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

Update convert-llama-h5-to-gguf.py

Browse source
This commit is contained in:
klosax 2023-07-31 23:04:03 +02:00 • committed by GitHub
parent f3de876a12
commit da4900e835
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 0 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

5
convert-llama-h5-to-gguf.py
View file

@ -104,11 +104,6 @@ if Path(dir_model + "/tokenizer.model").is_file():
print("Adding sentencepiece tokenizer vocab.") print("Adding sentencepiece tokenizer vocab.")
tokenizer = SentencePieceProcessor(dir_model + "/tokenizer.model") tokenizer = SentencePieceProcessor(dir_model + "/tokenizer.model")
# output vocab_size followed by all piece/score pairs
outbytes: bytes
outbytes = b""
outbytes += struct.pack("I", tokenizer.vocab_size())
for i in range(tokenizer.vocab_size()): for i in range(tokenizer.vocab_size()):
text: bytes text: bytes
if tokenizer.is_unknown(i): if tokenizer.is_unknown(i):