Functions — github.com/csslc/PiSA-SR

hub / github.com/csslc/PiSA-SR / functions

Functions373 in github.com/csslc/PiSA-SR

⨍ Functions 373◇ Types & classes 67

Methodforward

Args: x: input features with shape of (num_windows*B, N, C) mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh

ram/models/swin_transformer.py:116

Methodforward

(self, x, condition=None)

ram/models/swin_transformer.py:247

Methodforward

x: B, H*W, C

ram/models/swin_transformer.py:347

Methodforward

(self, x, condition=None)

ram/models/swin_transformer.py:428

Methodforward

(self, x)

ram/models/swin_transformer.py:480

Methodforward

(self, x, idx_to_group_img=None, image_atts=None, condition=None, **kwargs)

ram/models/swin_transformer.py:599

Methodforward

call function as forward Args: image: type: torch.Tensor shape: batch_size * 3 * 384 * 384 caption: type: l

ram/models/tag2text_lora.py:141

Methodforward

(self, x)

ram/models/swin_transformer_lora.py:31

Methodforward

Args: x: input features with shape of (num_windows*B, N, C) mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh

ram/models/swin_transformer_lora.py:122

Methodforward

(self, x)

ram/models/swin_transformer_lora.py:242

Methodforward

x: B, H*W, C

ram/models/swin_transformer_lora.py:316

Methodforward

(self, x)

ram/models/swin_transformer_lora.py:397

Methodforward

(self, x)

ram/models/swin_transformer_lora.py:449

Methodforward

(self, x, idx_to_group_img=None, image_atts=None, **kwargs)

ram/models/swin_transformer_lora.py:568

Methodfuse_qkv_projections

Enables fused QKV projections. For self-attention modules, all projection matrices (i.e., query, key, value) are fused. For cross-att

src/models/autoencoder_kl.py:453

Methodfuse_qkv_projections

Enables fused QKV projections. For self-attention modules, all projection matrices (i.e., query, key, value) are fused. For cross-att

src/models/unet_2d_condition.py:803

Methodgenerate

(self, image, sample=False, num_beams=3, m

ram/models/tag2text_lora.py:280

Methodgenerate_image_embeds

(self, image, condition=False )

ram/models/tag2text.py:230

Methodgenerate_image_embeds

(self, image, condition=False )

ram/models/ram_lora.py:183

Methodgenerate_image_embeds

(self, image, condition=False )

ram/models/tag2text_lora.py:230

Methodgenerate_tag

(self, image, threshold=0.68, tag_input=None,

ram/models/ram_lora.py:192

Methodgenerate_tag_openset

(self, image, threshold=0.68, tag_input=None,

ram/models/ram_lora.py:283

Functionget_PR

( pred_file: str, gt_file: str, taglist: List[str] )

ram/utils/metrics.py:61

Methodget_attention_map

(self)

ram/models/bert_lora.py:189

Methodget_attention_map

(self)

ram/models/vit.py:67

Methodget_attention_map

(self)

ram/models/bert.py:184

Methodget_attn_gradients

(self)

ram/models/bert_lora.py:183

Methodget_attn_gradients

(self)

ram/models/vit.py:61

Methodget_attn_gradients

(self)

ram/models/bert.py:178

Functionget_device_for

(task)

src/my_utils/devices.py:37

Methodget_input_embeddings

(self)

ram/models/bert_lora.py:674

Methodget_input_embeddings

(self)

ram/models/bert.py:669

Functionget_mAP

( preds: ndarray, gt_file: str, taglist: List[str] )

ram/utils/metrics.py:7

Methodget_output_embeddings

(self)

ram/models/bert_lora.py:903

Methodget_output_embeddings

(self)

ram/models/bert.py:898

Functionget_transform

(image_size=384)

ram/transform.py:7

Methodgroup_norm_func

(x, mean=mean, var=var, weight=weight, bias=bias)

src/my_utils/vaehook.py:531

Functioninference_ram

(image, model)

ram/inference.py:33

Functioninference_ram_openset

(image, model)

ram/inference.py:41

Functioninference_tag2text

(image, model, input_tag="None")

ram/inference.py:8

Functioninplace_nonlinearity

(x)

src/my_utils/vaehook.py:129

Functioninterpolate_relative_pos_embed

(rel_pos_bias, dst_num_pos, param_name='')

ram/models/swin_transformer_lora.py:602

Functionload_checkpoint_swinlarge_condition

(model, url_or_filename, kwargs)

ram/models/utils.py:204

Methodload_pretrained

(self, checkpoint_path, prefix='')

ram/models/vit.py:197

Methodmerge_and_unload

( self, progressbar: bool = False, safe_merge: bool = False, adapter_names: Optional[list[str]] = None

src/models/unet_2d_condition.py:1216

Methodno_weight_decay

(self)

ram/models/vit.py:177

Methodno_weight_decay

(self)

ram/models/swin_transformer.py:592

Methodno_weight_decay

(self)

ram/models/swin_transformer_lora.py:561

Methodno_weight_decay_keywords

(self)

ram/models/swin_transformer.py:596

Methodno_weight_decay_keywords

(self)

ram/models/swin_transformer_lora.py:565

Functionperfcount

(fn)

src/my_utils/vaehook.py:431

Methodprepare_inputs_for_generation

(self, input_ids, past=None, attention_mask=None, **model_kwargs)

ram/models/bert_lora.py:1015

Methodprepare_inputs_for_generation

(self, input_ids, past=None, attention_mask=None, **model_kwargs)

ram/models/bert.py:1010

Functionram

(pretrained='', pretrained_condition='', **kwargs)

ram/models/ram_lora.py:328

Functionrandn

(seed, shape)

src/my_utils/devices.py:83

Functionrandn_without_seed

(shape)

src/my_utils/devices.py:88

Methodsave_attn_gradients

(self, attn_gradients)

ram/models/bert_lora.py:180

Methodsave_attn_gradients

(self, attn_gradients)

ram/models/vit.py:58

Methodsave_attn_gradients

(self, attn_gradients)

ram/models/bert.py:175

Methodset_default_attn_processor

Disables custom attention processors and sets the default attention implementation.

src/models/autoencoder_kl.py:221

Methodset_default_attn_processor

Disables custom attention processors and sets the default attention implementation.

src/models/unet_2d_condition.py:687

Methodset_input_embeddings

(self, value)

ram/models/bert_lora.py:677

Methodset_input_embeddings

(self, value)

ram/models/bert.py:672

Methodset_output_embeddings

(self, new_embeddings)

ram/models/bert_lora.py:906

Methodset_output_embeddings

(self, new_embeddings)

ram/models/bert.py:901

Functiontag2text

(pretrained='', **kwargs)

ram/models/tag2text.py:409

Functiontag2text

(pretrained='', **kwargs)

ram/models/tag2text_lora.py:409

Functiontest_for_nans

(x, where)

src/my_utils/devices.py:107

Functiontorch_gc

()

src/my_utils/devices.py:41

Methodunfuse_qkv_projections

Disables the fused QKV projection if enabled. <Tip warning={true}> This API is 🧪 experimental. </Tip>

src/models/autoencoder_kl.py:477

Methodunfuse_qkv_projections

Disables the fused QKV projection if enabled. <Tip warning={true}> This API is 🧪 experimental. </Tip>

src/models/unet_2d_condition.py:826

Functionwithout_autocast

(disable=False)

src/my_utils/devices.py:99

Functionwrapper

(*args, **kwargs)

src/my_utils/vaehook.py:432

← previous301–373 of 373, ranked by callers