Functions — github.com/apple/ml-fastvlm

CodeHub MCPcopy Create free account

hub / github.com/apple/ml-fastvlm / functions

Functions294 in github.com/apple/ml-fastvlm

⨍ Functions 294◇ Types & classes 53 ↳ Endpoints 11

Functionclear_history

(request: gr.Request)

llava/serve/gradio_web_server.py:122

Methodcls_init_weights

Init. for classification

llava/model/multimodal_encoder/mobileclip/mci.py:1420

llava/model/multimodal_projector/builder.py:13

llava/model/multimodal_encoder/mobileclip_encoder.py:103

llava/model/multimodal_encoder/clip_encoder.py:91

Methodcreate_optimizer

Setup the optimizer. We provide a reasonable default that works well. If you want to use something else, you can pass a tuple in the

llava/train/llava_trainer.py:154

llava/model/multimodal_encoder/mobileclip_encoder.py:99

Functiondownvote_last_response

(state, model_selector, request: gr.Request)

llava/serve/gradio_web_server.py:100

llava/model/multimodal_encoder/mobileclip_encoder.py:95

llava/model/multimodal_encoder/clip_encoder.py:83

Methoddummy_feature

llava/model/multimodal_encoder/mobileclip_encoder.py:91

Methoddummy_feature

llava/model/multimodal_encoder/clip_encoder.py:79

Methodexpand2square

(pil_img, background_color=(122, 116, 104))

llava/conversation.py:148

Methodexpand2square

(pil_img, background_color)

llava/train/train.py:818

Methodexpand2square

(pil_img, background_color)

llava/train/train_qwen.py:942

Functionfastvithd

Instantiate FastViTHD model variant.

llava/model/multimodal_encoder/mobileclip/mci.py:1455

Functionflag_last_response

(state, model_selector, request: gr.Request)

llava/serve/gradio_web_server.py:106

llava/utils.py:88

Functionforward

( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None, position_i

llava/train/llama_flash_attn_monkey_patch.py:16

( self, input_ids: torch.LongTensor = None, attention_mask: Optional[torch.Tensor] = N

llava/model/language_model/llava_qwen.py:57

( self, input_ids: torch.LongTensor = None, attention_mask: Optional[torch.Tensor] = N

llava/model/language_model/llava_llama.py:57

( self, input_ids: Optional[torch.LongTensor] = None, past_key_values: Opt

llava/model/language_model/llava_mpt.py:60

( self, input_ids: torch.LongTensor = None, attention_mask: Optional[torch.Tensor] = N

llava/model/language_model/llava_mistral.py:57

(self, x, *args, **kwargs)

llava/model/multimodal_projector/builder.py:9

llava/model/multimodal_encoder/mobileclip_encoder.py:70

llava/model/multimodal_encoder/clip_encoder.py:58

llava/model/multimodal_encoder/clip_encoder.py:153

A forward function of the model.

llava/model/multimodal_encoder/mobileclip/__init__.py:55

Apply forward pass.

llava/model/multimodal_encoder/mobileclip/mci.py:72

Apply forward pass.

llava/model/multimodal_encoder/mobileclip/mci.py:194

Apply forward pass.

llava/model/multimodal_encoder/mobileclip/mci.py:442

llava/model/multimodal_encoder/mobileclip/mci.py:617

(self, x: torch.Tensor)

llava/model/multimodal_encoder/mobileclip/mci.py:661

(self, x: torch.Tensor)

llava/model/multimodal_encoder/mobileclip/mci.py:739

(self, x: torch.Tensor)

llava/model/multimodal_encoder/mobileclip/mci.py:808

(self, x: torch.Tensor)

llava/model/multimodal_encoder/mobileclip/mci.py:920

(self, x: torch.Tensor)

llava/model/multimodal_encoder/mobileclip/mci.py:992

llava/model/multimodal_encoder/mobileclip/mci.py:1106

llava/model/multimodal_encoder/mobileclip/mci.py:1185

(self, x: Tensor, *args, **kwargs)

llava/model/multimodal_encoder/mobileclip/mci.py:1290

(self, x: torch.Tensor, *args, **kwargs)

llava/model/multimodal_encoder/mobileclip/mci.py:1436

Methodforward_feature

llava/model/multimodal_encoder/clip_encoder.py:147

( self, inputs: Optional[torch.Tensor] = None, images: Optional[torch.Tensor] = None,

llava/model/language_model/llava_llama.py:106

( self, inputs: Optional[torch.Tensor] = None, images: Optional[torch.Tensor] = None,

llava/model/language_model/llava_mistral.py:105

Functiongenerate_stream

(request: Request)

llava/serve/model_worker.py:232

Functiongenerate_stream

(request: Request)

llava/serve/sglang_worker.py:202

Methodget_model

llava/model/language_model/llava_mpt.py:53

Functionget_status

(request: Request)

llava/serve/model_worker.py:248

Functionget_status

(request: Request)

llava/serve/sglang_worker.py:218

(weights, keyword)

llava/model/llava_arch.py:95

Functionget_worker_address

(request: Request)

llava/serve/controller.py:262

Functionheart_beat_controller

llava/serve/controller.py:51

Functionheart_beat_worker

llava/serve/model_worker.py:37

Functionheart_beat_worker

llava/serve/sglang_worker.py:38

Methodhidden_size

llava/model/multimodal_encoder/mobileclip_encoder.py:107

Methodhidden_size

llava/model/multimodal_encoder/clip_encoder.py:98

Methodhidden_size

llava/model/multimodal_encoder/clip_encoder.py:165

Functionhttp_bot

(state, model_selector, temperature, top_p, max_new_tokens, request: gr.Request)

llava/serve/gradio_web_server.py:154

llava/train/train.py:789

llava/train/train_qwen.py:913

Functionlist_models

llava/serve/controller.py:256

Functionload_demo

(url_params, request: gr.Request)

llava/serve/gradio_web_server.py:58

Functionload_demo_refresh_model_list

(request: gr.Request)

llava/serve/gradio_web_server.py:71

Methodload_model

(self, device_map=None)

llava/model/multimodal_encoder/clip_encoder.py:132

Functionload_model_config

( model_name: str, )

llava/model/multimodal_encoder/mobileclip/__init__.py:15

Functionmake_inputs_require_grad

(module, input, output)

llava/train/train.py:987

Functionmake_inputs_require_grad

(module, input, output)

llava/train/train_qwen.py:1109

Methodmodality_lengths

llava/train/train.py:797

Methodmodality_lengths

llava/train/train_qwen.py:921

Methodnum_patches

llava/model/multimodal_encoder/mobileclip_encoder.py:115

Methodnum_patches

llava/model/multimodal_encoder/clip_encoder.py:106

Methodnum_patches_per_side

llava/model/multimodal_encoder/mobileclip_encoder.py:111

Methodnum_patches_per_side

llava/model/multimodal_encoder/clip_encoder.py:102

Functionpipeline

(s, prompt, max_tokens)

llava/serve/sglang_worker.py:45

Methodprepare_inputs_for_generation

(self, input_ids, past_key_values=None, inputs_embeds=None, **kwargs)

llava/model/language_model/llava_qwen.py:145

Methodprepare_inputs_for_generation

(self, input_ids, past_key_values=None, inputs_embeds=None, **kwargs)

llava/model/language_model/llava_llama.py:145

Methodprepare_inputs_for_generation

(self, input_ids, past_key_values=None, inputs_embeds=None, **kwargs)

llava/model/language_model/llava_mpt.py:87

Methodprepare_inputs_for_generation

(self, input_ids, past_key_values=None, inputs_embeds=None, **kwargs)

llava/model/language_model/llava_mistral.py:144

Functionreceive_heart_beat

(request: Request)

llava/serve/controller.py:269

Functionrefresh_all_workers

llava/serve/controller.py:251

Functionregenerate

(state, image_process_mode, request: gr.Request)

llava/serve/gradio_web_server.py:112

Functionregister_worker

(request: Request)

llava/serve/controller.py:243

Functionrelease_model_semaphore

llava/serve/model_worker.py:225

Functionrelease_model_semaphore

llava/serve/sglang_worker.py:195

Methodreparameterize

Following works like `RepVGG: Making VGG-style ConvNets Great Again` - https://arxiv.org/pdf/2101.03697.pdf. We re-parameterize multi-branched

llava/model/multimodal_encoder/mobileclip/mci.py:219

Methodreparameterize

Following works like `RepVGG: Making VGG-style ConvNets Great Again` - https://arxiv.org/pdf/2101.03697.pdf. We re-parameterize multi

llava/model/multimodal_encoder/mobileclip/mci.py:469

Methodreparameterize

Reparameterize mixer and norm into a single convolutional layer for efficient inference.

llava/model/multimodal_encoder/mobileclip/mci.py:819

Functionreplace_llama_attn_with_flash_attn

llava/train/llama_flash_attn_monkey_patch.py:105

Functionupvote_last_response

(state, model_selector, request: gr.Request)

llava/serve/gradio_web_server.py:94

Functionworker_api_generate_stream

(request: Request)

llava/serve/controller.py:277

Functionworker_api_get_status

(request: Request)

llava/serve/controller.py:284

Methodwrap_inst

llava/conversation.py:108

llava/conversation.py:107

Functionxformers_forward

( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None, position_i

llava/train/llama_xformers_attn_monkey_patch.py:23

← previous201–294 of 294, ranked by callers