File size: 10,371 Bytes
94724ad
 
 
 
 
 
 
 
d605daf
94724ad
 
 
 
 
 
 
370ac60
94724ad
 
 
 
 
 
 
 
 
 
370ac60
94724ad
370ac60
 
 
 
 
 
 
 
94724ad
 
 
370ac60
94724ad
 
 
 
 
 
 
 
 
370ac60
 
 
 
 
94724ad
370ac60
 
 
 
 
 
 
94724ad
 
 
370ac60
94724ad
 
370ac60
94724ad
370ac60
 
 
 
 
 
 
 
 
 
 
 
94724ad
 
370ac60
94724ad
 
 
370ac60
94724ad
 
 
 
370ac60
 
94724ad
370ac60
94724ad
 
 
370ac60
 
 
 
 
 
 
94724ad
370ac60
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
94724ad
370ac60
 
 
 
 
 
 
94724ad
370ac60
 
 
 
94724ad
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
# my_custom_olmoe/modeling_custom.py

import torch
import torch.nn as nn
import torch.nn.functional as F

# 导入官方实现(注意根据你的 transformers 版本调整导入路径)
from transformers.models.olmoe.modeling_olmoe import OlmoeForCausalLM, OlmoeSparseMoeBlock, OlmoeMLP
from .configuration_densebackward_olmoe import DenseBackwardOLMoEConfig


class DenseBackwardOlmoeSparseMoeBlock(OlmoeSparseMoeBlock):
    def forward(self, hidden_states: torch.Tensor):
        batch_size, seq_length, hidden_dim = hidden_states.shape
        flat_hidden = hidden_states.view(-1, hidden_dim)  # (B*seq_len, hidden_dim)

        # 计算路由 logits 和 routing 权重
        router_logits = self.gate(flat_hidden)  # (B*seq_len, num_experts)
        routing_weights = F.softmax(router_logits, dim=1, dtype=torch.float)  # (B*seq_len, num_experts)

        # Top-k 选择
        routing_weights_topk, selected_experts = torch.topk(routing_weights, self.top_k, dim=-1)
        if self.norm_topk_prob:
            routing_weights_topk = routing_weights_topk / routing_weights_topk.sum(dim=-1, keepdim=True)
        routing_weights_topk = routing_weights_topk.to(flat_hidden.dtype)

        # ---------- 稀疏计算部分 ----------
        # 初始化稀疏输出
        sparse_output = torch.zeros((flat_hidden.size(0), hidden_dim), dtype=flat_hidden.dtype, device=flat_hidden.device)
        
        # 存储所有激活信息的数据结构
        num_tokens = flat_hidden.size(0)
        all_activated_outputs = {}  # {expert_idx: {token_idx: output_tensor}}
        all_routing_indices = {}    # {expert_idx: [token_indices]}
        token_activated_experts = {}  # {token_idx: [activated_expert_indices]}
        
        # one-hot 编码 top-k 专家
        expert_mask = F.one_hot(selected_experts, num_classes=self.num_experts)  # (B*seq_len, top_k, num_experts)
        expert_mask = expert_mask.permute(2, 1, 0)  # (num_experts, top_k, B*seq_len)

        # 稀疏计算,同时记录激活情况
        for expert_idx in range(self.num_experts):
            expert_layer = self.experts[expert_idx]
            idx, top_x = torch.where(expert_mask[expert_idx])
            if top_x.numel() > 0:
                current_state = flat_hidden[top_x]  # (n, hidden_dim)
                current_output = expert_layer(current_state)  # (n, hidden_dim)
                weight = routing_weights_topk[top_x, idx].unsqueeze(-1)  # (n, 1)
                weighted_output = current_output * weight
                sparse_output.index_add_(0, top_x, weighted_output.to(flat_hidden.dtype))
                
                # 记录该专家激活的token和对应输出
                all_activated_outputs[expert_idx] = {}
                all_routing_indices[expert_idx] = top_x.tolist()
                
                for pos, token_idx in enumerate(top_x.tolist()):
                    # 记录该专家对该token的输出
                    all_activated_outputs[expert_idx][token_idx] = current_output[pos]
                    
                    # 记录该token激活的专家
                    if token_idx not in token_activated_experts:
                        token_activated_experts[token_idx] = []
                    token_activated_experts[token_idx].append(expert_idx)
        # ---------- 稀疏计算结束 ----------

        # ---------- Dense估计部分 ----------
        # 将activated_experts 转换为list格式,与路由权重匹配
        all_routing = selected_experts.tolist()  # 长度为 (B*seq_len)
        
        # 使用已激活信息估计dense输出
        dense_outputs = []
        for token_idx in range(num_tokens):
            # 获取当前token的激活专家列表
            activated = all_routing[token_idx] if token_idx in token_activated_experts else []
            
            # 估计dense输出(只使用已经计算过的专家输出)
            dense_est = self.estimate_dense_output_efficient(
                token_idx=token_idx,
                activated=activated,
                gate_prob=routing_weights[token_idx],
                all_activated_outputs=all_activated_outputs,
                all_routing_indices=all_routing_indices,
                token_activated_experts=token_activated_experts
            )
            dense_outputs.append(dense_est.unsqueeze(0))
            
        dense_outputs = torch.cat(dense_outputs, dim=0)  # (B*seq_len, hidden_dim)
        # ---------- Dense估计结束 ----------

        # 使用直通梯度技巧
        final_flat = sparse_output.detach() + (dense_outputs - dense_outputs.detach())
        final_output = final_flat.view(batch_size, seq_length, hidden_dim)
        return final_output, router_logits

    def estimate_dense_output_efficient(self, token_idx, activated, gate_prob, 
                                       all_activated_outputs, all_routing_indices, token_activated_experts):
        """
        优化版本的dense输出估计,只使用已计算的专家输出
        """
        num_experts = gate_prob.size(0)
        dense_parts = {}
        
        # 对于激活的专家,直接使用其输出
        for expert_idx in activated:
            if expert_idx in all_activated_outputs and token_idx in all_activated_outputs[expert_idx]:
                dense_parts[expert_idx] = all_activated_outputs[expert_idx][token_idx]
        
        # 对于未激活的专家,使用其他token的激活输出估计
        non_activated = [i for i in range(num_experts) if i not in activated]
        for expert_idx in non_activated:
            # 如果该专家没有被任何token激活,跳过
            if expert_idx not in all_routing_indices or not all_routing_indices[expert_idx]:
                # 使用零向量或平均值作为估计
                dense_parts[expert_idx] = torch.zeros_like(next(iter(dense_parts.values()))) if dense_parts else 0
                continue
                
            # 找出激活了该专家的token,并且这些token也激活了当前token激活的某些专家
            candidate_tokens = []
            for other_token in all_routing_indices[expert_idx]:
                # 检查other_token是否与当前token共享某些激活专家
                if other_token in token_activated_experts:
                    common_experts = set(activated) & set(token_activated_experts[other_token])
                    if common_experts:
                        candidate_tokens.append(other_token)
            
            # 如果找到了候选token,使用它们的输出平均值
            if candidate_tokens:
                expert_outputs = [all_activated_outputs[expert_idx][t] for t in candidate_tokens]
                estimated = torch.stack(expert_outputs).mean(dim=0)
            else:
                # 找不到合适的候选,使用所有激活了该专家的token
                expert_outputs = [all_activated_outputs[expert_idx][t] for t in all_routing_indices[expert_idx]]
                estimated = torch.stack(expert_outputs).mean(dim=0)
                
            dense_parts[expert_idx] = estimated
            
        # 按路由权重加权求和
        estimated_dense = 0
        for expert_idx in range(num_experts):
            if expert_idx in dense_parts:
                estimated_dense += gate_prob[expert_idx] * dense_parts[expert_idx]
                
        return estimated_dense


class DenseBackwardOLMoEForCausalLM(OlmoeForCausalLM):
    """
    自定义的 Olmoe ForCausalLM 模型,使用新的 DenseBackwardOlmoeSparseMoeBlock 替换原版的 MoE 模块,
    以实现 dense backward 功能。
    
    配置类:DenseBackwardOLMoEConfig
    """
    config_class = DenseBackwardOLMoEConfig
    base_model_prefix = "olmoe"

    def __init__(self, config):
        # 首先调用父类初始化方法
        super().__init__(config)
        
        # 不要尝试重新赋值self,而是从预训练模型加载并更新当前模型
        pretrained_model = OlmoeForCausalLM.from_pretrained("allenai/OLMoE-1B-7B-0924", torch_dtype=torch.bfloat16)
        
        # 复制预训练模型的状态到当前模型
        self.config = pretrained_model.config
        self.model = pretrained_model.model
        self.vocab_size = pretrained_model.vocab_size
        self.router_aux_loss_coef = pretrained_model.router_aux_loss_coef
        self.num_experts = pretrained_model.num_experts
        self.lm_head = pretrained_model.lm_head
        
        # 遍历模型中所有 decoder 层,替换每个 OlmoeSparseMoeBlock 为 DenseBackward 版本
        # 此处假设官方模型在 self.model.layers 中组织 decoder 层,
        # 且每层中 mlp 模块包含属性 sparse_moe_block。
        for layer in self.model.layers:
            if hasattr(layer.mlp, "gate"):
                print("111")
                orig_block = layer.mlp
                # 通过直接复制原版属性创建新的块
                new_block = DenseBackwardOlmoeSparseMoeBlock(config)  # 或其他适当参数
                # 然后手动复制需要共享的属性:
                new_block.gate = orig_block.gate
                new_block.experts = orig_block.experts
                new_block.num_experts = orig_block.num_experts
                new_block.top_k = orig_block.top_k
                new_block.norm_topk_prob = orig_block.norm_topk_prob
                layer.mlp = new_block
                print(type(layer.mlp))
        # 在调用post_init()前
        test_param = self.model.layers[0].mlp.experts[0].up_proj.weight.data[0, 0].item()
        print(f"权重示例值(前): {test_param}")
        self.post_init()
        # 在调用post_init()后
        test_param_after = self.model.layers[0].mlp.experts[0].up_proj.weight.data[0, 0].item()
        print(f"权重示例值(后): {test_param_after}")

def main():
    config = DenseBackwardOLMoEConfig(        # 官方模型参数
    model_marker="DenseBackward_olmoe_marker",
    torch_dtype="bfloat16"
)
# 创建自定义模型实例
    model = DenseBackwardOLMoEForCausalLM(config)
    print(type(model))
    print(type(model.model))
    print(type(model.model.layers[0]))
    print(type(model.model.layers[0].mlp))
    print(type(model.model.layers[0].mlp.experts))
if __name__ == "__main__":
    main()