| 96 | |
| 97 | |
| 98 | class MLP(Module): |
| 99 | |
| 100 | def __init__( |
| 101 | self, |
| 102 | hidden_size, |
| 103 | ffn_hidden_size, |
| 104 | hidden_act, |
| 105 | bias=True, |
| 106 | dtype=None, |
| 107 | tp_group=None, |
| 108 | tp_size=1, |
| 109 | quant_mode=QuantMode(0), |
| 110 | inner_layernorm=False, |
| 111 | eps=1e-05, |
| 112 | is_expert=False, |
| 113 | ): |
| 114 | super().__init__() |
| 115 | if hidden_act not in ACT2FN: |
| 116 | raise ValueError( |
| 117 | 'unsupported activation function: {}'.format(hidden_act)) |
| 118 | fc_output_size = 2 * ffn_hidden_size if hidden_act in [ |
| 119 | 'swiglu', 'gegelu' |
| 120 | ] else ffn_hidden_size |
| 121 | self.inner_layernorm = LayerNorm(ffn_hidden_size, dtype=dtype, |
| 122 | eps=eps) if inner_layernorm else None |
| 123 | |
| 124 | self.fc = ColumnLinear(hidden_size, |
| 125 | fc_output_size, |
| 126 | bias=bias, |
| 127 | dtype=dtype, |
| 128 | tp_group=tp_group, |
| 129 | tp_size=tp_size, |
| 130 | gather_output=False) |
| 131 | self.proj = RowLinear(ffn_hidden_size, |
| 132 | hidden_size, |
| 133 | bias=bias, |
| 134 | dtype=dtype, |
| 135 | tp_group=tp_group, |
| 136 | tp_size=tp_size, |
| 137 | is_expert=is_expert) |
| 138 | |
| 139 | self.hidden_size = hidden_size |
| 140 | self.ffn_hidden_size = ffn_hidden_size |
| 141 | self.hidden_act = hidden_act |
| 142 | self.dtype = dtype |
| 143 | self.bias = bias |
| 144 | self.tp_group = tp_group |
| 145 | self.tp_size = tp_size |
| 146 | self.quant_mode = quant_mode |
| 147 | self.eps = eps |
| 148 | self.is_expert = is_expert |
| 149 | # see optimize_model's add_lora for LoRA initialization |
| 150 | self.lora = None |
| 151 | self.dora = None |
| 152 | |
| 153 | def forward(self, hidden_states, lora_layer_params=None, gegelu_limit=None): |
| 154 | if lora_layer_params is not None: |
| 155 | assert lora_layer_params.get_runtime_params( |
no outgoing calls
no test coverage detected