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

gguf-dump.py: element count autosizing

Browse source
This commit is contained in:
brian khuu 2024-06-15 22:03:39 +10:00
parent dfbf6e1458
commit 51032b15e1
1 changed files with 9 additions and 8 deletions
Download patch file Download diff file Expand all files Collapse all files

17
gguf-py/scripts/gguf-dump.py
View file

@ -14,7 +14,7 @@ import numpy as np
if "NO_LOCAL_GGUF" not in os.environ and (Path(__file__).parent.parent.parent / 'gguf-py').exists(): if "NO_LOCAL_GGUF" not in os.environ and (Path(__file__).parent.parent.parent / 'gguf-py').exists():
sys.path.insert(0, str(Path(__file__).parent.parent)) sys.path.insert(0, str(Path(__file__).parent.parent))
from gguf import GGUFReader, GGUFValueType # noqa: E402 from gguf import GGUFReader, GGUFValueType, ReaderTensor # noqa: E402
logger = logging.getLogger("gguf-dump") logger = logging.getLogger("gguf-dump")
@ -319,22 +319,23 @@ def dump_markdown_metadata(reader: GGUFReader, args: argparse.Namespace) -> None
group_percentage = group_elements / total_elements * 100 group_percentage = group_elements / total_elements * 100
markdown_content += f"### <a name=\"{group.replace('.', '_')}\">{translate_tensor_name(group)} Tensor Group : {element_count_rounded_notation(group_elements)} Elements</a>\n\n" markdown_content += f"### <a name=\"{group.replace('.', '_')}\">{translate_tensor_name(group)} Tensor Group : {element_count_rounded_notation(group_elements)} Elements</a>\n\n"
# Precalculate pretty shape column sizing for visual consistency # Precalculate column sizing for visual consistency
prettify_element_est_count_size: int = 1
prettify_element_count_size: int = 1
prettify_dimension_max_widths: dict[int, int] = {} prettify_dimension_max_widths: dict[int, int] = {}
for tensor in tensors: for tensor in tensors:
prettify_element_est_count_size = max(prettify_element_est_count_size, len(str(element_count_rounded_notation(tensor.n_elements))))
prettify_element_count_size = max(prettify_element_count_size, len(str(tensor.n_elements)))
for i, dimension_size in enumerate(list(tensor.shape) + [1] * (4 - len(tensor.shape))): for i, dimension_size in enumerate(list(tensor.shape) + [1] * (4 - len(tensor.shape))):
if i in prettify_dimension_max_widths: prettify_dimension_max_widths[i] = max(prettify_dimension_max_widths.get(i,1), len(str(dimension_size)))
prettify_dimension_max_widths[i] = max(prettify_dimension_max_widths[i], len(str(dimension_size)))
else:
prettify_dimension_max_widths[i] = len(str(dimension_size))
# Generate Tensor Layer Table Content # Generate Tensor Layer Table Content
tensor_dump_table: list[dict[str, str | int]] = [] tensor_dump_table: list[dict[str, str | int]] = []
for tensor in tensors: for tensor in tensors:
human_friendly_name = translate_tensor_name(tensor.name.replace(".weight", ".(W)").replace(".bias", ".(B)")) human_friendly_name = translate_tensor_name(tensor.name.replace(".weight", ".(W)").replace(".bias", ".(B)"))
pretty_dimension = ' x '.join(f'{str(d):>{prettify_dimension_max_widths[i]}}' for i, d in enumerate(list(tensor.shape) + [1] * (4 - len(tensor.shape)))) pretty_dimension = ' x '.join(f'{str(d):>{prettify_dimension_max_widths[i]}}' for i, d in enumerate(list(tensor.shape) + [1] * (4 - len(tensor.shape))))
element_count_est = f"({element_count_rounded_notation(tensor.n_elements):>5})" element_count_est = f"({element_count_rounded_notation(tensor.n_elements):>{prettify_element_est_count_size}})"
element_count_string = f"{element_count_est} {tensor.n_elements:>8}" element_count_string = f"{element_count_est} {tensor.n_elements:>{prettify_element_count_size}}"
type_name_string = f"{tensor.tensor_type.name}" type_name_string = f"{tensor.tensor_type.name}"
tensor_dump_table.append({"t_id":tensor_name_to_key[tensor.name], "layer_name":tensor.name, "human_layer_name":human_friendly_name, "element_count":element_count_string, "pretty_dimension":pretty_dimension, "tensor_type":type_name_string}) tensor_dump_table.append({"t_id":tensor_name_to_key[tensor.name], "layer_name":tensor.name, "human_layer_name":human_friendly_name, "element_count":element_count_string, "pretty_dimension":pretty_dimension, "tensor_type":type_name_string})