RWKV-Runner/backend-python/convert_safetensors.py

114 lines
3.4 KiB
Python
Raw Normal View History

import collections
import numpy
2023-08-16 23:07:58 +08:00
import os
import torch
from safetensors.torch import serialize_file, load_file
2023-08-16 23:07:58 +08:00
import argparse
parser = argparse.ArgumentParser()
parser.add_argument("--input", type=str, help="Path to input pth model")
parser.add_argument(
"--output",
type=str,
default="./converted.st",
help="Path to output safetensors model",
)
args = parser.parse_args()
def rename_key(rename, name):
for k, v in rename.items():
if k in name:
name = name.replace(k, v)
return name
2023-12-06 23:08:40 +08:00
def convert_file(pt_filename: str, sf_filename: str, rename={}, transpose_names=[]):
loaded: collections.OrderedDict = torch.load(pt_filename, map_location="cpu")
2023-08-16 23:07:58 +08:00
if "state_dict" in loaded:
loaded = loaded["state_dict"]
kk = list(loaded.keys())
version = 4
for x in kk:
if "ln_x" in x:
version = max(5, version)
if "gate.weight" in x:
version = max(5.1, version)
if int(version) == 5 and "att.time_decay" in x:
if len(loaded[x].shape) > 1:
if loaded[x].shape[1] > 1:
version = max(5.2, version)
if "time_maa" in x:
version = max(6, version)
print(f"Model detected: v{version:.1f}")
if version == 5.1:
_, n_emb = loaded["emb.weight"].shape
for k in kk:
if "time_decay" in k or "time_faaaa" in k:
# print(k, mm[k].shape)
loaded[k] = (
loaded[k].unsqueeze(1).repeat(1, n_emb // loaded[k].shape[0])
)
2024-02-28 23:25:46 +08:00
with torch.no_grad():
for k in kk:
new_k = rename_key(rename, k).lower()
v = loaded[k].half()
del loaded[k]
for transpose_name in transpose_names:
if transpose_name in new_k:
dims = len(v.shape)
v = v.transpose(dims - 2, dims - 1)
print(f"{new_k}\t{v.shape}\t{v.dtype}")
loaded[new_k] = {
"dtype": str(v.dtype).split(".")[-1],
"shape": v.shape,
"data": v.numpy().tobytes(),
}
2023-08-16 23:07:58 +08:00
dirname = os.path.dirname(sf_filename)
os.makedirs(dirname, exist_ok=True)
serialize_file(loaded, sf_filename, metadata={"format": "pt"})
# reloaded = load_file(sf_filename)
# for k in loaded:
# pt_tensor = torch.Tensor(
# numpy.frombuffer(
# bytearray(loaded[k]["data"]),
# dtype=getattr(numpy, loaded[k]["dtype"]),
# ).reshape(loaded[k]["shape"])
# )
# sf_tensor = reloaded[k]
# if not torch.equal(pt_tensor, sf_tensor):
# raise RuntimeError(f"The output tensors do not match for key {k}")
2023-08-16 23:07:58 +08:00
if __name__ == "__main__":
try:
convert_file(
args.input,
args.output,
2023-12-06 23:08:40 +08:00
rename={
"time_faaaa": "time_first",
"time_maa": "time_mix",
"lora_A": "lora.0",
"lora_B": "lora.1",
},
transpose_names=[
"time_mix_w1",
"time_mix_w2",
"time_decay_w1",
"time_decay_w2",
"time_state",
2024-04-30 21:52:47 +08:00
"lora.0",
2023-12-06 23:08:40 +08:00
],
)
2023-08-16 23:07:58 +08:00
print(f"Saved to {args.output}")
except Exception as e:
2023-11-08 22:57:38 +08:00
print(e)
2023-08-16 23:07:58 +08:00
with open("error.txt", "w") as f:
f.write(str(e))