{"record":{"id":"13a07d839e032eba","repo":"rohitg00/ai-engineering-from-scratch","slug":"hidden-must-divide-by-heads","errorCode":null,"errorMessage":"hidden must divide by heads","messagePattern":"hidden must divide by heads","errorType":"validation","errorClass":"ValueError","httpStatus":null,"severity":"error","filePath":"phases/19-capstone-projects/38-classifier-finetuning/code/main.py","lineNumber":63,"sourceCode":"        while len(raw) < max_len:\n            raw.append(self.PAD_ID)\n            attn.append(0)\n        return raw, attn\n\n    def decode(self, ids: Sequence[int]) -> str:\n        return bytes(i for i in ids if i < 256).decode(\"utf-8\", errors=\"replace\")\n\n\n# ---------------------------------------------------------------------------\n# Tiny transformer body\n# ---------------------------------------------------------------------------\n\n\nclass MultiHeadAttention(nn.Module):\n    def __init__(self, hidden: int, heads: int):\n        super().__init__()\n        if hidden % heads != 0:\n            raise ValueError(\"hidden must divide by heads\")\n        self.heads = heads\n        self.head_dim = hidden // heads\n        self.qkv = nn.Linear(hidden, hidden * 3, bias=False)\n        self.out = nn.Linear(hidden, hidden, bias=False)\n\n    def forward(self, x: torch.Tensor, mask: torch.Tensor) -> torch.Tensor:\n        B, T, D = x.shape\n        qkv = self.qkv(x).view(B, T, 3, self.heads, self.head_dim).permute(2, 0, 3, 1, 4)\n        q, k, v = qkv[0], qkv[1], qkv[2]\n        att = (q @ k.transpose(-2, -1)) / math.sqrt(self.head_dim)\n        # mask: B x T, 1 for real, 0 for pad. broadcast to B x 1 x 1 x T.\n        m = mask.view(B, 1, 1, T).to(att.dtype)\n        att = att.masked_fill(m == 0, float(\"-inf\"))\n        weights = F.softmax(att, dim=-1)\n        # Replace nan rows (all-pad keys, never happens for valid input) with zeros.\n        weights = torch.nan_to_num(weights, nan=0.0)\n        ctx = (weights @ v).transpose(1, 2).contiguous().view(B, T, D)\n        return self.out(ctx)","sourceCodeStart":45,"sourceCodeEnd":81,"githubUrl":"https://github.com/rohitg00/ai-engineering-from-scratch/blob/39ea8a1c6d0b61f071226eff7ede4d4105fed820/phases/19-capstone-projects/38-classifier-finetuning/code/main.py#L45-L81","documentation":"Error \"hidden must divide by heads\" thrown in rohitg00/ai-engineering-from-scratch.","triggerScenarios":"Thrown at phases/19-capstone-projects/38-classifier-finetuning/code/main.py:63 when the library encounters an invalid state.","commonSituations":"See trigger scenarios.","solutions":[],"exampleFix":null,"handlingStrategy":null,"validationCode":null,"typeGuard":null,"tryCatchPattern":null,"preventionTips":[],"tags":[],"backgroundTag":null,"analyzedSha":"39ea8a1c6d0b61f071226eff7ede4d4105fed820","analyzedAt":"2026-08-26T03:13:46.626Z","schemaVersion":2},"datasetVersion":"2026-08-26T07:17:17.940Z"}