diff --git "a/config.json" "b/config.json" --- "a/config.json" +++ "b/config.json" @@ -85,519 +85,3208 @@ "mode": "affine", "model.layers.0.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.0.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.0.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.1.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.1.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.1.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.2.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.2.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.2.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.3.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.3.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.3.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.4.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.4.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.4.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.5.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.5.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.5.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.6.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.6.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.6.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.7.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.7.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.7.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.8.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.8.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.8.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.9.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.9.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.9.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.10.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.10.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.10.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.11.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.11.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.11.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.12.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.12.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.12.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.13.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.13.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.13.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.14.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.14.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.14.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.15.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.15.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.15.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.16.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.16.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.16.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.17.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.17.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.17.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.18.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.18.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.18.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.19.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.19.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.19.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.20.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.20.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.20.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.21.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.21.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.21.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.22.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.22.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.22.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.23.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.23.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.23.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.24.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.24.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.24.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.25.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.25.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.25.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.26.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.26.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.26.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.27.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.27.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.27.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.28.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.28.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.28.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.29.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.29.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.29.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.30.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.30.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.30.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.31.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.31.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.31.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.32.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.32.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.32.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.33.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.33.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.33.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.34.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.34.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.34.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.35.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.35.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.35.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.36.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.36.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.36.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.37.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.37.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.37.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.38.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.38.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.38.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.39.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.39.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.39.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.40.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.40.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.40.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.41.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.41.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.41.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.42.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.42.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.42.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" + }, + "model.embed_tokens": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "lm_head": { + "group_size": 64, + "bits": 4, + "mode": "affine" } }, "quantization_config": { @@ -606,519 +3295,3208 @@ "mode": "affine", "model.layers.0.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.0.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.0.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.1.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.1.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.1.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.2.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.2.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.2.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.3.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.3.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.3.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.4.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.4.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.4.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.5.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.5.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.5.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.6.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.6.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.6.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.7.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.7.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.7.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.8.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.8.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.8.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.9.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.9.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.9.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.10.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.10.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.10.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.11.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.11.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.11.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.12.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.12.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.12.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.13.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.13.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.13.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.14.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.14.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.14.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.15.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.15.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.15.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.16.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.16.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.16.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.17.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.17.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.17.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.18.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.18.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.18.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.19.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.19.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.19.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.20.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.20.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.20.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.21.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.21.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.21.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.22.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.22.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.22.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.23.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.23.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.23.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.24.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.24.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.24.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.25.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.25.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.25.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.26.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.26.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.26.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.27.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.27.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.27.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.28.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.28.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.28.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.29.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.29.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.29.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.30.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.30.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.30.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.31.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.31.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.31.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.32.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.32.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.32.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.33.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.33.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.33.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.34.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.34.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.34.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.35.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.35.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.35.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.36.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.36.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.36.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.37.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.37.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.37.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.38.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.38.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.38.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.39.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.39.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.39.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.40.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.40.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.40.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.41.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.41.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.41.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.42.ffn.switch_mlp.gate_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.42.ffn.switch_mlp.up_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" }, "model.layers.42.ffn.switch_mlp.down_proj": { "group_size": 32, - "bits": 4 + "bits": 4, + "mode": "mxfp4" + }, + "model.embed_tokens": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.0.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.1.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.2.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.3.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.4.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.5.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.6.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.7.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.8.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.9.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.10.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.11.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.12.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.13.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.14.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.15.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.16.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.17.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.18.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.19.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.20.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.21.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.22.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.23.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.24.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.25.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.26.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.27.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.28.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.29.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.30.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.31.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.32.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.33.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.34.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.35.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.36.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.37.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.38.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.39.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.40.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.41.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.wq_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.wo_a": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.wo_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.indexer.wq_b": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.indexer.weights_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.indexer.compressor.wkv": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.attn.indexer.compressor.wgate": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.ffn.shared_experts.gate_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.ffn.shared_experts.up_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "model.layers.42.ffn.shared_experts.down_proj": { + "group_size": 64, + "bits": 4, + "mode": "affine" + }, + "lm_head": { + "group_size": 64, + "bits": 4, + "mode": "affine" } }, "rms_norm_eps": 1e-06,