Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/YihuaJerry/ICM-LoRA
/ functions
Functions
45 in github.com/YihuaJerry/ICM-LoRA
⨍
Functions
45
◇
Types & classes
5
↓ 5 callers
Function
load_vectors_from_dir
(directory, label, keyword)
ICL/hidden_state/plot_cluster.py:8
↓ 5 callers
Function
load_vectors_from_dir
(directory, label, keyword)
ICL/hidden_state/plot.py:8
↓ 4 callers
Method
device
(self)
model/CVAE_design.py:127
↓ 2 callers
Method
decode
(self, z, **kwargs)
model/VAEdesign.py:91
↓ 2 callers
Method
decode
解码器:接收潜在变量 z 和类别向量 c,返回重建结果。
model/CVAE_design.py:99
↓ 2 callers
Function
filter_annotations
(input_path, output_path, category)
data/utils/filter_one.py:6
↓ 2 callers
Method
pad_sequence
在序列末尾添加零以调整长度到 self.adjusted_input_length。
model/VAEdesign.py:57
↓ 2 callers
Method
pad_sequence
在序列末尾添加零以调整长度到 self.adjusted_input_length。
model/CVAE_design.py:60
↓ 1 callers
Method
_load_entries
(self)
data/florence_detection_dataset.py:31
↓ 1 callers
Method
encode
(self, input, **kwargs)
model/VAEdesign.py:70
↓ 1 callers
Method
encode
编码器:接收输入 x 和类别向量 c,返回潜在变量的均值和方差。
model/CVAE_design.py:74
↓ 1 callers
Method
encode_decode
(self, input, **kwargs)
model/VAEdesign.py:113
↓ 1 callers
Function
extract_classes
(dataset: DetectionDataset)
test.py:70
↓ 1 callers
Function
extract_classes
(dataset: DetectionDataset)
ICL/last_time_step_icl.py:79
↓ 1 callers
Function
main
()
train_lora/train_lora_arg.py:139
↓ 1 callers
Function
main
()
utils/ReconstructLora_cvae.py:13
↓ 1 callers
Function
main
()
data/utils/filter_one.py:34
↓ 1 callers
Function
parse_args
Parse command-line arguments.
TrainScript_CVAE.py:296
↓ 1 callers
Function
reformat
(source_path, target_path)
utils/reformat_lora_param.py:6
↓ 1 callers
Method
reparameterize
(self, mu, log_var, **kwargs)
model/VAEdesign.py:108
↓ 1 callers
Method
reparameterize
(self, mu, log_var)
model/CVAE_design.py:94
↓ 1 callers
Function
train
Main training loop.
TrainScript_CVAE.py:91
↓ 1 callers
Function
train_model
(train_loader, val_loader, model, processor, epochs=10, lr=1e-6)
train_lora/train_lora_arg.py:70
↓ 1 callers
Function
train_with_dataset
(dataset_name: str, lora_r: int, download_location: str)
train_lora/train_lora_arg.py:20
Method
__getitem__
Load and return a data sample.
TrainScript_CVAE.py:64
Method
__getitem__
(self, idx: int)
data/florence_detection_dataset.py:42
Method
__getitem__
(self, idx)
data/florence_detection_dataset.py:62
Method
__init__
Args: data_dir (str): Directory containing the normalized data files.
TrainScript_CVAE.py:51
Method
__init__
(self, latent_dim, input_length = 1929928, kernel_size=7, divide_slice_length=4096,kld_weight=0.005)
model/VAEdesign.py:7
Method
__init__
(self, latent_dim, condition_dim = 512, input_length=1929928, kernel_size=7, divide_slice_length=4096, kld_wei
model/CVAE_design.py:6
Method
__init__
(self, jsonl_file_path: str, image_directory_path: str)
data/florence_detection_dataset.py:26
Method
__init__
(self, jsonl_file_path: str, image_directory_path: str)
data/florence_detection_dataset.py:56
Method
__len__
(self)
TrainScript_CVAE.py:61
Method
__len__
(self)
data/florence_detection_dataset.py:39
Method
__len__
(self)
data/florence_detection_dataset.py:59
Function
collate_fn
(batch)
test.py:41
Function
collate_fn
(batch)
train_lora/train_lora_arg.py:32
Function
collate_fn
(batch)
ICL/last_time_step_icl.py:48
Method
device
(self)
model/VAEdesign.py:140
Method
forward
(self, x, **kwargs)
model/VAEdesign.py:124
Method
forward
前向传播:接收输入 x 和类别向量 c,返回总损失和各部分损失。
model/CVAE_design.py:111
Function
hook
(grad)
TrainScript_CVAE.py:88
Function
print_grad
(name)
TrainScript_CVAE.py:87
Function
replace_hidden_state_hook_decoder
(module, input, output)
ICL/last_time_step_icl.py:103
Method
sample
(self, batch=1)
model/VAEdesign.py:119