{ "created_at": "2026-08-02T21:26:21.502647Z", "plan_sha256": "e78d2b28e28cc14bcbfe9bedf61bd436d875f375a77f000f3a68837890e3d881", "records": [ { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.embed_tokens", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.0.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.0.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.0.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.0.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.0.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.0.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.0.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.0.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.1.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.1.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.1.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.1.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.1.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.1.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.1.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.1.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.2.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.2.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.2.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.2.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.2.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.2.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.2.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.2.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.3.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.3.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.3.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.3.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.3.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.3.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.3.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.3.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.3.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.4.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.4.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.4.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.4.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.4.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.4.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.4.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.4.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.5.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.5.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.5.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.5.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.5.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.5.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.5.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.5.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.6.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.6.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.6.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.6.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.6.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.6.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.6.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.6.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.7.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.7.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.7.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.7.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.7.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.7.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.7.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.7.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.7.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.8.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.8.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.8.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.8.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.8.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.8.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.8.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.8.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.9.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.9.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.9.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.9.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.10.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.10.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.10.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.10.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.9.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.9.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.9.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.9.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.10.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.10.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.10.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.10.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.11.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.11.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.11.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.11.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.11.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.11.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.11.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.11.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.11.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.12.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.12.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.12.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.12.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.12.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.12.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.12.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.12.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.13.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.13.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.13.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.13.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.13.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.13.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.13.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.13.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.14.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.14.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.14.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.14.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.14.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.14.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.14.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.14.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.15.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.15.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.15.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.15.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.15.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.15.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.15.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.15.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.15.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.16.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.16.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.16.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.16.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.16.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.16.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.16.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.16.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.17.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.17.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.17.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.17.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.17.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.17.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.17.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.17.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.18.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.18.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.18.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.18.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.18.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.18.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.18.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.18.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.19.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.19.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.19.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.19.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.19.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.19.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.19.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.19.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.19.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.20.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.20.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.20.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.20.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.20.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.20.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.20.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.20.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.21.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.21.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.21.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.21.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.21.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.21.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.21.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.21.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.22.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.22.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.22.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.22.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.22.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.22.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.22.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.22.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.23.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.23.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.23.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.23.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.23.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.23.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.23.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.23.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.23.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.24.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.24.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.24.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.24.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.24.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.24.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.24.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.24.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.25.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.25.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.25.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.25.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.25.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.25.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.25.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.25.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.26.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.26.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.26.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.26.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.26.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.26.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.26.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.26.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.27.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.27.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.27.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.27.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.27.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.27.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.27.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.27.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.27.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.28.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.28.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.28.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.28.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.28.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.28.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.28.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.28.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.29.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.29.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.29.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.29.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.29.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.29.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.29.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.29.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.30.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.30.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.30.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.30.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.30.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.30.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.30.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.30.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.31.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.31.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.31.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.31.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.31.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.31.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.31.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.31.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.31.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.32.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.32.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.32.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.32.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.32.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.32.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.32.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.32.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.33.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.33.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.33.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.33.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.33.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.33.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.33.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.33.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.34.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.34.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.34.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.34.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.34.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.34.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.34.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.34.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.35.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.35.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.35.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.35.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.35.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.35.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.35.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.35.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.35.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.36.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.36.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.36.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.36.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.36.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.36.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.36.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.36.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.37.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.37.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.37.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.37.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.37.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.37.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.37.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.37.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.38.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.38.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.38.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.38.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.38.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.38.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.38.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.38.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.39.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.39.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.39.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.39.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.39.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.39.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.39.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.39.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.39.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.40.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.40.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.40.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.40.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.40.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.40.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.40.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.40.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.41.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.41.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.41.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.41.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.41.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.41.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.41.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.41.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.42.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.42.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.42.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.42.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.42.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.42.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.42.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.42.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.43.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.43.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.43.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.43.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.43.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.43.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.43.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.43.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.43.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 6, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.44.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.44.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.44.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.44.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.44.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.44.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.44.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.44.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 6, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.45.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.45.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.45.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.45.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.45.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.45.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.45.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.45.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 6, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.46.linear_attn.in_proj_ba", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.46.linear_attn.in_proj_qkvz", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.46.linear_attn.out_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.46.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.46.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.46.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.46.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 6, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.46.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 8, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.47.mlp.gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.47.mlp.shared_expert.down_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.47.mlp.shared_expert.gate_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 64, "metadata": {}, "method": "affine", "module_path": "model.layers.47.mlp.shared_expert.up_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 6, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.47.mlp.shared_expert_gate", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.47.self_attn.k_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.47.self_attn.o_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.47.self_attn.q_proj", "note": "MLX-LM affine packing", "success": true }, { "bits": 4, "fallback": false, "group_size": 32, "metadata": {}, "method": "affine", "module_path": "model.layers.47.self_attn.v_proj", "note": "MLX-LM affine packing", "success": true } ], "schema_version": "axquant.quantizer-execution.v1" }