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

convert-hf : begin refactoring write_tensor

Browse source
This commit is contained in:
Francis Couture-Harpin 2024-04-30 14:07:28 -04:00
parent b8a7a5a90f
commit 47e02eb7bc
10 changed files with 386 additions and 852 deletions
Download patch file Download diff file Expand all files Collapse all files

8
gguf-py/gguf/gguf_reader.py
View file

@ -63,7 +63,7 @@ class ReaderTensor(NamedTuple):
class GGUFReader:
# I - same as host, S - swapped
byte_order: Literal['I' | 'S'] = 'I'
byte_order: Literal['I'] | Literal['S'] = 'I'
alignment: int = GGUF_DEFAULT_ALIGNMENT
# Note: Internal helper, API may change.
@ -81,7 +81,7 @@ class GGUFReader:
GGUFValueType.BOOL: np.bool_,
}
def __init__(self, path: os.PathLike[str] | str, mode: Literal['r' | 'r+' | 'c'] = 'r'):
def __init__(self, path: os.PathLike[str] | str, mode: Literal['r'] | Literal['r+'] | Literal['c'] = 'r'):
self.data = np.memmap(path, mode = mode)
offs = 0
if self._get(offs, np.uint32, override_order = '<')[0] != GGUF_MAGIC:
@ -126,7 +126,7 @@ class GGUFReader:
return self.tensors[idx]
def _get(
self, offset: int, dtype: npt.DTypeLike, count: int = 1, override_order: None | Literal['I' | 'S' | '<'] = None,
self, offset: int, dtype: npt.DTypeLike, count: int = 1, override_order: None | Literal['I'] | Literal['S'] | Literal['<'] = None,
) -> npt.NDArray[Any]:
count = int(count)
itemsize = int(np.empty([], dtype = dtype).itemsize)
@ -248,7 +248,7 @@ class GGUFReader:
raise ValueError(f'Found duplicated tensor with name {tensor_name}')
tensor_names.add(tensor_name)
ggml_type = GGMLQuantizationType(raw_dtype[0])
n_elems = np.prod(dims)
n_elems = int(np.prod(dims))
block_size, type_size = GGML_QUANT_SIZES[ggml_type]
n_bytes = n_elems * type_size // block_size
data_offs = int(start_offs + offset_tensor[0])