{ "architectures": [ "Qwen3_5ForConditionalGeneration" ], "dtype": "bfloat16", "image_token_id": 248056, "language_model_only": false, "model_type": "qwen3_5", "quantization_config": { "bits": 4.0, "checkpoint_format": "exl3", "codebook": "mcg", "desc_act": false, "format": "exl3", "group_size": -1, "head_bits": 6.0, "lm_head": false, "meta": { "act_group_aware": null, "auto_forward_data_parallel": true, "damp_auto_increment": null, "damp_percent": null, "dense_vram_strategy": "exclusive", "dense_vram_strategy_devices": null, "fallback": { "smooth": null, "strategy": "rtn", "threshold": "0.5%" }, "foem": null, "gc_mode": "interval", "gptaq": null, "moe_vram_strategy": "exclusive", "moe_vram_strategy_devices": null, "mse": null, "offload_to_disk": false, "offload_to_disk_path": null, "pack_impl": "cpu", "quantizer": [ "gptqmodel:7.3.4" ], "static_groups": null, "true_sequential": true, "uri": "https://github.com/modelcloud/gptqmodel", "wait_for_submodule_finalizers": false }, "method": "exl3", "out_scales": "auto", "pack_dtype": "int32", "quant_method": "exl3", "tensor_storage": { "model.language_model.layers.0.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.0.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.0.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.0.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.0.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.0.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.0.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.0.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.0.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.0.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.0.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.0.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.0.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.0.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.0.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.0.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.0.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.0.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.0.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.0.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.0.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.0.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.0.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.0.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.0.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.0.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.0.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.0.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.0.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.0.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.1.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.1.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.1.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.1.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.1.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.1.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.1.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.1.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.1.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.1.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.1.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.1.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.1.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.1.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.1.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.1.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.1.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.1.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.1.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.1.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.1.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.1.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.1.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.1.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.1.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.1.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.1.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.1.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.1.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.1.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.10.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.10.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.10.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.10.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.10.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.10.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.10.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.10.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.10.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.10.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.10.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.10.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.10.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.10.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.10.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.10.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.10.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.10.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.10.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.10.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.10.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.10.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.10.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.10.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.10.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.10.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.10.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.10.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.10.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.10.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.11.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.11.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.11.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.11.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.11.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.11.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.11.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.11.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.11.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.11.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.11.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.11.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.11.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.11.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.11.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.11.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.11.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.11.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.11.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.11.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.11.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.11.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.11.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.11.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.11.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.11.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.11.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.11.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.11.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.11.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.11.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.11.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.11.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.11.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.11.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.12.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.12.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.12.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.12.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.12.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.12.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.12.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.12.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.12.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.12.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.12.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.12.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.12.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.12.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.12.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.12.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.12.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.12.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.12.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.12.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.12.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.12.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.12.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.12.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.12.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.12.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.12.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.12.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.12.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.12.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.13.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.13.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.13.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.13.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.13.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.13.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.13.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.13.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.13.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.13.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.13.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.13.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.13.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.13.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.13.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.13.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.13.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.13.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.13.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.13.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.13.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.13.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.13.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.13.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.13.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.13.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.13.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.13.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.13.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.13.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.14.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.14.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.14.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.14.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.14.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.14.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.14.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.14.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.14.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.14.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.14.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.14.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.14.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.14.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.14.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.14.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.14.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.14.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.14.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.14.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.14.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.14.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.14.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.14.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.14.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.14.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.14.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.14.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.14.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.14.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.15.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.15.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.15.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.15.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.15.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.15.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.15.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.15.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.15.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.15.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.15.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.15.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.15.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.15.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.15.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.15.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.15.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.15.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.15.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.15.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.15.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.15.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.15.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.15.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.15.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.15.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.15.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.15.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.15.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.15.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.15.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.15.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.15.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.15.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.15.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.16.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.16.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.16.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.16.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.16.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.16.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.16.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.16.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.16.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.16.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.16.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.16.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.16.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.16.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.16.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.16.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.16.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.16.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.16.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.16.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.16.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.16.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.16.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.16.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.16.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.16.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.16.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.16.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.16.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.16.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.17.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.17.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.17.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.17.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.17.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.17.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.17.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.17.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.17.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.17.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.17.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.17.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.17.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.17.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.17.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.17.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.17.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.17.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.17.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.17.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.17.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.17.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.17.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.17.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.17.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.17.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.17.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.17.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.17.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.17.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.18.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.18.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.18.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.18.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.18.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.18.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.18.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.18.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.18.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.18.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.18.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.18.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.18.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.18.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.18.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.18.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.18.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.18.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.18.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.18.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.18.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.18.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.18.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.18.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.18.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.18.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.18.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.18.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.18.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.18.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.19.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.19.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.19.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.19.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.19.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.19.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.19.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.19.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.19.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.19.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.19.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.19.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.19.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.19.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.19.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.19.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.19.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.19.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.19.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.19.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.19.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.19.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.19.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.19.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.19.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.19.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.19.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.19.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.19.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.19.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.19.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.19.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.19.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.19.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.19.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.2.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.2.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.2.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.2.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.2.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.2.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.2.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.2.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.2.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.2.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.2.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.2.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.2.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.2.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.2.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.2.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.2.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.2.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.2.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.2.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.2.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.2.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.2.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.2.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.2.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.2.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.2.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.2.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.2.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.2.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.20.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.20.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.20.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.20.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.20.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.20.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.20.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.20.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.20.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.20.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.20.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.20.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.20.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.20.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.20.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.20.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.20.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.20.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.20.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.20.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.20.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.20.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.20.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.20.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.20.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.20.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.20.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.20.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.20.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.20.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.21.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.21.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.21.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.21.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.21.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.21.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.21.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.21.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.21.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.21.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.21.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.21.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.21.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.21.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.21.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.21.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.21.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.21.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.21.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.21.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.21.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.21.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.21.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.21.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.21.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.21.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.21.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.21.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.21.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.21.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.22.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.22.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.22.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.22.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.22.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.22.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.22.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.22.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.22.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.22.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.22.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.22.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.22.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.22.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.22.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.22.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.22.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.22.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.22.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.22.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.22.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.22.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.22.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.22.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.22.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.22.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.22.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.22.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.22.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.22.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.23.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.23.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.23.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.23.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.23.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.23.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.23.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.23.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.23.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.23.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.23.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.23.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.23.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.23.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.23.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.23.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.23.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.23.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.23.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.23.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.23.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.23.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.23.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.23.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.23.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.23.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.23.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.23.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.23.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.23.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.23.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.23.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.23.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.23.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.23.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.24.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.24.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.24.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.24.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.24.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.24.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.24.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.24.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.24.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.24.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.24.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.24.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.24.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.24.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.24.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.24.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.24.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.24.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.24.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.24.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.24.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.24.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.24.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.24.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.24.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.24.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.24.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.24.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.24.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.24.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.25.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.25.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.25.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.25.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.25.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.25.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.25.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.25.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.25.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.25.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.25.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.25.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.25.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.25.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.25.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.25.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.25.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.25.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.25.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.25.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.25.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.25.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.25.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.25.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.25.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.25.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.25.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.25.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.25.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.25.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.26.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.26.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.26.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.26.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.26.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.26.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.26.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.26.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.26.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.26.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.26.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.26.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.26.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.26.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.26.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.26.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.26.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.26.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.26.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.26.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.26.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.26.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.26.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.26.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.26.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.26.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.26.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.26.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.26.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.26.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.27.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.27.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.27.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.27.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.27.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.27.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.27.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.27.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.27.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.27.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.27.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.27.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.27.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.27.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.27.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.27.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.27.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.27.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.27.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.27.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.27.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.27.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.27.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.27.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.27.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.27.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.27.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.27.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.27.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.27.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.27.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.27.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.27.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.27.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.27.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.28.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.28.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.28.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.28.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.28.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.28.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.28.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.28.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.28.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.28.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.28.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.28.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.28.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.28.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.28.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.28.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.28.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.28.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.28.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.28.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.28.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.28.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.28.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.28.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.28.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.28.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.28.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.28.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.28.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.28.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.29.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.29.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.29.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.29.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.29.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.29.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.29.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.29.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.29.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.29.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.29.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.29.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.29.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.29.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.29.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.29.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.29.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.29.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.29.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.29.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.29.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.29.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.29.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.29.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.29.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.29.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.29.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.29.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.29.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.29.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.3.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.3.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.3.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.3.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.3.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.3.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.3.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.3.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.3.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.3.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.3.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.3.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.3.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.3.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.3.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.3.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.3.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.3.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.3.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.3.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.3.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.3.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.3.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.3.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.3.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.3.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.3.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.3.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.3.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.3.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.3.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.3.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.3.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.3.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.3.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.30.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.30.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.30.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.30.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.30.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.30.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.30.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.30.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.30.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.30.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.30.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.30.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.30.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.30.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.30.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.30.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.30.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.30.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.30.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.30.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.30.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.30.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.30.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.30.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.30.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.30.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.30.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.30.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.30.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.30.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.31.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.31.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.31.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.31.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.31.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.31.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.31.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.31.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.31.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.31.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.31.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.31.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.31.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.31.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.31.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.31.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.31.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.31.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.31.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.31.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.31.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.31.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.31.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.31.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.31.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.31.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.31.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.31.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.31.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.31.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.31.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.31.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.31.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.31.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.31.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.32.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.32.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.32.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.32.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.32.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.32.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.32.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.32.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.32.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.32.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.32.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.32.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.32.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.32.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.32.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.32.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.32.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.32.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.32.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.32.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.32.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.32.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.32.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.32.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.32.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.32.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.32.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.32.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.32.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.32.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.33.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.33.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.33.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.33.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.33.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.33.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.33.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.33.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.33.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.33.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.33.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.33.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.33.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.33.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.33.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.33.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.33.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.33.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.33.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.33.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.33.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.33.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.33.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.33.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.33.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.33.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.33.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.33.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.33.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.33.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.34.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.34.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.34.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.34.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.34.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.34.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.34.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.34.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.34.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.34.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.34.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.34.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.34.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.34.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.34.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.34.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.34.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.34.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.34.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.34.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.34.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.34.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.34.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.34.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.34.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.34.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.34.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.34.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.34.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.34.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.35.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.35.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.35.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.35.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.35.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.35.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.35.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.35.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.35.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.35.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.35.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.35.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.35.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.35.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.35.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.35.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.35.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.35.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.35.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.35.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.35.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.35.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.35.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.35.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.35.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.35.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.35.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.35.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.35.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.35.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.35.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.35.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.35.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.35.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.35.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.36.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.36.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.36.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.36.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.36.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.36.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.36.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.36.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.36.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.36.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.36.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.36.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.36.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.36.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.36.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.36.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.36.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.36.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.36.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.36.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.36.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.36.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.36.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.36.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.36.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.36.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.36.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.36.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.36.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.36.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.37.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.37.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.37.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.37.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.37.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.37.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.37.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.37.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.37.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.37.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.37.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.37.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.37.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.37.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.37.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.37.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.37.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.37.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.37.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.37.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.37.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.37.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.37.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.37.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.37.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.37.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.37.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.37.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.37.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.37.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.38.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.38.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.38.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.38.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.38.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.38.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.38.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.38.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.38.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.38.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.38.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.38.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.38.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.38.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.38.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.38.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.38.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.38.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.38.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.38.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.38.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.38.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.38.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.38.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.38.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.38.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.38.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.38.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.38.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.38.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.39.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.39.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.39.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.39.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.39.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.39.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.39.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.39.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.39.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.39.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.39.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.39.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.39.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.39.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.39.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.39.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.39.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.39.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.39.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.39.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.39.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.39.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.39.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.39.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.39.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.39.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.39.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.39.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.39.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.39.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.39.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.39.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.39.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.39.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.39.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.4.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.4.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.4.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.4.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.4.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.4.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.4.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.4.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.4.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.4.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.4.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.4.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.4.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.4.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.4.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.4.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.4.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.4.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.4.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.4.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.4.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.4.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.4.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.4.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.4.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.4.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.4.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.4.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.4.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.4.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.40.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.40.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.40.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.40.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.40.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.40.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.40.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.40.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.40.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.40.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.40.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.40.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.40.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.40.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.40.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.40.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.40.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.40.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.40.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.40.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.40.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.40.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.40.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.40.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.40.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.40.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.40.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.40.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.40.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.40.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.41.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.41.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.41.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.41.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.41.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.41.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.41.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.41.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.41.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.41.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.41.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.41.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.41.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.41.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.41.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.41.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.41.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.41.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.41.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.41.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.41.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.41.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.41.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.41.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.41.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.41.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.41.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.41.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.41.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.41.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.42.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.42.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.42.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.42.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.42.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.42.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.42.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.42.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.42.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.42.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.42.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.42.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.42.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.42.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.42.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.42.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.42.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.42.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.42.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.42.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.42.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.42.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.42.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.42.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.42.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.42.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.42.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.42.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.42.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.42.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.43.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.43.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.43.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.43.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.43.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.43.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.43.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.43.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.43.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.43.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.43.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.43.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.43.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.43.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.43.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.43.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.43.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.43.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.43.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.43.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.43.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.43.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.43.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.43.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.43.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.43.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.43.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.43.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.43.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.43.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.43.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.43.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.43.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.43.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.43.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.44.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.44.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.44.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.44.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.44.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.44.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.44.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.44.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.44.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.44.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.44.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.44.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.44.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.44.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.44.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.44.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.44.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.44.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.44.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.44.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.44.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.44.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.44.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.44.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.44.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.44.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.44.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.44.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.44.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.44.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.45.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.45.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.45.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.45.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.45.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.45.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.45.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.45.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.45.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.45.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.45.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.45.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.45.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.45.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.45.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.45.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.45.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.45.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.45.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.45.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.45.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.45.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.45.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.45.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.45.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.45.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.45.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.45.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.45.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.45.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.46.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.46.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.46.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.46.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.46.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.46.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.46.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.46.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.46.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.46.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.46.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.46.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.46.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.46.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.46.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.46.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.46.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.46.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.46.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.46.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.46.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.46.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.46.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.46.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.46.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.46.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.46.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.46.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.46.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.46.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.47.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.47.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.47.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.47.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.47.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.47.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.47.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.47.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.47.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.47.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.47.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.47.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.47.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.47.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.47.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.47.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.47.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.47.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.47.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.47.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.47.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.47.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.47.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.47.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.47.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.47.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.47.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.47.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.47.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.47.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.47.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.47.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.47.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.47.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.47.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.48.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.48.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.48.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.48.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.48.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.48.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.48.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.48.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.48.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.48.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.48.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.48.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.48.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.48.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.48.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.48.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.48.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.48.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.48.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.48.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.48.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.48.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.48.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.48.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.48.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.48.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.48.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.48.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.48.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.48.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.49.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.49.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.49.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.49.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.49.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.49.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.49.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.49.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.49.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.49.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.49.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.49.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.49.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.49.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.49.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.49.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.49.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.49.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.49.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.49.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.49.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.49.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.49.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.49.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.49.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.49.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.49.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.49.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.49.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.49.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.5.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.5.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.5.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.5.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.5.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.5.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.5.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.5.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.5.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.5.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.5.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.5.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.5.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.5.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.5.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.5.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.5.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.5.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.5.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.5.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.5.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.5.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.5.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.5.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.5.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.5.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.5.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.5.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.5.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.5.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.50.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.50.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.50.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.50.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.50.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.50.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.50.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.50.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.50.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.50.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.50.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.50.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.50.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.50.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.50.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.50.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.50.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.50.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.50.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.50.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.50.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.50.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.50.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.50.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.50.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.50.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.50.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.50.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.50.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.50.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.51.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.51.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.51.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.51.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.51.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.51.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.51.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.51.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.51.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.51.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.51.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.51.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.51.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.51.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.51.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.51.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.51.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.51.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.51.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.51.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.51.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.51.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.51.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.51.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.51.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.51.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.51.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.51.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.51.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.51.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.51.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.51.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.51.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.51.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.51.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.52.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.52.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.52.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.52.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.52.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.52.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.52.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.52.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.52.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.52.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.52.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.52.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.52.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.52.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.52.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.52.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.52.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.52.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.52.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.52.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.52.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.52.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.52.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.52.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.52.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.52.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.52.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.52.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.52.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.52.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.53.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.53.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.53.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.53.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.53.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.53.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.53.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.53.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.53.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.53.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.53.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.53.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.53.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.53.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.53.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.53.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.53.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.53.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.53.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.53.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.53.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.53.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.53.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.53.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.53.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.53.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.53.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.53.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.53.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.53.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.54.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.54.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.54.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.54.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.54.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.54.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.54.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.54.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.54.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.54.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.54.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.54.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.54.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.54.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.54.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.54.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.54.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.54.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.54.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.54.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.54.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.54.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.54.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.54.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.54.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.54.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.54.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.54.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.54.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.54.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.55.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.55.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.55.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.55.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.55.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.55.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.55.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.55.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.55.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.55.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.55.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.55.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.55.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.55.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.55.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.55.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.55.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.55.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.55.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.55.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.55.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.55.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.55.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.55.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.55.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.55.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.55.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.55.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.55.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.55.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.55.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.55.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.55.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.55.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.55.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.56.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.56.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.56.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.56.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.56.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.56.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.56.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.56.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.56.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.56.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.56.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.56.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.56.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.56.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.56.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.56.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.56.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.56.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.56.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.56.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.56.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.56.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.56.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.56.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.56.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.56.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.56.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.56.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.56.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.56.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.57.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.57.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.57.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.57.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.57.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.57.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.57.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.57.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.57.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.57.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.57.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.57.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.57.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.57.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.57.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.57.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.57.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.57.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.57.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.57.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.57.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.57.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.57.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.57.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.57.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.57.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.57.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.57.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.57.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.57.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.58.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.58.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.58.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.58.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.58.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.58.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.58.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.58.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.58.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.58.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.58.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.58.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.58.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.58.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.58.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.58.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.58.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.58.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.58.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.58.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.58.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.58.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.58.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.58.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.58.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.58.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.58.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.58.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.58.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.58.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.59.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.59.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.59.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.59.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.59.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.59.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.59.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.59.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.59.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.59.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.59.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.59.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.59.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.59.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.59.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.59.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.59.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.59.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.59.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.59.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.59.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.59.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.59.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.59.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.59.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.59.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.59.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.59.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.59.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.59.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.59.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.59.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.59.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.59.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.59.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.6.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.6.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.6.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.6.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.6.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.6.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.6.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.6.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.6.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.6.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.6.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.6.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.6.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.6.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.6.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.6.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.6.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.6.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.6.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.6.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.6.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.6.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.6.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.6.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.6.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.6.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.6.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.6.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.6.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.6.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.60.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.60.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.60.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.60.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.60.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.60.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.60.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.60.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.60.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.60.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.60.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.60.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.60.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.60.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.60.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.60.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.60.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.60.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.60.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.60.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.60.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.60.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.60.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.60.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.60.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.60.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.60.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.60.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.60.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.60.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.61.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.61.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.61.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.61.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.61.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.61.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.61.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.61.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.61.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.61.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.61.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.61.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.61.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.61.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.61.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.61.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.61.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.61.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.61.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.61.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.61.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.61.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.61.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.61.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.61.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.61.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.61.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.61.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.61.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.61.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.62.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.62.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.62.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.62.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.62.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.62.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.62.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.62.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.62.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.62.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.62.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.62.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.62.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.62.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.62.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.62.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.62.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.62.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.62.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.62.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.62.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.62.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.62.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.62.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.62.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.62.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.62.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.62.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.62.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.62.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.63.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.63.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.63.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.63.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.63.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.63.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.63.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.63.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.63.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.63.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.63.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.63.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.63.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.63.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.63.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.63.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.63.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.63.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.63.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.63.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.63.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.63.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.63.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.63.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.63.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.63.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.63.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.63.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.63.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.63.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.63.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.63.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.63.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.63.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.63.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.7.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.7.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.7.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.7.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.7.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.7.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.7.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.7.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.7.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.7.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.7.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.7.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.7.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.7.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.7.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.7.self_attn.k_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.7.self_attn.k_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.7.self_attn.k_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.7.self_attn.k_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.7.self_attn.k_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.7.self_attn.o_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.7.self_attn.o_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.7.self_attn.o_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.7.self_attn.o_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.7.self_attn.o_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.7.self_attn.q_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.7.self_attn.q_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.7.self_attn.q_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.7.self_attn.q_proj.svh": { "shape": [ 12288 ], "torch_dtype": "float16" }, "model.language_model.layers.7.self_attn.q_proj.trellis": { "shape": [ 320, 768, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.7.self_attn.v_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.7.self_attn.v_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.7.self_attn.v_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.7.self_attn.v_proj.svh": { "shape": [ 1024 ], "torch_dtype": "float16" }, "model.language_model.layers.7.self_attn.v_proj.trellis": { "shape": [ 320, 64, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.8.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.8.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.8.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.8.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.8.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.8.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.8.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.8.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.8.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.8.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.8.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.8.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.8.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.8.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.8.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.8.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.8.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.8.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.8.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.8.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.8.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.8.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.8.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.8.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.8.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.8.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.8.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.8.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.8.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.8.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.9.linear_attn.in_proj_qkv": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.9.linear_attn.in_proj_qkv.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.9.linear_attn.in_proj_qkv.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.9.linear_attn.in_proj_qkv.svh": { "shape": [ 10240 ], "torch_dtype": "float16" }, "model.language_model.layers.9.linear_attn.in_proj_qkv.trellis": { "shape": [ 320, 640, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.9.linear_attn.in_proj_z": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.9.linear_attn.in_proj_z.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.9.linear_attn.in_proj_z.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.9.linear_attn.in_proj_z.svh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.9.linear_attn.in_proj_z.trellis": { "shape": [ 320, 384, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.9.linear_attn.out_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.9.linear_attn.out_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.9.linear_attn.out_proj.suh": { "shape": [ 6144 ], "torch_dtype": "float16" }, "model.language_model.layers.9.linear_attn.out_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.9.linear_attn.out_proj.trellis": { "shape": [ 384, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.9.mlp.down_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.9.mlp.down_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.9.mlp.down_proj.suh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.9.mlp.down_proj.svh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.9.mlp.down_proj.trellis": { "shape": [ 1088, 320, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.9.mlp.gate_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.9.mlp.gate_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.9.mlp.gate_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.9.mlp.gate_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.9.mlp.gate_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } }, "model.language_model.layers.9.mlp.up_proj": { "bits_per_weight": 4, "mcg_multiplier": 3417055213, "quant_format": "exl3", "stored_tensors": { "model.language_model.layers.9.mlp.up_proj.mcg": { "shape": [], "torch_dtype": "int32" }, "model.language_model.layers.9.mlp.up_proj.suh": { "shape": [ 5120 ], "torch_dtype": "float16" }, "model.language_model.layers.9.mlp.up_proj.svh": { "shape": [ 17408 ], "torch_dtype": "float16" }, "model.language_model.layers.9.mlp.up_proj.trellis": { "shape": [ 320, 1088, 64 ], "torch_dtype": "int16" } } } } }, "rope_parameters": { "rope_theta": 10000.0, "rope_type": "default" }, "text_config": { "attention_bias": false, "attention_dropout": 0.0, "attn_output_gate": true, "bos_token_id": 248044, "dtype": "bfloat16", "eos_token_id": 248044, "full_attention_interval": 4, "head_dim": 256, "hidden_act": "silu", "hidden_size": 5120, "initializer_range": 0.02, "intermediate_size": 17408, "layer_types": [ "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention", "linear_attention", "linear_attention", "linear_attention", "full_attention" ], "linear_conv_kernel_dim": 4, "linear_key_head_dim": 128, "linear_num_key_heads": 16, "linear_num_value_heads": 48, "linear_value_head_dim": 128, "mamba_ssm_dtype": "float32", "max_position_embeddings": 262144, "model_type": "qwen3_5_text", "mtp_num_hidden_layers": 1, "mtp_use_dedicated_embeddings": false, "num_attention_heads": 24, "num_hidden_layers": 64, "num_key_value_heads": 4, "output_gate_type": "swish", "pad_token_id": 248044, "partial_rotary_factor": 0.25, "rms_norm_eps": 1e-06, "rope_parameters": { "mrope_interleaved": true, "mrope_section": [ 11, 11, 10 ], "partial_rotary_factor": 0.25, "rope_theta": 10000000, "rope_type": "default" }, "tie_word_embeddings": false, "use_cache": true, "vocab_size": 248320 }, "tie_word_embeddings": false, "transformers_version": "5.15.0", "use_cache": false, "video_token_id": 248057, "vision_config": { "deepstack_visual_indexes": [], "depth": 27, "dtype": "bfloat16", "hidden_act": "gelu_pytorch_tanh", "hidden_size": 1152, "in_channels": 3, "initializer_range": 0.02, "intermediate_size": 4304, "model_type": "qwen3_5_vision", "num_heads": 16, "num_position_embeddings": 2304, "out_hidden_size": 5120, "patch_size": 16, "spatial_merge_size": 2, "temporal_patch_size": 2 }, "vision_end_token_id": 248054, "vision_start_token_id": 248053 }