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

convert : fix python 3.8 support, modernize type annotations (#2916)

Browse source 
* convert : fix python 3.8 support

* convert : sort imports

* convert : fix required parameters in convert-llama-ggmlv3-to-gguf

* convert : fix mypy errors in convert-llama-ggmlv3-to-gguf

* convert : use PEP 585 generics and PEP 604 unions

Now that we have `from __future__ import annotations`, we can use this
modern syntax in Python 3.7 instead of restricting support to Python 3.9
or 3.10 respectively.

* gguf.py : a tuple is already a tuple

* add mypy.ini

* convert : add necessary `type: ignore` comments

* gguf-py: bump version
This commit is contained in:
Cebtenzzre 2023-08-31 01:02:23 -04:00 • committed by GitHub
parent 8afe228000
commit 92d0b751a7
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
10 changed files with 193 additions and 168 deletions
Download patch file Download diff file Expand all files Collapse all files

18
convert-llama-ggmlv3-to-gguf.py
View file

@ -1,10 +1,14 @@
#!/usr/bin/env python3
import sys, struct, math, argparse
from __future__ import annotations
import argparse
import math
import struct
import sys
from pathlib import Path
import numpy as np
import gguf
import numpy as np
# Note: Does not support GGML_QKK_64
QK_K = 256
@ -72,7 +76,7 @@ class Vocab:
class Tensor:
def __init__(self):
self.name = None
self.dims = ()
self.dims: tuple[int, ...] = ()
self.dtype = None
self.start_offset = 0
self.len_bytes = np.int64(0)
@ -119,7 +123,7 @@ class GGMLV3Model:
offset += hp.load(data, offset)
vocab = Vocab()
offset += vocab.load(data, offset, hp.n_vocab)
tensors = []
tensors: list[Tensor] = []
tensor_map = {}
while offset < len(data):
tensor = Tensor()
@ -305,8 +309,8 @@ def handle_metadata(cfg, hp):
def handle_args():
parser = argparse.ArgumentParser(description = 'Convert GGMLv3 models to GGUF')
parser.add_argument('--input', '-i', type = Path, help = 'Input GGMLv3 filename')
parser.add_argument('--output', '-o', type = Path, help ='Output GGUF filename')
parser.add_argument('--input', '-i', type = Path, required = True, help = 'Input GGMLv3 filename')
parser.add_argument('--output', '-o', type = Path, required = True, help ='Output GGUF filename')
parser.add_argument('--name', help = 'Set model name')
parser.add_argument('--desc', help = 'Set model description')
parser.add_argument('--gqa', type = int, default = 1, help = 'grouped-query attention factor (use 8 for LLaMA2 70B)')