| { |
| "block_size": 16, |
| "scheme": "nvfp4_e2m1_block_fp8scale", |
| "awq_linears": 360, |
| "plain_linears": 144, |
| "tasks_calibrated": [ |
| "x2t_image", |
| "t2i" |
| ], |
| "fp4_lut_pos": [ |
| 0.0, |
| 0.5, |
| 1.0, |
| 1.5, |
| 2.0, |
| 3.0, |
| 4.0, |
| 6.0 |
| ], |
| "per_weight": { |
| "language_model.model.layers.0.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.0.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.0.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.0.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.0.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.0.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.0.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.0.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.0.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.0.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.0.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.0.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.0.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.0.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.1.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.1.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.1.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.1.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.1.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.1.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.1.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.1.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.1.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.1.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.1.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.1.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.1.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.1.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.2.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.2.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.2.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.2.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.2.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.2.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.2.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.2.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.2.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.2.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.2.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.2.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.2.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.2.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.3.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.3.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.3.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.3.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.3.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.3.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.3.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.3.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.3.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.3.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.3.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.3.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.3.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.3.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.4.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.4.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.4.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.4.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.4.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.4.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.4.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.4.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.4.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.4.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.4.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.4.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.4.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.4.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.5.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.5.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.5.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.5.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.5.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.5.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.5.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.5.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.5.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.5.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.5.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.5.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.5.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.5.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.6.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.6.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.6.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.6.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.6.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.6.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.6.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.6.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.6.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.6.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.6.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.6.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.6.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.6.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.7.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.7.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.7.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.7.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.7.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.7.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.7.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.7.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.7.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.7.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.7.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.7.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.7.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.7.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.8.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.8.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.8.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.8.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.8.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.8.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.8.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.8.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.8.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.8.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.8.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.8.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.8.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.8.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.9.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.9.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.9.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.9.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.9.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.9.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.9.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.9.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.9.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.9.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.9.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.9.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.9.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.9.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.10.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.10.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.10.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.10.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.10.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.10.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.10.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.10.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.10.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.10.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.10.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.10.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.10.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.10.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.11.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.11.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.11.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.11.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.11.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.11.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.11.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.11.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.11.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.11.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.11.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.11.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.11.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.11.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.12.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.12.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.12.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.12.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.12.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.12.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.12.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.12.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.12.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.12.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.12.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.12.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.12.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.12.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.13.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.13.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.13.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.13.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.13.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.13.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.13.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.13.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.13.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.13.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.13.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.13.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.13.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.13.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.14.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.14.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.14.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.14.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.14.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.14.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.14.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.14.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.14.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.14.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.14.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.14.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.14.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.14.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.15.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.15.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.15.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.15.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.15.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.15.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.15.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.15.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.15.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.15.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.15.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.15.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.15.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.15.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.16.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.16.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.16.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.16.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.16.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.16.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.16.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.16.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.16.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.16.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.16.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.16.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.16.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.16.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.17.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.17.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.17.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.17.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.17.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.17.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.17.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.17.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.17.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.17.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.17.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.17.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.17.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.17.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.18.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.18.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.18.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.18.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.18.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.18.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.18.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.18.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.18.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.18.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.18.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.18.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.18.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.18.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.19.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.19.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.19.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.19.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.19.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.19.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.19.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.19.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.19.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.19.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.19.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.19.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.19.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.19.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.20.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.20.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.20.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.20.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.20.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.20.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.20.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.20.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.20.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.20.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.20.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.20.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.20.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.20.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.21.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.21.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.21.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.21.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.21.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.21.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.21.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.21.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.21.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.21.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.21.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.21.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.21.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.21.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.22.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.22.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.22.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.22.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.22.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.22.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.22.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.22.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.22.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.22.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.22.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.22.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.22.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.22.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.23.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.23.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.23.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.23.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.23.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.23.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.23.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.23.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.23.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.23.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.23.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.23.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.23.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.23.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.24.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.24.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.24.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.24.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.24.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.24.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.24.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.24.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.24.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.24.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.24.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.24.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.24.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.24.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.25.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.25.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.25.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.25.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.25.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.25.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.25.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.25.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.25.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.25.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.25.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.25.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.25.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.25.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.26.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.26.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.26.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.26.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.26.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.26.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.26.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.26.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.26.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.26.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.26.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.26.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.26.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.26.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.27.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.27.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.27.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.27.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.27.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.27.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.27.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.27.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.27.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.27.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.27.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.27.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.27.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.27.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.28.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.28.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.28.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.28.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.28.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.28.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.28.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.28.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.28.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.28.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.28.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.28.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.28.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.28.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.29.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.29.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.29.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.29.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.29.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.29.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.29.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.29.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.29.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.29.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.29.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.29.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.29.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.29.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.30.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.30.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.30.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.30.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.30.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.30.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.30.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.30.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.30.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.30.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.30.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.30.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.30.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.30.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.31.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.31.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.31.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.31.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.31.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.31.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.31.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.31.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.31.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.31.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.31.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.31.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.31.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.31.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.32.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.32.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.32.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.32.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.32.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.32.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.32.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.32.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.32.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.32.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.32.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.32.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.32.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.32.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.33.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.33.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.33.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.33.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.33.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.33.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.33.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.33.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.33.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.33.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.33.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.33.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.33.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.33.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.34.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.34.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.34.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.34.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.34.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.34.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.34.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.34.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.34.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.34.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.34.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.34.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.34.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.34.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.35.self_attn.q_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.35.self_attn.k_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.35.self_attn.v_proj.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.35.self_attn.q_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.35.self_attn.k_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.35.self_attn.v_proj_moe_gen.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.35.mlp.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.35.mlp.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.35.mlp_moe_gen.gate_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.35.mlp_moe_gen.up_proj.weight": { |
| "shape": [ |
| 11008, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_awq" |
| }, |
| "language_model.model.layers.35.self_attn.o_proj.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.35.self_attn.o_proj_moe_gen.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.35.mlp.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| }, |
| "language_model.model.layers.35.mlp_moe_gen.down_proj.weight": { |
| "shape": [ |
| 2048, |
| 11008 |
| ], |
| "block_size": 16, |
| "scheme": "nvfp4_minmax" |
| } |
| } |
| } |