root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-25 19:04:35 +00:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

llama : add Minerva 7B model support (#10673)

Browse Source 
* Support for Minerva 7B

* Update convert_hf_to_gguf_update.py
This commit is contained in:
Riccardo Orlando 2024-12-05 19:30:59 +01:00 committed by GitHub
parent 0cd182ebcc
commit 6fe6247831
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
5 changed files with 9 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
convert_hf_to_gguf.py
View File

@ -658,6 +658,9 @@ class Model:
if chkhsh == "60824e3c0d9401f89943cbb2fff727f0e2d4c545ba4df2d6e4f09a6db0f5b450":
# ref: https://huggingface.co/facebook/chameleon-7b
res = "chameleon"
if chkhsh == "1431a23e583c97432bc230bff598d103ddb5a1f89960c8f1d1051aaa944d0b35":
# ref: https://huggingface.co/sapienzanlp/Minerva-7B-base-v1.0
res = "minerva-7b"
if res is None:
logger.warning("\n")

1
convert_hf_to_gguf_update.py
View File

@ -102,6 +102,7 @@ models = [
{"name": "exaone", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/LGAI-EXAONE/EXAONE-3.0-7.8B-Instruct", },
{"name": "phi-2", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/microsoft/phi-2", },
{"name": "chameleon", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/facebook/chameleon-7b", },
{"name": "minerva-7b", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/sapienzanlp/Minerva-7B-base-v1.0", },
]

1
include/llama.h
View File

@ -104,6 +104,7 @@ extern "C" {
LLAMA_VOCAB_PRE_TYPE_GPT3_FINNISH = 24,
LLAMA_VOCAB_PRE_TYPE_EXAONE = 25,
LLAMA_VOCAB_PRE_TYPE_CHAMELEON = 26,
LLAMA_VOCAB_PRE_TYPE_MINERVA = 27,
};
enum llama_rope_type {

1
src/llama-vocab.cpp
View File

@ -418,6 +418,7 @@ struct llm_tokenizer_bpe : llm_tokenizer {
case LLAMA_VOCAB_PRE_TYPE_SMOLLM:
case LLAMA_VOCAB_PRE_TYPE_CODESHELL:
case LLAMA_VOCAB_PRE_TYPE_EXAONE:
case LLAMA_VOCAB_PRE_TYPE_MINERVA:
regex_exprs = {
"\\p{N}",
"'s|'t|'re|'ve|'m|'ll|'d| ?\\p{L}+| ?\\p{N}+| ?[^\\s\\p{L}\\p{N}]+|\\s+(?!\\S)",

3
src/llama.cpp
View File

@ -6479,6 +6479,9 @@ static void llm_load_vocab(
vocab.type_pre = LLAMA_VOCAB_PRE_TYPE_CHAMELEON;
vocab.tokenizer_add_bos = true;
vocab.tokenizer_clean_spaces = false;
} else if (
tokenizer_pre == "minerva-7b") {
vocab.type_pre = LLAMA_VOCAB_PRE_TYPE_MINERVA;
} else {
throw std::runtime_error(format("unknown pre-tokenizer type: '%s'", tokenizer_pre.c_str()));
}