# bart-large-mnli

Hugging Face: `facebook/bart-large-mnli`
Reconstructed from `config.json` alone. No weights were read.

| | |
|---|---|
| Derived from structure | 405,185,536 |
| In the published checkpoint | 407,344,133 (safetensors.total, 2026-09-06) |
| Delta | -0.53% |
| Comparability | comparable |
| Layers | 136 |
| Priced on | A10G (24GB) |
| Est. one run | $14.56 |

## Structure

| # | Layer | Type | Output shape |
|---|---|---|---|
| 1 | Input | Input | 1 × 1024 |
| 2 | Encoder_Embedding | Embedding | 1 × 1024 × 1024 |
| 3 | Encoder_Position_Embedding | Learned Pos Embed | 1 × 1024 × 1024 |
| 4 | Enc_Attn_1 | Multi-Head Attention | 1 × 1024 × 1024 |
| 5 | Enc_Add_1 | Add | 1 × 1024 × 1024 |
| 6 | Enc_Norm_1 | LayerNorm | 1 × 1024 × 1024 |
| 7 | Enc_FFN_1 | Feed Forward | 1 × 1024 × 1024 |
| 8 | Enc_Attn_2 | Multi-Head Attention | 1 × 1024 × 1024 |
| 9 | Enc_Add_2 | Add | 1 × 1024 × 1024 |
| 10 | Enc_Norm_2 | LayerNorm | 1 × 1024 × 1024 |
| 11 | Enc_FFN_2 | Feed Forward | 1 × 1024 × 1024 |
| 12 | Enc_Attn_3 | Multi-Head Attention | 1 × 1024 × 1024 |
| 13 | Enc_Add_3 | Add | 1 × 1024 × 1024 |
| 14 | Enc_Norm_3 | LayerNorm | 1 × 1024 × 1024 |
| 15 | Enc_FFN_3 | Feed Forward | 1 × 1024 × 1024 |
| 16 | Enc_Attn_4 | Multi-Head Attention | 1 × 1024 × 1024 |
| 17 | Enc_Add_4 | Add | 1 × 1024 × 1024 |
| 18 | Enc_Norm_4 | LayerNorm | 1 × 1024 × 1024 |
| 19 | Enc_FFN_4 | Feed Forward | 1 × 1024 × 1024 |
| 20 | Enc_Attn_5 | Multi-Head Attention | 1 × 1024 × 1024 |
| 21 | Enc_Add_5 | Add | 1 × 1024 × 1024 |
| 22 | Enc_Norm_5 | LayerNorm | 1 × 1024 × 1024 |
| 23 | Enc_FFN_5 | Feed Forward | 1 × 1024 × 1024 |
| 24 | Enc_Attn_6 | Multi-Head Attention | 1 × 1024 × 1024 |
| 25 | Enc_Add_6 | Add | 1 × 1024 × 1024 |
| 26 | Enc_Norm_6 | LayerNorm | 1 × 1024 × 1024 |
| 27 | Enc_FFN_6 | Feed Forward | 1 × 1024 × 1024 |
| 28 | Enc_Attn_7 | Multi-Head Attention | 1 × 1024 × 1024 |
| 29 | Enc_Add_7 | Add | 1 × 1024 × 1024 |
| 30 | Enc_Norm_7 | LayerNorm | 1 × 1024 × 1024 |
| 31 | Enc_FFN_7 | Feed Forward | 1 × 1024 × 1024 |
| 32 | Enc_Attn_8 | Multi-Head Attention | 1 × 1024 × 1024 |
| 33 | Enc_Add_8 | Add | 1 × 1024 × 1024 |
| 34 | Enc_Norm_8 | LayerNorm | 1 × 1024 × 1024 |
| 35 | Enc_FFN_8 | Feed Forward | 1 × 1024 × 1024 |
| 36 | Enc_Attn_9 | Multi-Head Attention | 1 × 1024 × 1024 |
| 37 | Enc_Add_9 | Add | 1 × 1024 × 1024 |
| 38 | Enc_Norm_9 | LayerNorm | 1 × 1024 × 1024 |
| 39 | Enc_FFN_9 | Feed Forward | 1 × 1024 × 1024 |
| 40 | Enc_Attn_10 | Multi-Head Attention | 1 × 1024 × 1024 |
| 41 | Enc_Add_10 | Add | 1 × 1024 × 1024 |
| 42 | Enc_Norm_10 | LayerNorm | 1 × 1024 × 1024 |
| 43 | Enc_FFN_10 | Feed Forward | 1 × 1024 × 1024 |
| 44 | Enc_Attn_11 | Multi-Head Attention | 1 × 1024 × 1024 |
| 45 | Enc_Add_11 | Add | 1 × 1024 × 1024 |
| 46 | Enc_Norm_11 | LayerNorm | 1 × 1024 × 1024 |
| 47 | Enc_FFN_11 | Feed Forward | 1 × 1024 × 1024 |
| 48 | Enc_Attn_12 | Multi-Head Attention | 1 × 1024 × 1024 |
| 49 | Enc_Add_12 | Add | 1 × 1024 × 1024 |
| 50 | Enc_Norm_12 | LayerNorm | 1 × 1024 × 1024 |
| 51 | Enc_FFN_12 | Feed Forward | 1 × 1024 × 1024 |
| 52 | Decoder_Embedding | Embedding | 1 × 1024 × 1024 |
| 53 | Dec_SelfAttn_1 | Multi-Head Attention | 1 × 1024 × 1024 |
| 54 | Dec_Add_1 | Add | 1 × 1024 × 1024 |
| 55 | Dec_Norm1_1 | LayerNorm | 1 × 1024 × 1024 |
| 56 | Dec_CrossAttn_1 | Cross-Attention | 1 × 1024 × 1024 |
| 57 | Dec_Add2_1 | Add | 1 × 1024 × 1024 |
| 58 | Dec_Norm2_1 | LayerNorm | 1 × 1024 × 1024 |
| 59 | Dec_FFN_1 | Feed Forward | 1 × 1024 × 1024 |
| 60 | Dec_SelfAttn_2 | Multi-Head Attention | 1 × 1024 × 1024 |
| 61 | Dec_Add_2 | Add | 1 × 1024 × 1024 |
| 62 | Dec_Norm1_2 | LayerNorm | 1 × 1024 × 1024 |
| 63 | Dec_CrossAttn_2 | Cross-Attention | 1 × 1024 × 1024 |
| 64 | Dec_Add2_2 | Add | 1 × 1024 × 1024 |
| 65 | Dec_Norm2_2 | LayerNorm | 1 × 1024 × 1024 |
| 66 | Dec_FFN_2 | Feed Forward | 1 × 1024 × 1024 |
| 67 | Dec_SelfAttn_3 | Multi-Head Attention | 1 × 1024 × 1024 |
| 68 | Dec_Add_3 | Add | 1 × 1024 × 1024 |
| 69 | Dec_Norm1_3 | LayerNorm | 1 × 1024 × 1024 |
| 70 | Dec_CrossAttn_3 | Cross-Attention | 1 × 1024 × 1024 |
| 71 | Dec_Add2_3 | Add | 1 × 1024 × 1024 |
| 72 | Dec_Norm2_3 | LayerNorm | 1 × 1024 × 1024 |
| 73 | Dec_FFN_3 | Feed Forward | 1 × 1024 × 1024 |
| 74 | Dec_SelfAttn_4 | Multi-Head Attention | 1 × 1024 × 1024 |
| 75 | Dec_Add_4 | Add | 1 × 1024 × 1024 |
| 76 | Dec_Norm1_4 | LayerNorm | 1 × 1024 × 1024 |
| 77 | Dec_CrossAttn_4 | Cross-Attention | 1 × 1024 × 1024 |
| 78 | Dec_Add2_4 | Add | 1 × 1024 × 1024 |
| 79 | Dec_Norm2_4 | LayerNorm | 1 × 1024 × 1024 |
| 80 | Dec_FFN_4 | Feed Forward | 1 × 1024 × 1024 |
| 81 | Dec_SelfAttn_5 | Multi-Head Attention | 1 × 1024 × 1024 |
| 82 | Dec_Add_5 | Add | 1 × 1024 × 1024 |
| 83 | Dec_Norm1_5 | LayerNorm | 1 × 1024 × 1024 |
| 84 | Dec_CrossAttn_5 | Cross-Attention | 1 × 1024 × 1024 |
| 85 | Dec_Add2_5 | Add | 1 × 1024 × 1024 |
| 86 | Dec_Norm2_5 | LayerNorm | 1 × 1024 × 1024 |
| 87 | Dec_FFN_5 | Feed Forward | 1 × 1024 × 1024 |
| 88 | Dec_SelfAttn_6 | Multi-Head Attention | 1 × 1024 × 1024 |
| 89 | Dec_Add_6 | Add | 1 × 1024 × 1024 |
| 90 | Dec_Norm1_6 | LayerNorm | 1 × 1024 × 1024 |
| 91 | Dec_CrossAttn_6 | Cross-Attention | 1 × 1024 × 1024 |
| 92 | Dec_Add2_6 | Add | 1 × 1024 × 1024 |
| 93 | Dec_Norm2_6 | LayerNorm | 1 × 1024 × 1024 |
| 94 | Dec_FFN_6 | Feed Forward | 1 × 1024 × 1024 |
| 95 | Dec_SelfAttn_7 | Multi-Head Attention | 1 × 1024 × 1024 |
| 96 | Dec_Add_7 | Add | 1 × 1024 × 1024 |
| 97 | Dec_Norm1_7 | LayerNorm | 1 × 1024 × 1024 |
| 98 | Dec_CrossAttn_7 | Cross-Attention | 1 × 1024 × 1024 |
| 99 | Dec_Add2_7 | Add | 1 × 1024 × 1024 |
| 100 | Dec_Norm2_7 | LayerNorm | 1 × 1024 × 1024 |
| 101 | Dec_FFN_7 | Feed Forward | 1 × 1024 × 1024 |
| 102 | Dec_SelfAttn_8 | Multi-Head Attention | 1 × 1024 × 1024 |
| 103 | Dec_Add_8 | Add | 1 × 1024 × 1024 |
| 104 | Dec_Norm1_8 | LayerNorm | 1 × 1024 × 1024 |
| 105 | Dec_CrossAttn_8 | Cross-Attention | 1 × 1024 × 1024 |
| 106 | Dec_Add2_8 | Add | 1 × 1024 × 1024 |
| 107 | Dec_Norm2_8 | LayerNorm | 1 × 1024 × 1024 |
| 108 | Dec_FFN_8 | Feed Forward | 1 × 1024 × 1024 |
| 109 | Dec_SelfAttn_9 | Multi-Head Attention | 1 × 1024 × 1024 |
| 110 | Dec_Add_9 | Add | 1 × 1024 × 1024 |
| 111 | Dec_Norm1_9 | LayerNorm | 1 × 1024 × 1024 |
| 112 | Dec_CrossAttn_9 | Cross-Attention | 1 × 1024 × 1024 |
| 113 | Dec_Add2_9 | Add | 1 × 1024 × 1024 |
| 114 | Dec_Norm2_9 | LayerNorm | 1 × 1024 × 1024 |
| 115 | Dec_FFN_9 | Feed Forward | 1 × 1024 × 1024 |
| 116 | Dec_SelfAttn_10 | Multi-Head Attention | 1 × 1024 × 1024 |
| 117 | Dec_Add_10 | Add | 1 × 1024 × 1024 |
| 118 | Dec_Norm1_10 | LayerNorm | 1 × 1024 × 1024 |
| 119 | Dec_CrossAttn_10 | Cross-Attention | 1 × 1024 × 1024 |
| 120 | Dec_Add2_10 | Add | 1 × 1024 × 1024 |
| 121 | Dec_Norm2_10 | LayerNorm | 1 × 1024 × 1024 |
| 122 | Dec_FFN_10 | Feed Forward | 1 × 1024 × 1024 |
| 123 | Dec_SelfAttn_11 | Multi-Head Attention | 1 × 1024 × 1024 |
| 124 | Dec_Add_11 | Add | 1 × 1024 × 1024 |
| 125 | Dec_Norm1_11 | LayerNorm | 1 × 1024 × 1024 |
| 126 | Dec_CrossAttn_11 | Cross-Attention | 1 × 1024 × 1024 |
| 127 | Dec_Add2_11 | Add | 1 × 1024 × 1024 |
| 128 | Dec_Norm2_11 | LayerNorm | 1 × 1024 × 1024 |
| 129 | Dec_FFN_11 | Feed Forward | 1 × 1024 × 1024 |
| 130 | Dec_SelfAttn_12 | Multi-Head Attention | 1 × 1024 × 1024 |
| 131 | Dec_Add_12 | Add | 1 × 1024 × 1024 |
| 132 | Dec_Norm1_12 | LayerNorm | 1 × 1024 × 1024 |
| 133 | Dec_CrossAttn_12 | Cross-Attention | 1 × 1024 × 1024 |
| 134 | Dec_Add2_12 | Add | 1 × 1024 × 1024 |
| 135 | Dec_Norm2_12 | LayerNorm | 1 × 1024 × 1024 |
| 136 | Dec_FFN_12 | Feed Forward | 1 × 1024 × 1024 |
| 137 | LM_Head | Linear | 1 × 1024 × 50265 |
| 138 | Output | Output | 1 × 1024 × 50265 |

## Do this to your own model

```
pip install neurarch-trace
neurarch-trace facebook/bart-large-mnli --plan --share
```

- Page: https://neurarch.com/m/bart-large-mnli.html
- All models: https://neurarch.com/m/index.json
- The checks: https://neurarch.com/r/index.json