MCPcopy Create free account

hub / github.com/BICLab/SpikingBrain-7B / functions

Functions395 in github.com/BICLab/SpikingBrain-7B

⨍ Functions 395◇ Types & classes 101↳ Endpoints 1
Method_init_weights
( self, module: nn.Module, rescale_prenorm_residual: bool = True, num_residual
hf_7B_model/modeling_gla_swa.py:138
Method_initialize_weights
(self, module: nn.Module)
W8ASpike/window_attention.py:131
Method_initialize_weights
(self, module: nn.Module)
W8ASpike/gla_attention.py:92
Method_initialize_weights
(self, module: nn.Module)
hf_7B_model/window_attention.py:131
Method_initialize_weights
(self, module: nn.Module)
hf_7B_model/gla_attention.py:93
Method_initialize_weights
(self, module: nn.Module)
hf_7B_model/window_attention_sdpa.py:131
Function_layer_norm_bwd_kernel
( X, # pointer to the input W, # pointer to the weights B, # pointer to the biases Y, # po
hf_7B_model/layernorm.py:212
Function_layer_norm_fwd_1pass_kernel
( X, # pointer to the input Y, # pointer to the output W, # pointer to the weights B, # po
hf_7B_model/layernorm.py:69
Method_repeat_interleave_samples
(x, lengths, repeat_times)
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:1977
Methodadvance_step
Update metadata in-place to advance one decode step.
vllm_hymeta/attention/meta_attention.py:338
Methodbackward
(ctx, dout)
W8ASpike/activations.py:36
Methodbackward
(ctx, dy)
W8ASpike/activations.py:144
Methodbackward
(ctx, dout)
W8ASpike/activations.py:179
Methodbackward
(ctx, grad_output)
W8ASpike/activations.py:225
Methodbackward
(ctx, grad_output)
W8ASpike/activations.py:263
Methodbackward
(ctx, grad_output)
W8ASpike/activations.py:296
Methodbackward
(ctx, dout)
W8ASpike/activations.py:392
Methodbackward
(ctx, dout, *args)
W8ASpike/activations.py:419
Methodbackward
(ctx, dy, *args)
hf_7B_model/layernorm.py:432
Methodbackward
(ctx, dout, *args)
hf_7B_model/layernorm.py:730
Methodbackward
(ctx, dout)
hf_7B_model/activations.py:36
Methodbackward
(ctx, dy)
hf_7B_model/activations.py:144
Methodbackward
(ctx, dout)
hf_7B_model/activations.py:179
Methodbackward
(ctx, grad_output)
hf_7B_model/activations.py:225
Methodbackward
(ctx, grad_output)
hf_7B_model/activations.py:263
Methodbackward
(ctx, grad_output)
hf_7B_model/activations.py:296
Methodbackward
(ctx, dout)
hf_7B_model/activations.py:392
Methodbackward
(ctx, dout, *args)
hf_7B_model/activations.py:419
Methodbuild
Build attention metadata with on-device tensors. Args: seq_lens: The maybe padded sequence lengths of the input sequences.
vllm_hymeta/attention/meta_attention.py:521
Methodcache
(self)
vllm_hymeta/model_for_7B/gla_cache.py:26
Methodcompute_logits
(self, hidden_states: torch.Tensor, sampling_metadata: SamplingMetadata)
vllm_hymeta/model_for_7B/modeling_gla_swa.py:579
Methodcopy_blocks
Copy blocks from source KV caches to destination KV caches.
vllm_hymeta/attention/meta_attention.py:115
Methodcopy_inputs_before_cuda_graphs
(self, input_buffers, **kwargs)
vllm_hymeta/model_for_7B/modeling_gla_swa.py:553
Methoddecode
This method forwards all its arguments to Qwen2TokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to the docstring of this
hf_7B_VLM/SpikingBrain_VL/processing_spikingbrain_vl.py:115
Methoddecode_metadata
(self)
vllm_hymeta/attention/meta_attention.py:287
Methodextra_repr
(self)
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:117
Methodextra_repr
(self)
W8ASpike/norms.py:20
Methodextra_repr
(self)
hf_7B_model/norms.py:20
Functionfake_quant
Simulate spike-based quantization-dequantization (fake quantization) on float input. Args: x (torch.Tensor): Input float tensor.
W8ASpike/Int2Spike/neuron.py:427
Methodforward
(self, q,k,v,hidden_states,initial_state)
hf_7B_VLM/SpikingBrain_VL/gla.py:44
Methodforward
(self, hidden_state)
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:59
Methodforward
(self, hidden_states: torch.Tensor)
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:80
Methodforward
(self, seqlen: int)
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:95
Methodforward
(self, hidden_states)
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:110
Methodforward
(self, x: torch.Tensor)
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:132
Methodforward
( self, hidden_states: torch.Tensor, cu_seqlens: torch.Tensor, rotary_pos_emb:
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:156
Methodforward
( self, hidden_states: torch.Tensor, cu_seqlens: torch.Tensor, rotary_pos_emb:
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:218
Methodforward
( self, hidden_states: torch.Tensor, cu_seqlens: torch.Tensor, rotary_pos_emb:
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:267
Methodforward
( self, hidden_states: torch.Tensor, cu_seqlens: torch.Tensor, rotary_pos_emb:
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:320
Methodforward
Args: hidden_states (`torch.Tensor` of shape `(seq_len, hidden_size)`): The final hidden states of the model.
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:483
Methodforward
(self, x, position_ids)
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:580
Methodforward
(self, x)
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:615
Methodforward
( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None,
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:715
Methodforward
( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None,
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:801
Methodforward
( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None,
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:915
Methodforward
Args: hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` attention_mask (
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:987
Methodforward
( self, input_ids: torch.LongTensor = None, attention_mask: Optional[torch.Tensor] = N
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:1083
Methodforward
r""" Args: labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): Labels for computing
hf_7B_VLM/SpikingBrain_VL/modeling_spikingbrain_vl.py:1681
Methodforward
(self, hidden_state)
W8ASpike/modeling_gla_swa.py:51
Methodforward
( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None,
W8ASpike/modeling_gla_swa.py:94
Methodforward
( self, input_ids: Optional[torch.LongTensor] = None, position_ids: Optional[torch.Lon
W8ASpike/modeling_gla_swa.py:195
Methodforward
( self, input_ids: torch.LongTensor = None, attention_mask: Optional[torch.Tensor] = N
W8ASpike/modeling_gla_swa.py:390
Methodforward
(self, x, position_ids)
W8ASpike/window_attention.py:34
Methodforward
( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None,
W8ASpike/window_attention.py:140
Methodforward
(self, hidden_states)
W8ASpike/norms.py:13
Methodforward
( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None,
W8ASpike/gla_attention.py:101
Methodforward
(ctx, x)
W8ASpike/activations.py:31
Methodforward
(ctx, x)
W8ASpike/activations.py:135
Methodforward
(ctx, x)
W8ASpike/activations.py:174
Methodforward
(ctx, input, bias)
W8ASpike/activations.py:220
Methodforward
(ctx, input)
W8ASpike/activations.py:258
Methodforward
(ctx, input)
W8ASpike/activations.py:291
Methodforward
(ctx, x, y)
W8ASpike/activations.py:387
Methodforward
(ctx, x, y, weight, bias)
W8ASpike/activations.py:409
Methodforward
(self, x)
W8ASpike/quant_linear.py:29
Methodforward
(self, weight)
W8ASpike/quant_linear.py:68
Methodforward
Forward pass placeholder for subclasses. Args: x (torch.Tensor): Input spike count tensor. T (int | None): O
W8ASpike/Int2Spike/neuron.py:18
Methodforward
(self, x: torch.Tensor, T: int | None = None)
W8ASpike/Int2Spike/neuron.py:112
Methodforward
(self, x: torch.Tensor, T: int | None = None)
W8ASpike/Int2Spike/neuron.py:162
Methodforward
(self, x: torch.Tensor, T: int | None = None)
W8ASpike/Int2Spike/neuron.py:219
Methodforward
( ctx, x, weight, bias, residual=None, eps=1e-6, preno
hf_7B_model/layernorm.py:387
Methodforward
(self, x, residual=None, prenorm=False, residual_in_fp32=False)
hf_7B_model/layernorm.py:566
Methodforward
(self, x, residual=None, prenorm=False, residual_in_fp32=False)
hf_7B_model/layernorm.py:613
Methodforward
(self, x, residual=None, prenorm=False, residual_in_fp32=False)
hf_7B_model/layernorm.py:656
Methodforward
( ctx, x, norm_weight, norm_bias, linear_weight, linear_bias,
hf_7B_model/layernorm.py:672
Methodforward
(self, x, weight, bias, residual=None, prenorm=False, residual_in_fp32=False)
hf_7B_model/layernorm.py:832
Methodforward
(self, x, weight, bias, residual=None, prenorm=False, residual_in_fp32=False)
hf_7B_model/layernorm.py:882
Methodforward
(self, x, weight, bias, residual=None, prenorm=False, residual_in_fp32=False)
hf_7B_model/layernorm.py:928
Methodforward
(self, hidden_state)
hf_7B_model/modeling_gla_swa.py:50
Methodforward
( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None,
hf_7B_model/modeling_gla_swa.py:93
Methodforward
( self, input_ids: Optional[torch.LongTensor] = None, position_ids: Optional[torch.Lon
hf_7B_model/modeling_gla_swa.py:193
Methodforward
( self, input_ids: torch.LongTensor = None, attention_mask: Optional[torch.Tensor] = N
hf_7B_model/modeling_gla_swa.py:386
Methodforward
(self, x, position_ids)
hf_7B_model/window_attention.py:34
Methodforward
( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None,
hf_7B_model/window_attention.py:140
Methodforward
(self, hidden_states)
hf_7B_model/norms.py:13
Methodforward
( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None,
hf_7B_model/gla_attention.py:102
Methodforward
(ctx, x)
hf_7B_model/activations.py:31
Methodforward
(ctx, x)
hf_7B_model/activations.py:135
Methodforward
(ctx, x)
hf_7B_model/activations.py:174
Methodforward
(ctx, input, bias)
hf_7B_model/activations.py:220
← previousnext →201–300 of 395, ranked by callers