f922a9c542
commit 9394bbd484
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
16 changed files with 162 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

7
src/llama-vocab.cpp
View file

@ -382,6 +382,13 @@ struct llm_tokenizer_bpe : llm_tokenizer {
"\\p{N}+",
};
break;
case LLAMA_VOCAB_PRE_TYPE_DEEPSEEK3_LLM:
regex_exprs = {
"\\p{N}{1,3}",
"[一-龥぀-ゟ゠-ヿ]+",
"[!\"#$%&'()*+,\\-./:;<=>?@\\[\\\\\\]^_`{|}~][A-Za-z]+|[^\r\n\\p{L}\\p{P}\\p{S}]?[\\p{L}\\p{M}]+| ?[\\p{P}\\p{S}]+[\r\n]*|\\s*[\r\n]+|\\s+(?!\\S)|\\s+",
};
break;
case LLAMA_VOCAB_PRE_TYPE_DEEPSEEK_CODER:
regex_exprs = {
"[\r\n]",