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

Shard dataclass and un-negative dont_add_architecture

Browse source
This commit is contained in:
Christian Zhou-Zheng 2024-06-06 08:37:35 -04:00
parent 6a05183b97
commit 3328b0a991
2 changed files with 31 additions and 22 deletions
Download patch file Download diff file Expand all files Collapse all files

49
gguf-py/gguf/gguf_manager.py
View file

@ -5,6 +5,7 @@ from enum import IntEnum
from typing import TYPE_CHECKING, Any, Sequence from typing import TYPE_CHECKING, Any, Sequence
from argparse import Namespace from argparse import Namespace
from collections import deque from collections import deque
from dataclasses import dataclass
import numpy as np import numpy as np
@ -28,7 +29,14 @@ LLM_KV_SPLIT_TENSORS_COUNT = "split.tensors.count"
KVTempData: TypeAlias = dict[str, tuple[Any, GGUFValueType]] # {key: (value, type)} KVTempData: TypeAlias = dict[str, tuple[Any, GGUFValueType]] # {key: (value, type)}
TensorTempData: TypeAlias = tuple[str, np.ndarray[Any, Any], GGMLQuantizationType] # (tensor name, tensor data, tensor dtype) TensorTempData: TypeAlias = tuple[str, np.ndarray[Any, Any], GGMLQuantizationType] # (tensor name, tensor data, tensor dtype)
Shard: TypeAlias = list[os.PathLike[str], int, int, deque[TensorTempData]] # [shard filename, shard tensor count, shard size, [tensor data]]
@dataclass
class Shard:
path: str
tensor_count: int
size: int
tensors: deque[TensorTempData]
class SplitStyle(IntEnum): class SplitStyle(IntEnum):
@ -73,11 +81,11 @@ class GGUFManager(GGUFWriter):
self.state = WriterState.EMPTY self.state = WriterState.EMPTY
if self.split_arguments.small_first_shard: if self.split_arguments.small_first_shard:
self.shards.append(["", 0, METADATA_ONLY_INDICATOR, None]) self.shards.append(Shard("", 0, METADATA_ONLY_INDICATOR, deque()))
def init_shards(self) -> None: def init_shards(self) -> None:
self.total_tensors = sum(shard[1] for shard in self.shards) self.total_tensors = sum(shard.tensor_count for shard in self.shards)
total_size = sum(shard[2] for shard in self.shards) total_size = sum(shard.size for shard in self.shards)
# check if we need to split # check if we need to split
if self.split_arguments.split_max_tensors and self.total_tensors < self.split_arguments.split_max_tensors: if self.split_arguments.split_max_tensors and self.total_tensors < self.split_arguments.split_max_tensors:
@ -90,19 +98,20 @@ class GGUFManager(GGUFWriter):
# no shards are created when writing vocab so make one # no shards are created when writing vocab so make one
if not self.shards: if not self.shards:
self.shards.append(["", 0, METADATA_ONLY_INDICATOR, None]) self.shards.append(Shard("", 0, METADATA_ONLY_INDICATOR, deque()))
# format shard names # format shard names
if len(self.shards) == 1: if len(self.shards) == 1:
self.shards[0][0] = self.path self.shards[0].path = self.path
else: else:
for i in range(len(self.shards)): for i in range(len(self.shards)):
self.shards[i][0] = self.path.with_name(SHARD_NAME_FORMAT.format(self.path.stem, i + 1, len(self.shards))) # TODO with_name is not explicit - import pathlib
self.shards[i].path = self.path.with_name(SHARD_NAME_FORMAT.format(self.path.stem, i + 1, len(self.shards)))
# print shard info # print shard info
print("\nWriting the following files:") print("\nWriting the following files:")
for (path, tensor_ct, size, _) in self.shards: for shard in self.shards:
print(f" {path}: n_tensors = {tensor_ct}, total_size = {GGUFManager.format_n_bytes_to_str(size)}") print(f" {shard.path}: n_tensors = {shard.tensor_count}, total_size = {GGUFManager.format_n_bytes_to_str(shard.size)}")
print() print()
if self.split_arguments.dry_run: if self.split_arguments.dry_run:
@ -110,10 +119,10 @@ class GGUFManager(GGUFWriter):
exit() exit()
# we don't want to initialize GGUFWriters until now because they create files # we don't want to initialize GGUFWriters until now because they create files
for i, (path, _, _, tensors) in enumerate(self.shards): for i, shard in enumerate(self.shards):
# dont_add_architecture is used for consistency - examples/gguf_split doesn't add arch to all shards # add_architecture is used for consistency - examples/gguf_split doesn't add arch to all shards
writer = GGUFWriter(path, self.arch, use_temp_file=self.use_temp_file, writer = GGUFWriter(shard.path, self.arch, use_temp_file=self.use_temp_file,
endianess=self.endianess, dont_add_architecture=not (i == 0)) endianess=self.endianess, add_architecture=(i == 0))
# only the first shard needs all the KV data # only the first shard needs all the KV data
if i == 0: if i == 0:
@ -130,7 +139,7 @@ class GGUFManager(GGUFWriter):
# add tensors, deque popleft() ensures references to eager tensors are not kept # add tensors, deque popleft() ensures references to eager tensors are not kept
while True: while True:
try: try:
(name, tensor, dtype) = tensors.popleft() (name, tensor, dtype) = shard.tensors.popleft()
writer.add_tensor(name, tensor, raw_dtype=dtype) writer.add_tensor(name, tensor, raw_dtype=dtype)
except: except:
break break
@ -199,17 +208,17 @@ class GGUFManager(GGUFWriter):
if (len(self.shards) == self.split_arguments.small_first_shard \ if (len(self.shards) == self.split_arguments.small_first_shard \
# or split when over tensor limit # or split when over tensor limit
or (self.split_arguments.split_style == SplitStyle.TENSORS \ or (self.split_arguments.split_style == SplitStyle.TENSORS \
and self.shards[-1][1] >= self.split_arguments.split_max_tensors) \ and self.shards[-1].tensor_count >= self.split_arguments.split_max_tensors) \
# or split when over size limit # or split when over size limit
or (self.split_arguments.split_style == SplitStyle.SIZE \ or (self.split_arguments.split_style == SplitStyle.SIZE \
and self.shards[-1][2] + GGUFManager.get_tensor_size(tensor) > self.split_arguments.split_max_size)): and self.shards[-1].size + GGUFManager.get_tensor_size(tensor) > self.split_arguments.split_max_size)):
# we fill in the name later when we know how many shards there are # we fill in the name later when we know how many shards there are
self.shards.append(["", 1, GGUFManager.get_tensor_size(tensor), deque([(name, tensor, raw_dtype)])]) self.shards.append(Shard("", 1, GGUFManager.get_tensor_size(tensor), deque([(name, tensor, raw_dtype)])))
else: else:
self.shards[-1][1] += 1 self.shards[-1].tensor_count += 1
self.shards[-1][2] += GGUFManager.get_tensor_size(tensor) self.shards[-1].size += GGUFManager.get_tensor_size(tensor)
self.shards[-1][3].append((name, tensor, raw_dtype)) self.shards[-1].tensors.append((name, tensor, raw_dtype))
def close(self) -> None: def close(self) -> None:
for writer in self.shard_writers: for writer in self.shard_writers:

4
gguf-py/gguf/gguf_writer.py
View file

@ -57,7 +57,7 @@ class GGUFWriter:
def __init__( def __init__(
self, path: os.PathLike[str] | str, arch: str, use_temp_file: bool = True, self, path: os.PathLike[str] | str, arch: str, use_temp_file: bool = True,
endianess: GGUFEndian = GGUFEndian.LITTLE, dont_add_architecture: bool = False endianess: GGUFEndian = GGUFEndian.LITTLE, add_architecture: bool = True
): ):
self.fout = open(path, "wb") self.fout = open(path, "wb")
self.arch = arch self.arch = arch
@ -77,7 +77,7 @@ class GGUFWriter:
)) ))
self.state = WriterState.EMPTY self.state = WriterState.EMPTY
if not dont_add_architecture: if add_architecture:
self.add_architecture() self.add_architecture()
def write_header_to_file(self) -> None: def write_header_to_file(self) -> None: