ComfyUI Node
Model Merge ERNIE Image
Block-wise merge of two ERNIE Image. 0.0 keeps the first, 1.0 uses the second, in-between values blend linearly. Weights not covered by any widget are kept from the first.
Model Merge ERNIE Image
- model1
- model2
- MODEL
◄x_embedder0.00►
◄text_proj0.00►
◄time_embedding0.00►
◄layers.00.00►
◄layers.0.self_attention0.00►
◄layers.0.self_attention.to_q0.00►
◄layers.0.self_attention.to_k0.00►
◄layers.0.self_attention.to_v0.00►
◄layers.0.self_attention.to_out0.00►
◄layers.0.self_attention.norm_q0.00►
◄layers.0.self_attention.norm_k0.00►
◄layers.0.mlp0.00►
◄layers.0.mlp.gate_proj0.00►
◄layers.0.mlp.up_proj0.00►
◄layers.0.mlp.linear_fc20.00►
◄layers.0.adaLN_mlp_ln0.00►
◄layers.0.adaLN_sa_ln0.00►
◄layers.10.00►
◄layers.1.self_attention0.00►
◄layers.1.self_attention.to_q0.00►
◄layers.1.self_attention.to_k0.00►
◄layers.1.self_attention.to_v0.00►
◄layers.1.self_attention.to_out0.00►
◄layers.1.self_attention.norm_q0.00►
◄layers.1.self_attention.norm_k0.00►
◄layers.1.mlp0.00►
◄layers.1.mlp.gate_proj0.00►
◄layers.1.mlp.up_proj0.00►
◄layers.1.mlp.linear_fc20.00►
◄layers.1.adaLN_mlp_ln0.00►
◄layers.1.adaLN_sa_ln0.00►
◄layers.20.00►
◄layers.2.self_attention0.00►
◄layers.2.self_attention.to_q0.00►
◄layers.2.self_attention.to_k0.00►
◄layers.2.self_attention.to_v0.00►
◄layers.2.self_attention.to_out0.00►
◄layers.2.self_attention.norm_q0.00►
◄layers.2.self_attention.norm_k0.00►
◄layers.2.mlp0.00►
◄layers.2.mlp.gate_proj0.00►
◄layers.2.mlp.up_proj0.00►
◄layers.2.mlp.linear_fc20.00►
◄layers.2.adaLN_mlp_ln0.00►
◄layers.2.adaLN_sa_ln0.00►
◄layers.30.00►
◄layers.3.self_attention0.00►
◄layers.3.self_attention.to_q0.00►
◄layers.3.self_attention.to_k0.00►
◄layers.3.self_attention.to_v0.00►
◄layers.3.self_attention.to_out0.00►
◄layers.3.self_attention.norm_q0.00►
◄layers.3.self_attention.norm_k0.00►
◄layers.3.mlp0.00►
◄layers.3.mlp.gate_proj0.00►
◄layers.3.mlp.up_proj0.00►
◄layers.3.mlp.linear_fc20.00►
◄layers.3.adaLN_mlp_ln0.00►
◄layers.3.adaLN_sa_ln0.00►
◄layers.40.00►
◄layers.4.self_attention0.00►
◄layers.4.self_attention.to_q0.00►
◄layers.4.self_attention.to_k0.00►
◄layers.4.self_attention.to_v0.00►
◄layers.4.self_attention.to_out0.00►
◄layers.4.self_attention.norm_q0.00►
◄layers.4.self_attention.norm_k0.00►
◄layers.4.mlp0.00►
◄layers.4.mlp.gate_proj0.00►
◄layers.4.mlp.up_proj0.00►
◄layers.4.mlp.linear_fc20.00►
◄layers.4.adaLN_mlp_ln0.00►
◄layers.4.adaLN_sa_ln0.00►
◄layers.50.00►
◄layers.5.self_attention0.00►
◄layers.5.self_attention.to_q0.00►
◄layers.5.self_attention.to_k0.00►
◄layers.5.self_attention.to_v0.00►
◄layers.5.self_attention.to_out0.00►
◄layers.5.self_attention.norm_q0.00►
◄layers.5.self_attention.norm_k0.00►
◄layers.5.mlp0.00►
◄layers.5.mlp.gate_proj0.00►
◄layers.5.mlp.up_proj0.00►
◄layers.5.mlp.linear_fc20.00►
◄layers.5.adaLN_mlp_ln0.00►
◄layers.5.adaLN_sa_ln0.00►
◄layers.60.00►
◄layers.6.self_attention0.00►
◄layers.6.self_attention.to_q0.00►
◄layers.6.self_attention.to_k0.00►
◄layers.6.self_attention.to_v0.00►
◄layers.6.self_attention.to_out0.00►
◄layers.6.self_attention.norm_q0.00►
◄layers.6.self_attention.norm_k0.00►
◄layers.6.mlp0.00►
◄layers.6.mlp.gate_proj0.00►
◄layers.6.mlp.up_proj0.00►
◄layers.6.mlp.linear_fc20.00►
◄layers.6.adaLN_mlp_ln0.00►
◄layers.6.adaLN_sa_ln0.00►
◄layers.70.00►
◄layers.7.self_attention0.00►
◄layers.7.self_attention.to_q0.00►
◄layers.7.self_attention.to_k0.00►
◄layers.7.self_attention.to_v0.00►
◄layers.7.self_attention.to_out0.00►
◄layers.7.self_attention.norm_q0.00►
◄layers.7.self_attention.norm_k0.00►
◄layers.7.mlp0.00►
◄layers.7.mlp.gate_proj0.00►
◄layers.7.mlp.up_proj0.00►
◄layers.7.mlp.linear_fc20.00►
◄layers.7.adaLN_mlp_ln0.00►
◄layers.7.adaLN_sa_ln0.00►
◄layers.80.00►
◄layers.8.self_attention0.00►
◄layers.8.self_attention.to_q0.00►
◄layers.8.self_attention.to_k0.00►
◄layers.8.self_attention.to_v0.00►
◄layers.8.self_attention.to_out0.00►
◄layers.8.self_attention.norm_q0.00►
◄layers.8.self_attention.norm_k0.00►
◄layers.8.mlp0.00►
◄layers.8.mlp.gate_proj0.00►
◄layers.8.mlp.up_proj0.00►
◄layers.8.mlp.linear_fc20.00►
◄layers.8.adaLN_mlp_ln0.00►
◄layers.8.adaLN_sa_ln0.00►
◄layers.90.00►
◄layers.9.self_attention0.00►
◄layers.9.self_attention.to_q0.00►
◄layers.9.self_attention.to_k0.00►
◄layers.9.self_attention.to_v0.00►
◄layers.9.self_attention.to_out0.00►
◄layers.9.self_attention.norm_q0.00►
◄layers.9.self_attention.norm_k0.00►
◄layers.9.mlp0.00►
◄layers.9.mlp.gate_proj0.00►
◄layers.9.mlp.up_proj0.00►
◄layers.9.mlp.linear_fc20.00►
◄layers.9.adaLN_mlp_ln0.00►
◄layers.9.adaLN_sa_ln0.00►
◄layers.100.00►
◄layers.10.self_attention0.00►
◄layers.10.self_attention.to_q0.00►
◄layers.10.self_attention.to_k0.00►
◄layers.10.self_attention.to_v0.00►
◄layers.10.self_attention.to_out0.00►
◄layers.10.self_attention.norm_q0.00►
◄layers.10.self_attention.norm_k0.00►
◄layers.10.mlp0.00►
◄layers.10.mlp.gate_proj0.00►
◄layers.10.mlp.up_proj0.00►
◄layers.10.mlp.linear_fc20.00►
◄layers.10.adaLN_mlp_ln0.00►
◄layers.10.adaLN_sa_ln0.00►
◄layers.110.00►
◄layers.11.self_attention0.00►
◄layers.11.self_attention.to_q0.00►
◄layers.11.self_attention.to_k0.00►
◄layers.11.self_attention.to_v0.00►
◄layers.11.self_attention.to_out0.00►
◄layers.11.self_attention.norm_q0.00►
◄layers.11.self_attention.norm_k0.00►
◄layers.11.mlp0.00►
◄layers.11.mlp.gate_proj0.00►
◄layers.11.mlp.up_proj0.00►
◄layers.11.mlp.linear_fc20.00►
◄layers.11.adaLN_mlp_ln0.00►
◄layers.11.adaLN_sa_ln0.00►
◄layers.120.00►
◄layers.12.self_attention0.00►
◄layers.12.self_attention.to_q0.00►
◄layers.12.self_attention.to_k0.00►
◄layers.12.self_attention.to_v0.00►
◄layers.12.self_attention.to_out0.00►
◄layers.12.self_attention.norm_q0.00►
◄layers.12.self_attention.norm_k0.00►
◄layers.12.mlp0.00►
◄layers.12.mlp.gate_proj0.00►
◄layers.12.mlp.up_proj0.00►
◄layers.12.mlp.linear_fc20.00►
◄layers.12.adaLN_mlp_ln0.00►
◄layers.12.adaLN_sa_ln0.00►
◄layers.130.00►
◄layers.13.self_attention0.00►
◄layers.13.self_attention.to_q0.00►
◄layers.13.self_attention.to_k0.00►
◄layers.13.self_attention.to_v0.00►
◄layers.13.self_attention.to_out0.00►
◄layers.13.self_attention.norm_q0.00►
◄layers.13.self_attention.norm_k0.00►
◄layers.13.mlp0.00►
◄layers.13.mlp.gate_proj0.00►
◄layers.13.mlp.up_proj0.00►
◄layers.13.mlp.linear_fc20.00►
◄layers.13.adaLN_mlp_ln0.00►
◄layers.13.adaLN_sa_ln0.00►
◄layers.140.00►
◄layers.14.self_attention0.00►
◄layers.14.self_attention.to_q0.00►
◄layers.14.self_attention.to_k0.00►
◄layers.14.self_attention.to_v0.00►
◄layers.14.self_attention.to_out0.00►
◄layers.14.self_attention.norm_q0.00►
◄layers.14.self_attention.norm_k0.00►
◄layers.14.mlp0.00►
◄layers.14.mlp.gate_proj0.00►
◄layers.14.mlp.up_proj0.00►
◄layers.14.mlp.linear_fc20.00►
◄layers.14.adaLN_mlp_ln0.00►
◄layers.14.adaLN_sa_ln0.00►
◄layers.150.00►
◄layers.15.self_attention0.00►
◄layers.15.self_attention.to_q0.00►
◄layers.15.self_attention.to_k0.00►
◄layers.15.self_attention.to_v0.00►
◄layers.15.self_attention.to_out0.00►
◄layers.15.self_attention.norm_q0.00►
◄layers.15.self_attention.norm_k0.00►
◄layers.15.mlp0.00►
◄layers.15.mlp.gate_proj0.00►
◄layers.15.mlp.up_proj0.00►
◄layers.15.mlp.linear_fc20.00►
◄layers.15.adaLN_mlp_ln0.00►
◄layers.15.adaLN_sa_ln0.00►
◄layers.160.00►
◄layers.16.self_attention0.00►
◄layers.16.self_attention.to_q0.00►
◄layers.16.self_attention.to_k0.00►
◄layers.16.self_attention.to_v0.00►
◄layers.16.self_attention.to_out0.00►
◄layers.16.self_attention.norm_q0.00►
◄layers.16.self_attention.norm_k0.00►
◄layers.16.mlp0.00►
◄layers.16.mlp.gate_proj0.00►
◄layers.16.mlp.up_proj0.00►
◄layers.16.mlp.linear_fc20.00►
◄layers.16.adaLN_mlp_ln0.00►
◄layers.16.adaLN_sa_ln0.00►
◄layers.170.00►
◄layers.17.self_attention0.00►
◄layers.17.self_attention.to_q0.00►
◄layers.17.self_attention.to_k0.00►
◄layers.17.self_attention.to_v0.00►
◄layers.17.self_attention.to_out0.00►
◄layers.17.self_attention.norm_q0.00►
◄layers.17.self_attention.norm_k0.00►
◄layers.17.mlp0.00►
◄layers.17.mlp.gate_proj0.00►
◄layers.17.mlp.up_proj0.00►
◄layers.17.mlp.linear_fc20.00►
◄layers.17.adaLN_mlp_ln0.00►
◄layers.17.adaLN_sa_ln0.00►
◄layers.180.00►
◄layers.18.self_attention0.00►
◄layers.18.self_attention.to_q0.00►
◄layers.18.self_attention.to_k0.00►
◄layers.18.self_attention.to_v0.00►
◄layers.18.self_attention.to_out0.00►
◄layers.18.self_attention.norm_q0.00►
◄layers.18.self_attention.norm_k0.00►
◄layers.18.mlp0.00►
◄layers.18.mlp.gate_proj0.00►
◄layers.18.mlp.up_proj0.00►
◄layers.18.mlp.linear_fc20.00►
◄layers.18.adaLN_mlp_ln0.00►
◄layers.18.adaLN_sa_ln0.00►
◄layers.190.00►
◄layers.19.self_attention0.00►
◄layers.19.self_attention.to_q0.00►
◄layers.19.self_attention.to_k0.00►
◄layers.19.self_attention.to_v0.00►
◄layers.19.self_attention.to_out0.00►
◄layers.19.self_attention.norm_q0.00►
◄layers.19.self_attention.norm_k0.00►
◄layers.19.mlp0.00►
◄layers.19.mlp.gate_proj0.00►
◄layers.19.mlp.up_proj0.00►
◄layers.19.mlp.linear_fc20.00►
◄layers.19.adaLN_mlp_ln0.00►
◄layers.19.adaLN_sa_ln0.00►
◄layers.200.00►
◄layers.20.self_attention0.00►
◄layers.20.self_attention.to_q0.00►
◄layers.20.self_attention.to_k0.00►
◄layers.20.self_attention.to_v0.00►
◄layers.20.self_attention.to_out0.00►
◄layers.20.self_attention.norm_q0.00►
◄layers.20.self_attention.norm_k0.00►
◄layers.20.mlp0.00►
◄layers.20.mlp.gate_proj0.00►
◄layers.20.mlp.up_proj0.00►
◄layers.20.mlp.linear_fc20.00►
◄layers.20.adaLN_mlp_ln0.00►
◄layers.20.adaLN_sa_ln0.00►
◄layers.210.00►
◄layers.21.self_attention0.00►
◄layers.21.self_attention.to_q0.00►
◄layers.21.self_attention.to_k0.00►
◄layers.21.self_attention.to_v0.00►
◄layers.21.self_attention.to_out0.00►
◄layers.21.self_attention.norm_q0.00►
◄layers.21.self_attention.norm_k0.00►
◄layers.21.mlp0.00►
◄layers.21.mlp.gate_proj0.00►
◄layers.21.mlp.up_proj0.00►
◄layers.21.mlp.linear_fc20.00►
◄layers.21.adaLN_mlp_ln0.00►
◄layers.21.adaLN_sa_ln0.00►
◄layers.220.00►
◄layers.22.self_attention0.00►
◄layers.22.self_attention.to_q0.00►
◄layers.22.self_attention.to_k0.00►
◄layers.22.self_attention.to_v0.00►
◄layers.22.self_attention.to_out0.00►
◄layers.22.self_attention.norm_q0.00►
◄layers.22.self_attention.norm_k0.00►
◄layers.22.mlp0.00►
◄layers.22.mlp.gate_proj0.00►
◄layers.22.mlp.up_proj0.00►
◄layers.22.mlp.linear_fc20.00►
◄layers.22.adaLN_mlp_ln0.00►
◄layers.22.adaLN_sa_ln0.00►
◄layers.230.00►
◄layers.23.self_attention0.00►
◄layers.23.self_attention.to_q0.00►
◄layers.23.self_attention.to_k0.00►
◄layers.23.self_attention.to_v0.00►
◄layers.23.self_attention.to_out0.00►
◄layers.23.self_attention.norm_q0.00►
◄layers.23.self_attention.norm_k0.00►
◄layers.23.mlp0.00►
◄layers.23.mlp.gate_proj0.00►
◄layers.23.mlp.up_proj0.00►
◄layers.23.mlp.linear_fc20.00►
◄layers.23.adaLN_mlp_ln0.00►
◄layers.23.adaLN_sa_ln0.00►
◄layers.240.00►
◄layers.24.self_attention0.00►
◄layers.24.self_attention.to_q0.00►
◄layers.24.self_attention.to_k0.00►
◄layers.24.self_attention.to_v0.00►
◄layers.24.self_attention.to_out0.00►
◄layers.24.self_attention.norm_q0.00►
◄layers.24.self_attention.norm_k0.00►
◄layers.24.mlp0.00►
◄layers.24.mlp.gate_proj0.00►
◄layers.24.mlp.up_proj0.00►
◄layers.24.mlp.linear_fc20.00►
◄layers.24.adaLN_mlp_ln0.00►
◄layers.24.adaLN_sa_ln0.00►
◄layers.250.00►
◄layers.25.self_attention0.00►
◄layers.25.self_attention.to_q0.00►
◄layers.25.self_attention.to_k0.00►
◄layers.25.self_attention.to_v0.00►
◄layers.25.self_attention.to_out0.00►
◄layers.25.self_attention.norm_q0.00►
◄layers.25.self_attention.norm_k0.00►
◄layers.25.mlp0.00►
◄layers.25.mlp.gate_proj0.00►
◄layers.25.mlp.up_proj0.00►
◄layers.25.mlp.linear_fc20.00►
◄layers.25.adaLN_mlp_ln0.00►
◄layers.25.adaLN_sa_ln0.00►
◄layers.260.00►
◄layers.26.self_attention0.00►
◄layers.26.self_attention.to_q0.00►
◄layers.26.self_attention.to_k0.00►
◄layers.26.self_attention.to_v0.00►
◄layers.26.self_attention.to_out0.00►
◄layers.26.self_attention.norm_q0.00►
◄layers.26.self_attention.norm_k0.00►
◄layers.26.mlp0.00►
◄layers.26.mlp.gate_proj0.00►
◄layers.26.mlp.up_proj0.00►
◄layers.26.mlp.linear_fc20.00►
◄layers.26.adaLN_mlp_ln0.00►
◄layers.26.adaLN_sa_ln0.00►
◄layers.270.00►
◄layers.27.self_attention0.00►
◄layers.27.self_attention.to_q0.00►
◄layers.27.self_attention.to_k0.00►
◄layers.27.self_attention.to_v0.00►
◄layers.27.self_attention.to_out0.00►
◄layers.27.self_attention.norm_q0.00►
◄layers.27.self_attention.norm_k0.00►
◄layers.27.mlp0.00►
◄layers.27.mlp.gate_proj0.00►
◄layers.27.mlp.up_proj0.00►
◄layers.27.mlp.linear_fc20.00►
◄layers.27.adaLN_mlp_ln0.00►
◄layers.27.adaLN_sa_ln0.00►
◄layers.280.00►
◄layers.28.self_attention0.00►
◄layers.28.self_attention.to_q0.00►
◄layers.28.self_attention.to_k0.00►
◄layers.28.self_attention.to_v0.00►
◄layers.28.self_attention.to_out0.00►
◄layers.28.self_attention.norm_q0.00►
◄layers.28.self_attention.norm_k0.00►
◄layers.28.mlp0.00►
◄layers.28.mlp.gate_proj0.00►
◄layers.28.mlp.up_proj0.00►
◄layers.28.mlp.linear_fc20.00►
◄layers.28.adaLN_mlp_ln0.00►
◄layers.28.adaLN_sa_ln0.00►
◄layers.290.00►
◄layers.29.self_attention0.00►
◄layers.29.self_attention.to_q0.00►
◄layers.29.self_attention.to_k0.00►
◄layers.29.self_attention.to_v0.00►
◄layers.29.self_attention.to_out0.00►
◄layers.29.self_attention.norm_q0.00►
◄layers.29.self_attention.norm_k0.00►
◄layers.29.mlp0.00►
◄layers.29.mlp.gate_proj0.00►
◄layers.29.mlp.up_proj0.00►
◄layers.29.mlp.linear_fc20.00►
◄layers.29.adaLN_mlp_ln0.00►
◄layers.29.adaLN_sa_ln0.00►
◄layers.300.00►
◄layers.30.self_attention0.00►
◄layers.30.self_attention.to_q0.00►
◄layers.30.self_attention.to_k0.00►
◄layers.30.self_attention.to_v0.00►
◄layers.30.self_attention.to_out0.00►
◄layers.30.self_attention.norm_q0.00►
◄layers.30.self_attention.norm_k0.00►
◄layers.30.mlp0.00►
◄layers.30.mlp.gate_proj0.00►
◄layers.30.mlp.up_proj0.00►
◄layers.30.mlp.linear_fc20.00►
◄layers.30.adaLN_mlp_ln0.00►
◄layers.30.adaLN_sa_ln0.00►
◄layers.310.00►
◄layers.31.self_attention0.00►
◄layers.31.self_attention.to_q0.00►
◄layers.31.self_attention.to_k0.00►
◄layers.31.self_attention.to_v0.00►
◄layers.31.self_attention.to_out0.00►
◄layers.31.self_attention.norm_q0.00►
◄layers.31.self_attention.norm_k0.00►
◄layers.31.mlp0.00►
◄layers.31.mlp.gate_proj0.00►
◄layers.31.mlp.up_proj0.00►
◄layers.31.mlp.linear_fc20.00►
◄layers.31.adaLN_mlp_ln0.00►
◄layers.31.adaLN_sa_ln0.00►
◄layers.320.00►
◄layers.32.self_attention0.00►
◄layers.32.self_attention.to_q0.00►
◄layers.32.self_attention.to_k0.00►
◄layers.32.self_attention.to_v0.00►
◄layers.32.self_attention.to_out0.00►
◄layers.32.self_attention.norm_q0.00►
◄layers.32.self_attention.norm_k0.00►
◄layers.32.mlp0.00►
◄layers.32.mlp.gate_proj0.00►
◄layers.32.mlp.up_proj0.00►
◄layers.32.mlp.linear_fc20.00►
◄layers.32.adaLN_mlp_ln0.00►
◄layers.32.adaLN_sa_ln0.00►
◄layers.330.00►
◄layers.33.self_attention0.00►
◄layers.33.self_attention.to_q0.00►
◄layers.33.self_attention.to_k0.00►
◄layers.33.self_attention.to_v0.00►
◄layers.33.self_attention.to_out0.00►
◄layers.33.self_attention.norm_q0.00►
◄layers.33.self_attention.norm_k0.00►
◄layers.33.mlp0.00►
◄layers.33.mlp.gate_proj0.00►
◄layers.33.mlp.up_proj0.00►
◄layers.33.mlp.linear_fc20.00►
◄layers.33.adaLN_mlp_ln0.00►
◄layers.33.adaLN_sa_ln0.00►
◄layers.340.00►
◄layers.34.self_attention0.00►
◄layers.34.self_attention.to_q0.00►
◄layers.34.self_attention.to_k0.00►
◄layers.34.self_attention.to_v0.00►
◄layers.34.self_attention.to_out0.00►
◄layers.34.self_attention.norm_q0.00►
◄layers.34.self_attention.norm_k0.00►
◄layers.34.mlp0.00►
◄layers.34.mlp.gate_proj0.00►
◄layers.34.mlp.up_proj0.00►
◄layers.34.mlp.linear_fc20.00►
◄layers.34.adaLN_mlp_ln0.00►
◄layers.34.adaLN_sa_ln0.00►
◄layers.350.00►
◄layers.35.self_attention0.00►
◄layers.35.self_attention.to_q0.00►
◄layers.35.self_attention.to_k0.00►
◄layers.35.self_attention.to_v0.00►
◄layers.35.self_attention.to_out0.00►
◄layers.35.self_attention.norm_q0.00►
◄layers.35.self_attention.norm_k0.00►
◄layers.35.mlp0.00►
◄layers.35.mlp.gate_proj0.00►
◄layers.35.mlp.up_proj0.00►
◄layers.35.mlp.linear_fc20.00►
◄layers.35.adaLN_mlp_ln0.00►
◄layers.35.adaLN_sa_ln0.00►
◄adaLN_modulation0.00►
◄final_norm0.00►
◄final_linear0.00►
Categoryadvanced/model_merging/model_specific
Inputs (512)
| Name | Type | Default | Description |
|---|---|---|---|
| model1 | MODEL | — | |
| model2 | MODEL | — | |
| x_embedder | FLOAT | 0.000–1 | — |
| text_proj | FLOAT | 0.000–1 | — |
| time_embedding | FLOAT | 0.000–1 | — |
| layers.0 | FLOAT | 0.000–1 | — |
| layers.0.self_attention | FLOAT | 0.000–1 | — |
| layers.0.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.0.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.0.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.0.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.0.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.0.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.0.mlp | FLOAT | 0.000–1 | — |
| layers.0.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.0.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.0.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.0.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.0.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.1 | FLOAT | 0.000–1 | — |
| layers.1.self_attention | FLOAT | 0.000–1 | — |
| layers.1.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.1.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.1.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.1.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.1.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.1.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.1.mlp | FLOAT | 0.000–1 | — |
| layers.1.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.1.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.1.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.1.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.1.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.2 | FLOAT | 0.000–1 | — |
| layers.2.self_attention | FLOAT | 0.000–1 | — |
| layers.2.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.2.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.2.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.2.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.2.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.2.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.2.mlp | FLOAT | 0.000–1 | — |
| layers.2.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.2.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.2.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.2.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.2.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.3 | FLOAT | 0.000–1 | — |
| layers.3.self_attention | FLOAT | 0.000–1 | — |
| layers.3.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.3.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.3.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.3.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.3.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.3.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.3.mlp | FLOAT | 0.000–1 | — |
| layers.3.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.3.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.3.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.3.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.3.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.4 | FLOAT | 0.000–1 | — |
| layers.4.self_attention | FLOAT | 0.000–1 | — |
| layers.4.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.4.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.4.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.4.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.4.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.4.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.4.mlp | FLOAT | 0.000–1 | — |
| layers.4.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.4.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.4.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.4.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.4.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.5 | FLOAT | 0.000–1 | — |
| layers.5.self_attention | FLOAT | 0.000–1 | — |
| layers.5.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.5.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.5.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.5.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.5.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.5.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.5.mlp | FLOAT | 0.000–1 | — |
| layers.5.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.5.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.5.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.5.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.5.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.6 | FLOAT | 0.000–1 | — |
| layers.6.self_attention | FLOAT | 0.000–1 | — |
| layers.6.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.6.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.6.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.6.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.6.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.6.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.6.mlp | FLOAT | 0.000–1 | — |
| layers.6.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.6.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.6.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.6.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.6.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.7 | FLOAT | 0.000–1 | — |
| layers.7.self_attention | FLOAT | 0.000–1 | — |
| layers.7.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.7.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.7.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.7.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.7.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.7.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.7.mlp | FLOAT | 0.000–1 | — |
| layers.7.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.7.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.7.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.7.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.7.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.8 | FLOAT | 0.000–1 | — |
| layers.8.self_attention | FLOAT | 0.000–1 | — |
| layers.8.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.8.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.8.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.8.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.8.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.8.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.8.mlp | FLOAT | 0.000–1 | — |
| layers.8.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.8.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.8.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.8.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.8.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.9 | FLOAT | 0.000–1 | — |
| layers.9.self_attention | FLOAT | 0.000–1 | — |
| layers.9.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.9.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.9.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.9.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.9.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.9.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.9.mlp | FLOAT | 0.000–1 | — |
| layers.9.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.9.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.9.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.9.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.9.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.10 | FLOAT | 0.000–1 | — |
| layers.10.self_attention | FLOAT | 0.000–1 | — |
| layers.10.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.10.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.10.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.10.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.10.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.10.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.10.mlp | FLOAT | 0.000–1 | — |
| layers.10.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.10.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.10.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.10.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.10.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.11 | FLOAT | 0.000–1 | — |
| layers.11.self_attention | FLOAT | 0.000–1 | — |
| layers.11.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.11.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.11.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.11.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.11.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.11.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.11.mlp | FLOAT | 0.000–1 | — |
| layers.11.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.11.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.11.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.11.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.11.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.12 | FLOAT | 0.000–1 | — |
| layers.12.self_attention | FLOAT | 0.000–1 | — |
| layers.12.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.12.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.12.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.12.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.12.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.12.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.12.mlp | FLOAT | 0.000–1 | — |
| layers.12.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.12.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.12.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.12.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.12.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.13 | FLOAT | 0.000–1 | — |
| layers.13.self_attention | FLOAT | 0.000–1 | — |
| layers.13.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.13.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.13.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.13.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.13.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.13.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.13.mlp | FLOAT | 0.000–1 | — |
| layers.13.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.13.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.13.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.13.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.13.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.14 | FLOAT | 0.000–1 | — |
| layers.14.self_attention | FLOAT | 0.000–1 | — |
| layers.14.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.14.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.14.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.14.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.14.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.14.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.14.mlp | FLOAT | 0.000–1 | — |
| layers.14.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.14.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.14.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.14.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.14.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.15 | FLOAT | 0.000–1 | — |
| layers.15.self_attention | FLOAT | 0.000–1 | — |
| layers.15.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.15.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.15.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.15.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.15.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.15.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.15.mlp | FLOAT | 0.000–1 | — |
| layers.15.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.15.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.15.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.15.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.15.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.16 | FLOAT | 0.000–1 | — |
| layers.16.self_attention | FLOAT | 0.000–1 | — |
| layers.16.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.16.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.16.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.16.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.16.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.16.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.16.mlp | FLOAT | 0.000–1 | — |
| layers.16.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.16.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.16.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.16.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.16.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.17 | FLOAT | 0.000–1 | — |
| layers.17.self_attention | FLOAT | 0.000–1 | — |
| layers.17.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.17.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.17.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.17.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.17.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.17.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.17.mlp | FLOAT | 0.000–1 | — |
| layers.17.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.17.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.17.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.17.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.17.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.18 | FLOAT | 0.000–1 | — |
| layers.18.self_attention | FLOAT | 0.000–1 | — |
| layers.18.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.18.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.18.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.18.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.18.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.18.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.18.mlp | FLOAT | 0.000–1 | — |
| layers.18.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.18.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.18.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.18.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.18.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.19 | FLOAT | 0.000–1 | — |
| layers.19.self_attention | FLOAT | 0.000–1 | — |
| layers.19.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.19.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.19.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.19.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.19.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.19.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.19.mlp | FLOAT | 0.000–1 | — |
| layers.19.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.19.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.19.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.19.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.19.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.20 | FLOAT | 0.000–1 | — |
| layers.20.self_attention | FLOAT | 0.000–1 | — |
| layers.20.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.20.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.20.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.20.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.20.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.20.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.20.mlp | FLOAT | 0.000–1 | — |
| layers.20.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.20.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.20.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.20.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.20.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.21 | FLOAT | 0.000–1 | — |
| layers.21.self_attention | FLOAT | 0.000–1 | — |
| layers.21.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.21.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.21.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.21.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.21.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.21.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.21.mlp | FLOAT | 0.000–1 | — |
| layers.21.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.21.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.21.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.21.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.21.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.22 | FLOAT | 0.000–1 | — |
| layers.22.self_attention | FLOAT | 0.000–1 | — |
| layers.22.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.22.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.22.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.22.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.22.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.22.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.22.mlp | FLOAT | 0.000–1 | — |
| layers.22.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.22.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.22.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.22.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.22.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.23 | FLOAT | 0.000–1 | — |
| layers.23.self_attention | FLOAT | 0.000–1 | — |
| layers.23.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.23.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.23.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.23.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.23.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.23.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.23.mlp | FLOAT | 0.000–1 | — |
| layers.23.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.23.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.23.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.23.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.23.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.24 | FLOAT | 0.000–1 | — |
| layers.24.self_attention | FLOAT | 0.000–1 | — |
| layers.24.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.24.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.24.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.24.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.24.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.24.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.24.mlp | FLOAT | 0.000–1 | — |
| layers.24.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.24.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.24.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.24.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.24.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.25 | FLOAT | 0.000–1 | — |
| layers.25.self_attention | FLOAT | 0.000–1 | — |
| layers.25.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.25.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.25.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.25.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.25.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.25.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.25.mlp | FLOAT | 0.000–1 | — |
| layers.25.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.25.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.25.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.25.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.25.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.26 | FLOAT | 0.000–1 | — |
| layers.26.self_attention | FLOAT | 0.000–1 | — |
| layers.26.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.26.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.26.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.26.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.26.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.26.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.26.mlp | FLOAT | 0.000–1 | — |
| layers.26.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.26.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.26.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.26.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.26.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.27 | FLOAT | 0.000–1 | — |
| layers.27.self_attention | FLOAT | 0.000–1 | — |
| layers.27.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.27.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.27.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.27.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.27.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.27.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.27.mlp | FLOAT | 0.000–1 | — |
| layers.27.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.27.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.27.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.27.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.27.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.28 | FLOAT | 0.000–1 | — |
| layers.28.self_attention | FLOAT | 0.000–1 | — |
| layers.28.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.28.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.28.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.28.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.28.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.28.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.28.mlp | FLOAT | 0.000–1 | — |
| layers.28.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.28.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.28.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.28.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.28.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.29 | FLOAT | 0.000–1 | — |
| layers.29.self_attention | FLOAT | 0.000–1 | — |
| layers.29.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.29.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.29.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.29.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.29.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.29.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.29.mlp | FLOAT | 0.000–1 | — |
| layers.29.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.29.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.29.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.29.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.29.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.30 | FLOAT | 0.000–1 | — |
| layers.30.self_attention | FLOAT | 0.000–1 | — |
| layers.30.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.30.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.30.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.30.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.30.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.30.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.30.mlp | FLOAT | 0.000–1 | — |
| layers.30.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.30.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.30.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.30.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.30.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.31 | FLOAT | 0.000–1 | — |
| layers.31.self_attention | FLOAT | 0.000–1 | — |
| layers.31.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.31.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.31.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.31.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.31.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.31.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.31.mlp | FLOAT | 0.000–1 | — |
| layers.31.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.31.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.31.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.31.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.31.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.32 | FLOAT | 0.000–1 | — |
| layers.32.self_attention | FLOAT | 0.000–1 | — |
| layers.32.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.32.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.32.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.32.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.32.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.32.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.32.mlp | FLOAT | 0.000–1 | — |
| layers.32.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.32.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.32.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.32.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.32.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.33 | FLOAT | 0.000–1 | — |
| layers.33.self_attention | FLOAT | 0.000–1 | — |
| layers.33.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.33.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.33.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.33.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.33.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.33.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.33.mlp | FLOAT | 0.000–1 | — |
| layers.33.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.33.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.33.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.33.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.33.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.34 | FLOAT | 0.000–1 | — |
| layers.34.self_attention | FLOAT | 0.000–1 | — |
| layers.34.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.34.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.34.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.34.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.34.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.34.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.34.mlp | FLOAT | 0.000–1 | — |
| layers.34.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.34.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.34.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.34.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.34.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| layers.35 | FLOAT | 0.000–1 | — |
| layers.35.self_attention | FLOAT | 0.000–1 | — |
| layers.35.self_attention.to_q | FLOAT | 0.000–1 | — |
| layers.35.self_attention.to_k | FLOAT | 0.000–1 | — |
| layers.35.self_attention.to_v | FLOAT | 0.000–1 | — |
| layers.35.self_attention.to_out | FLOAT | 0.000–1 | — |
| layers.35.self_attention.norm_q | FLOAT | 0.000–1 | — |
| layers.35.self_attention.norm_k | FLOAT | 0.000–1 | — |
| layers.35.mlp | FLOAT | 0.000–1 | — |
| layers.35.mlp.gate_proj | FLOAT | 0.000–1 | — |
| layers.35.mlp.up_proj | FLOAT | 0.000–1 | — |
| layers.35.mlp.linear_fc2 | FLOAT | 0.000–1 | — |
| layers.35.adaLN_mlp_ln | FLOAT | 0.000–1 | — |
| layers.35.adaLN_sa_ln | FLOAT | 0.000–1 | — |
| adaLN_modulation | FLOAT | 0.000–1 | — |
| final_norm | FLOAT | 0.000–1 | — |
| final_linear | FLOAT | 0.000–1 | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| MODEL | MODEL | — |