Upload bertmodel_arch.txt
Browse files- bertmodel_arch.txt +46 -0
bertmodel_arch.txt
ADDED
|
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
RobertModel(
|
| 2 |
+
(bert): RobertaForSequenceClassification(
|
| 3 |
+
(roberta): RobertaModel(
|
| 4 |
+
(embeddings): RobertaEmbeddings(
|
| 5 |
+
(word_embeddings): Embedding(50265, 768, padding_idx=1)
|
| 6 |
+
(position_embeddings): Embedding(514, 768, padding_idx=1)
|
| 7 |
+
(token_type_embeddings): Embedding(1, 768)
|
| 8 |
+
(LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
|
| 9 |
+
(dropout): Dropout(p=0.1, inplace=False)
|
| 10 |
+
)
|
| 11 |
+
(encoder): RobertaEncoder(
|
| 12 |
+
(layer): ModuleList(
|
| 13 |
+
(0-11): 12 x RobertaLayer(
|
| 14 |
+
(attention): RobertaAttention(
|
| 15 |
+
(self): RobertaSdpaSelfAttention(
|
| 16 |
+
(query): Linear(in_features=768, out_features=768, bias=True)
|
| 17 |
+
(key): Linear(in_features=768, out_features=768, bias=True)
|
| 18 |
+
(value): Linear(in_features=768, out_features=768, bias=True)
|
| 19 |
+
(dropout): Dropout(p=0.1, inplace=False)
|
| 20 |
+
)
|
| 21 |
+
(output): RobertaSelfOutput(
|
| 22 |
+
(dense): Linear(in_features=768, out_features=768, bias=True)
|
| 23 |
+
(LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
|
| 24 |
+
(dropout): Dropout(p=0.1, inplace=False)
|
| 25 |
+
)
|
| 26 |
+
)
|
| 27 |
+
(intermediate): RobertaIntermediate(
|
| 28 |
+
(dense): Linear(in_features=768, out_features=3072, bias=True)
|
| 29 |
+
(intermediate_act_fn): ReLU()
|
| 30 |
+
)
|
| 31 |
+
(output): RobertaOutput(
|
| 32 |
+
(dense): Linear(in_features=3072, out_features=768, bias=True)
|
| 33 |
+
(LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
|
| 34 |
+
(dropout): Dropout(p=0.1, inplace=False)
|
| 35 |
+
)
|
| 36 |
+
)
|
| 37 |
+
)
|
| 38 |
+
)
|
| 39 |
+
)
|
| 40 |
+
(classifier): RobertaClassificationHead(
|
| 41 |
+
(dense): Linear(in_features=768, out_features=768, bias=True)
|
| 42 |
+
(dropout): Dropout(p=0.1, inplace=False)
|
| 43 |
+
(out_proj): Linear(in_features=768, out_features=5, bias=True)
|
| 44 |
+
)
|
| 45 |
+
)
|
| 46 |
+
)
|