|
{ |
|
"attention_probs_dropout_prob": 0.1, |
|
"hidden_act": "gelu", |
|
"hidden_dropout_prob": 0.1, |
|
"hidden_size": 768, |
|
"initializer_range": 0.02, |
|
"intermediate_size": 3072, |
|
"layer_norm_eps": 1e-12, |
|
"max_position_embeddings": 512, |
|
"model": "bert", |
|
"num_attention_heads": 12, |
|
"num_hidden_layers": 12, |
|
"bos_token_id": 0, |
|
"eos_token_id": 2, |
|
"pad_token_id": 0, |
|
"torch_dtype": "float32", |
|
"type_vocab_size": 2, |
|
"vocab_size": 21128, |
|
"with_pool": true, |
|
"pool_strategy": "cls", |
|
"norm_mode": "torch_buildin", |
|
"mapping": { |
|
"embeddings.word_embeddings.weight": "embeddings.word_embeddings.weight", |
|
"embeddings.position_embeddings.weight": "embeddings.position_embeddings.weight", |
|
"embeddings.segment_embeddings.weight": "embeddings.token_type_embeddings.weight", |
|
"embeddings.layerNorm.weight": "embeddings.LayerNorm.weight", |
|
"embeddings.layerNorm.bias": "embeddings.LayerNorm.bias", |
|
"pooler.weight": "pooler.dense.weight", |
|
"pooler.bias": "pooler.dense.bias", |
|
"encoderLayer.0.multiHeadAttention.q.weight": "encoder.layer.0.attention.self.query.weight", |
|
"encoderLayer.0.multiHeadAttention.q.bias": "encoder.layer.0.attention.self.query.bias", |
|
"encoderLayer.0.multiHeadAttention.k.weight": "encoder.layer.0.attention.self.key.weight", |
|
"encoderLayer.0.multiHeadAttention.k.bias": "encoder.layer.0.attention.self.key.bias", |
|
"encoderLayer.0.multiHeadAttention.v.weight": "encoder.layer.0.attention.self.value.weight", |
|
"encoderLayer.0.multiHeadAttention.v.bias": "encoder.layer.0.attention.self.value.bias", |
|
"encoderLayer.0.multiHeadAttention.o.weight": "encoder.layer.0.attention.output.dense.weight", |
|
"encoderLayer.0.multiHeadAttention.o.bias": "encoder.layer.0.attention.output.dense.bias", |
|
"encoderLayer.0.attnLayerNorm.weight": "encoder.layer.0.attention.output.LayerNorm.weight", |
|
"encoderLayer.0.attnLayerNorm.bias": "encoder.layer.0.attention.output.LayerNorm.bias", |
|
"encoderLayer.0.feedForward.intermediateDense.weight": "encoder.layer.0.intermediate.dense.weight", |
|
"encoderLayer.0.feedForward.intermediateDense.bias": "encoder.layer.0.intermediate.dense.bias", |
|
"encoderLayer.0.feedForward.outputDense.weight": "encoder.layer.0.output.dense.weight", |
|
"encoderLayer.0.feedForward.outputDense.bias": "encoder.layer.0.output.dense.bias", |
|
"encoderLayer.0.ffnLayerNorm.weight": "encoder.layer.0.output.LayerNorm.weight", |
|
"encoderLayer.0.ffnLayerNorm.bias": "encoder.layer.0.output.LayerNorm.bias", |
|
"encoderLayer.1.multiHeadAttention.q.weight": "encoder.layer.1.attention.self.query.weight", |
|
"encoderLayer.1.multiHeadAttention.q.bias": "encoder.layer.1.attention.self.query.bias", |
|
"encoderLayer.1.multiHeadAttention.k.weight": "encoder.layer.1.attention.self.key.weight", |
|
"encoderLayer.1.multiHeadAttention.k.bias": "encoder.layer.1.attention.self.key.bias", |
|
"encoderLayer.1.multiHeadAttention.v.weight": "encoder.layer.1.attention.self.value.weight", |
|
"encoderLayer.1.multiHeadAttention.v.bias": "encoder.layer.1.attention.self.value.bias", |
|
"encoderLayer.1.multiHeadAttention.o.weight": "encoder.layer.1.attention.output.dense.weight", |
|
"encoderLayer.1.multiHeadAttention.o.bias": "encoder.layer.1.attention.output.dense.bias", |
|
"encoderLayer.1.attnLayerNorm.weight": "encoder.layer.1.attention.output.LayerNorm.weight", |
|
"encoderLayer.1.attnLayerNorm.bias": "encoder.layer.1.attention.output.LayerNorm.bias", |
|
"encoderLayer.1.feedForward.intermediateDense.weight": "encoder.layer.1.intermediate.dense.weight", |
|
"encoderLayer.1.feedForward.intermediateDense.bias": "encoder.layer.1.intermediate.dense.bias", |
|
"encoderLayer.1.feedForward.outputDense.weight": "encoder.layer.1.output.dense.weight", |
|
"encoderLayer.1.feedForward.outputDense.bias": "encoder.layer.1.output.dense.bias", |
|
"encoderLayer.1.ffnLayerNorm.weight": "encoder.layer.1.output.LayerNorm.weight", |
|
"encoderLayer.1.ffnLayerNorm.bias": "encoder.layer.1.output.LayerNorm.bias", |
|
"encoderLayer.2.multiHeadAttention.q.weight": "encoder.layer.2.attention.self.query.weight", |
|
"encoderLayer.2.multiHeadAttention.q.bias": "encoder.layer.2.attention.self.query.bias", |
|
"encoderLayer.2.multiHeadAttention.k.weight": "encoder.layer.2.attention.self.key.weight", |
|
"encoderLayer.2.multiHeadAttention.k.bias": "encoder.layer.2.attention.self.key.bias", |
|
"encoderLayer.2.multiHeadAttention.v.weight": "encoder.layer.2.attention.self.value.weight", |
|
"encoderLayer.2.multiHeadAttention.v.bias": "encoder.layer.2.attention.self.value.bias", |
|
"encoderLayer.2.multiHeadAttention.o.weight": "encoder.layer.2.attention.output.dense.weight", |
|
"encoderLayer.2.multiHeadAttention.o.bias": "encoder.layer.2.attention.output.dense.bias", |
|
"encoderLayer.2.attnLayerNorm.weight": "encoder.layer.2.attention.output.LayerNorm.weight", |
|
"encoderLayer.2.attnLayerNorm.bias": "encoder.layer.2.attention.output.LayerNorm.bias", |
|
"encoderLayer.2.feedForward.intermediateDense.weight": "encoder.layer.2.intermediate.dense.weight", |
|
"encoderLayer.2.feedForward.intermediateDense.bias": "encoder.layer.2.intermediate.dense.bias", |
|
"encoderLayer.2.feedForward.outputDense.weight": "encoder.layer.2.output.dense.weight", |
|
"encoderLayer.2.feedForward.outputDense.bias": "encoder.layer.2.output.dense.bias", |
|
"encoderLayer.2.ffnLayerNorm.weight": "encoder.layer.2.output.LayerNorm.weight", |
|
"encoderLayer.2.ffnLayerNorm.bias": "encoder.layer.2.output.LayerNorm.bias", |
|
"encoderLayer.3.multiHeadAttention.q.weight": "encoder.layer.3.attention.self.query.weight", |
|
"encoderLayer.3.multiHeadAttention.q.bias": "encoder.layer.3.attention.self.query.bias", |
|
"encoderLayer.3.multiHeadAttention.k.weight": "encoder.layer.3.attention.self.key.weight", |
|
"encoderLayer.3.multiHeadAttention.k.bias": "encoder.layer.3.attention.self.key.bias", |
|
"encoderLayer.3.multiHeadAttention.v.weight": "encoder.layer.3.attention.self.value.weight", |
|
"encoderLayer.3.multiHeadAttention.v.bias": "encoder.layer.3.attention.self.value.bias", |
|
"encoderLayer.3.multiHeadAttention.o.weight": "encoder.layer.3.attention.output.dense.weight", |
|
"encoderLayer.3.multiHeadAttention.o.bias": "encoder.layer.3.attention.output.dense.bias", |
|
"encoderLayer.3.attnLayerNorm.weight": "encoder.layer.3.attention.output.LayerNorm.weight", |
|
"encoderLayer.3.attnLayerNorm.bias": "encoder.layer.3.attention.output.LayerNorm.bias", |
|
"encoderLayer.3.feedForward.intermediateDense.weight": "encoder.layer.3.intermediate.dense.weight", |
|
"encoderLayer.3.feedForward.intermediateDense.bias": "encoder.layer.3.intermediate.dense.bias", |
|
"encoderLayer.3.feedForward.outputDense.weight": "encoder.layer.3.output.dense.weight", |
|
"encoderLayer.3.feedForward.outputDense.bias": "encoder.layer.3.output.dense.bias", |
|
"encoderLayer.3.ffnLayerNorm.weight": "encoder.layer.3.output.LayerNorm.weight", |
|
"encoderLayer.3.ffnLayerNorm.bias": "encoder.layer.3.output.LayerNorm.bias", |
|
"encoderLayer.4.multiHeadAttention.q.weight": "encoder.layer.4.attention.self.query.weight", |
|
"encoderLayer.4.multiHeadAttention.q.bias": "encoder.layer.4.attention.self.query.bias", |
|
"encoderLayer.4.multiHeadAttention.k.weight": "encoder.layer.4.attention.self.key.weight", |
|
"encoderLayer.4.multiHeadAttention.k.bias": "encoder.layer.4.attention.self.key.bias", |
|
"encoderLayer.4.multiHeadAttention.v.weight": "encoder.layer.4.attention.self.value.weight", |
|
"encoderLayer.4.multiHeadAttention.v.bias": "encoder.layer.4.attention.self.value.bias", |
|
"encoderLayer.4.multiHeadAttention.o.weight": "encoder.layer.4.attention.output.dense.weight", |
|
"encoderLayer.4.multiHeadAttention.o.bias": "encoder.layer.4.attention.output.dense.bias", |
|
"encoderLayer.4.attnLayerNorm.weight": "encoder.layer.4.attention.output.LayerNorm.weight", |
|
"encoderLayer.4.attnLayerNorm.bias": "encoder.layer.4.attention.output.LayerNorm.bias", |
|
"encoderLayer.4.feedForward.intermediateDense.weight": "encoder.layer.4.intermediate.dense.weight", |
|
"encoderLayer.4.feedForward.intermediateDense.bias": "encoder.layer.4.intermediate.dense.bias", |
|
"encoderLayer.4.feedForward.outputDense.weight": "encoder.layer.4.output.dense.weight", |
|
"encoderLayer.4.feedForward.outputDense.bias": "encoder.layer.4.output.dense.bias", |
|
"encoderLayer.4.ffnLayerNorm.weight": "encoder.layer.4.output.LayerNorm.weight", |
|
"encoderLayer.4.ffnLayerNorm.bias": "encoder.layer.4.output.LayerNorm.bias", |
|
"encoderLayer.5.multiHeadAttention.q.weight": "encoder.layer.5.attention.self.query.weight", |
|
"encoderLayer.5.multiHeadAttention.q.bias": "encoder.layer.5.attention.self.query.bias", |
|
"encoderLayer.5.multiHeadAttention.k.weight": "encoder.layer.5.attention.self.key.weight", |
|
"encoderLayer.5.multiHeadAttention.k.bias": "encoder.layer.5.attention.self.key.bias", |
|
"encoderLayer.5.multiHeadAttention.v.weight": "encoder.layer.5.attention.self.value.weight", |
|
"encoderLayer.5.multiHeadAttention.v.bias": "encoder.layer.5.attention.self.value.bias", |
|
"encoderLayer.5.multiHeadAttention.o.weight": "encoder.layer.5.attention.output.dense.weight", |
|
"encoderLayer.5.multiHeadAttention.o.bias": "encoder.layer.5.attention.output.dense.bias", |
|
"encoderLayer.5.attnLayerNorm.weight": "encoder.layer.5.attention.output.LayerNorm.weight", |
|
"encoderLayer.5.attnLayerNorm.bias": "encoder.layer.5.attention.output.LayerNorm.bias", |
|
"encoderLayer.5.feedForward.intermediateDense.weight": "encoder.layer.5.intermediate.dense.weight", |
|
"encoderLayer.5.feedForward.intermediateDense.bias": "encoder.layer.5.intermediate.dense.bias", |
|
"encoderLayer.5.feedForward.outputDense.weight": "encoder.layer.5.output.dense.weight", |
|
"encoderLayer.5.feedForward.outputDense.bias": "encoder.layer.5.output.dense.bias", |
|
"encoderLayer.5.ffnLayerNorm.weight": "encoder.layer.5.output.LayerNorm.weight", |
|
"encoderLayer.5.ffnLayerNorm.bias": "encoder.layer.5.output.LayerNorm.bias", |
|
"encoderLayer.6.multiHeadAttention.q.weight": "encoder.layer.6.attention.self.query.weight", |
|
"encoderLayer.6.multiHeadAttention.q.bias": "encoder.layer.6.attention.self.query.bias", |
|
"encoderLayer.6.multiHeadAttention.k.weight": "encoder.layer.6.attention.self.key.weight", |
|
"encoderLayer.6.multiHeadAttention.k.bias": "encoder.layer.6.attention.self.key.bias", |
|
"encoderLayer.6.multiHeadAttention.v.weight": "encoder.layer.6.attention.self.value.weight", |
|
"encoderLayer.6.multiHeadAttention.v.bias": "encoder.layer.6.attention.self.value.bias", |
|
"encoderLayer.6.multiHeadAttention.o.weight": "encoder.layer.6.attention.output.dense.weight", |
|
"encoderLayer.6.multiHeadAttention.o.bias": "encoder.layer.6.attention.output.dense.bias", |
|
"encoderLayer.6.attnLayerNorm.weight": "encoder.layer.6.attention.output.LayerNorm.weight", |
|
"encoderLayer.6.attnLayerNorm.bias": "encoder.layer.6.attention.output.LayerNorm.bias", |
|
"encoderLayer.6.feedForward.intermediateDense.weight": "encoder.layer.6.intermediate.dense.weight", |
|
"encoderLayer.6.feedForward.intermediateDense.bias": "encoder.layer.6.intermediate.dense.bias", |
|
"encoderLayer.6.feedForward.outputDense.weight": "encoder.layer.6.output.dense.weight", |
|
"encoderLayer.6.feedForward.outputDense.bias": "encoder.layer.6.output.dense.bias", |
|
"encoderLayer.6.ffnLayerNorm.weight": "encoder.layer.6.output.LayerNorm.weight", |
|
"encoderLayer.6.ffnLayerNorm.bias": "encoder.layer.6.output.LayerNorm.bias", |
|
"encoderLayer.7.multiHeadAttention.q.weight": "encoder.layer.7.attention.self.query.weight", |
|
"encoderLayer.7.multiHeadAttention.q.bias": "encoder.layer.7.attention.self.query.bias", |
|
"encoderLayer.7.multiHeadAttention.k.weight": "encoder.layer.7.attention.self.key.weight", |
|
"encoderLayer.7.multiHeadAttention.k.bias": "encoder.layer.7.attention.self.key.bias", |
|
"encoderLayer.7.multiHeadAttention.v.weight": "encoder.layer.7.attention.self.value.weight", |
|
"encoderLayer.7.multiHeadAttention.v.bias": "encoder.layer.7.attention.self.value.bias", |
|
"encoderLayer.7.multiHeadAttention.o.weight": "encoder.layer.7.attention.output.dense.weight", |
|
"encoderLayer.7.multiHeadAttention.o.bias": "encoder.layer.7.attention.output.dense.bias", |
|
"encoderLayer.7.attnLayerNorm.weight": "encoder.layer.7.attention.output.LayerNorm.weight", |
|
"encoderLayer.7.attnLayerNorm.bias": "encoder.layer.7.attention.output.LayerNorm.bias", |
|
"encoderLayer.7.feedForward.intermediateDense.weight": "encoder.layer.7.intermediate.dense.weight", |
|
"encoderLayer.7.feedForward.intermediateDense.bias": "encoder.layer.7.intermediate.dense.bias", |
|
"encoderLayer.7.feedForward.outputDense.weight": "encoder.layer.7.output.dense.weight", |
|
"encoderLayer.7.feedForward.outputDense.bias": "encoder.layer.7.output.dense.bias", |
|
"encoderLayer.7.ffnLayerNorm.weight": "encoder.layer.7.output.LayerNorm.weight", |
|
"encoderLayer.7.ffnLayerNorm.bias": "encoder.layer.7.output.LayerNorm.bias", |
|
"encoderLayer.8.multiHeadAttention.q.weight": "encoder.layer.8.attention.self.query.weight", |
|
"encoderLayer.8.multiHeadAttention.q.bias": "encoder.layer.8.attention.self.query.bias", |
|
"encoderLayer.8.multiHeadAttention.k.weight": "encoder.layer.8.attention.self.key.weight", |
|
"encoderLayer.8.multiHeadAttention.k.bias": "encoder.layer.8.attention.self.key.bias", |
|
"encoderLayer.8.multiHeadAttention.v.weight": "encoder.layer.8.attention.self.value.weight", |
|
"encoderLayer.8.multiHeadAttention.v.bias": "encoder.layer.8.attention.self.value.bias", |
|
"encoderLayer.8.multiHeadAttention.o.weight": "encoder.layer.8.attention.output.dense.weight", |
|
"encoderLayer.8.multiHeadAttention.o.bias": "encoder.layer.8.attention.output.dense.bias", |
|
"encoderLayer.8.attnLayerNorm.weight": "encoder.layer.8.attention.output.LayerNorm.weight", |
|
"encoderLayer.8.attnLayerNorm.bias": "encoder.layer.8.attention.output.LayerNorm.bias", |
|
"encoderLayer.8.feedForward.intermediateDense.weight": "encoder.layer.8.intermediate.dense.weight", |
|
"encoderLayer.8.feedForward.intermediateDense.bias": "encoder.layer.8.intermediate.dense.bias", |
|
"encoderLayer.8.feedForward.outputDense.weight": "encoder.layer.8.output.dense.weight", |
|
"encoderLayer.8.feedForward.outputDense.bias": "encoder.layer.8.output.dense.bias", |
|
"encoderLayer.8.ffnLayerNorm.weight": "encoder.layer.8.output.LayerNorm.weight", |
|
"encoderLayer.8.ffnLayerNorm.bias": "encoder.layer.8.output.LayerNorm.bias", |
|
"encoderLayer.9.multiHeadAttention.q.weight": "encoder.layer.9.attention.self.query.weight", |
|
"encoderLayer.9.multiHeadAttention.q.bias": "encoder.layer.9.attention.self.query.bias", |
|
"encoderLayer.9.multiHeadAttention.k.weight": "encoder.layer.9.attention.self.key.weight", |
|
"encoderLayer.9.multiHeadAttention.k.bias": "encoder.layer.9.attention.self.key.bias", |
|
"encoderLayer.9.multiHeadAttention.v.weight": "encoder.layer.9.attention.self.value.weight", |
|
"encoderLayer.9.multiHeadAttention.v.bias": "encoder.layer.9.attention.self.value.bias", |
|
"encoderLayer.9.multiHeadAttention.o.weight": "encoder.layer.9.attention.output.dense.weight", |
|
"encoderLayer.9.multiHeadAttention.o.bias": "encoder.layer.9.attention.output.dense.bias", |
|
"encoderLayer.9.attnLayerNorm.weight": "encoder.layer.9.attention.output.LayerNorm.weight", |
|
"encoderLayer.9.attnLayerNorm.bias": "encoder.layer.9.attention.output.LayerNorm.bias", |
|
"encoderLayer.9.feedForward.intermediateDense.weight": "encoder.layer.9.intermediate.dense.weight", |
|
"encoderLayer.9.feedForward.intermediateDense.bias": "encoder.layer.9.intermediate.dense.bias", |
|
"encoderLayer.9.feedForward.outputDense.weight": "encoder.layer.9.output.dense.weight", |
|
"encoderLayer.9.feedForward.outputDense.bias": "encoder.layer.9.output.dense.bias", |
|
"encoderLayer.9.ffnLayerNorm.weight": "encoder.layer.9.output.LayerNorm.weight", |
|
"encoderLayer.9.ffnLayerNorm.bias": "encoder.layer.9.output.LayerNorm.bias", |
|
"encoderLayer.10.multiHeadAttention.q.weight": "encoder.layer.10.attention.self.query.weight", |
|
"encoderLayer.10.multiHeadAttention.q.bias": "encoder.layer.10.attention.self.query.bias", |
|
"encoderLayer.10.multiHeadAttention.k.weight": "encoder.layer.10.attention.self.key.weight", |
|
"encoderLayer.10.multiHeadAttention.k.bias": "encoder.layer.10.attention.self.key.bias", |
|
"encoderLayer.10.multiHeadAttention.v.weight": "encoder.layer.10.attention.self.value.weight", |
|
"encoderLayer.10.multiHeadAttention.v.bias": "encoder.layer.10.attention.self.value.bias", |
|
"encoderLayer.10.multiHeadAttention.o.weight": "encoder.layer.10.attention.output.dense.weight", |
|
"encoderLayer.10.multiHeadAttention.o.bias": "encoder.layer.10.attention.output.dense.bias", |
|
"encoderLayer.10.attnLayerNorm.weight": "encoder.layer.10.attention.output.LayerNorm.weight", |
|
"encoderLayer.10.attnLayerNorm.bias": "encoder.layer.10.attention.output.LayerNorm.bias", |
|
"encoderLayer.10.feedForward.intermediateDense.weight": "encoder.layer.10.intermediate.dense.weight", |
|
"encoderLayer.10.feedForward.intermediateDense.bias": "encoder.layer.10.intermediate.dense.bias", |
|
"encoderLayer.10.feedForward.outputDense.weight": "encoder.layer.10.output.dense.weight", |
|
"encoderLayer.10.feedForward.outputDense.bias": "encoder.layer.10.output.dense.bias", |
|
"encoderLayer.10.ffnLayerNorm.weight": "encoder.layer.10.output.LayerNorm.weight", |
|
"encoderLayer.10.ffnLayerNorm.bias": "encoder.layer.10.output.LayerNorm.bias", |
|
"encoderLayer.11.multiHeadAttention.q.weight": "encoder.layer.11.attention.self.query.weight", |
|
"encoderLayer.11.multiHeadAttention.q.bias": "encoder.layer.11.attention.self.query.bias", |
|
"encoderLayer.11.multiHeadAttention.k.weight": "encoder.layer.11.attention.self.key.weight", |
|
"encoderLayer.11.multiHeadAttention.k.bias": "encoder.layer.11.attention.self.key.bias", |
|
"encoderLayer.11.multiHeadAttention.v.weight": "encoder.layer.11.attention.self.value.weight", |
|
"encoderLayer.11.multiHeadAttention.v.bias": "encoder.layer.11.attention.self.value.bias", |
|
"encoderLayer.11.multiHeadAttention.o.weight": "encoder.layer.11.attention.output.dense.weight", |
|
"encoderLayer.11.multiHeadAttention.o.bias": "encoder.layer.11.attention.output.dense.bias", |
|
"encoderLayer.11.attnLayerNorm.weight": "encoder.layer.11.attention.output.LayerNorm.weight", |
|
"encoderLayer.11.attnLayerNorm.bias": "encoder.layer.11.attention.output.LayerNorm.bias", |
|
"encoderLayer.11.feedForward.intermediateDense.weight": "encoder.layer.11.intermediate.dense.weight", |
|
"encoderLayer.11.feedForward.intermediateDense.bias": "encoder.layer.11.intermediate.dense.bias", |
|
"encoderLayer.11.feedForward.outputDense.weight": "encoder.layer.11.output.dense.weight", |
|
"encoderLayer.11.feedForward.outputDense.bias": "encoder.layer.11.output.dense.bias", |
|
"encoderLayer.11.ffnLayerNorm.weight": "encoder.layer.11.output.LayerNorm.weight", |
|
"encoderLayer.11.ffnLayerNorm.bias": "encoder.layer.11.output.LayerNorm.bias" |
|
} |
|
} |
|
|