Datasets:
itazap
/
transformers_code_embeddings_v3

Dataset card Data Studio Files
xet
 Community
1
identifier
stringlengths
24
117
embedding
listlengths
2.56k
2.56k
tokens
listlengths
4
444
table_transformer/modeling_table_transformer.py:TableTransformerEncoder
[ -0.000011108153557870537, 0.03829123452305794, 0.021623285487294197, 0.02015920914709568, -0.00009502421016804874, 0.05067957565188408, 0.0036883470602333546, -0.01745629869401455, 0.006700966507196426, -0.003308250103145838, 0.019483482465147972, 0.022411635145545006, -0.002562134061008692,...
[ "BaseModelOutput", "False", "LayerNorm", "ModelEncoder", "ModelEncoderLayer", "ModelPreTrainedModel", "ModuleList", "None", "True", "_", "__init__", "all_attentions", "attention_mask", "attentions", "class", "config", "create_bidirectional_mask", "d_model", "def", "dropout", ...
table_transformer/modeling_table_transformer.py:TableTransformerDecoder
[ -0.00011335760063957423, 0.06163841485977173, 0.012260195799171925, -0.006130097899585962, -0.0006819030968472362, 0.032168954610824585, 0.030594250187277794, -0.02384551800787449, 0.003655562875792384, -0.015522082336246967, -0.0024885947350412607, 0.02474534884095192, -0.004639753140509129...
[ "False", "LayerNorm", "ModelDecoder", "ModelDecoderLayer", "ModelDecoderOutput", "ModelPreTrainedModel", "ModuleList", "None", "_", "__init__", "all_cross_attentions", "all_hidden_states", "all_self_attns", "and", "attention_mask", "attentions", "auxiliary_loss", "class", "config...
table_transformer/modeling_table_transformer.py:TableTransformerModel
[ -0.00004775000343215652, 0.03170425817370415, 0.00708880415186286, 0.011889097280800343, -0.00020408221462275833, 0.033490415662527084, 0.010493663139641285, -0.016298668459057808, 0.009042412042617798, -0.001269844826310873, 0.004632840398699045, 0.016298668459057808, -0.0003174612065777182...
[ "BaseModelOutput", "Conv2d", "Embedding", "False", "ModelConvEncoder", "ModelConvModel", "ModelDecoder", "ModelEncoder", "ModelModel", "ModelModelOutput", "ModelPreTrainedModel", "None", "True", "__init__", "and", "attention_mask", "attentions", "auto_docstring", "backbone", "b...
table_transformer/modeling_table_transformer.py:TableTransformerForObjectDetection
[ -0.00020076845248695463, 0.0120601961389184, -0.0033250071574002504, 0.004057636018842459, -0.0005741273052990437, 0.03403905779123306, 0.03223566338419914, -0.018259361386299133, 0.005607427563518286, 0.016005119308829308, 0.009467816911637783, 0.03990008682012558, -0.0024233104195445776, ...
[ "Linear", "ModelForObjectDetection", "ModelMLPPredictionHead", "ModelModel", "ModelObjectDetectionOutput", "ModelPreTrainedModel", "None", "__init__", "auto_docstring", "auxiliary_loss", "auxiliary_outputs", "bbox_predictor", "class", "class_labels_classifier", "config", "cross_attenti...
table_transformer/modeling_table_transformer.py:TableTransformerMLPPredictionHead
[ -0.0002314915764145553, -0.0008084397995844483, 0.027921445667743683, 0.042395010590553284, -0.0006731062894687057, 0.04057157039642334, 0.04604189470410347, -0.009630049578845501, -0.0012678615748882294, -0.012992019765079021, 0.021767331287264824, -0.05926184356212616, -0.00289186416193842...
[ "Linear", "ModelMLPPredictionHead", "Module", "ModuleList", "__init__", "class", "def", "else", "enumerate", "for", "forward", "functional", "h", "hidden_dim", "i", "if", "in", "input_dim", "k", "layer", "layers", "n", "nn", "num_layers", "output_dim", "relu", "re...
videoprism/modeling_videoprism.py:BaseModelOutputWithSpatialAndTemporalStates
[ -0.00027354489429853857, 0.0153040224686265, 0.027593616396188736, 0.005159310530871153, -0.0014564907178282738, 0.05217280238866806, 0.03176743909716606, -0.0556509904563427, 0.015535901300609112, 0.017043115571141243, 0.031071802601218224, -0.0033767397981137037, 0.001449244562536478, -0...
[ "ModelModelOutput", "ModelModelOutputWithSpatialAndTemporalStates", "None", "class", "last_spatial_hidden_state", "last_temporal_hidden_state", "r" ]
videoprism/modeling_videoprism.py:VideoPrismClipOutput
[ -0.00019192349282093346, 0.026044728234410286, 0.015012681484222412, 0.002672712318599224, -0.0007712480728514493, 0.042763397097587585, 0.05186199024319649, -0.0243387408554554, 0.014330286532640457, 0.000778356334194541, 0.024111276492476463, 0.034574661403894424, -0.002288865391165018, ...
[ "ModelClipOutput", "ModelOutput", "Model_embeds", "Model_model_output", "None", "class", "def", "else", "for", "getattr", "if", "in", "k", "keys", "logits_per_Model", "logits_per_text", "loss", "not", "r", "return", "self", "text_embeds", "text_model_output", "to_tuple"...
videoprism/modeling_videoprism.py:VideoPrismTubeletEmbeddings
[ -0.00021339273371268064, 0.014620501548051834, -0.006941904779523611, 0.0124670946970582, -0.0006091875839047134, 0.03558788821101189, 0.026180898770689964, -0.04102807492017746, 0.003329280996695161, 0.018813978880643845, 0.013997146859765053, 0.014960513450205326, -0.003300946671515703, ...
[ "Conv3d", "False", "Image", "Iterable", "ModelTubeletEmbeddings", "Module", "Set", "True", "ValueError", "__init__", "and", "automatically", "batch_size", "class", "config", "def", "doesn", "else", "embeddings", "f", "flatten", "forward", "height", "hidden_size", "hid...
videoprism/modeling_videoprism.py:VideoPrismSpatialEmbeddings
[ -0.00017743211355991662, 0.022824302315711975, -0.0050281258299946785, -0.00420893682166934, -0.0007097284542396665, 0.018643613904714584, 0.03706124424934387, -0.02316327765583992, 0.005960306618362665, 0.020790454000234604, 0.014688907191157341, 0.02960379794239998, -0.0012146596563979983,...
[ "Dropout", "False", "ModelSpatialEmbeddings", "ModelTubeletEmbeddings", "Module", "Parameter", "True", "__init__", "and", "antialias", "batch", "bilinear", "channel", "class", "config", "def", "dim", "dropout", "else", "embeddings", "forward", "frames", "functional", "h...
videoprism/modeling_videoprism.py:VideoPrismTemporalEmbeddings
[ -0.0003376822860445827, 0.02496320754289627, 0.020109249278903008, 0.007223150227218866, -0.0014951920602470636, 0.02149609476327896, 0.03721366822719574, -0.03975621983408928, 0.0058363052085042, 0.03305313363671303, 0.008667780086398125, 0.024732066318392754, 0.0001146675058407709, 0.002...
[ "Dropout", "False", "ModelTemporalEmbeddings", "Module", "None", "Parameter", "True", "_", "__init__", "and", "antialias", "batch", "bilinear", "channel", "class", "config", "def", "dim", "dropout", "else", "embeddings", "features", "forward", "frames", "functional", ...
videoprism/modeling_videoprism.py:create_sinusoidal_positions
[ -0.0001407179079251364, 0.03895071521401405, 0.01981308124959469, 0.024316053837537766, -0.0006473023677244782, 0.011370006017386913, 0.05673745647072792, 0, 0.013227482326328754, 0.021389121189713478, 0.014353225938975811, -0.001280532917007804, -0.0007000715704634786, 0.01756159402430057...
[ "Model_sinusoidal_positions", "arange", "cat", "cos", "def", "dim", "dtype", "einsum", "float", "i", "int64", "inv_freq", "j", "num_pos", "return", "sin", "sinusoid_inp", "torch" ]
videoprism/modeling_videoprism.py:VideoPrismTextEmbeddings
[ -0.00020157844119239599, 0.031007714569568634, 0.024670371785759926, 0.01946469582617283, -0.000862897140905261, 0.025802040472626686, 0.030328713357448578, 0.0095625976100564, 0.007921678945422173, -0.007214386016130447, 0.0330447182059288, 0.03508172184228897, 0.0005127872573211789, 0.01...
[ "Embedding", "ModelTextEmbeddings", "Module", "None", "Parameter", "__init__", "arange", "cat", "class", "cls_emb", "config", "create_sinusoidal_positions", "def", "dim", "dtype", "embed_dim", "embeddings", "expand", "forward", "hidden_size", "if", "input_ids", "inputs_em...
videoprism/modeling_videoprism.py:eager_attention_forward
[ -0.00012093633267795667, 0.0348014160990715, 0.028473885729908943, -0.02734396979212761, -0.00046962135820649564, 0.020451480522751808, 0.059207603335380554, -0.019434556365013123, 0.013671984896063805, 0.014575917273759842, 0.020564472302794456, 0.017174724489450455, 0.0031213932670652866, ...
[ "Model_attention_forward", "None", "attention_mask", "attn_output", "attn_weights", "contiguous", "def", "dim", "dropout", "dtype", "float32", "functional", "head_dim", "if", "is", "key", "key_states", "kwargs", "matmul", "module", "nn", "not", "num_key_value_groups", "...
videoprism/modeling_videoprism.py:repeat_kv
[ -0.0002072998322546482, -0.00195862608961761, -0.003717100480571389, -0.009378530085086823, -0.00046463756007142365, 0.03248174116015434, 0.010007577948272228, -0.05787239596247673, 0.011380046606063843, 0.05192503333091736, 0.006519222632050514, -0.021959485486149788, 0.0005682875053025782,...
[ "Model_kv", "None", "batch", "def", "expand", "head_dim", "hidden_states", "if", "n_rep", "num_key_value_heads", "reshape", "return", "shape", "slen" ]
videoprism/modeling_videoprism.py:VideoPrismAttention
[ -0.00011360530334059149, 0.04306081309914589, 0.0405808724462986, -0.0007010063854977489, -0.0003874909598380327, 0.024123072624206543, 0.04554075375199318, -0.0081725362688303, 0.002043134067207575, 0.013865130953490734, 0.027955710887908936, 0.016457797959446907, -0.0010004312498494983, ...
[ "ALL_ATTENTION_FUNCTIONS", "False", "Linear", "ModelAttention", "Module", "None", "__init__", "_attn_implementation", "attention_dropout", "attention_interface", "attention_mask", "attention_probs_dropout_prob", "attn_logit_softcapping", "attn_output", "attn_weights", "class", "confi...
videoprism/modeling_videoprism.py:VideoPrismLayerNorm
[ -0.00008592988160671666, 0.03158694505691528, 0.029781976714730263, 0.04016054421663284, -0.0005499512772075832, 0.035196881741285324, 0.017372818663716316, -0.02684890292584896, 0.01895216666162014, 0.03113570250570774, 0.019516218453645706, 0.007445494178682566, 0.004456015303730965, -0....
[ "F", "LayerNorm", "ModelLayerNorm", "bias", "class", "def", "eps", "forward", "hidden_states", "layer_norm", "nn", "normalized_shape", "return", "self", "weight" ]
videoprism/modeling_videoprism.py:VideoPrismMLP
[ -0.00019217404769733548, 0.02812858670949936, 0.043046984821558, 0.019929159432649612, -0.0005444931448437274, 0.0466911755502224, 0.030747847631573677, -0.027900824323296547, 0.0044982959516346455, -0.009565996937453747, 0.04395803436636925, -0.03484756126999855, 0.001217102282680571, -0....
[ "ACT2FN", "Linear", "ModelMLP", "Module", "__init__", "activation_fn", "class", "config", "def", "fc1", "fc2", "forward", "hidden_act", "hidden_size", "hidden_states", "intermediate_size", "nn", "return", "self", "super" ]
videoprism/modeling_videoprism.py:VideoPrismLayer
[ -0.000020983308786526322, 0.014118361286818981, 0.036898981779813766, 0.005202981643378735, 0.00017929193563759327, 0.05242355540394783, 0.028236722573637962, -0.0027421119157224894, 0.009449739940464497, 0.009449739940464497, 0.010855951346457005, -0.0010335652623325586, 0.00162417395040392...
[ "Dropout", "GradientCheckpointingLayer", "ModelAttention", "ModelLayer", "ModelLayerNorm", "ModelMLP", "None", "_", "__init__", "attention", "attention_mask", "class", "config", "def", "dropout", "eps", "forward", "hidden_dropout_prob", "hidden_size", "hidden_states", "kwargs...
videoprism/modeling_videoprism.py:VideoPrismPreTrainedModel
[ -0.00014538095274474472, 0.05436014011502266, -0.012236670590937138, 0.020187687128782272, -0.0007260236307047307, 0.027067288756370544, 0.017480958253145218, -0.019849345088005066, -0.003157850354909897, 0.0072179436683654785, 0.01770651899278164, 0.014548667706549168, -0.003369313664734363...
[ "Conv3d", "False", "Linear", "Model", "ModelAttention", "ModelConfig", "ModelLayer", "ModelLayerNorm", "ModelMultiheadAttentionPoolingHead", "ModelPreTrainedModel", "ModelSpatialEmbeddings", "ModelTemporalEmbeddings", "ModelTextEmbeddings", "ModelTextModel", "PreTrainedModel", "True", ...
videoprism/modeling_videoprism.py:VideoPrismVisionModel
[ 0, 0.042303942143917084, -0.00793198961764574, 0.029702767729759216, 0.00038675477844662964, 0.039153650403022766, 0.03487825021147728, -0.04365406930446625, 0.015638956800103188, 0.03487825021147728, 0.037353482097387314, 0.021827034652233124, 0.002151763066649437, -0.008269520476460457, ...
[ "BaseModelOutputWithSpatialAndTemporalStates", "False", "Model", "ModelLayer", "ModelLayerNorm", "ModelPreTrainedModel", "ModelSpatialEmbeddings", "ModelTemporalEmbeddings", "ModelVisionConfig", "ModelVisionModel", "ModuleList", "None", "_", "__init__", "auto_docstring", "base_model_pr...
videoprism/modeling_videoprism.py:VideoPrismMultiheadAttentionPoolingHead
[ -0.00028416121494956315, 0.023104630410671234, 0.04735305532813072, 0.017042525112628937, -0.0011151988292112947, 0.011723884381353855, 0.04666677862405777, -0.007720606867223978, -0.00406046723946929, 0.02218959666788578, 0.01223859190940857, 0.007720606867223978, -0.001515526557341218, -...
[ "ALL_ATTENTION_FUNCTIONS", "False", "Linear", "ModelMultiheadAttentionPoolingHead", "Module", "None", "Parameter", "_R_SOFTPLUS_0", "__init__", "_attn_implementation", "attention_dropout", "attention_interface", "attention_mask", "attention_probs_dropout_prob", "attn_output", "attn_wei...
videoprism/modeling_videoprism.py:l2norm
[ -0.000021997486328473315, 0.033171769231557846, 0.03182696923613548, 0.03362003713846207, 0.00018736167112365365, 0.03630964085459709, 0.009077410213649273, 0.011486846022307873, 0.011710979975759983, 0.03989577665925026, 0.015016949735581875, -0.004790855571627617, 0.001645980984903872, 0...
[ "Model", "True", "def", "dim", "eps", "inv_norm", "keepdim", "return", "rsqrt", "sum", "torch", "x" ]
videoprism/modeling_videoprism.py:VideoPrismTextModel
[ 0.000090712659584824, 0.041979026049375534, 0.024450549855828285, 0.015183903276920319, 0.00019363663159310818, 0.036396708339452744, 0.01496061123907566, -0.008150183595716953, 0.008875885047018528, 0.01507225725799799, 0.03974610194563866, 0.0006698780925944448, -0.0010676182573661208, 0...
[ "BaseModelOutputWithPooling", "ModelLayer", "ModelLayerNorm", "ModelPreTrainedModel", "ModelTextConfig", "ModelTextEmbeddings", "ModelTextModel", "ModuleList", "None", "ValueError", "You", "_", "__init__", "_input_embed_layer", "_no_split_modules", "apply_l2norm", "attention_mask", ...
videoprism/modeling_videoprism.py:VideoPrismVideoModel
[ 0, 0.038542091846466064, 0.006302304565906525, 0.028906570747494698, 0.0002871049800887704, 0.04862578213214874, 0.023416562005877495, -0.018262678757309914, 0.0067224581725895405, 0.028682487085461617, 0.02139982581138611, 0.029802897945046425, 0.001687617041170597, 0.005265925545245409, ...
[ "BaseModelOutputWithPooling", "False", "ModelLayer", "ModelLayerNorm", "ModelModelModel", "ModelMultiheadAttentionPoolingHead", "ModelPreTrainedModel", "ModelVisionConfig", "ModelVisionModel", "Model_embeddings", "ModuleList", "_", "__init__", "_from_config", "apply_l2norm", "attention...
videoprism/modeling_videoprism.py:VideoPrismClipModel
[ -0.000011014500159944873, 0.053602445870637894, 0.023674413561820984, 0.020770946517586708, 0.00009422304719919339, 0.031044749543070793, 0.026466207578778267, -0.028141284361481667, 0.010832160711288452, 0.01574571803212166, 0.04042517766356468, 0.03260815516114235, 0.0009910868247970939, ...
[ "False", "ModelClipModel", "ModelClipOutput", "ModelModelModel", "ModelPreTrainedModel", "ModelTextModel", "Model_emb_dim", "Model_embeddings", "Model_embeds", "Model_model", "Model_model_output", "Model_model_outputs", "None", "T", "True", "__init__", "_from_config", "attention_ma...
videoprism/modeling_videoprism.py:VideoPrismForVideoClassification
[ 0, 0.04150530695915222, -0.000833313912153244, 0.03369516879320145, 0.0003835335955955088, 0.040166426450014114, 0.024546150118112564, -0.005132376682013273, 0.0033890423364937305, 0.016624437645077705, 0.03034796752035618, 0.016959158703684807, 0.0016805744962766767, 0.006917551159858704,...
[ "False", "ImageClassifierOutput", "Linear", "Model", "ModelForModelClassification", "ModelLayerNorm", "ModelMultiheadAttentionPoolingHead", "ModelPreTrainedModel", "ModelVisionConfig", "ModelVisionModel", "None", "__init__", "_from_config", "attentions", "auto_docstring", "base_model_p...
video_llama_3/modeling_video_llama_3.py:VideoLlama3VisionRotaryEmbedding
[ -0.000260195869486779, 0.03766918182373047, 0.006000654771924019, -0.0051106056198477745, -0.0013494295999407768, 0.04478957876563072, 0.06982580572366714, -0.008900493383407593, 0.00430669030174613, 0.021131493151187897, -0.005340295843780041, 0.0029141937848180532, -0.0017226760974153876, ...
[ "False", "ModelVisionRotaryEmbedding", "Module", "Tensor", "__init__", "arange", "class", "def", "dim", "dtype", "flatten", "float", "forward", "inv_freq", "nn", "persistent", "position_ids", "register_buffer", "return", "self", "super", "theta", "torch", "unsqueeze" ]
video_llama_3/modeling_video_llama_3.py:VideoLlama3VisionEmbeddings
[ 0.000021584271962638013, 0.005301184952259064, 0.01638038083910942, 0.030741263180971146, 0.0002752268628682941, 0.010714564472436905, 0.0209803506731987, -0.01828768663108349, 0.014697465114295483, 0.006310934666544199, 0.02961931936442852, 0.006395080126821995, 0.00069770886329934, -0.01...
[ "Conv2d", "ModelVisionEmbeddings", "Module", "__init__", "class", "config", "def", "embed_dim", "embeddings", "forward", "hidden_size", "hidden_states", "in_channels", "kernel_size", "nn", "num_channels", "out_channels", "padding", "patch_embedding", "patch_embeds", "patch_si...
video_llama_3/modeling_video_llama_3.py:VideoLlama3VisionMLP
[ -0.00018684289534576237, 0.03735434636473656, 0.031887855380773544, 0.033937789499759674, -0.0004697764234151691, 0.04737624153494835, 0.031432315707206726, -0.033937789499759674, 0.0033169062808156013, 0.0001565921411383897, 0.04942617565393448, -0.02072710543870926, 0.0018079275032505393, ...
[ "ACT2FN", "Linear", "ModelVisionMLP", "Module", "__init__", "activation_fn", "class", "config", "def", "fc1", "fc2", "forward", "hidden_act", "hidden_size", "hidden_states", "intermediate_size", "nn", "return", "self", "super" ]
video_llama_3/modeling_video_llama_3.py:eager_attention_forward
[ -0.000020769030015799217, 0.020362488925457, 0.014366867020726204, -0.018552489578723907, 0, 0.03823622688651085, 0.0527162179350853, -0.03461622819304466, 0.020362488925457, 0.007296558003872633, 0.028054984286427498, 0.019909989088773727, 0.002474607899785042, -0.017760613933205605, -0...
[ "Model_attention_forward", "None", "attention_mask", "attn_output", "attn_weights", "contiguous", "def", "dim", "dropout", "dtype", "float32", "functional", "if", "is", "key", "key_states", "kwargs", "matmul", "module", "nn", "not", "num_key_value_groups", "p", "query",...
video_llama_3/modeling_video_llama_3.py:rotate_half
[ 0, 0.01424520555883646, 0.034771762788295746, 0.003014487447217107, 0.0002628913498483598, 0.028714744374155998, 0.019853554666042328, -0.01873188465833664, 0.014581706374883652, 0.01873188465833664, -0.001486212364397943, -0.01278703473508358, 0.00042237876914441586, 0.010263278149068356,...
[ "Model_half", "cat", "def", "dim", "return", "shape", "torch", "x", "x1", "x2" ]
video_llama_3/modeling_video_llama_3.py:repeat_kv
[ -0.0002072998322546482, -0.00195862608961761, -0.003717100480571389, -0.009378530085086823, -0.00046463756007142365, 0.03248174116015434, 0.010007577948272228, -0.05787239596247673, 0.011380046606063843, 0.05192503333091736, 0.006519222632050514, -0.021959485486149788, 0.0005682875053025782,...
[ "Model_kv", "None", "batch", "def", "expand", "head_dim", "hidden_states", "if", "n_rep", "num_key_value_heads", "reshape", "return", "shape", "slen" ]
video_llama_3/modeling_video_llama_3.py:apply_rotary_pos_emb_vision
[ -0.00007567541615571827, 0.03580022603273392, 0.01869315654039383, -0.0019684459548443556, -0.00036996870767325163, 0.016540611162781715, 0.06027126684784889, -0.009913037531077862, 0.01688048616051674, 0.047809164971113205, 0.00923328660428524, 0.020052658393979073, 0.0014161481522023678, ...
[ "Model_rotary_pos_emb_vision", "cos", "def", "dtype", "float", "k", "k_embed", "orig_q_dtype", "q", "q_embed", "return", "rotate_half", "sin", "unsqueeze" ]
video_llama_3/modeling_video_llama_3.py:VideoLlama3VisionAttention
[ -0.0001208942849189043, 0.03817788511514664, 0.020557323470711708, 0.023155225440859795, -0.0005153449601493776, 0.023494083434343338, 0.03795197978615761, -0.0228163693100214, 0.004376902710646391, 0.028351033106446266, 0.029819412156939507, 0.027334462851285934, 0.00011912939953617752, -...
[ "ALL_ATTENTION_FUNCTIONS", "False", "Linear", "ModelVisionAttention", "Module", "None", "Tensor", "ValueError", "__init__", "_attn_implementation", "and", "append", "apply_rotary_pos_emb_vision", "attention_dropout", "attention_interface", "attention_mask", "attn_output", "attn_out...
video_llama_3/modeling_video_llama_3.py:VideoLlama3VisionEncoderLayer
[ -0.00009837093239184469, 0.03080415353178978, 0.019336914643645287, 0.030129609629511833, 0, 0.04451987147331238, 0.02878052368760109, -0.016413891687989235, 0.009218761697411537, 0.03237808868288994, 0.020910849794745445, 0.029005371034145355, 0.0008466926519758999, -0.01742570661008358, ...
[ "GradientCheckpointingLayer", "LayerNorm", "ModelVisionAttention", "ModelVisionEncoderLayer", "ModelVisionMLP", "Tensor", "_", "__init__", "auto_docstring", "class", "config", "cu_seqlens", "def", "embed_dim", "eps", "forward", "hidden_size", "hidden_states", "kwargs", "layer_n...
video_llama_3/modeling_video_llama_3.py:VideoLlama3VisionEncoder
[ -0.0001258246775250882, 0.028494451195001602, -0.0021962125319987535, 0.047077786177396774, -0.00019621610408648849, 0.03784243389964104, 0.01723182201385498, -0.03288687393069267, 0.010586870834231377, 0.02883232943713665, 0.023876773193478584, 0.01892121694982052, 0.0010840280447155237, ...
[ "BaseModelOutput", "False", "ModelVisionEncoder", "ModelVisionEncoderLayer", "Module", "ModuleList", "Tensor", "_", "__init__", "auto_docstring", "can_return_tuple", "class", "config", "cu_seqlens", "def", "encoder_layer", "for", "forward", "gradient_checkpointing", "hidden_sta...
video_llama_3/modeling_video_llama_3.py:VideoLlama3PreTrainedModel
[ -0.00031900990870781243, 0.045241404324769974, 0.005394167732447386, 0.01740054041147232, -0.001674802042543888, 0.0361931249499321, 0.027724862098693848, -0.012412386015057564, -0.003581611206755042, 0.025752799585461617, 0.006467200815677643, -0.008004249073565006, -0.004060125909745693, ...
[ "Model", "ModelConfig", "ModelPreTrainedModel", "ModelVisionEncoderLayer", "ModelVisionRotaryEmbedding", "PreTrainedModel", "True", "_can_compile_fullgraph", "_init_weights", "_no_split_modules", "_skip_keys_device_placement", "_supports_attention_backend", "_supports_flash_attn", "_suppor...
video_llama_3/modeling_video_llama_3.py:VideoLlama3VisionModel
[ -0.00016826613864395767, 0.045544035732746124, -0.008076774887740612, 0.03589677810668945, -0.0005153150414116681, 0.024791212752461433, 0.035448066890239716, -0.047338876873254776, 0.002664214000105858, 0.03634548559784889, 0.01806056685745716, 0.0045431856997311115, -0.0011918852105736732,...
[ "BaseModelOutput", "False", "LayerNorm", "ModelPreTrainedModel", "ModelVisionAttention", "ModelVisionConfig", "ModelVisionEmbeddings", "ModelVisionEncoder", "ModelVisionEncoderLayer", "ModelVisionModel", "ModelVisionRotaryEmbedding", "__init__", "_can_record_outputs", "append", "attentio...
video_llama_3/modeling_video_llama_3.py:VideoLlama3Projector
[ -0.00026639134739525616, 0.03707309067249298, 0.018879815936088562, 0.04325193911790848, -0.0009940508753061295, 0.03158078342676163, 0.03661540150642395, -0.03959039971232414, -0.002803366631269455, 0.008581734262406826, 0.017049046233296394, 0.005978608038276434, -0.0024887030012905598, ...
[ "GELU", "Linear", "ModelProjector", "Module", "Sequential", "__init__", "class", "config", "def", "forward", "hidden_size", "hidden_states", "in_hidden_size", "nn", "out_hidden_size", "readout", "return", "self", "super", "text_config", "vision_config" ]
video_llama_3/modeling_video_llama_3.py:VideoLlama3ModelOutputWithPast
[ -0.00016389242955483496, 0.010090072639286518, 0.021662304177880287, 0.020294157788157463, -0.0010332348756492138, 0.03192339465022087, 0.039448194205760956, -0.044464729726314545, 0.018697988241910934, -0.0032635971438139677, 0.014935588464140892, 0.0265648253262043, -0.0019382061436772346,...
[ "ModelModelOutputWithPast", "ModelOutput", "Model_hidden_states", "None", "attentions", "class", "hidden_states", "image_hidden_states", "last_hidden_state", "past_key_values", "r" ]
video_llama_3/modeling_video_llama_3.py:VideoLlama3Model
[ -0.000047064837417565286, 0.03726140782237053, -0.007641935255378485, 0.01573011465370655, -0.00020133292127866298, 0.04841751605272293, 0.017626652494072914, -0.04395507276058197, 0.0009482693276368082, 0.017180409282445908, 0.03302208334207535, 0.011937037110328674, 0.0005299152107909322, ...
[ "AutoModel", "False", "Image", "Model", "ModelModel", "ModelModelOutputWithPast", "ModelPreTrainedModel", "ModelProjector", "Model_compression_mask", "Model_embeds", "Model_features", "Model_grid_thw", "Model_hidden_states", "Model_mask", "Model_merge_sizes", "Model_token_id", "None"...
video_llama_3/modeling_video_llama_3.py:VideoLlama3CausalLMOutputWithPast
[ -0.00019051949493587017, 0.02381671778857708, 0.028944719582796097, 0.013446759432554245, -0.0010897002648562193, 0.037149522453546524, 0.03623787686228752, -0.03327503055334091, 0.018574761226773262, -0.003162267617881298, 0.02381671778857708, 0.015839826315641403, -0.0002528389450162649, ...
[ "ModelCausalLMOutputWithPast", "ModelOutput", "Model_hidden_states", "None", "attentions", "class", "hidden_states", "image_hidden_states", "logits", "loss", "past_key_values", "r" ]
video_llama_3/modeling_video_llama_3.py:VideoLlama3ForConditionalGeneration
[ -0.00026622298173606396, 0.026011761277914047, -0.003350859973579645, 0.027147645130753517, -0.0009797005914151669, 0.04180055856704712, 0.021695399656891823, -0.044981036335229874, -0.0048559075221419334, -0.0020587912295013666, 0.03612113371491432, 0.0004792013787664473, 0.0020445925183594...
[ "False", "GenerationMixin", "Linear", "Model", "ModelCausalLMOutputWithPast", "ModelForConditionalGeneration", "ModelModel", "ModelPreTrainedModel", "Model_compression_mask", "Model_grid_thw", "Model_hidden_states", "Model_idx", "Model_mask", "Model_merge_sizes", "Model_nums", "Model_t...
ministral3/modeling_ministral3.py:rotate_half
[ 0, 0.01424520555883646, 0.034771762788295746, 0.003014487447217107, 0.0002628913498483598, 0.028714744374155998, 0.019853554666042328, -0.01873188465833664, 0.014581706374883652, 0.01873188465833664, -0.001486212364397943, -0.01278703473508358, 0.00042237876914441586, 0.010263278149068356,...
[ "Model_half", "cat", "def", "dim", "return", "shape", "torch", "x", "x1", "x2" ]
ministral3/modeling_ministral3.py:apply_rotary_pos_emb
[ -0.00015594261640217155, 0.02687883749604225, 0.02733248844742775, 0.002977086463943124, -0.0005706082447431982, 0.02052772045135498, 0.04604560136795044, -0.0014531017513945699, 0.012078464962542057, 0.035838451236486435, 0.007485245820134878, 0.001984724309295416, -0.0008080662810243666, ...
[ "Model_rotary_pos_emb", "cos", "def", "k", "k_embed", "q", "q_embed", "return", "rotate_half", "sin", "unsqueeze", "unsqueeze_dim" ]
ministral3/modeling_ministral3.py:repeat_kv
[ -0.0002072998322546482, -0.00195862608961761, -0.003717100480571389, -0.009378530085086823, -0.00046463756007142365, 0.03248174116015434, 0.010007577948272228, -0.05787239596247673, 0.011380046606063843, 0.05192503333091736, 0.006519222632050514, -0.021959485486149788, 0.0005682875053025782,...
[ "Model_kv", "None", "batch", "def", "expand", "head_dim", "hidden_states", "if", "n_rep", "num_key_value_heads", "reshape", "return", "shape", "slen" ]
ministral3/modeling_ministral3.py:eager_attention_forward
[ -0.000020769030015799217, 0.020362488925457, 0.014366867020726204, -0.018552489578723907, 0, 0.03823622688651085, 0.0527162179350853, -0.03461622819304466, 0.020362488925457, 0.007296558003872633, 0.028054984286427498, 0.019909989088773727, 0.002474607899785042, -0.017760613933205605, -0...
[ "Model_attention_forward", "None", "attention_mask", "attn_output", "attn_weights", "contiguous", "def", "dim", "dropout", "dtype", "float32", "functional", "if", "is", "key", "key_states", "kwargs", "matmul", "module", "nn", "not", "num_key_value_groups", "p", "query",...
ministral3/modeling_ministral3.py:get_llama_4_attn_scale
[ -0.00018276748596690595, 0.022372160106897354, 0.026574037969112396, -0.025892652571201324, -0.0008304386865347624, 0.037021949887275696, 0.046788476407527924, 0.018283845856785774, 0.012321722693741322, 0.022031467407941818, 0.020555131137371063, 0.02498413808643818, 0.001973179168999195, ...
[ "Model_llama_4_attn_scale", "None", "beta", "def", "floor", "log", "max_position_embeddings", "positions_ids", "return", "scaling", "torch" ]
ministral3/modeling_ministral3.py:Ministral3Attention
[ -0.000055549939133925363, 0.033725179731845856, 0.025743553414940834, -0.004777733702212572, -0.00018970413657370955, 0.029902992770075798, 0.04519174247980118, -0.0014543983852490783, 0.0035692481324076653, 0.008375086821615696, 0.01753709279000759, 0.031027166172862053, -0.0015316852368414...
[ "ALL_ATTENTION_FUNCTIONS", "Linear", "ModelAttention", "Module", "None", "Tensor", "True", "__init__", "_attn_implementation", "apply_rotary_pos_emb", "attention_dropout", "attention_interface", "attention_mask", "attn_output", "attn_weights", "class", "config", "contiguous", "co...
ministral3/modeling_ministral3.py:Ministral3MLP
[ -0.00025481736520305276, 0.025214267894625664, 0.0227853711694479, 0.026023900136351585, -0.0010409560054540634, 0.06060677394270897, 0.03608647361397743, -0.005089118145406246, -0.00044638218241743743, -0.0052626109682023525, 0.028221474960446358, -0.047421328723430634, -0.00168432469945400...
[ "ACT2FN", "Linear", "ModelMLP", "Module", "__init__", "act_fn", "class", "config", "def", "down_proj", "forward", "gate_proj", "hidden_act", "hidden_size", "intermediate_size", "nn", "return", "self", "super", "up_proj", "x" ]
ministral3/modeling_ministral3.py:Ministral3RMSNorm
[ -0.00009835940727498382, 0.04200431704521179, 0.031616151332855225, 0.04855337738990784, -0.00046577368630096316, 0.03839104250073433, 0.02461543306708336, -0.029357856139540672, 0.008355697616934776, 0.04358512535691261, 0.019873009994626045, 0.005194082390516996, 0.002427669009193778, 0....
[ "ModelRMSNorm", "Module", "Parameter", "True", "__init__", "class", "def", "dtype", "eps", "extra_repr", "f", "float32", "forward", "hidden_size", "hidden_states", "input_dtype", "keepdim", "mean", "nn", "ones", "pow", "return", "rsqrt", "self", "shape", "super", ...
ministral3/modeling_ministral3.py:Ministral3DecoderLayer
[ -0.00018933851970359683, 0.04348769411444664, 0.0117212925106287, 0.0033974761608988047, -0.0007573540788143873, 0.03963722288608551, 0.04348769411444664, -0.03352176398038864, 0.0073611983098089695, -0.00600220775231719, -0.003652286948636174, 0.022423341870307922, -0.0011466481955721974, ...
[ "False", "GradientCheckpointingLayer", "ModelAttention", "ModelDecoderLayer", "ModelMLP", "ModelRMSNorm", "None", "Tensor", "_", "__init__", "attention_mask", "class", "config", "def", "eps", "forward", "hidden_size", "hidden_states", "input_layernorm", "kwargs", "layer_idx",...
ministral3/modeling_ministral3.py:Ministral3PreTrainedModel
[ -0.00034938708995468915, 0.027718042954802513, 0.016304731369018555, 0.010831000283360481, -0.0018415611702948809, 0.029232053086161613, 0.030746065080165863, -0.0333082377910614, -0.003843258135020733, 0.004017951432615519, 0.006871279329061508, 0.004250876139849424, -0.004250876139849424, ...
[ "ModelAttention", "ModelConfig", "ModelDecoderLayer", "ModelPreTrainedModel", "PreTrainedModel", "True", "_can_compile_fullgraph", "_can_record_outputs", "_no_split_modules", "_skip_keys_device_placement", "_supports_attention_backend", "_supports_flash_attn", "_supports_flex_attn", "_supp...
ministral3/modeling_ministral3.py:Ministral3RotaryEmbedding
[ -0.0003635344619397074, 0.0484125092625618, 0.00013219435641076416, -0.008989215828478336, -0.0019241622649133205, 0.038306985050439835, 0.04324224218726158, 0.003260793862864375, -0.004758996423333883, 0.023618724197149277, -0.0009070001542568207, -0.000745429249946028, -0.00192416226491332...
[ "False", "ModelRotaryEmbedding", "Module", "None", "ROPE_INIT_FUNCTIONS", "Tensor", "__init__", "and", "arange", "attention_factor", "attention_scaling", "base", "cat", "class", "clone", "compute_default_rope_parameters", "config", "cos", "cpu", "def", "default", "deprecate...
ministral3/modeling_ministral3.py:Ministral3Model
[ -0.0001460221246816218, 0.0464438870549202, -0.0059470827691257, -0.006768346764147282, -0.0009557812009006739, 0.04123310744762421, 0.04213932901620865, -0.01914394274353981, 0.011950804851949215, 0.0057205273769795895, 0.020956387743353844, -0.0008920624386519194, -0.001614208216778934, ...
[ "BaseModelOutputWithPast", "DynamicCache", "Embedding", "False", "ModelDecoderLayer", "ModelModel", "ModelPreTrainedModel", "ModelRMSNorm", "ModelRotaryEmbedding", "ModuleList", "None", "ValueError", "You", "__init__", "and", "arange", "attention_mask", "auto_docstring", "capture...
ministral3/modeling_ministral3.py:Ministral3ForCausalLM
[ -0.0002956408425234258, 0.037842027842998505, 0.0035191944334656, -0.0007551307789981365, -0.0012680498184636235, 0.02564595267176628, 0.039893701672554016, -0.008947587572038174, 0.0022938880138099194, 0.03282681852579117, 0.026557808741927147, 0.0004523661045823246, 0.0011754394508898258, ...
[ "CausalLMOutputWithPast", "GenerationMixin", "Linear", "ModelForCausalLM", "ModelModel", "ModelPreTrainedModel", "None", "__init__", "_fsdp_plan", "_pp_plan", "_tied_weights_keys", "_tp_plan", "attention_mask", "attentions", "auto_docstring", "can_return_tuple", "class", "colwise_g...
ministral3/modeling_ministral3.py:Ministral3ForTokenClassification
[ -0.0001748590439092368, 0.024540623649954796, -0.007157681975513697, -0.024540623649954796, -0.0008379032369703054, 0.035674795508384705, 0.034993112087249756, -0.008521050214767456, 0.0015337889781221747, -0.018746308982372284, 0.03999212756752968, 0.023631710559129715, -0.00431733205914497...
[ "GenericForTokenClassification", "ModelForTokenClassification", "ModelPreTrainedModel", "class", "pass" ]
ministral3/modeling_ministral3.py:Ministral3ForSequenceClassification
[ -0.00020226075139362365, 0.015783434733748436, -0.017940884456038475, 0.02781972475349903, -0.0007238805992528796, 0.028046824038028717, 0.0147614860907197, -0.01044659037142992, 0.002554872538894415, -0.013569212518632412, 0.028614573180675507, 0.008232367224991322, -0.0036335967015475035, ...
[ "GenericForSequenceClassification", "ModelForSequenceClassification", "ModelPreTrainedModel", "class", "pass" ]
ministral3/modeling_ministral3.py:Ministral3ForQuestionAnswering
[ -0.00017194084648508579, 0.02392711117863655, -0.011117077432572842, 0.025394339114427567, -0.0004585089045576751, 0.03521348536014557, 0.04288821667432785, 0.026071522384881973, 0.0049660042859613895, 0.014390125870704651, 0.009367689490318298, 0.019638288766145706, -0.0027933772653341293, ...
[ "GenericForQuestionAnswering", "ModelForQuestionAnswering", "ModelPreTrainedModel", "class", "pass" ]
granite_swa/modeling_granite_swa.py:rotate_half
[ 0, 0.014133960008621216, 0.03477402776479721, 0.002930553164333105, 0.00026290849200449884, 0.028492268174886703, 0.019854847341775894, -0.018733104690909386, 0.014470482245087624, 0.01862093061208725, -0.0018648974364623427, -0.013012217357754707, 0.0003610609855968505, 0.0099835116416215...
[ "Model_half", "cat", "def", "dim", "return", "shape", "torch", "x", "x1", "x2" ]
granite_swa/modeling_granite_swa.py:apply_rotary_pos_emb
[ -0.00015525084745604545, 0.02736673317849636, 0.027593841776251793, 0.002838872605934739, -0.000663586484733969, 0.021348321810364723, 0.04633040353655815, -0.0014549222541972995, 0.013115591369569302, 0.03656468167901039, 0.007892065681517124, 0.0013200758257880807, -0.0007345583289861679, ...
[ "Model_rotary_pos_emb", "cos", "def", "k", "k_embed", "q", "q_embed", "return", "rotate_half", "sin", "unsqueeze", "unsqueeze_dim" ]
granite_swa/modeling_granite_swa.py:repeat_kv
[ -0.0002073117793770507, -0.001958739012479782, -0.003831693669781089, -0.009035934694111347, -0.00046823869342915714, 0.03202609717845917, 0.010179723612964153, -0.0571894571185112, 0.011380702257156372, 0.052156783640384674, 0.006176461465656757, -0.02207513153553009, 0.0006791247869841754,...
[ "Model_kv", "None", "batch", "def", "expand", "head_dim", "hidden_states", "if", "n_rep", "num_key_value_heads", "reshape", "return", "shape", "slen" ]
granite_swa/modeling_granite_swa.py:eager_attention_forward
[ 0.00006307986768661067, 0.03275177627801895, -0.0006716672796756029, -0.012793662957847118, 0.00036071022623218596, 0.031842004507780075, 0.06459378451108932, -0.03593597933650017, 0.02467755414545536, 0.003752807853743434, 0.028771527111530304, 0.017854267731308937, 0.0022033529821783304, ...
[ "F", "Model_attention_forward", "None", "attention_mask", "attn_output", "attn_weights", "contiguous", "def", "dim", "dropout", "dtype", "float32", "if", "is", "key", "key_states", "kwargs", "logsumexp", "lse", "matmul", "module", "not", "num_key_value_groups", "p", "...
granite_swa/modeling_granite_swa.py:GraniteSWAAttention
[ -0.00004743336103274487, 0.03800290450453758, 0.01630302146077156, -0.0025438335724174976, -0.00020290937391109765, 0.02855839766561985, 0.04407437518239021, -0.014728937298059464, 0.0016654380597174168, 0.0006710942252539098, 0.02372370846569538, 0.02529779262840748, -0.0014686774229630828,...
[ "ALL_ATTENTION_FUNCTIONS", "Linear", "ModelAttention", "Module", "None", "Parameter", "Tensor", "True", "__init__", "_attn_implementation", "apply_rotary_pos_emb", "attention_dropout", "attention_interface", "attention_mask", "attention_multiplier", "attn_output", "attn_weights", "...
granite_swa/modeling_granite_swa.py:GraniteSWARMSNorm
[ -0.00010937952902168036, 0.04177592322230339, 0.032291658222675323, 0.04855039715766907, -0.0003740074171219021, 0.03861450031399727, 0.024952644482254982, -0.030033500865101814, 0.008581000380218029, 0.042001739144325256, 0.0195330660790205, 0.005617167800664902, 0.002427519764751196, 0.0...
[ "ModelRMSNorm", "Module", "Parameter", "True", "__init__", "class", "def", "dtype", "eps", "extra_repr", "f", "float32", "forward", "hidden_size", "hidden_states", "input_dtype", "keepdim", "mean", "nn", "ones", "pow", "return", "rsqrt", "self", "shape", "super", ...
granite_swa/modeling_granite_swa.py:GraniteSWAMLP
[ -0.00024346320424228907, 0.025046052411198616, 0.022045142948627472, 0.02608482912182808, -0.0009449979406781495, 0.060479868203401566, 0.03601091355085373, -0.005799834616482258, -0.00044544751290231943, -0.004155105445533991, 0.028162380680441856, -0.04709119349718094, -0.00169522536452859...
[ "ACT2FN", "Linear", "ModelMLP", "Module", "__init__", "act_fn", "class", "config", "def", "down_proj", "forward", "gate_proj", "hidden_act", "hidden_size", "intermediate_size", "nn", "return", "self", "super", "up_proj", "x" ]
granite_swa/modeling_granite_swa.py:GraniteSWADecoderLayer
[ -0.00014368224947247654, 0.040844544768333435, 0.019181139767169952, 0.0008215450798161328, -0.0005676770815625787, 0.039490580558776855, 0.039941899478435516, -0.033397749066352844, 0.004202926065772772, 0.0009379012626595795, -0.005895379465073347, 0.021437743678689003, -0.0009520050371065...
[ "False", "GradientCheckpointingLayer", "ModelAttention", "ModelDecoderLayer", "ModelMLP", "ModelRMSNorm", "None", "Tensor", "_", "__init__", "attention_mask", "class", "config", "def", "eps", "forward", "hidden_size", "hidden_states", "input_layernorm", "kwargs", "layer_idx",...
granite_swa/modeling_granite_swa.py:GraniteSWAPreTrainedModel
[ -0.00017047161236405373, 0.0371171198785305, 0.002910319482907653, 0.002066748682409525, -0.0005834698677062988, 0.0230576042085886, 0.031718265265226364, -0.029018837958574295, -0.0009068386862054467, 0.00008040284592425451, 0.001968332100659609, 0.001349713420495391, -0.0024322960525751114...
[ "False", "ModelAttention", "ModelConfig", "ModelDecoderLayer", "ModelPreTrainedModel", "PreTrainedModel", "True", "_can_compile_fullgraph", "_can_record_outputs", "_compatible_flash_implementations", "_init_weights", "_no_split_modules", "_skip_keys_device_placement", "_supports_attention_...
granite_swa/modeling_granite_swa.py:GraniteSWARotaryEmbedding
[ -0.0003748119343072176, 0.04797592759132385, 0.00013228657189756632, -0.009054280817508698, -0.0019108060514554381, 0.0378633551299572, 0.04374275729060173, 0.0032630686182528734, -0.00473291939124465, 0.024105552583932877, -0.0011317851021885872, -0.0009921492310240865, -0.00199899706058204...
[ "False", "ModelRotaryEmbedding", "Module", "None", "ROPE_INIT_FUNCTIONS", "Tensor", "__init__", "and", "arange", "attention_factor", "attention_scaling", "base", "cat", "class", "clone", "compute_default_rope_parameters", "config", "cos", "cpu", "def", "default", "deprecate...
granite_swa/modeling_granite_swa.py:GraniteSWAModel
[ -0.0002218089357484132, 0.051287949085235596, 0.004207214806228876, -0.005008589010685682, -0.0011448203586041927, 0.05861480161547661, 0.04693763330578804, -0.003033773973584175, 0.007441332098096609, 0.0014811112778261304, 0.013680603355169296, -0.00046150569687597454, -0.00048654864076524...
[ "BaseModelOutputWithPast", "DynamicCache", "Embedding", "False", "ModelDecoderLayer", "ModelModel", "ModelPreTrainedModel", "ModelRMSNorm", "ModelRotaryEmbedding", "ModuleList", "None", "ValueError", "You", "__init__", "and", "append", "arange", "attention_mask", "auto_docstring"...
granite_swa/modeling_granite_swa.py:GraniteSWAForCausalLM
[ -0.000315315235638991, 0.04264059662818909, 0.0031068348325788975, 0.000559372769203037, -0.0012541351607069373, 0.026564862579107285, 0.037852078676223755, -0.0036768962163478136, 0.0027220433112233877, 0.03534381091594696, 0.023144494742155075, 0.0026792888529598713, 0.0017956935334950686,...
[ "CausalLMOutputWithPast", "GenerationMixin", "Linear", "ModelForCausalLM", "ModelModel", "ModelPreTrainedModel", "None", "__init__", "_fsdp_plan", "_pp_plan", "_tied_weights_keys", "_tp_plan", "attention_mask", "attentions", "auto_docstring", "can_return_tuple", "class", "colwise_g...
cosmos3_omni/modeling_cosmos3_omni.py:Cosmos3OmniPreTrainedModel
[ -0.00030907895416021347, 0.04209224134683609, 0.0006073760450817645, 0.01282318215817213, -0.0014663280453532934, 0.026451407000422478, 0.021046120673418045, -0.03289175033569336, -0.004370232578366995, 0.01644587516784668, 0.008855471387505531, 0.000679254881106317, -0.004887760151177645, ...
[ "ModelConfig", "ModelPreTrainedModel", "None", "PreTrainedModel", "True", "_Model_DROPPED_UNIFIED_CHECKPOINT_KEYS", "_can_compile_fullgraph", "_can_record_outputs", "_init_weights", "_keys_to_ignore_on_load_unexpected", "_no_split_modules", "_skip_keys_device_placement", "_supports_attention...
cosmos3_omni/modeling_cosmos3_omni.py:BaseModelOutputWithDeepstackFeatures
[ -0.00030385711579583585, 0.013973788358271122, 0.015371167100965977, 0.022241612896323204, -0.0012954865815117955, 0.05147011950612068, 0.06334783881902695, -0.029810747131705284, 0.00896651390939951, -0.011644823476672173, 0.03120812587440014, 0.01606985554099083, -0.000513100007083267, 0...
[ "ModelModelOutputWithDeepstackFeatures", "ModelModelOutputWithPooling", "None", "class", "deepstack_features", "r" ]
cosmos3_omni/modeling_cosmos3_omni.py:Cosmos3OmniModelOutputWithPast
[ -0.0001941061345860362, 0.017829008400440216, 0.00724662933498621, 0.006383934989571571, -0.0010424219071865082, 0.06993572413921356, 0.05728287622332573, -0.025420714169740677, 0.012250253930687904, 0.0007225061999633908, 0.023350249975919724, 0.023235224187374115, -0.0014593906234949827, ...
[ "BaseModelOutputWithPast", "ModelModelOutputWithPast", "None", "class", "r", "rope_deltas" ]
cosmos3_omni/modeling_cosmos3_omni.py:Cosmos3OmniModel
[ -0.00012053254613419995, 0.039189793169498444, -0.015090322121977806, 0.0020974422805011272, -0.0004117462085559964, 0.06216311827301979, 0.028040971606969833, -0.0488746277987957, 0.008727611973881721, 0.03040587343275547, 0.037838421761989594, 0.016441695392131805, 0.0017244071932509542, ...
[ "AutoModel", "False", "H_grid", "Image", "M", "ModelModel", "ModelModelOutputWithPast", "ModelPreTrainedModel", "Multimodal", "None", "Please", "RoPE", "T_grid", "Tensor", "True", "ValueError", "Video", "W_grid", "You", "_", "__init__", "accepts_loss_kwargs", "accepts_pre...
cosmos3_omni/modeling_cosmos3_omni.py:Cosmos3OmniCausalLMOutputWithPast
[ -0.00028211696189828217, 0.022714966908097267, 0.00961017794907093, 0.003625658107921481, -0.0015434528468176723, 0.05731160566210747, 0.051720231771469116, -0.028655802831053734, 0.01304654497653246, 0.010425587184727192, 0.01560925878584385, 0.02143360860645771, -0.001863792072981596, 0....
[ "CausalLMOutputWithPast", "ModelCausalLMOutputWithPast", "None", "class", "r", "rope_deltas" ]
cosmos3_omni/modeling_cosmos3_omni.py:Cosmos3OmniForConditionalGeneration
[ -0.0002540634013712406, 0.03116511180996895, -0.00818648748099804, 0.011856292374432087, -0.0007868908578529954, 0.03997264429926872, 0.03319761902093887, -0.05216768756508827, -0.0006951457471586764, 0.0088075315579772, 0.0406501479446888, 0.005476478021591902, 0.0007445469382219017, 0.00...
[ "False", "GenerationMixin", "Linear", "ModelCausalLMOutputWithPast", "ModelForConditionalGeneration", "ModelModel", "ModelPreTrainedModel", "None", "Tensor", "__init__", "_expand_dict_for_generation", "_expand_dict_for_generation_visual", "_expand_inputs_for_generation", "_get_image_nums_a...
jetmoe/modeling_jetmoe.py:JetMoeRMSNorm
[ -0.00010937952902168036, 0.04177592322230339, 0.032291658222675323, 0.04855039715766907, -0.0003740074171219021, 0.03861450031399727, 0.024952644482254982, -0.030033500865101814, 0.008581000380218029, 0.042001739144325256, 0.0195330660790205, 0.005617167800664902, 0.002427519764751196, 0.0...
[ "ModelRMSNorm", "Module", "Parameter", "True", "__init__", "class", "def", "dtype", "eps", "extra_repr", "f", "float32", "forward", "hidden_size", "hidden_states", "input_dtype", "keepdim", "mean", "nn", "ones", "pow", "return", "rsqrt", "self", "shape", "super", ...
jetmoe/modeling_jetmoe.py:JetMoeRotaryEmbedding
[ -0.0003748119343072176, 0.04797592759132385, 0.00013228657189756632, -0.009054280817508698, -0.0019108060514554381, 0.0378633551299572, 0.04374275729060173, 0.0032630686182528734, -0.00473291939124465, 0.024105552583932877, -0.0011317851021885872, -0.0009921492310240865, -0.00199899706058204...
[ "False", "ModelRotaryEmbedding", "Module", "None", "ROPE_INIT_FUNCTIONS", "Tensor", "__init__", "and", "arange", "attention_factor", "attention_scaling", "base", "cat", "class", "clone", "compute_default_rope_parameters", "config", "cos", "cpu", "def", "default", "deprecate...
jetmoe/modeling_jetmoe.py:JetMoeParallelExperts
[ -0.0002687142405193299, 0.01731160283088684, -0.017425494268536568, 0.0394066758453846, -0.000793685729149729, 0.04737912490963936, 0.03963445872068405, -0.01218645740300417, -0.007032838650047779, -0.005808498244732618, 0.00950999278575182, -0.03462320566177368, -0.0015090706292539835, 0....
[ "F", "ModelParallelExperts", "Module", "Parameter", "__init__", "append", "cat", "class", "def", "dim", "empty", "expert_size", "for", "forward", "i", "in", "input_list", "input_size", "inputs", "linear", "nn", "num_experts", "output_list", "output_size", "range", "...
jetmoe/modeling_jetmoe.py:JetMoeTopKGating
[ -0.00023362308274954557, 0.011698988266289234, -0.01940320059657097, 0.0060777668841183186, -0.0009986941004171968, 0.031958211213350296, 0.04748077318072319, -0.015294287353754044, -0.002111524809151888, -0.008217825554311275, 0.009986940771341324, -0.041545674204826355, -0.0018832517089322...
[ "Linear", "ModelTopKGating", "Module", "_", "__init__", "batch_gates", "batch_index", "class", "def", "device", "dim", "div", "dtype", "expert_size", "flatten", "float", "forward", "gates", "hidden_states", "index_sorted_experts", "input_size", "layer", "logits", "long"...
jetmoe/modeling_jetmoe.py:JetMoeMoE
[ -0.00042667306843213737, 0.03522730991244316, -0.004166989121586084, 0.006265259813517332, -0.0016697508981451392, 0.06856322288513184, 0.03995580971240997, -0.023287853226065636, -0.003930564038455486, -0.006442578509449959, 0.0263613760471344, -0.040901508182287216, 0.0005836740019731224, ...
[ "ACT2FN", "ModelMoE", "ModelParallelExperts", "ModelTopKGating", "Module", "None", "Parameter", "_", "__init__", "activation", "activation_function", "batch_gates", "batch_index", "bias", "bsz", "chunk", "chunked_hidden_states", "class", "config", "def", "device", "dim", ...
jetmoe/modeling_jetmoe.py:JetMoeMoA
[ -0.000506267708260566, 0.0216007549315691, -0.020520716905593872, 0.006690233945846558, -0.0019050666596740484, 0.06864240020513535, 0.028200985863804817, -0.0024300848599523306, -0.008460295386612415, 0.011820413172245026, 0.024240847676992416, -0.038401342928409576, -0.0010800377931445837,...
[ "ModelMoA", "ModelParallelExperts", "ModelTopKGating", "Module", "None", "NotImplementedError", "Parameter", "This", "__init__", "and", "batch_gates", "batch_index", "bias", "bsz", "call", "class", "config", "def", "device", "doesn", "dtype", "emb_size", "empty", "exper...
jetmoe/modeling_jetmoe.py:rotate_half
[ 0, 0.014133960008621216, 0.03477402776479721, 0.002930553164333105, 0.00026290849200449884, 0.028492268174886703, 0.019854847341775894, -0.018733104690909386, 0.014470482245087624, 0.01862093061208725, -0.0018648974364623427, -0.013012217357754707, 0.0003610609855968505, 0.0099835116416215...
[ "Model_half", "cat", "def", "dim", "return", "shape", "torch", "x", "x1", "x2" ]
jetmoe/modeling_jetmoe.py:apply_rotary_pos_emb
[ -0.00015525084745604545, 0.02736673317849636, 0.027593841776251793, 0.002838872605934739, -0.000663586484733969, 0.021348321810364723, 0.04633040353655815, -0.0014549222541972995, 0.013115591369569302, 0.03656468167901039, 0.007892065681517124, 0.0013200758257880807, -0.0007345583289861679, ...
[ "Model_rotary_pos_emb", "cos", "def", "k", "k_embed", "q", "q_embed", "return", "rotate_half", "sin", "unsqueeze", "unsqueeze_dim" ]
jetmoe/modeling_jetmoe.py:repeat_kv
[ -0.0002073117793770507, -0.001958739012479782, -0.003831693669781089, -0.009035934694111347, -0.00046823869342915714, 0.03202609717845917, 0.010179723612964153, -0.0571894571185112, 0.011380702257156372, 0.052156783640384674, 0.006176461465656757, -0.02207513153553009, 0.0006791247869841754,...
[ "Model_kv", "None", "batch", "def", "expand", "head_dim", "hidden_states", "if", "n_rep", "num_key_value_heads", "reshape", "return", "shape", "slen" ]
jetmoe/modeling_jetmoe.py:eager_attention_forward
[ -0.000020655716070905328, 0.020133469253778458, 0.01408211700618267, -0.01877615600824356, 0, 0.03845718875527382, 0.05293518677353859, -0.03461147099733353, 0.020359687507152557, 0.007861101999878883, 0.02805112488567829, 0.020246578380465508, 0.002516683656722307, -0.01832371950149536, ...
[ "Model_attention_forward", "None", "attention_mask", "attn_output", "attn_weights", "contiguous", "def", "dim", "dropout", "dtype", "float32", "functional", "if", "is", "key", "key_states", "kwargs", "matmul", "module", "nn", "not", "num_key_value_groups", "p", "query",...
jetmoe/modeling_jetmoe.py:JetMoeAttention
[ -0.00038803473580628633, 0.03754281625151634, 0.0025067408569157124, -0.009152518585324287, -0.0014136852696537971, 0.0561976321041584, 0.04873570427298546, -0.009210814721882343, -0.0082780746743083, -0.0011732130078598857, 0.02017051912844181, 0.0165561493486166, -0.001588574144989252, -...
[ "ALL_ATTENTION_FUNCTIONS", "Instantiating", "Linear", "ModelAttention", "ModelMoA", "Module", "None", "Please", "True", "__class__", "__init__", "__name__", "_attn_implementation", "a", "and", "apply_rotary_pos_emb", "attention_dropout", "attention_interface", "attention_mask", ...
jetmoe/modeling_jetmoe.py:JetMoeDecoderLayer
[ -0.00019376467389520258, 0.05186876654624939, 0.0119479363784194, 0.0014085541479289532, -0.0007750586955808103, 0.04031720757484436, 0.0450737327337265, -0.028199395164847374, 0.007078161463141441, -0.0058040926232934, 0.003510768059641123, 0.014382824301719666, -0.0016775242984294891, 0....
[ "False", "GradientCheckpointingLayer", "ModelAttention", "ModelDecoderLayer", "ModelMoE", "ModelRMSNorm", "None", "Tensor", "_", "__init__", "attention_mask", "class", "config", "def", "forward", "hidden_size", "hidden_states", "input_layernorm", "kwargs", "layer_idx", "mlp",...
jetmoe/modeling_jetmoe.py:JetMoePreTrainedModel
[ -0.0004239208356011659, 0.03899337723851204, -0.014152715913951397, 0.0014681241009384394, -0.0017690894892439246, 0.04815446957945824, 0.034765180200338364, -0.006782733369618654, -0.007135082967579365, -0.0073993452824652195, 0.009395994246006012, -0.01027686893939972, -0.00387584767304360...
[ "False", "ModelAttention", "ModelConfig", "ModelDecoderLayer", "ModelMoA", "ModelMoE", "ModelParallelExperts", "ModelPreTrainedModel", "ModelTopKGating", "OutputRecorder", "PreTrainedModel", "True", "_can_compile_fullgraph", "_can_record_outputs", "_init_weights", "_no_split_modules", ...
jetmoe/modeling_jetmoe.py:JetMoeModel
[ -0.0001949134311871603, 0.05491040647029877, -0.004984087776392698, -0.006322671193629503, -0.0010751389199867845, 0.0471637099981308, 0.039872702211141586, -0.013442796654999256, 0.008487189188599586, -0.004215114284306765, 0.020961647853255272, -0.006778359413146973, -0.001765791093930602,...
[ "DynamicCache", "Embedding", "False", "ModelDecoderLayer", "ModelModel", "ModelPreTrainedModel", "ModelRMSNorm", "ModelRotaryEmbedding", "ModuleList", "MoeModelOutputWithPast", "None", "ValueError", "You", "__init__", "_attn_implementation", "and", "arange", "attention_mask", "au...
jetmoe/modeling_jetmoe.py:load_balancing_loss_func
[ -0.00025320221902802587, 0.01882968656718731, 0.00164046511054039, -0.028872186318039894, -0.0009414842934347689, 0.05774437263607979, 0.03811585158109665, -0.009414843283593655, 0.0002219977177446708, -0.022823862731456757, 0.03195340558886528, -0.005905674304813147, -0.00099854392465204, ...
[ "Model_balancing_loss_func", "None", "_", "attention_mask", "batch_size", "cat", "compute_device", "concatenated_gate_logits", "def", "device", "dim", "else", "expand", "expert_attention_mask", "expert_mask", "float", "for", "functional", "gate_logits", "if", "in", "is", ...
jetmoe/modeling_jetmoe.py:JetMoeForCausalLM
[ -0.0004201026458758861, 0.04516012594103813, 0.004481094889342785, -0.017575202509760857, -0.001600390998646617, 0.04888467118144035, 0.03910773620009422, -0.00744909280911088, -0.002967997919768095, -0.001105724717490375, 0.027235744521021843, 0.00270611559972167, -0.0004473820445127785, ...
[ "False", "GenerationMixin", "Linear", "ModelForCausalLM", "ModelModel", "ModelPreTrainedModel", "MoeCausalLMOutputWithPast", "None", "__init__", "_tied_weights_keys", "attention_mask", "attentions", "auto_docstring", "aux_loss", "aux_loss_coef", "can_return_tuple", "class", "config...
jetmoe/modeling_jetmoe.py:JetMoeForSequenceClassification
[ -0.00028425012715160847, 0.024524668231606483, -0.011053371243178844, 0.024064110592007637, -0.001129804295487702, 0.035002343356609344, 0.03246928006410599, -0.004317723214626312, 0.010650384239852428, -0.012319903820753098, 0.027518289163708687, 0.010477675125002861, -0.005785749293863773,...
[ "GenericForSequenceClassification", "ModelForSequenceClassification", "ModelPreTrainedModel", "class" ]
nanochat/modeling_nanochat.py:NanoChatRMSNorm
[ -0.00011879759404109791, 0.041079144924879074, 0.024057399481534958, 0.06173219159245491, -0.000461005576653406, 0.05469653755426407, 0.011177612468600273, -0.0155465267598629, 0.008454132825136185, 0.038809578865766525, 0.016340874135494232, 0.013220221735537052, 0.0015532341785728931, 0....
[ "ModelRMSNorm", "Module", "True", "__init__", "_norm", "class", "def", "eps", "extra_repr", "f", "float", "forward", "keepdim", "mean", "nn", "pow", "return", "rsqrt", "self", "super", "torch", "type_as", "x" ]
nanochat/modeling_nanochat.py:NanoChatRotaryEmbedding
[ -0.0003748119343072176, 0.04797592759132385, 0.00013228657189756632, -0.009054280817508698, -0.0019108060514554381, 0.0378633551299572, 0.04374275729060173, 0.0032630686182528734, -0.00473291939124465, 0.024105552583932877, -0.0011317851021885872, -0.0009921492310240865, -0.00199899706058204...
[ "False", "ModelRotaryEmbedding", "Module", "None", "ROPE_INIT_FUNCTIONS", "Tensor", "__init__", "and", "arange", "attention_factor", "attention_scaling", "base", "cat", "class", "clone", "compute_default_rope_parameters", "config", "cos", "cpu", "def", "default", "deprecate...
nanochat/modeling_nanochat.py:apply_rotary_pos_emb
[ -0.00015525084745604545, 0.02736673317849636, 0.027593841776251793, 0.002838872605934739, -0.000663586484733969, 0.021348321810364723, 0.04633040353655815, -0.0014549222541972995, 0.013115591369569302, 0.03656468167901039, 0.007892065681517124, 0.0013200758257880807, -0.0007345583289861679, ...
[ "Model_rotary_pos_emb", "cos", "def", "k", "k_embed", "q", "q_embed", "return", "rotate_half", "sin", "unsqueeze", "unsqueeze_dim" ]
nanochat/modeling_nanochat.py:repeat_kv
[ -0.0002073117793770507, -0.001958739012479782, -0.003831693669781089, -0.009035934694111347, -0.00046823869342915714, 0.03202609717845917, 0.010179723612964153, -0.0571894571185112, 0.011380702257156372, 0.052156783640384674, 0.006176461465656757, -0.02207513153553009, 0.0006791247869841754,...
[ "Model_kv", "None", "batch", "def", "expand", "head_dim", "hidden_states", "if", "n_rep", "num_key_value_heads", "reshape", "return", "shape", "slen" ]
nanochat/modeling_nanochat.py:eager_attention_forward
[ -0.000020655716070905328, 0.020133469253778458, 0.01408211700618267, -0.01877615600824356, 0, 0.03845718875527382, 0.05293518677353859, -0.03461147099733353, 0.020359687507152557, 0.007861101999878883, 0.02805112488567829, 0.020246578380465508, 0.002516683656722307, -0.01832371950149536, ...
[ "Model_attention_forward", "None", "attention_mask", "attn_output", "attn_weights", "contiguous", "def", "dim", "dropout", "dtype", "float32", "functional", "if", "is", "key", "key_states", "kwargs", "matmul", "module", "nn", "not", "num_key_value_groups", "p", "query",...
nanochat/modeling_nanochat.py:rotate_half
[ 0, 0.02056076005101204, 0.03415558114647865, 0.0009058531722985208, 0.0002668404777068645, 0.03213321045041084, 0.019661929458379745, -0.012976872734725475, 0.015055419877171516, 0.016403665766119957, -0.003595324233174324, -0.016291312873363495, 0.0018959717126563191, 0.013089226558804512...
[ "Model_half", "cat", "def", "dim", "return", "shape", "torch", "x", "x1", "x2" ]