赞
踩
大家好,我是微学AI,今天给大家介绍一下大模型的实践应用4-ChatGLM大模型的结构与核心代码解读,最全的ChatGLM模型架构介绍与源码解读,本文介绍将ChatGLM-6B的模型结构,与设计原理。 主要代码来自:https://huggingface.co/THUDM/chatglm-6b/blob/main/modeling_chatglm.py
ChatGLM-6B是有清华团队开发的开源大语言模型,可以用中文和英文进行问答对话。它有着62亿个参数,它采用了General Language Model (GLM)架构,并且通过模型量化技术,可以在普通的显卡上运行(只需6GB显存)。为了优化中文问答和对话,ChatGLM-6B经过了大约1T的中英双语训练,并结合了监督微调、反馈自助和人类反馈强化学习等技术。现在,这个具有62亿参数的ChatGLM-6B已经可以生成非常符合人类喜好的回答了。对于学术研究人员来说,ChatGLM-6B的权重是完全开放的,目前已经发展开发出ChatGLM2-6B,对模型有些升级与改造。
ChatGLM模型引入了一种全新的自回归空格填充的任务,例如下图: 对原始的数据
x
1
,
x
2
,
x
3
,
x
4
,
x
5
,
x
6
x_1,x_2,x_3,x_4,x_5,x_6
x1,x2,x3,x4,x5,x6,随机
m
a
s
k
mask
mask了
x
3
x_3
x3和
x
5
,
x
6
x_5,x_6
x5,x6,目标就是利用未
m
a
s
k
mask
mask的来自回归式预测被
m
a
s
k
mask
mask的信息。图
(
c
)
(c)
(c)可以看到,不同于
M
L
M
MLM
MLM的结构,这里通过两种位置编码,就能自回归式预测被
m
a
s
k
mask
mask的信息。这里有
p
o
s
i
t
i
o
n
1
position1
position1,
p
o
s
i
t
o
n
2
positon2
positon2两种
p
o
s
i
t
i
o
n
position
position,
p
o
s
i
t
i
o
n
1
position1
position1标记的是整体的位置信息;
p
o
s
i
t
i
o
n
2
position2
position2标记的是每个被
m
a
s
k
mask
mask的块内部的相对位置信息。在
(
d
)
(d)
(d)中就很清晰地展示出对于未被
m
a
s
k
mask
mask的信息(用来做
p
r
o
m
p
t
prompt
prompt的),在计算self-attention的时候,全部没有
m
a
s
k
mask
mask,也就是上下文都可见,对于第一块遮挡的信息
x
5
,
x
6
x_5,x_6
x5,x6,自己区域内呈下三角形状,也就是自回归预测形式,第二块
m
a
s
k
mask
mask的信息
x
3
x_3
x3,由于这时候
x
5
x_5
x5和
x
6
x_6
x6已经预测出来了,因此对于
x
3
x_3
x3也变得可见。
ChatGLM-6B使用的激活函数为GELU,其可以近似实现为:
G
E
L
U
(
x
)
≈
0.5
x
(
1
+
tanh
(
2
π
(
x
+
0.044715
x
3
)
)
)
GELU(x)\approx 0.5x(1+ \tanh(\sqrt{\frac{2}{\pi}}(x+0.044715x^{3})))
GELU(x)≈0.5x(1+tanh(π2
(x+0.044715x3)))
@torch.jit.script
def gelu_impl(x):
return 0.5 * x * (1.0 + torch.tanh(0.7978845608028654 * x *
(1.0 + 0.044715 * x * x)))
def gelu(x):
return gelu_impl(x)
ChatGLM2-6B(升级版)模型则使用的 SwiGLU 激活函数:
其实在大模型LLaMA中全连接层也使用了SwiGLU 激活函数,它的计算公式如下:
F
F
N
S
w
i
G
L
U
(
x
,
W
,
V
,
W
2
)
=
S
w
i
G
L
U
(
x
,
W
,
V
)
W
2
FFN_{SwiGLU}(x,W,V,W_{2})=SwiGLU(x,W,V)W_{2}
FFNSwiGLU(x,W,V,W2)=SwiGLU(x,W,V)W2
S
w
i
G
L
U
(
x
,
W
,
V
)
=
S
w
i
s
h
β
(
x
W
)
⊗
x
V
SwiGLU(x,W,V)=Swish_{\beta}(xW)\otimes xV
SwiGLU(x,W,V)=Swishβ(xW)⊗xV
S
w
i
s
h
β
(
x
)
=
x
σ
(
β
x
)
Swish_{\beta}(x)=x \sigma(\beta x)
Swishβ(x)=xσ(βx)
其中,
σ
(
x
)
σ(x)
σ(x)是
S
i
g
m
o
i
d
Sigmoid
Sigmoid函数。
ChatGLM定义了一个名为GLU模块。GLU通过将输入数据与由另一层计算出的“门”值相乘,来实现对输入数据的选择性过滤。设定特定版本的GLU模型首先将输入hidden_states通过一个线性变换(self.dense_h_to_4h)扩展到4倍的维度,然后对其应用激活函数。其中的激活函数是GELU。然后,它再次将结果投影回原始维度(self.dense_4h_to_h)。
class GLU(torch.nn.Module): def __init__(self, hidden_size, inner_hidden_size=None, layer_id=None, bias=True, activation_func=gelu, params_dtype=torch.float, empty_init=True): super(GLU, self).__init__() if empty_init: init_method = skip_init else: init_method = default_init self.layer_id = layer_id self.activation_func = activation_func self.hidden_size = hidden_size if inner_hidden_size is None: inner_hidden_size = 4 * hidden_size self.inner_hidden_size = inner_hidden_size self.dense_h_to_4h = init_method( torch.nn.Linear, self.hidden_size, self.inner_hidden_size, bias=bias, dtype=params_dtype, ) # Project back to h. self.dense_4h_to_h = init_method( torch.nn.Linear, self.inner_hidden_size, self.hidden_size, bias=bias, dtype=params_dtype, ) def forward(self, hidden_states): intermediate_parallel = self.dense_h_to_4h(hidden_states) intermediate_parallel = self.activation_func(intermediate_parallel) output = self.dense_4h_to_h(intermediate_parallel) return output
在位置编码上,ChatGLM使用旋转位置嵌入(Rotary Positional Embeddings,RoPE)代替原有的绝对位置编码。
RoPE借助了复数的思想,出发点是通过绝对位置编码的方式实现相对位置编码。其目标是通过下述运算来给
q
,
k
q,k
q,k添加绝对位置信息:
q
~
m
=
f
(
q
,
m
)
,
k
~
n
=
f
(
k
,
n
)
\tilde{q}_{m}=f(q,m), \tilde{k}_{n}=f(k,n)
q~m=f(q,m),k~n=f(k,n)
经过上述操作后,
q
~
m
\tilde{q}_{m}
q~m和
k
~
n
\tilde{k}_{n}
k~n就带有位置
m
m
m和
n
n
n 的绝对位置信息。
最终可以得到二维情况下用复数表示的 RoPE:
f
(
q
,
m
)
=
R
f
(
q
,
m
)
e
i
θ
f
(
q
,
m
)
=
∣
q
∣
e
i
(
θ
(
q
)
+
m
θ
)
=
q
e
i
m
θ
f(q,m)=R_{f}(q,m)e^{i \theta _{f}(q,m)}=|q|e^{i(\theta(q)+m \theta)}=qe^{im \theta}
f(q,m)=Rf(q,m)eiθf(q,m)=∣q∣ei(θ(q)+mθ)=qeimθ
根据复数乘法的几何意义,上述变换实际上是对应向量旋转,所以位置向量称为“旋转式位置编码”。还可以使用矩阵形式表示:
f
(
q
,
m
)
=
(
cos
m
θ
−
sin
cos
m
θ
sin
m
θ
cos
m
θ
)
(
q
0
q
1
)
f ( q , m ) = \left(
ChatGLM模型中定义了一个名为RotaryEmbedding的模块,用于实现旋转嵌入(Rotary Embedding)。它可以捕获序列中单词的位置信息。
RotaryEmbedding模型中定义各个方法的功能:
1 __init__
: 初始化函数。定义了embedding维度(dim),基数(base),精度(precision)等参数,并根据是否可学习(learnable)设置inverse frequency (inv_freq
)为参数或缓冲区。
2._load_from_state_dict
: 这是一个PyTorch内部函数,用于从状态字典加载模型参数。在这里没有进行实现。
3.forward
: 前向传播函数。首先计算出输入序列长度(seq_len),然后根据seq_len和inv_freq计算频率(freqs)。接着将freqs复制并拼接到emb上,并根据精度将其转换为相应类型。最后计算cosine和sine值,并缓存起来以供后续使用。
4._apply
: PyTorch内部函数,对缓存的cosine和sine值应用给定操作(fn)。
数学计算过程:
inv_freq:inverse frequency(逆频率)是通过对等差数列[0, 2, …, dim-2]除以dim做归一化后取base的负指数得到。
freqs:通过将时间步长t(一个长度为seq_len、元素值从0到seq_len-1的向量)与inv_freq做外积得到。
emb:emb是由两份freqs拼接而成。
cos_cached 和 sin_cached: 是emb中每个元素分别取余弦和正弦得到。
通过对位置索引生成周期性信号(余弦和正弦),进而构建了能够捕获相对位置关系的embedding,也就实现了所谓“旋转”的效果,代码如下:
class RotaryEmbedding(torch.nn.Module): def __init__(self, dim, base=10000, precision=torch.half, learnable=False): super().__init__() inv_freq = 1. / (base ** (torch.arange(0, dim, 2).float() / dim)) inv_freq = inv_freq.half() self.learnable = learnable if learnable: self.inv_freq = torch.nn.Parameter(inv_freq) self.max_seq_len_cached = None else: self.register_buffer('inv_freq', inv_freq) self.max_seq_len_cached = None self.cos_cached = None self.sin_cached = None self.precision = precision def _load_from_state_dict(self, state_dict, prefix, local_metadata, strict, missing_keys, unexpected_keys, error_msgs): pass def forward(self, x, seq_dim=1, seq_len=None): if seq_len is None: seq_len = x.shape[seq_dim] if self.max_seq_len_cached is None or (seq_len > self.max_seq_len_cached): self.max_seq_len_cached = None if self.learnable else seq_len t = torch.arange(seq_len, device=x.device, dtype=self.inv_freq.dtype) freqs = torch.einsum('i,j->ij', t, self.inv_freq) emb = torch.cat((freqs, freqs), dim=-1).to(x.device) if self.precision == torch.bfloat16: emb = emb.float() # [sx, 1 (b * np), hn] cos_cached = emb.cos()[:, None, :] sin_cached = emb.sin()[:, None, :] if self.precision == torch.bfloat16: cos_cached = cos_cached.bfloat16() sin_cached = sin_cached.bfloat16() if self.learnable: return cos_cached, sin_cached self.cos_cached, self.sin_cached = cos_cached, sin_cached return self.cos_cached[:seq_len, ...], self.sin_cached[:seq_len, ...] def _apply(self, fn): if self.cos_cached is not None: self.cos_cached = fn(self.cos_cached) if self.sin_cached is not None: self.sin_cached = fn(self.sin_cached) return super()._apply(fn)
ChatGLM采用标准的自注意力机制,在自注意力机制中,输入是一组查询(query) Q Q Q, 键(key) K K K, 值(value) V V V. 这三者都是由输入序列经过线性变换得到。然后计算查询和键之间点积作为权重,并通过softmax函数进行归一化: Attention ( Q , K , V ) = s o f t m a x ( Q K T d ) V \text{Attention}(Q, K, V) = softmax(\frac{QK^T}{\sqrt{d}})V Attention(Q,K,V)=softmax(d QKT)V
代码中,函数attention_fn实现了自注意力机制:
def attention_fn( self, query_layer, key_layer, value_layer, attention_mask, hidden_size_per_partition, layer_id, layer_past=None, scaling_attention_score=True, use_cache=False, ): if layer_past is not None: past_key, past_value = layer_past[0], layer_past[1] key_layer = torch.cat((past_key, key_layer), dim=0) value_layer = torch.cat((past_value, value_layer), dim=0) # seqlen, batch, num_attention_heads, hidden_size_per_attention_head seq_len, b, nh, hidden_size = key_layer.shape if use_cache: present = (key_layer, value_layer) else: present = None query_key_layer_scaling_coeff = float(layer_id + 1) if scaling_attention_score: query_layer = query_layer / (math.sqrt(hidden_size) * query_key_layer_scaling_coeff) # [b, np, sq, sk] output_size = (query_layer.size(1), query_layer.size(2), query_layer.size(0), key_layer.size(0)) # [sq, b, np, hn] -> [sq, b * np, hn] query_layer = query_layer.view(output_size[2], output_size[0] * output_size[1], -1) # [sk, b, np, hn] -> [sk, b * np, hn] key_layer = key_layer.view(output_size[3], output_size[0] * output_size[1], -1) matmul_result = torch.zeros( 1, 1, 1, dtype=query_layer.dtype, device=query_layer.device, ) matmul_result = torch.baddbmm( matmul_result, query_layer.transpose(0, 1), # [b * np, sq, hn] key_layer.transpose(0, 1).transpose(1, 2), # [b * np, hn, sk] beta=0.0, alpha=1.0, ) # change view to [b, np, sq, sk] attention_scores = matmul_result.view(*output_size) if self.scale_mask_softmax: self.scale_mask_softmax.scale = query_key_layer_scaling_coeff attention_probs = self.scale_mask_softmax(attention_scores, attention_mask.contiguous()) else: if not (attention_mask == 0).all(): # if auto-regressive, skip attention_scores.masked_fill_(attention_mask, -10000.0) dtype = attention_scores.dtype attention_scores = attention_scores.float() attention_scores = attention_scores * query_key_layer_scaling_coeff attention_probs = F.softmax(attention_scores, dim=-1) attention_probs = attention_probs.type(dtype) # context layer shape: [b, np, sq, hn] output_size = (value_layer.size(1), value_layer.size(2), query_layer.size(0), value_layer.size(3)) # change view [sk, b * np, hn] value_layer = value_layer.view(value_layer.size(0), output_size[0] * output_size[1], -1) attention_probs = attention_probs.view(output_size[0] * output_size[1], output_size[2], -1) # 注意力分数乘以value,得到最终的输出context context_layer = torch.bmm(attention_probs, value_layer.transpose(0, 1)) context_layer = context_layer.view(*output_size) context_layer = context_layer.permute(2, 0, 1, 3).contiguous() new_context_layer_shape = context_layer.size()[:-2] + (hidden_size_per_partition,) context_layer = context_layer.view(*new_context_layer_shape) outputs = (context_layer, present, attention_probs) return outputs
下面是SelfAttention模块,模块中调用attention_fn实现注意力机制,代码如下:
class SelfAttention(torch.nn.Module): def __init__(self, hidden_size, num_attention_heads, layer_id, hidden_size_per_attention_head=None, bias=True, params_dtype=torch.float, position_encoding_2d=True, empty_init=True): if empty_init: init_method = skip_init else: init_method = default_init super(SelfAttention, self).__init__() self.layer_id = layer_id self.hidden_size = hidden_size self.hidden_size_per_partition = hidden_size self.num_attention_heads = num_attention_heads self.num_attention_heads_per_partition = num_attention_heads self.position_encoding_2d = position_encoding_2d self.rotary_emb = RotaryEmbedding( self.hidden_size // (self.num_attention_heads * 2) if position_encoding_2d else self.hidden_size // self.num_attention_heads, base=10000, precision=torch.half, learnable=False, ) self.scale_mask_softmax = None if hidden_size_per_attention_head is None: self.hidden_size_per_attention_head = hidden_size // num_attention_heads else: self.hidden_size_per_attention_head = hidden_size_per_attention_head self.inner_hidden_size = num_attention_heads * self.hidden_size_per_attention_head # Strided linear layer. self.query_key_value = init_method( torch.nn.Linear, hidden_size, 3 * self.inner_hidden_size, bias=bias, dtype=params_dtype, ) self.dense = init_method( torch.nn.Linear, self.inner_hidden_size, hidden_size, bias=bias, dtype=params_dtype, ) @staticmethod def attention_mask_func(attention_scores, attention_mask): attention_scores.masked_fill_(attention_mask, -10000.0) return attention_scores def split_tensor_along_last_dim(self, tensor, num_partitions, contiguous_split_chunks=False): # Get the size and dimension. last_dim = tensor.dim() - 1 last_dim_size = tensor.size()[last_dim] // num_partitions # Split. tensor_list = torch.split(tensor, last_dim_size, dim=last_dim) # Note: torch.split does not create contiguous tensors by default. if contiguous_split_chunks: return tuple(chunk.contiguous() for chunk in tensor_list) return tensor_list def forward( self, hidden_states: torch.Tensor, position_ids, attention_mask: torch.Tensor, layer_id, layer_past: Optional[Tuple[torch.Tensor, torch.Tensor]] = None, use_cache: bool = False, output_attentions: bool = False, ): # [seq_len, batch, 3 * hidden_size] mixed_raw_layer = self.query_key_value(hidden_states) # [seq_len, batch, 3 * hidden_size] --> [seq_len, batch, num_attention_heads, 3 * hidden_size_per_attention_head] new_tensor_shape = mixed_raw_layer.size()[:-1] + ( self.num_attention_heads_per_partition, 3 * self.hidden_size_per_attention_head, ) mixed_raw_layer = mixed_raw_layer.view(*new_tensor_shape) (query_layer, key_layer, value_layer) = self.split_tensor_along_last_dim(mixed_raw_layer, 3) if self.position_encoding_2d: q1, q2 = query_layer.chunk(2, dim=(query_layer.ndim - 1)) k1, k2 = key_layer.chunk(2, dim=(key_layer.ndim - 1)) cos, sin = self.rotary_emb(q1, seq_len=position_ids.max() + 1) position_ids, block_position_ids = position_ids[:, 0, :].transpose(0, 1).contiguous(), \ position_ids[:, 1, :].transpose(0, 1).contiguous() q1, k1 = apply_rotary_pos_emb_index(q1, k1, cos, sin, position_ids) q2, k2 = apply_rotary_pos_emb_index(q2, k2, cos, sin, block_position_ids) query_layer = torch.concat([q1, q2], dim=(q1.ndim - 1)) key_layer = torch.concat([k1, k2], dim=(k1.ndim - 1)) else: position_ids = position_ids.transpose(0, 1) cos, sin = self.rotary_emb(value_layer, seq_len=position_ids.max() + 1) # [seq_len, batch, num_attention_heads, hidden_size_per_attention_head] query_layer, key_layer = apply_rotary_pos_emb_index(query_layer, key_layer, cos, sin, position_ids) # [seq_len, batch, hidden_size] context_layer, present, attention_probs = attention_fn( self=self, query_layer=query_layer, key_layer=key_layer, value_layer=value_layer, attention_mask=attention_mask, hidden_size_per_partition=self.hidden_size_per_partition, layer_id=layer_id, layer_past=layer_past, use_cache=use_cache ) output = self.dense(context_layer) outputs = (output, present) if output_attentions: outputs += (attention_probs,) return outputs # output, present, attention_probs
GLMBlock是基于Transformer模型的一种变体,主要包含以下几个部分:
1.Layer Norm: 这是一种常见的归一化方法,主要用于神经网络中的深度学习。它将每个样本在特征维度上进行归一化,使得输出在每个特征维度上都有均值为0和方差为1。这种方法可以加速模型收敛速度,并有助于解决梯度消失和梯度爆炸问题。
2.Self Attention: 自注意力机制是Transformer模型的核心组成部分。给定一个输入序列,自注意力机制能够根据序列中每个元素与其他元素之间的关系,计算出一个权重向量,并用这个权重向量对输入序列进行加权平均。这样可以让模型更好地捕获序列中长距离依赖关系。 在GLMBlock中,Self Attention后面接了一个残差连接(Residual Connection)。残差连接可以让信息直接从前层传递到后层,在深层网络中有助于解决梯度消失问题。
3.Layer Normalization: GLMBlock在Self Attention和GLU之间又添加了一次Layer Norm操作。
4.GLU: GLU是一种非线性激活函数,主要由两部分组成:线性变换和门控机制。线性变换负责提取输入特征,而门控机制则负责控制信息流动。通过这种方式,GLU能够更好地处理复杂任务。 同样地,在GLU后面也接了一个残差连接。
class GLMBlock(torch.nn.Module): def __init__( self, hidden_size, num_attention_heads, layernorm_epsilon, layer_id, inner_hidden_size=None, hidden_size_per_attention_head=None, layernorm=LayerNorm, use_bias=True, params_dtype=torch.float, num_layers=28, position_encoding_2d=True, empty_init=True ): super(GLMBlock, self).__init__() self.layer_id = layer_id # LayerNorm层 self.input_layernorm = layernorm(hidden_size, eps=layernorm_epsilon) # 是否使用2维位置编码 self.position_encoding_2d = position_encoding_2d # 自注意力层 self.attention = SelfAttention( hidden_size, num_attention_heads, layer_id, hidden_size_per_attention_head=hidden_size_per_attention_head, bias=use_bias, params_dtype=params_dtype, position_encoding_2d=self.position_encoding_2d, empty_init=empty_init ) # Post Layer Norm层 self.post_attention_layernorm = layernorm(hidden_size, eps=layernorm_epsilon) self.num_layers = num_layers # mlp层 self.mlp = GLU( hidden_size, inner_hidden_size=inner_hidden_size, bias=use_bias, layer_id=layer_id, params_dtype=params_dtype, empty_init=empty_init ) def forward( self, hidden_states: torch.Tensor, position_ids, attention_mask: torch.Tensor, layer_id, layer_past: Optional[Tuple[torch.Tensor, torch.Tensor]] = None, use_cache: bool = False, output_attentions: bool = False, ): # 对输入进行Layer Norm # [seq_len, batch, hidden_size] attention_input = self.input_layernorm(hidden_states) # 自注意力 attention_outputs = self.attention( attention_input, position_ids, attention_mask=attention_mask, layer_id=layer_id, layer_past=layer_past, use_cache=use_cache, output_attentions=output_attentions ) attention_output = attention_outputs[0] outputs = attention_outputs[1:] # 自注意力的输出和输入残差连接 alpha = (2 * self.num_layers) ** 0.5 hidden_states = attention_input * alpha + attention_output # Layer Norm mlp_input = self.post_attention_layernorm(hidden_states) # 全连接层投影 mlp_output = self.mlp(mlp_input) # MLP层的输出和输入残差连接 output = mlp_input * alpha + mlp_output if use_cache: outputs = (output,) + outputs else: outputs = (output,) + outputs[1:] return outputs
ChatGLM的预训练模型目的是获取注意力mask和position ids,下面具体介绍ChatGLMPreTrainedModel中的get_masks函数实现与获取position_ids函数:
def get_masks(self, input_ids, device): batch_size, seq_length = input_ids.shape # context_lengths记录了batch中每个样本的真实长度 context_lengths = [seq.tolist().index(self.config.bos_token_id) for seq in input_ids] # 生成causal mask,即下三角以及对角线为1,上三角为0 attention_mask = torch.ones((batch_size, seq_length, seq_length), device=device) attention_mask.tril_() # 将前缀部分的注意力改为双向 for i, context_length in enumerate(context_lengths): attention_mask[i, :, :context_length] = 1 attention_mask.unsqueeze_(1) attention_mask = (attention_mask < 0.5).bool() return attention_mask def get_position_ids(self, input_ids, mask_positions, device, use_gmasks=None): batch_size, seq_length = input_ids.shape if use_gmasks is None: use_gmasks = [False] * batch_size # context_lengths:未被padding前,batch中各个样本的长度 context_lengths = [seq.tolist().index(self.config.bos_token_id) for seq in input_ids] # 2维位置编码 if self.position_encoding_2d: # [0,1,2,...,seq_length-1] position_ids = torch.arange(seq_length, dtype=torch.long, device=device).unsqueeze(0).repeat(batch_size, 1) # 将原始输入后所有位置的postion id都设置为[Mask]或者[gMask]的位置id # (该操作见注意力层对位置编码的介绍) for i, context_length in enumerate(context_lengths): position_ids[i, context_length:] = mask_positions[i] # 原始输入的位置编码全部设置为0,待生成的位置添加顺序的位置id # 例如:[0,0,0,0,1,2,3,4,5] block_position_ids = [torch.cat(( torch.zeros(context_length, dtype=torch.long, device=device), torch.arange(seq_length - context_length, dtype=torch.long, device=device) + 1 )) for context_length in context_lengths] block_position_ids = torch.stack(block_position_ids, dim=0) # 将postion_ids和block_position_ids堆叠在一起,用于后续的参数传入; # 在注意力层中,还有将这个position_ids拆分为两部分 position_ids = torch.stack((position_ids, block_position_ids), dim=1) else: position_ids = torch.arange(seq_length, dtype=torch.long, device=device).unsqueeze(0).repeat(batch_size, 1) for i, context_length in enumerate(context_lengths): if not use_gmasks[i]: position_ids[i, context_length:] = mask_positions[i] return position_ids
ChatGLMModel是将以上各种组件与模型块集成加载后的组合模型,代码如下:
class ChatGLMModel(ChatGLMPreTrainedModel): def __init__(self, config: ChatGLMConfig, empty_init=True): super().__init__(config) if empty_init: init_method = skip_init else: init_method = default_init self.max_sequence_length = config.max_sequence_length self.hidden_size = config.hidden_size self.params_dtype = torch.half self.num_attention_heads = config.num_attention_heads self.vocab_size = config.vocab_size self.num_layers = config.num_layers self.layernorm_epsilon = config.layernorm_epsilon self.inner_hidden_size = config.inner_hidden_size self.hidden_size_per_attention_head = self.hidden_size // self.num_attention_heads self.position_encoding_2d = config.position_encoding_2d self.pre_seq_len = config.pre_seq_len self.prefix_projection = config.prefix_projection # 初始化embedding层 self.word_embeddings = init_method( torch.nn.Embedding, num_embeddings=self.vocab_size, embedding_dim=self.hidden_size, dtype=self.params_dtype ) self.gradient_checkpointing = False def get_layer(layer_id): return GLMBlock( self.hidden_size, self.num_attention_heads, self.layernorm_epsilon, layer_id, inner_hidden_size=self.inner_hidden_size, hidden_size_per_attention_head=self.hidden_size_per_attention_head, layernorm=LayerNorm, use_bias=True, params_dtype=self.params_dtype, position_encoding_2d=self.position_encoding_2d, empty_init=empty_init ) # 堆叠GLMBlock self.layers = torch.nn.ModuleList( [get_layer(layer_id) for layer_id in range(self.num_layers)] ) # 最后的Layer Norm层 self.final_layernorm = LayerNorm(self.hidden_size, eps=self.layernorm_epsilon) def get_input_embeddings(self): return self.word_embeddings def set_input_embeddings(self, new_embeddings: torch.Tensor): self.word_embeddings = new_embeddings @add_start_docstrings_to_model_forward(CHATGLM_6B_INPUTS_DOCSTRING.format("batch_size, sequence_length")) @add_code_sample_docstrings( checkpoint=_CHECKPOINT_FOR_DOC, output_type=BaseModelOutputWithPastAndCrossAttentions, config_class=_CONFIG_FOR_DOC, ) def forward( self, input_ids: Optional[torch.LongTensor] = None, position_ids: Optional[torch.LongTensor] = None, attention_mask: Optional[torch.Tensor] = None, past_key_values: Optional[Tuple[Tuple[torch.Tensor, torch.Tensor], ...]] = None, inputs_embeds: Optional[torch.LongTensor] = None, use_cache: Optional[bool] = None, output_attentions: Optional[bool] = None, output_hidden_states: Optional[bool] = None, return_dict: Optional[bool] = None, ) -> Union[Tuple[torch.Tensor, ...], BaseModelOutputWithPast]: output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions output_hidden_states = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) use_cache = use_cache if use_cache is not None else self.config.use_cache return_dict = return_dict if return_dict is not None else self.config.use_return_dict if self.gradient_checkpointing and self.training: if use_cache: logger.warning_once( "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..." ) use_cache = False if input_ids is not None and inputs_embeds is not None: raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time") elif input_ids is not None: batch_size, seq_length = input_ids.shape[:2] elif inputs_embeds is not None: batch_size, seq_length = inputs_embeds.shape[:2] else: raise ValueError("You have to specify either input_ids or inputs_embeds") ### (结束)一些输入输出和参数设置,可以忽略 # embedding层 if inputs_embeds is None: inputs_embeds = self.word_embeddings(input_ids) if past_key_values is None: past_key_values = tuple([None] * len(self.layers)) # 获得注意力mask,该功能继承自ChatGLMPreTrainedModel if attention_mask is None: attention_mask = self.get_masks( input_ids, device=input_ids.device ) if position_ids is None: MASK, gMASK = self.config.mask_token_id, self.config.gmask_token_id seqs = input_ids.tolist() # 记录input_ids中是否使用了mask以及mask的位置 # mask_positions记录每个样本中mask的位置 # use_gmasks记录是否使用了gMask mask_positions, use_gmasks = [], [] for seq in seqs: mask_token = gMASK if gMASK in seq else MASK use_gmask = mask_token == gMASK mask_positions.append(seq.index(mask_token)) use_gmasks.append(use_gmask) # 获得position_ids,该功能继承自ChatGLMPreTrainedModel position_ids = self.get_position_ids( input_ids, mask_positions=mask_positions, device=input_ids.device, use_gmasks=use_gmasks ) # [seq_len, batch, hidden_size] hidden_states = inputs_embeds.transpose(0, 1) presents = () if use_cache else None all_self_attentions = () if output_attentions else None all_hidden_states = () if output_hidden_states else None if attention_mask is None: attention_mask = torch.zeros(1, 1, device=input_ids.device).bool() else: attention_mask = attention_mask.to(hidden_states.device) # 模型的前向传播 for i, layer in enumerate(self.layers): if output_hidden_states: all_hidden_states = all_hidden_states + (hidden_states,) layer_past = past_key_values[i] if self.gradient_checkpointing and self.training: layer_ret = torch.utils.checkpoint.checkpoint( layer, hidden_states, position_ids, attention_mask, torch.tensor(i), layer_past, use_cache, output_attentions ) else: layer_ret = layer( hidden_states, position_ids=position_ids, attention_mask=attention_mask, layer_id=torch.tensor(i), layer_past=layer_past, use_cache=use_cache, output_attentions=output_attentions ) hidden_states = layer_ret[0] if use_cache: presents = presents + (layer_ret[1],) if output_attentions: all_self_attentions = all_self_attentions + (layer_ret[2 if use_cache else 1],) # 最终的Layer Norm hidden_states = self.final_layernorm(hidden_states) if output_hidden_states: all_hidden_states = all_hidden_states + (hidden_states,) if not return_dict: return tuple(v for v in [hidden_states, presents, all_hidden_states, all_self_attentions] if v is not None) return BaseModelOutputWithPast( last_hidden_state=hidden_states, past_key_values=presents, hidden_states=all_hidden_states, attentions=all_self_attentions, )
到此为止,我已经详细介绍了ChaGLM的详细源码与原理介绍,相信大家对ChaGLM的模型架构有了大致的了解了。更多细节内容请持续关注“微学AI”。
Copyright © 2003-2013 www.wpsshop.cn 版权所有,并保留所有权利。