File size: 7,934 Bytes
11a7660
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
"""
HuggingFace PreTrainedModel wrapper for InterpGPT / TaskGPT.

Weights map 1:1 to the original gpt_model.TaskGPT state dict, so the same
.pt checkpoints produced during Phase 1 load here without remapping.

Usage (after upload):
    from transformers import AutoModel, AutoTokenizer
    model = AutoModel.from_pretrained("connaaa/interpgpt-standard-23M",
                                      trust_remote_code=True)
    # Or for the analysis pipeline:
    from transformer_lens import HookedTransformer
    hooked = HookedTransformer.from_pretrained("connaaa/interpgpt-standard-23M",
                                               hf_model=model,
                                               ...)
"""
import math
import torch
import torch.nn as nn
import torch.nn.functional as F
from transformers import PreTrainedModel

from .configuration_interpgpt import InterpGPTConfig


class RMSNorm(nn.Module):
    def __init__(self, d_model: int, eps: float = 1e-6):
        super().__init__()
        self.weight = nn.Parameter(torch.ones(d_model))
        self.eps = eps

    def forward(self, x):
        norm = torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)
        return x * norm * self.weight


class RotaryPositionalEncoding(nn.Module):
    def __init__(self, d_model: int, max_seq_len: int = 512, base: float = 10000.0):
        super().__init__()
        assert d_model % 2 == 0
        inv_freq = 1.0 / (base ** (torch.arange(0, d_model, 2).float() / d_model))
        self.register_buffer("inv_freq", inv_freq)
        t = torch.arange(max_seq_len, dtype=torch.float)
        freqs = torch.einsum("i,j->ij", t, inv_freq)
        self.register_buffer("cos_cached", freqs.cos())
        self.register_buffer("sin_cached", freqs.sin())

    def forward(self, seq_len: int):
        return self.cos_cached[:seq_len], self.sin_cached[:seq_len]


def apply_rotary_emb(x, cos, sin):
    d_half = x.shape[-1] // 2
    x1, x2 = x[..., :d_half], x[..., d_half:]
    cos = cos[: x.shape[2]].unsqueeze(0).unsqueeze(0)
    sin = sin[: x.shape[2]].unsqueeze(0).unsqueeze(0)
    out1 = x1 * cos - x2 * sin
    out2 = x2 * cos + x1 * sin
    return torch.cat([out1, out2], dim=-1)


class CausalSelfAttention(nn.Module):
    def __init__(self, config: InterpGPTConfig):
        super().__init__()
        assert config.d_model % config.n_heads == 0
        self.n_heads = config.n_heads
        self.head_dim = config.d_model // config.n_heads
        self.qkv = nn.Linear(config.d_model, 3 * config.d_model, bias=config.bias)
        self.out_proj = nn.Linear(config.d_model, config.d_model, bias=config.bias)
        self.attn_dropout = nn.Dropout(config.dropout)
        self.resid_dropout = nn.Dropout(config.dropout)
        self.rope = RotaryPositionalEncoding(self.head_dim, config.max_seq_len)
        mask = torch.tril(torch.ones(config.max_seq_len, config.max_seq_len))
        self.register_buffer("causal_mask", mask.view(1, 1, config.max_seq_len, config.max_seq_len))

    def forward(self, x, kv_cache=None):
        B, T, D = x.shape
        qkv = self.qkv(x)
        q, k, v = qkv.chunk(3, dim=-1)
        q = q.view(B, T, self.n_heads, self.head_dim).transpose(1, 2)
        k = k.view(B, T, self.n_heads, self.head_dim).transpose(1, 2)
        v = v.view(B, T, self.n_heads, self.head_dim).transpose(1, 2)
        cos, sin = self.rope(T)
        q = apply_rotary_emb(q, cos, sin)
        k = apply_rotary_emb(k, cos, sin)
        if kv_cache is not None:
            if "k" in kv_cache:
                k = torch.cat([kv_cache["k"], k], dim=2)
                v = torch.cat([kv_cache["v"], v], dim=2)
            kv_cache["k"] = k
            kv_cache["v"] = v
        if hasattr(F, "scaled_dot_product_attention") and kv_cache is None:
            out = F.scaled_dot_product_attention(
                q, k, v,
                attn_mask=None,
                dropout_p=self.attn_dropout.p if self.training else 0.0,
                is_causal=True,
            )
        else:
            scale = 1.0 / math.sqrt(self.head_dim)
            attn = torch.matmul(q, k.transpose(-2, -1)) * scale
            T_k = k.size(2)
            causal = self.causal_mask[:, :, T_k - T : T_k, :T_k]
            attn = attn.masked_fill(causal == 0, float("-inf"))
            attn = F.softmax(attn, dim=-1)
            attn = self.attn_dropout(attn)
            out = torch.matmul(attn, v)
        out = out.transpose(1, 2).contiguous().view(B, T, D)
        return self.resid_dropout(self.out_proj(out))


class FeedForward(nn.Module):
    def __init__(self, config: InterpGPTConfig):
        super().__init__()
        hidden = int(2 * config.d_ff / 3)
        hidden = 64 * ((hidden + 63) // 64)
        self.gate_proj = nn.Linear(config.d_model, hidden, bias=config.bias)
        self.up_proj = nn.Linear(config.d_model, hidden, bias=config.bias)
        self.down_proj = nn.Linear(hidden, config.d_model, bias=config.bias)
        self.dropout = nn.Dropout(config.dropout)

    def forward(self, x):
        return self.dropout(self.down_proj(F.silu(self.gate_proj(x)) * self.up_proj(x)))


class TransformerBlock(nn.Module):
    def __init__(self, config: InterpGPTConfig):
        super().__init__()
        self.ln1 = RMSNorm(config.d_model)
        self.attn = CausalSelfAttention(config)
        self.ln2 = RMSNorm(config.d_model)
        self.ffn = FeedForward(config)

    def forward(self, x, kv_cache=None):
        x = x + self.attn(self.ln1(x), kv_cache)
        x = x + self.ffn(self.ln2(x))
        return x


class InterpGPTModel(PreTrainedModel):
    """
    HF-wrapped InterpGPT / TaskGPT. State dict parameter names match the
    original gpt_model.TaskGPT exactly so Phase 1 .pt checkpoints load
    via state_dict without remapping.
    """
    config_class = InterpGPTConfig
    base_model_prefix = "interpgpt"
    supports_gradient_checkpointing = False

    def __init__(self, config: InterpGPTConfig):
        super().__init__(config)
        self.config = config
        self.token_embedding = nn.Embedding(config.vocab_size, config.d_model, padding_idx=config.pad_id)
        self.drop = nn.Dropout(config.dropout)
        self.blocks = nn.ModuleList([TransformerBlock(config) for _ in range(config.n_layers)])
        self.ln_final = RMSNorm(config.d_model)
        self.lm_head = nn.Linear(config.d_model, config.vocab_size, bias=False)
        self.lm_head.weight = self.token_embedding.weight
        self.post_init()

    def _init_weights(self, module):
        if isinstance(module, nn.Linear):
            nn.init.normal_(module.weight, mean=0.0, std=0.02)
            if module.bias is not None:
                nn.init.zeros_(module.bias)
        elif isinstance(module, nn.Embedding):
            nn.init.normal_(module.weight, mean=0.0, std=0.02)
            if module.padding_idx is not None:
                nn.init.zeros_(module.weight[module.padding_idx])

    def forward(self, input_ids, attention_mask=None, labels=None, loss_mask=None, **kwargs):
        B, T = input_ids.shape
        x = self.drop(self.token_embedding(input_ids))
        for block in self.blocks:
            x = block(x)
        x = self.ln_final(x)
        logits = self.lm_head(x)
        output = {"logits": logits}
        if labels is not None:
            shift_logits = logits[:, :-1].contiguous()
            shift_labels = labels[:, 1:].contiguous()
            loss = F.cross_entropy(
                shift_logits.view(-1, self.config.vocab_size),
                shift_labels.view(-1),
                ignore_index=self.config.pad_id,
                reduction="none",
            ).view(B, T - 1)
            if loss_mask is not None:
                shift_mask = loss_mask[:, 1:].contiguous().float()
                loss = (loss * shift_mask).sum() / shift_mask.sum().clamp(min=1.0)
            else:
                loss = loss.mean()
            output["loss"] = loss
        return output