decoder_pos_embed 1x257x1152
trainable_cls_token 1x1x1152
decoder_embed.weight 1152x768
decoder_embed.bias 1152
decoder_layers.0.attention.attention.query.weight 1152x1152
decoder_layers.0.attention.attention.query.bias 1152
decoder_layers.0.attention.attention.key.weight 1152x1152
decoder_layers.0.attention.attention.key.bias 1152
decoder_layers.0.attention.attention.value.weight 1152x1152
decoder_layers.0.attention.attention.value.bias 1152
decoder_layers.0.attention.output.dense.weight 1152x1152
decoder_layers.0.attention.output.dense.bias 1152
decoder_layers.0.intermediate.dense.weight 4096x1152
decoder_layers.0.intermediate.dense.bias 4096
decoder_layers.0.output.dense.weight 1152x4096
decoder_layers.0.output.dense.bias 1152
decoder_layers.0.layernorm_before.weight 1152
decoder_layers.0.layernorm_before.bias 1152
decoder_layers.0.layernorm_after.weight 1152
decoder_layers.0.layernorm_after.bias 1152
decoder_layers.1.attention.attention.query.weight 1152x1152
decoder_layers.1.attention.attention.query.bias 1152
decoder_layers.1.attention.attention.key.weight 1152x1152
decoder_layers.1.attention.attention.key.bias 1152
decoder_layers.1.attention.attention.value.weight 1152x1152
decoder_layers.1.attention.attention.value.bias 1152
decoder_layers.1.attention.output.dense.weight 1152x1152
decoder_layers.1.attention.output.dense.bias 1152
decoder_layers.1.intermediate.dense.weight 4096x1152
decoder_layers.1.intermediate.dense.bias 4096
decoder_layers.1.output.dense.weight 1152x4096
decoder_layers.1.output.dense.bias 1152
decoder_layers.1.layernorm_before.weight 1152
decoder_layers.1.layernorm_before.bias 1152
decoder_layers.1.layernorm_after.weight 1152
decoder_layers.1.layernorm_after.bias 1152
decoder_layers.2.attention.attention.query.weight 1152x1152
decoder_layers.2.attention.attention.query.bias 1152
decoder_layers.2.attention.attention.key.weight 1152x1152
decoder_layers.2.attention.attention.key.bias 1152
decoder_layers.2.attention.attention.value.weight 1152x1152
decoder_layers.2.attention.attention.value.bias 1152
decoder_layers.2.attention.output.dense.weight 1152x1152
decoder_layers.2.attention.output.dense.bias 1152
decoder_layers.2.intermediate.dense.weight 4096x1152
decoder_layers.2.intermediate.dense.bias 4096
decoder_layers.2.output.dense.weight 1152x4096
decoder_layers.2.output.dense.bias 1152
decoder_layers.2.layernorm_before.weight 1152
decoder_layers.2.layernorm_before.bias 1152
decoder_layers.2.layernorm_after.weight 1152
decoder_layers.2.layernorm_after.bias 1152
decoder_layers.3.attention.attention.query.weight 1152x1152
decoder_layers.3.attention.attention.query.bias 1152
decoder_layers.3.attention.attention.key.weight 1152x1152
decoder_layers.3.attention.attention.key.bias 1152
decoder_layers.3.attention.attention.value.weight 1152x1152
decoder_layers.3.attention.attention.value.bias 1152
decoder_layers.3.attention.output.dense.weight 1152x1152
decoder_layers.3.attention.output.dense.bias 1152
decoder_layers.3.intermediate.dense.weight 4096x1152
decoder_layers.3.intermediate.dense.bias 4096
decoder_layers.3.output.dense.weight 1152x4096
decoder_layers.3.output.dense.bias 1152
decoder_layers.3.layernorm_before.weight 1152
decoder_layers.3.layernorm_before.bias 1152
decoder_layers.3.layernorm_after.weight 1152
decoder_layers.3.layernorm_after.bias 1152
decoder_layers.4.attention.attention.query.weight 1152x1152
decoder_layers.4.attention.attention.query.bias 1152
decoder_layers.4.attention.attention.key.weight 1152x1152
decoder_layers.4.attention.attention.key.bias 1152
decoder_layers.4.attention.attention.value.weight 1152x1152
decoder_layers.4.attention.attention.value.bias 1152
decoder_layers.4.attention.output.dense.weight 1152x1152
decoder_layers.4.attention.output.dense.bias 1152
decoder_layers.4.intermediate.dense.weight 4096x1152
decoder_layers.4.intermediate.dense.bias 4096
decoder_layers.4.output.dense.weight 1152x4096
decoder_layers.4.output.dense.bias 1152
decoder_layers.4.layernorm_before.weight 1152
decoder_layers.4.layernorm_before.bias 1152
decoder_layers.4.layernorm_after.weight 1152
decoder_layers.4.layernorm_after.bias 1152
decoder_layers.5.attention.attention.query.weight 1152x1152
decoder_layers.5.attention.attention.query.bias 1152
decoder_layers.5.attention.attention.key.weight 1152x1152
decoder_layers.5.attention.attention.key.bias 1152
decoder_layers.5.attention.attention.value.weight 1152x1152
decoder_layers.5.attention.attention.value.bias 1152
decoder_layers.5.attention.output.dense.weight 1152x1152
decoder_layers.5.attention.output.dense.bias 1152
decoder_layers.5.intermediate.dense.weight 4096x1152
decoder_layers.5.intermediate.dense.bias 4096
decoder_layers.5.output.dense.weight 1152x4096
decoder_layers.5.output.dense.bias 1152
decoder_layers.5.layernorm_before.weight 1152
decoder_layers.5.layernorm_before.bias 1152
decoder_layers.5.layernorm_after.weight 1152
decoder_layers.5.layernorm_after.bias 1152
decoder_layers.6.attention.attention.query.weight 1152x1152
decoder_layers.6.attention.attention.query.bias 1152
decoder_layers.6.attention.attention.key.weight 1152x1152
decoder_layers.6.attention.attention.key.bias 1152
decoder_layers.6.attention.attention.value.weight 1152x1152
decoder_layers.6.attention.attention.value.bias 1152
decoder_layers.6.attention.output.dense.weight 1152x1152
decoder_layers.6.attention.output.dense.bias 1152
decoder_layers.6.intermediate.dense.weight 4096x1152
decoder_layers.6.intermediate.dense.bias 4096
decoder_layers.6.output.dense.weight 1152x4096
decoder_layers.6.output.dense.bias 1152
decoder_layers.6.layernorm_before.weight 1152
decoder_layers.6.layernorm_before.bias 1152
decoder_layers.6.layernorm_after.weight 1152
decoder_layers.6.layernorm_after.bias 1152
decoder_layers.7.attention.attention.query.weight 1152x1152
decoder_layers.7.attention.attention.query.bias 1152
decoder_layers.7.attention.attention.key.weight 1152x1152
decoder_layers.7.attention.attention.key.bias 1152
decoder_layers.7.attention.attention.value.weight 1152x1152
decoder_layers.7.attention.attention.value.bias 1152
decoder_layers.7.attention.output.dense.weight 1152x1152
decoder_layers.7.attention.output.dense.bias 1152
decoder_layers.7.intermediate.dense.weight 4096x1152
decoder_layers.7.intermediate.dense.bias 4096
decoder_layers.7.output.dense.weight 1152x4096
decoder_layers.7.output.dense.bias 1152
decoder_layers.7.layernorm_before.weight 1152
decoder_layers.7.layernorm_before.bias 1152
decoder_layers.7.layernorm_after.weight 1152
decoder_layers.7.layernorm_after.bias 1152
decoder_layers.8.attention.attention.query.weight 1152x1152
decoder_layers.8.attention.attention.query.bias 1152
decoder_layers.8.attention.attention.key.weight 1152x1152
decoder_layers.8.attention.attention.key.bias 1152
decoder_layers.8.attention.attention.value.weight 1152x1152
decoder_layers.8.attention.attention.value.bias 1152
decoder_layers.8.attention.output.dense.weight 1152x1152
decoder_layers.8.attention.output.dense.bias 1152
decoder_layers.8.intermediate.dense.weight 4096x1152
decoder_layers.8.intermediate.dense.bias 4096
decoder_layers.8.output.dense.weight 1152x4096
decoder_layers.8.output.dense.bias 1152
decoder_layers.8.layernorm_before.weight 1152
decoder_layers.8.layernorm_before.bias 1152
decoder_layers.8.layernorm_after.weight 1152
decoder_layers.8.layernorm_after.bias 1152
decoder_layers.9.attention.attention.query.weight 1152x1152
decoder_layers.9.attention.attention.query.bias 1152
decoder_layers.9.attention.attention.key.weight 1152x1152
decoder_layers.9.attention.attention.key.bias 1152
decoder_layers.9.attention.attention.value.weight 1152x1152
decoder_layers.9.attention.attention.value.bias 1152
decoder_layers.9.attention.output.dense.weight 1152x1152
decoder_layers.9.attention.output.dense.bias 1152
decoder_layers.9.intermediate.dense.weight 4096x1152
decoder_layers.9.intermediate.dense.bias 4096
decoder_layers.9.output.dense.weight 1152x4096
decoder_layers.9.output.dense.bias 1152
decoder_layers.9.layernorm_before.weight 1152
decoder_layers.9.layernorm_before.bias 1152
decoder_layers.9.layernorm_after.weight 1152
decoder_layers.9.layernorm_after.bias 1152
decoder_layers.10.attention.attention.query.weight 1152x1152
decoder_layers.10.attention.attention.query.bias 1152
decoder_layers.10.attention.attention.key.weight 1152x1152
decoder_layers.10.attention.attention.key.bias 1152
decoder_layers.10.attention.attention.value.weight 1152x1152
decoder_layers.10.attention.attention.value.bias 1152
decoder_layers.10.attention.output.dense.weight 1152x1152
decoder_layers.10.attention.output.dense.bias 1152
decoder_layers.10.intermediate.dense.weight 4096x1152
decoder_layers.10.intermediate.dense.bias 4096
decoder_layers.10.output.dense.weight 1152x4096
decoder_layers.10.output.dense.bias 1152
decoder_layers.10.layernorm_before.weight 1152
decoder_layers.10.layernorm_before.bias 1152
decoder_layers.10.layernorm_after.weight 1152
decoder_layers.10.layernorm_after.bias 1152
decoder_layers.11.attention.attention.query.weight 1152x1152
decoder_layers.11.attention.attention.query.bias 1152
decoder_layers.11.attention.attention.key.weight 1152x1152
decoder_layers.11.attention.attention.key.bias 1152
decoder_layers.11.attention.attention.value.weight 1152x1152
decoder_layers.11.attention.attention.value.bias 1152
decoder_layers.11.attention.output.dense.weight 1152x1152
decoder_layers.11.attention.output.dense.bias 1152
decoder_layers.11.intermediate.dense.weight 4096x1152
decoder_layers.11.intermediate.dense.bias 4096
decoder_layers.11.output.dense.weight 1152x4096
decoder_layers.11.output.dense.bias 1152
decoder_layers.11.layernorm_before.weight 1152
decoder_layers.11.layernorm_before.bias 1152
decoder_layers.11.layernorm_after.weight 1152
decoder_layers.11.layernorm_after.bias 1152
decoder_layers.12.attention.attention.query.weight 1152x1152
decoder_layers.12.attention.attention.query.bias 1152
decoder_layers.12.attention.attention.key.weight 1152x1152
decoder_layers.12.attention.attention.key.bias 1152
decoder_layers.12.attention.attention.value.weight 1152x1152
decoder_layers.12.attention.attention.value.bias 1152
decoder_layers.12.attention.output.dense.weight 1152x1152
decoder_layers.12.attention.output.dense.bias 1152
decoder_layers.12.intermediate.dense.weight 4096x1152
decoder_layers.12.intermediate.dense.bias 4096
decoder_layers.12.output.dense.weight 1152x4096
decoder_layers.12.output.dense.bias 1152
decoder_layers.12.layernorm_before.weight 1152
decoder_layers.12.layernorm_before.bias 1152
decoder_layers.12.layernorm_after.weight 1152
decoder_layers.12.layernorm_after.bias 1152
decoder_layers.13.attention.attention.query.weight 1152x1152
decoder_layers.13.attention.attention.query.bias 1152
decoder_layers.13.attention.attention.key.weight 1152x1152
decoder_layers.13.attention.attention.key.bias 1152
decoder_layers.13.attention.attention.value.weight 1152x1152
decoder_layers.13.attention.attention.value.bias 1152
decoder_layers.13.attention.output.dense.weight 1152x1152
decoder_layers.13.attention.output.dense.bias 1152
decoder_layers.13.intermediate.dense.weight 4096x1152
decoder_layers.13.intermediate.dense.bias 4096
decoder_layers.13.output.dense.weight 1152x4096
decoder_layers.13.output.dense.bias 1152
decoder_layers.13.layernorm_before.weight 1152
decoder_layers.13.layernorm_before.bias 1152
decoder_layers.13.layernorm_after.weight 1152
decoder_layers.13.layernorm_after.bias 1152
decoder_layers.14.attention.attention.query.weight 1152x1152
decoder_layers.14.attention.attention.query.bias 1152
decoder_layers.14.attention.attention.key.weight 1152x1152
decoder_layers.14.attention.attention.key.bias 1152
decoder_layers.14.attention.attention.value.weight 1152x1152
decoder_layers.14.attention.attention.value.bias 1152
decoder_layers.14.attention.output.dense.weight 1152x1152
decoder_layers.14.attention.output.dense.bias 1152
decoder_layers.14.intermediate.dense.weight 4096x1152
decoder_layers.14.intermediate.dense.bias 4096
decoder_layers.14.output.dense.weight 1152x4096
decoder_layers.14.output.dense.bias 1152
decoder_layers.14.layernorm_before.weight 1152
decoder_layers.14.layernorm_before.bias 1152
decoder_layers.14.layernorm_after.weight 1152
decoder_layers.14.layernorm_after.bias 1152
decoder_layers.15.attention.attention.query.weight 1152x1152
decoder_layers.15.attention.attention.query.bias 1152
decoder_layers.15.attention.attention.key.weight 1152x1152
decoder_layers.15.attention.attention.key.bias 1152
decoder_layers.15.attention.attention.value.weight 1152x1152
decoder_layers.15.attention.attention.value.bias 1152
decoder_layers.15.attention.output.dense.weight 1152x1152
decoder_layers.15.attention.output.dense.bias 1152
decoder_layers.15.intermediate.dense.weight 4096x1152
decoder_layers.15.intermediate.dense.bias 4096
decoder_layers.15.output.dense.weight 1152x4096
decoder_layers.15.output.dense.bias 1152
decoder_layers.15.layernorm_before.weight 1152
decoder_layers.15.layernorm_before.bias 1152
decoder_layers.15.layernorm_after.weight 1152
decoder_layers.15.layernorm_after.bias 1152
decoder_layers.16.attention.attention.query.weight 1152x1152
decoder_layers.16.attention.attention.query.bias 1152
decoder_layers.16.attention.attention.key.weight 1152x1152
decoder_layers.16.attention.attention.key.bias 1152
decoder_layers.16.attention.attention.value.weight 1152x1152
decoder_layers.16.attention.attention.value.bias 1152
decoder_layers.16.attention.output.dense.weight 1152x1152
decoder_layers.16.attention.output.dense.bias 1152
decoder_layers.16.intermediate.dense.weight 4096x1152
decoder_layers.16.intermediate.dense.bias 4096
decoder_layers.16.output.dense.weight 1152x4096
decoder_layers.16.output.dense.bias 1152
decoder_layers.16.layernorm_before.weight 1152
decoder_layers.16.layernorm_before.bias 1152
decoder_layers.16.layernorm_after.weight 1152
decoder_layers.16.layernorm_after.bias 1152
decoder_layers.17.attention.attention.query.weight 1152x1152
decoder_layers.17.attention.attention.query.bias 1152
decoder_layers.17.attention.attention.key.weight 1152x1152
decoder_layers.17.attention.attention.key.bias 1152
decoder_layers.17.attention.attention.value.weight 1152x1152
decoder_layers.17.attention.attention.value.bias 1152
decoder_layers.17.attention.output.dense.weight 1152x1152
decoder_layers.17.attention.output.dense.bias 1152
decoder_layers.17.intermediate.dense.weight 4096x1152
decoder_layers.17.intermediate.dense.bias 4096
decoder_layers.17.output.dense.weight 1152x4096
decoder_layers.17.output.dense.bias 1152
decoder_layers.17.layernorm_before.weight 1152
decoder_layers.17.layernorm_before.bias 1152
decoder_layers.17.layernorm_after.weight 1152
decoder_layers.17.layernorm_after.bias 1152
decoder_layers.18.attention.attention.query.weight 1152x1152
decoder_layers.18.attention.attention.query.bias 1152
decoder_layers.18.attention.attention.key.weight 1152x1152
decoder_layers.18.attention.attention.key.bias 1152
decoder_layers.18.attention.attention.value.weight 1152x1152
decoder_layers.18.attention.attention.value.bias 1152
decoder_layers.18.attention.output.dense.weight 1152x1152
decoder_layers.18.attention.output.dense.bias 1152
decoder_layers.18.intermediate.dense.weight 4096x1152
decoder_layers.18.intermediate.dense.bias 4096
decoder_layers.18.output.dense.weight 1152x4096
decoder_layers.18.output.dense.bias 1152
decoder_layers.18.layernorm_before.weight 1152
decoder_layers.18.layernorm_before.bias 1152
decoder_layers.18.layernorm_after.weight 1152
decoder_layers.18.layernorm_after.bias 1152
decoder_layers.19.attention.attention.query.weight 1152x1152
decoder_layers.19.attention.attention.query.bias 1152
decoder_layers.19.attention.attention.key.weight 1152x1152
decoder_layers.19.attention.attention.key.bias 1152
decoder_layers.19.attention.attention.value.weight 1152x1152
decoder_layers.19.attention.attention.value.bias 1152
decoder_layers.19.attention.output.dense.weight 1152x1152
decoder_layers.19.attention.output.dense.bias 1152
decoder_layers.19.intermediate.dense.weight 4096x1152
decoder_layers.19.intermediate.dense.bias 4096
decoder_layers.19.output.dense.weight 1152x4096
decoder_layers.19.output.dense.bias 1152
decoder_layers.19.layernorm_before.weight 1152
decoder_layers.19.layernorm_before.bias 1152
decoder_layers.19.layernorm_after.weight 1152
decoder_layers.19.layernorm_after.bias 1152
decoder_layers.20.attention.attention.query.weight 1152x1152
decoder_layers.20.attention.attention.query.bias 1152
decoder_layers.20.attention.attention.key.weight 1152x1152
decoder_layers.20.attention.attention.key.bias 1152
decoder_layers.20.attention.attention.value.weight 1152x1152
decoder_layers.20.attention.attention.value.bias 1152
decoder_layers.20.attention.output.dense.weight 1152x1152
decoder_layers.20.attention.output.dense.bias 1152
decoder_layers.20.intermediate.dense.weight 4096x1152
decoder_layers.20.intermediate.dense.bias 4096
decoder_layers.20.output.dense.weight 1152x4096
decoder_layers.20.output.dense.bias 1152
decoder_layers.20.layernorm_before.weight 1152
decoder_layers.20.layernorm_before.bias 1152
decoder_layers.20.layernorm_after.weight 1152
decoder_layers.20.layernorm_after.bias 1152
decoder_layers.21.attention.attention.query.weight 1152x1152
decoder_layers.21.attention.attention.query.bias 1152
decoder_layers.21.attention.attention.key.weight 1152x1152
decoder_layers.21.attention.attention.key.bias 1152
decoder_layers.21.attention.attention.value.weight 1152x1152
decoder_layers.21.attention.attention.value.bias 1152
decoder_layers.21.attention.output.dense.weight 1152x1152
decoder_layers.21.attention.output.dense.bias 1152
decoder_layers.21.intermediate.dense.weight 4096x1152
decoder_layers.21.intermediate.dense.bias 4096
decoder_layers.21.output.dense.weight 1152x4096
decoder_layers.21.output.dense.bias 1152
decoder_layers.21.layernorm_before.weight 1152
decoder_layers.21.layernorm_before.bias 1152
decoder_layers.21.layernorm_after.weight 1152
decoder_layers.21.layernorm_after.bias 1152
decoder_layers.22.attention.attention.query.weight 1152x1152
decoder_layers.22.attention.attention.query.bias 1152
decoder_layers.22.attention.attention.key.weight 1152x1152
decoder_layers.22.attention.attention.key.bias 1152
decoder_layers.22.attention.attention.value.weight 1152x1152
decoder_layers.22.attention.attention.value.bias 1152
decoder_layers.22.attention.output.dense.weight 1152x1152
decoder_layers.22.attention.output.dense.bias 1152
decoder_layers.22.intermediate.dense.weight 4096x1152
decoder_layers.22.intermediate.dense.bias 4096
decoder_layers.22.output.dense.weight 1152x4096
decoder_layers.22.output.dense.bias 1152
decoder_layers.22.layernorm_before.weight 1152
decoder_layers.22.layernorm_before.bias 1152
decoder_layers.22.layernorm_after.weight 1152
decoder_layers.22.layernorm_after.bias 1152
decoder_layers.23.attention.attention.query.weight 1152x1152
decoder_layers.23.attention.attention.query.bias 1152
decoder_layers.23.attention.attention.key.weight 1152x1152
decoder_layers.23.attention.attention.key.bias 1152
decoder_layers.23.attention.attention.value.weight 1152x1152
decoder_layers.23.attention.attention.value.bias 1152
decoder_layers.23.attention.output.dense.weight 1152x1152
decoder_layers.23.attention.output.dense.bias 1152
decoder_layers.23.intermediate.dense.weight 4096x1152
decoder_layers.23.intermediate.dense.bias 4096
decoder_layers.23.output.dense.weight 1152x4096
decoder_layers.23.output.dense.bias 1152
decoder_layers.23.layernorm_before.weight 1152
decoder_layers.23.layernorm_before.bias 1152
decoder_layers.23.layernorm_after.weight 1152
decoder_layers.23.layernorm_after.bias 1152
decoder_layers.24.attention.attention.query.weight 1152x1152
decoder_layers.24.attention.attention.query.bias 1152
decoder_layers.24.attention.attention.key.weight 1152x1152
decoder_layers.24.attention.attention.key.bias 1152
decoder_layers.24.attention.attention.value.weight 1152x1152
decoder_layers.24.attention.attention.value.bias 1152
decoder_layers.24.attention.output.dense.weight 1152x1152
decoder_layers.24.attention.output.dense.bias 1152
decoder_layers.24.intermediate.dense.weight 4096x1152
decoder_layers.24.intermediate.dense.bias 4096
decoder_layers.24.output.dense.weight 1152x4096
decoder_layers.24.output.dense.bias 1152
decoder_layers.24.layernorm_before.weight 1152
decoder_layers.24.layernorm_before.bias 1152
decoder_layers.24.layernorm_after.weight 1152
decoder_layers.24.layernorm_after.bias 1152
decoder_layers.25.attention.attention.query.weight 1152x1152
decoder_layers.25.attention.attention.query.bias 1152
decoder_layers.25.attention.attention.key.weight 1152x1152
decoder_layers.25.attention.attention.key.bias 1152
decoder_layers.25.attention.attention.value.weight 1152x1152
decoder_layers.25.attention.attention.value.bias 1152
decoder_layers.25.attention.output.dense.weight 1152x1152
decoder_layers.25.attention.output.dense.bias 1152
decoder_layers.25.intermediate.dense.weight 4096x1152
decoder_layers.25.intermediate.dense.bias 4096
decoder_layers.25.output.dense.weight 1152x4096
decoder_layers.25.output.dense.bias 1152
decoder_layers.25.layernorm_before.weight 1152
decoder_layers.25.layernorm_before.bias 1152
decoder_layers.25.layernorm_after.weight 1152
decoder_layers.25.layernorm_after.bias 1152
decoder_layers.26.attention.attention.query.weight 1152x1152
decoder_layers.26.attention.attention.query.bias 1152
decoder_layers.26.attention.attention.key.weight 1152x1152
decoder_layers.26.attention.attention.key.bias 1152
decoder_layers.26.attention.attention.value.weight 1152x1152
decoder_layers.26.attention.attention.value.bias 1152
decoder_layers.26.attention.output.dense.weight 1152x1152
decoder_layers.26.attention.output.dense.bias 1152
decoder_layers.26.intermediate.dense.weight 4096x1152
decoder_layers.26.intermediate.dense.bias 4096
decoder_layers.26.output.dense.weight 1152x4096
decoder_layers.26.output.dense.bias 1152
decoder_layers.26.layernorm_before.weight 1152
decoder_layers.26.layernorm_before.bias 1152
decoder_layers.26.layernorm_after.weight 1152
decoder_layers.26.layernorm_after.bias 1152
decoder_layers.27.attention.attention.query.weight 1152x1152
decoder_layers.27.attention.attention.query.bias 1152
decoder_layers.27.attention.attention.key.weight 1152x1152
decoder_layers.27.attention.attention.key.bias 1152
decoder_layers.27.attention.attention.value.weight 1152x1152
decoder_layers.27.attention.attention.value.bias 1152
decoder_layers.27.attention.output.dense.weight 1152x1152
decoder_layers.27.attention.output.dense.bias 1152
decoder_layers.27.intermediate.dense.weight 4096x1152
decoder_layers.27.intermediate.dense.bias 4096
decoder_layers.27.output.dense.weight 1152x4096
decoder_layers.27.output.dense.bias 1152
decoder_layers.27.layernorm_before.weight 1152
decoder_layers.27.layernorm_before.bias 1152
decoder_layers.27.layernorm_after.weight 1152
decoder_layers.27.layernorm_after.bias 1152
decoder_norm.weight 1152
decoder_norm.bias 1152
decoder_pred.weight 768x1152
decoder_pred.bias 768
