SamuelBang_AesCoder-4B-EXL3 / quantization_config.json
ArtusDev's picture
Upload folder using huggingface_hub
fc514de verified
Raw
History Blame
349 kB
{
"quant_method": "exl3",
"version": "0.0.12",
"bits": 4.0,
"head_bits": 6,
"calibration": {
"rows": 250,
"cols": 2048
},
"out_scales": "auto",
"codebook": "mcg",
"tensor_storage": {
"model.embed_tokens": {
"stored_tensors": {
"model.embed_tokens.weight": {
"shape": [
151936,
2560
],
"n_bytes": 777912320,
"dtype": "torch.float16"
}
}
},
"model.layers.0.input_layernorm": {
"stored_tensors": {
"model.layers.0.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.0.self_attn.q_proj": {
"stored_tensors": {
"model.layers.0.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.0.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.0.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.0.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.0.self_attn.k_proj": {
"stored_tensors": {
"model.layers.0.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.0.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.0.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.0.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.0.self_attn.v_proj": {
"stored_tensors": {
"model.layers.0.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.0.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.0.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.0.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.0.self_attn.o_proj": {
"stored_tensors": {
"model.layers.0.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.0.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.0.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.0.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.0.self_attn.q_norm": {
"stored_tensors": {
"model.layers.0.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.0.self_attn.k_norm": {
"stored_tensors": {
"model.layers.0.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.0.post_attention_layernorm": {
"stored_tensors": {
"model.layers.0.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.0.mlp.up_proj": {
"stored_tensors": {
"model.layers.0.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.0.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.0.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.0.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.0.mlp.gate_proj": {
"stored_tensors": {
"model.layers.0.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.0.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.0.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.0.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.0.mlp.down_proj": {
"stored_tensors": {
"model.layers.0.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.0.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.0.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.0.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.1.input_layernorm": {
"stored_tensors": {
"model.layers.1.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.1.self_attn.q_proj": {
"stored_tensors": {
"model.layers.1.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.1.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.1.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.1.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.1.self_attn.k_proj": {
"stored_tensors": {
"model.layers.1.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.1.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.1.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.1.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.1.self_attn.v_proj": {
"stored_tensors": {
"model.layers.1.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.1.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.1.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.1.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.1.self_attn.o_proj": {
"stored_tensors": {
"model.layers.1.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.1.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.1.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.1.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.1.self_attn.q_norm": {
"stored_tensors": {
"model.layers.1.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.1.self_attn.k_norm": {
"stored_tensors": {
"model.layers.1.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.1.post_attention_layernorm": {
"stored_tensors": {
"model.layers.1.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.1.mlp.up_proj": {
"stored_tensors": {
"model.layers.1.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.1.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.1.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.1.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.1.mlp.gate_proj": {
"stored_tensors": {
"model.layers.1.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.1.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.1.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.1.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.1.mlp.down_proj": {
"stored_tensors": {
"model.layers.1.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.1.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.1.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.1.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.2.input_layernorm": {
"stored_tensors": {
"model.layers.2.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.2.self_attn.q_proj": {
"stored_tensors": {
"model.layers.2.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.2.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.2.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.2.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.2.self_attn.k_proj": {
"stored_tensors": {
"model.layers.2.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.2.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.2.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.2.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.2.self_attn.v_proj": {
"stored_tensors": {
"model.layers.2.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.2.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.2.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.2.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.2.self_attn.o_proj": {
"stored_tensors": {
"model.layers.2.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.2.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.2.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.2.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.2.self_attn.q_norm": {
"stored_tensors": {
"model.layers.2.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.2.self_attn.k_norm": {
"stored_tensors": {
"model.layers.2.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.2.post_attention_layernorm": {
"stored_tensors": {
"model.layers.2.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.2.mlp.up_proj": {
"stored_tensors": {
"model.layers.2.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.2.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.2.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.2.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.2.mlp.gate_proj": {
"stored_tensors": {
"model.layers.2.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.2.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.2.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.2.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.2.mlp.down_proj": {
"stored_tensors": {
"model.layers.2.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.2.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.2.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.2.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.3.input_layernorm": {
"stored_tensors": {
"model.layers.3.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.3.self_attn.q_proj": {
"stored_tensors": {
"model.layers.3.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.3.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.3.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.3.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.3.self_attn.k_proj": {
"stored_tensors": {
"model.layers.3.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.3.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.3.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.3.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.3.self_attn.v_proj": {
"stored_tensors": {
"model.layers.3.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.3.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.3.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.3.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.3.self_attn.o_proj": {
"stored_tensors": {
"model.layers.3.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.3.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.3.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.3.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.3.self_attn.q_norm": {
"stored_tensors": {
"model.layers.3.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.3.self_attn.k_norm": {
"stored_tensors": {
"model.layers.3.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.3.post_attention_layernorm": {
"stored_tensors": {
"model.layers.3.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.3.mlp.up_proj": {
"stored_tensors": {
"model.layers.3.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.3.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.3.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.3.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.3.mlp.gate_proj": {
"stored_tensors": {
"model.layers.3.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.3.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.3.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.3.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.3.mlp.down_proj": {
"stored_tensors": {
"model.layers.3.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.3.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.3.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.3.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.4.input_layernorm": {
"stored_tensors": {
"model.layers.4.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.4.self_attn.q_proj": {
"stored_tensors": {
"model.layers.4.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.4.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.4.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.4.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.4.self_attn.k_proj": {
"stored_tensors": {
"model.layers.4.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.4.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.4.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.4.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.4.self_attn.v_proj": {
"stored_tensors": {
"model.layers.4.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.4.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.4.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.4.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.4.self_attn.o_proj": {
"stored_tensors": {
"model.layers.4.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.4.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.4.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.4.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.4.self_attn.q_norm": {
"stored_tensors": {
"model.layers.4.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.4.self_attn.k_norm": {
"stored_tensors": {
"model.layers.4.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.4.post_attention_layernorm": {
"stored_tensors": {
"model.layers.4.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.4.mlp.up_proj": {
"stored_tensors": {
"model.layers.4.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.4.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.4.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.4.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.4.mlp.gate_proj": {
"stored_tensors": {
"model.layers.4.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.4.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.4.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.4.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.4.mlp.down_proj": {
"stored_tensors": {
"model.layers.4.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.4.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.4.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.4.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.5.input_layernorm": {
"stored_tensors": {
"model.layers.5.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.5.self_attn.q_proj": {
"stored_tensors": {
"model.layers.5.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.5.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.5.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.5.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.5.self_attn.k_proj": {
"stored_tensors": {
"model.layers.5.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.5.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.5.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.5.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.5.self_attn.v_proj": {
"stored_tensors": {
"model.layers.5.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.5.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.5.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.5.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.5.self_attn.o_proj": {
"stored_tensors": {
"model.layers.5.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.5.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.5.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.5.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.5.self_attn.q_norm": {
"stored_tensors": {
"model.layers.5.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.5.self_attn.k_norm": {
"stored_tensors": {
"model.layers.5.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.5.post_attention_layernorm": {
"stored_tensors": {
"model.layers.5.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.5.mlp.up_proj": {
"stored_tensors": {
"model.layers.5.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.5.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.5.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.5.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.5.mlp.gate_proj": {
"stored_tensors": {
"model.layers.5.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.5.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.5.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.5.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.5.mlp.down_proj": {
"stored_tensors": {
"model.layers.5.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.5.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.5.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.5.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.6.input_layernorm": {
"stored_tensors": {
"model.layers.6.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.6.self_attn.q_proj": {
"stored_tensors": {
"model.layers.6.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.6.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.6.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.6.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.6.self_attn.k_proj": {
"stored_tensors": {
"model.layers.6.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.6.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.6.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.6.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.6.self_attn.v_proj": {
"stored_tensors": {
"model.layers.6.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.6.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.6.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.6.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.6.self_attn.o_proj": {
"stored_tensors": {
"model.layers.6.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.6.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.6.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.6.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.6.self_attn.q_norm": {
"stored_tensors": {
"model.layers.6.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.6.self_attn.k_norm": {
"stored_tensors": {
"model.layers.6.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.6.post_attention_layernorm": {
"stored_tensors": {
"model.layers.6.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.6.mlp.up_proj": {
"stored_tensors": {
"model.layers.6.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.6.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.6.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.6.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.6.mlp.gate_proj": {
"stored_tensors": {
"model.layers.6.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.6.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.6.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.6.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.6.mlp.down_proj": {
"stored_tensors": {
"model.layers.6.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.6.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.6.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.6.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.7.input_layernorm": {
"stored_tensors": {
"model.layers.7.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.7.self_attn.q_proj": {
"stored_tensors": {
"model.layers.7.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.7.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.7.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.7.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.7.self_attn.k_proj": {
"stored_tensors": {
"model.layers.7.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.7.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.7.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.7.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.7.self_attn.v_proj": {
"stored_tensors": {
"model.layers.7.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.7.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.7.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.7.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.7.self_attn.o_proj": {
"stored_tensors": {
"model.layers.7.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.7.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.7.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.7.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.7.self_attn.q_norm": {
"stored_tensors": {
"model.layers.7.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.7.self_attn.k_norm": {
"stored_tensors": {
"model.layers.7.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.7.post_attention_layernorm": {
"stored_tensors": {
"model.layers.7.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.7.mlp.up_proj": {
"stored_tensors": {
"model.layers.7.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.7.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.7.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.7.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.7.mlp.gate_proj": {
"stored_tensors": {
"model.layers.7.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.7.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.7.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.7.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.7.mlp.down_proj": {
"stored_tensors": {
"model.layers.7.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.7.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.7.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.7.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.8.input_layernorm": {
"stored_tensors": {
"model.layers.8.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.8.self_attn.q_proj": {
"stored_tensors": {
"model.layers.8.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.8.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.8.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.8.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.8.self_attn.k_proj": {
"stored_tensors": {
"model.layers.8.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.8.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.8.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.8.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.8.self_attn.v_proj": {
"stored_tensors": {
"model.layers.8.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.8.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.8.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.8.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.8.self_attn.o_proj": {
"stored_tensors": {
"model.layers.8.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.8.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.8.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.8.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.8.self_attn.q_norm": {
"stored_tensors": {
"model.layers.8.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.8.self_attn.k_norm": {
"stored_tensors": {
"model.layers.8.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.8.post_attention_layernorm": {
"stored_tensors": {
"model.layers.8.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.8.mlp.up_proj": {
"stored_tensors": {
"model.layers.8.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.8.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.8.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.8.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.8.mlp.gate_proj": {
"stored_tensors": {
"model.layers.8.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.8.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.8.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.8.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.8.mlp.down_proj": {
"stored_tensors": {
"model.layers.8.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.8.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.8.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.8.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.9.input_layernorm": {
"stored_tensors": {
"model.layers.9.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.9.self_attn.q_proj": {
"stored_tensors": {
"model.layers.9.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.9.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.9.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.9.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.9.self_attn.k_proj": {
"stored_tensors": {
"model.layers.9.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.9.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.9.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.9.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.9.self_attn.v_proj": {
"stored_tensors": {
"model.layers.9.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.9.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.9.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.9.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.9.self_attn.o_proj": {
"stored_tensors": {
"model.layers.9.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.9.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.9.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.9.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.9.self_attn.q_norm": {
"stored_tensors": {
"model.layers.9.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.9.self_attn.k_norm": {
"stored_tensors": {
"model.layers.9.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.9.post_attention_layernorm": {
"stored_tensors": {
"model.layers.9.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.9.mlp.up_proj": {
"stored_tensors": {
"model.layers.9.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.9.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.9.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.9.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.9.mlp.gate_proj": {
"stored_tensors": {
"model.layers.9.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.9.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.9.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.9.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.9.mlp.down_proj": {
"stored_tensors": {
"model.layers.9.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.9.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.9.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.9.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.10.input_layernorm": {
"stored_tensors": {
"model.layers.10.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.10.self_attn.q_proj": {
"stored_tensors": {
"model.layers.10.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.10.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.10.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.10.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.10.self_attn.k_proj": {
"stored_tensors": {
"model.layers.10.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.10.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.10.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.10.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.10.self_attn.v_proj": {
"stored_tensors": {
"model.layers.10.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.10.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.10.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.10.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.10.self_attn.o_proj": {
"stored_tensors": {
"model.layers.10.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.10.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.10.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.10.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.10.self_attn.q_norm": {
"stored_tensors": {
"model.layers.10.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.10.self_attn.k_norm": {
"stored_tensors": {
"model.layers.10.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.10.post_attention_layernorm": {
"stored_tensors": {
"model.layers.10.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.10.mlp.up_proj": {
"stored_tensors": {
"model.layers.10.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.10.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.10.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.10.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.10.mlp.gate_proj": {
"stored_tensors": {
"model.layers.10.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.10.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.10.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.10.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.10.mlp.down_proj": {
"stored_tensors": {
"model.layers.10.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.10.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.10.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.10.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.11.input_layernorm": {
"stored_tensors": {
"model.layers.11.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.11.self_attn.q_proj": {
"stored_tensors": {
"model.layers.11.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.11.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.11.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.11.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.11.self_attn.k_proj": {
"stored_tensors": {
"model.layers.11.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.11.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.11.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.11.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.11.self_attn.v_proj": {
"stored_tensors": {
"model.layers.11.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.11.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.11.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.11.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.11.self_attn.o_proj": {
"stored_tensors": {
"model.layers.11.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.11.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.11.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.11.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.11.self_attn.q_norm": {
"stored_tensors": {
"model.layers.11.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.11.self_attn.k_norm": {
"stored_tensors": {
"model.layers.11.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.11.post_attention_layernorm": {
"stored_tensors": {
"model.layers.11.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.11.mlp.up_proj": {
"stored_tensors": {
"model.layers.11.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.11.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.11.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.11.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.11.mlp.gate_proj": {
"stored_tensors": {
"model.layers.11.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.11.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.11.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.11.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.11.mlp.down_proj": {
"stored_tensors": {
"model.layers.11.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.11.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.11.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.11.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.12.input_layernorm": {
"stored_tensors": {
"model.layers.12.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.12.self_attn.q_proj": {
"stored_tensors": {
"model.layers.12.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.12.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.12.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.12.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.12.self_attn.k_proj": {
"stored_tensors": {
"model.layers.12.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.12.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.12.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.12.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.12.self_attn.v_proj": {
"stored_tensors": {
"model.layers.12.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.12.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.12.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.12.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.12.self_attn.o_proj": {
"stored_tensors": {
"model.layers.12.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.12.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.12.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.12.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.12.self_attn.q_norm": {
"stored_tensors": {
"model.layers.12.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.12.self_attn.k_norm": {
"stored_tensors": {
"model.layers.12.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.12.post_attention_layernorm": {
"stored_tensors": {
"model.layers.12.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.12.mlp.up_proj": {
"stored_tensors": {
"model.layers.12.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.12.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.12.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.12.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.12.mlp.gate_proj": {
"stored_tensors": {
"model.layers.12.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.12.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.12.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.12.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.12.mlp.down_proj": {
"stored_tensors": {
"model.layers.12.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.12.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.12.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.12.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.13.input_layernorm": {
"stored_tensors": {
"model.layers.13.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.13.self_attn.q_proj": {
"stored_tensors": {
"model.layers.13.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.13.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.13.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.13.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.13.self_attn.k_proj": {
"stored_tensors": {
"model.layers.13.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.13.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.13.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.13.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.13.self_attn.v_proj": {
"stored_tensors": {
"model.layers.13.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.13.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.13.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.13.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.13.self_attn.o_proj": {
"stored_tensors": {
"model.layers.13.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.13.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.13.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.13.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.13.self_attn.q_norm": {
"stored_tensors": {
"model.layers.13.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.13.self_attn.k_norm": {
"stored_tensors": {
"model.layers.13.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.13.post_attention_layernorm": {
"stored_tensors": {
"model.layers.13.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.13.mlp.up_proj": {
"stored_tensors": {
"model.layers.13.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.13.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.13.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.13.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.13.mlp.gate_proj": {
"stored_tensors": {
"model.layers.13.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.13.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.13.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.13.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.13.mlp.down_proj": {
"stored_tensors": {
"model.layers.13.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.13.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.13.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.13.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.14.input_layernorm": {
"stored_tensors": {
"model.layers.14.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.14.self_attn.q_proj": {
"stored_tensors": {
"model.layers.14.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.14.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.14.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.14.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.14.self_attn.k_proj": {
"stored_tensors": {
"model.layers.14.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.14.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.14.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.14.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.14.self_attn.v_proj": {
"stored_tensors": {
"model.layers.14.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.14.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.14.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.14.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.14.self_attn.o_proj": {
"stored_tensors": {
"model.layers.14.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.14.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.14.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.14.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.14.self_attn.q_norm": {
"stored_tensors": {
"model.layers.14.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.14.self_attn.k_norm": {
"stored_tensors": {
"model.layers.14.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.14.post_attention_layernorm": {
"stored_tensors": {
"model.layers.14.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.14.mlp.up_proj": {
"stored_tensors": {
"model.layers.14.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.14.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.14.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.14.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.14.mlp.gate_proj": {
"stored_tensors": {
"model.layers.14.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.14.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.14.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.14.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.14.mlp.down_proj": {
"stored_tensors": {
"model.layers.14.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.14.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.14.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.14.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.15.input_layernorm": {
"stored_tensors": {
"model.layers.15.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.15.self_attn.q_proj": {
"stored_tensors": {
"model.layers.15.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.15.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.15.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.15.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.15.self_attn.k_proj": {
"stored_tensors": {
"model.layers.15.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.15.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.15.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.15.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.15.self_attn.v_proj": {
"stored_tensors": {
"model.layers.15.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.15.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.15.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.15.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.15.self_attn.o_proj": {
"stored_tensors": {
"model.layers.15.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.15.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.15.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.15.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.15.self_attn.q_norm": {
"stored_tensors": {
"model.layers.15.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.15.self_attn.k_norm": {
"stored_tensors": {
"model.layers.15.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.15.post_attention_layernorm": {
"stored_tensors": {
"model.layers.15.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.15.mlp.up_proj": {
"stored_tensors": {
"model.layers.15.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.15.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.15.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.15.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.15.mlp.gate_proj": {
"stored_tensors": {
"model.layers.15.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.15.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.15.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.15.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.15.mlp.down_proj": {
"stored_tensors": {
"model.layers.15.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.15.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.15.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.15.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.16.input_layernorm": {
"stored_tensors": {
"model.layers.16.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.16.self_attn.q_proj": {
"stored_tensors": {
"model.layers.16.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.16.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.16.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.16.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.16.self_attn.k_proj": {
"stored_tensors": {
"model.layers.16.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.16.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.16.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.16.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.16.self_attn.v_proj": {
"stored_tensors": {
"model.layers.16.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.16.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.16.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.16.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.16.self_attn.o_proj": {
"stored_tensors": {
"model.layers.16.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.16.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.16.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.16.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.16.self_attn.q_norm": {
"stored_tensors": {
"model.layers.16.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.16.self_attn.k_norm": {
"stored_tensors": {
"model.layers.16.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.16.post_attention_layernorm": {
"stored_tensors": {
"model.layers.16.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.16.mlp.up_proj": {
"stored_tensors": {
"model.layers.16.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.16.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.16.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.16.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.16.mlp.gate_proj": {
"stored_tensors": {
"model.layers.16.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.16.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.16.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.16.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.16.mlp.down_proj": {
"stored_tensors": {
"model.layers.16.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.16.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.16.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.16.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.17.input_layernorm": {
"stored_tensors": {
"model.layers.17.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.17.self_attn.q_proj": {
"stored_tensors": {
"model.layers.17.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.17.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.17.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.17.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.17.self_attn.k_proj": {
"stored_tensors": {
"model.layers.17.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.17.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.17.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.17.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.17.self_attn.v_proj": {
"stored_tensors": {
"model.layers.17.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.17.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.17.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.17.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.17.self_attn.o_proj": {
"stored_tensors": {
"model.layers.17.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.17.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.17.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.17.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.17.self_attn.q_norm": {
"stored_tensors": {
"model.layers.17.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.17.self_attn.k_norm": {
"stored_tensors": {
"model.layers.17.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.17.post_attention_layernorm": {
"stored_tensors": {
"model.layers.17.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.17.mlp.up_proj": {
"stored_tensors": {
"model.layers.17.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.17.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.17.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.17.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.17.mlp.gate_proj": {
"stored_tensors": {
"model.layers.17.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.17.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.17.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.17.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.17.mlp.down_proj": {
"stored_tensors": {
"model.layers.17.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.17.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.17.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.17.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.18.input_layernorm": {
"stored_tensors": {
"model.layers.18.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.18.self_attn.q_proj": {
"stored_tensors": {
"model.layers.18.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.18.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.18.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.18.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.18.self_attn.k_proj": {
"stored_tensors": {
"model.layers.18.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.18.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.18.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.18.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.18.self_attn.v_proj": {
"stored_tensors": {
"model.layers.18.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.18.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.18.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.18.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.18.self_attn.o_proj": {
"stored_tensors": {
"model.layers.18.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.18.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.18.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.18.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.18.self_attn.q_norm": {
"stored_tensors": {
"model.layers.18.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.18.self_attn.k_norm": {
"stored_tensors": {
"model.layers.18.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.18.post_attention_layernorm": {
"stored_tensors": {
"model.layers.18.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.18.mlp.up_proj": {
"stored_tensors": {
"model.layers.18.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.18.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.18.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.18.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.18.mlp.gate_proj": {
"stored_tensors": {
"model.layers.18.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.18.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.18.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.18.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.18.mlp.down_proj": {
"stored_tensors": {
"model.layers.18.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.18.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.18.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.18.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.19.input_layernorm": {
"stored_tensors": {
"model.layers.19.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.19.self_attn.q_proj": {
"stored_tensors": {
"model.layers.19.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.19.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.19.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.19.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.19.self_attn.k_proj": {
"stored_tensors": {
"model.layers.19.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.19.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.19.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.19.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.19.self_attn.v_proj": {
"stored_tensors": {
"model.layers.19.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.19.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.19.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.19.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.19.self_attn.o_proj": {
"stored_tensors": {
"model.layers.19.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.19.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.19.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.19.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.19.self_attn.q_norm": {
"stored_tensors": {
"model.layers.19.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.19.self_attn.k_norm": {
"stored_tensors": {
"model.layers.19.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.19.post_attention_layernorm": {
"stored_tensors": {
"model.layers.19.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.19.mlp.up_proj": {
"stored_tensors": {
"model.layers.19.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.19.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.19.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.19.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.19.mlp.gate_proj": {
"stored_tensors": {
"model.layers.19.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.19.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.19.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.19.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.19.mlp.down_proj": {
"stored_tensors": {
"model.layers.19.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.19.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.19.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.19.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.20.input_layernorm": {
"stored_tensors": {
"model.layers.20.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.20.self_attn.q_proj": {
"stored_tensors": {
"model.layers.20.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.20.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.20.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.20.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.20.self_attn.k_proj": {
"stored_tensors": {
"model.layers.20.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.20.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.20.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.20.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.20.self_attn.v_proj": {
"stored_tensors": {
"model.layers.20.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.20.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.20.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.20.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.20.self_attn.o_proj": {
"stored_tensors": {
"model.layers.20.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.20.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.20.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.20.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.20.self_attn.q_norm": {
"stored_tensors": {
"model.layers.20.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.20.self_attn.k_norm": {
"stored_tensors": {
"model.layers.20.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.20.post_attention_layernorm": {
"stored_tensors": {
"model.layers.20.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.20.mlp.up_proj": {
"stored_tensors": {
"model.layers.20.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.20.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.20.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.20.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.20.mlp.gate_proj": {
"stored_tensors": {
"model.layers.20.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.20.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.20.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.20.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.20.mlp.down_proj": {
"stored_tensors": {
"model.layers.20.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.20.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.20.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.20.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.21.input_layernorm": {
"stored_tensors": {
"model.layers.21.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.21.self_attn.q_proj": {
"stored_tensors": {
"model.layers.21.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.21.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.21.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.21.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.21.self_attn.k_proj": {
"stored_tensors": {
"model.layers.21.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.21.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.21.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.21.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.21.self_attn.v_proj": {
"stored_tensors": {
"model.layers.21.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.21.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.21.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.21.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.21.self_attn.o_proj": {
"stored_tensors": {
"model.layers.21.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.21.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.21.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.21.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.21.self_attn.q_norm": {
"stored_tensors": {
"model.layers.21.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.21.self_attn.k_norm": {
"stored_tensors": {
"model.layers.21.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.21.post_attention_layernorm": {
"stored_tensors": {
"model.layers.21.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.21.mlp.up_proj": {
"stored_tensors": {
"model.layers.21.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.21.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.21.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.21.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.21.mlp.gate_proj": {
"stored_tensors": {
"model.layers.21.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.21.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.21.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.21.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.21.mlp.down_proj": {
"stored_tensors": {
"model.layers.21.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.21.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.21.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.21.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.22.input_layernorm": {
"stored_tensors": {
"model.layers.22.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.22.self_attn.q_proj": {
"stored_tensors": {
"model.layers.22.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.22.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.22.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.22.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.22.self_attn.k_proj": {
"stored_tensors": {
"model.layers.22.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.22.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.22.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.22.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.22.self_attn.v_proj": {
"stored_tensors": {
"model.layers.22.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.22.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.22.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.22.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.22.self_attn.o_proj": {
"stored_tensors": {
"model.layers.22.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.22.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.22.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.22.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.22.self_attn.q_norm": {
"stored_tensors": {
"model.layers.22.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.22.self_attn.k_norm": {
"stored_tensors": {
"model.layers.22.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.22.post_attention_layernorm": {
"stored_tensors": {
"model.layers.22.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.22.mlp.up_proj": {
"stored_tensors": {
"model.layers.22.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.22.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.22.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.22.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.22.mlp.gate_proj": {
"stored_tensors": {
"model.layers.22.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.22.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.22.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.22.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.22.mlp.down_proj": {
"stored_tensors": {
"model.layers.22.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.22.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.22.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.22.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.23.input_layernorm": {
"stored_tensors": {
"model.layers.23.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.23.self_attn.q_proj": {
"stored_tensors": {
"model.layers.23.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.23.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.23.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.23.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.23.self_attn.k_proj": {
"stored_tensors": {
"model.layers.23.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.23.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.23.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.23.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.23.self_attn.v_proj": {
"stored_tensors": {
"model.layers.23.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.23.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.23.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.23.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.23.self_attn.o_proj": {
"stored_tensors": {
"model.layers.23.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.23.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.23.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.23.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.23.self_attn.q_norm": {
"stored_tensors": {
"model.layers.23.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.23.self_attn.k_norm": {
"stored_tensors": {
"model.layers.23.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.23.post_attention_layernorm": {
"stored_tensors": {
"model.layers.23.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.23.mlp.up_proj": {
"stored_tensors": {
"model.layers.23.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.23.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.23.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.23.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.23.mlp.gate_proj": {
"stored_tensors": {
"model.layers.23.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.23.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.23.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.23.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.23.mlp.down_proj": {
"stored_tensors": {
"model.layers.23.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.23.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.23.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.23.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.24.input_layernorm": {
"stored_tensors": {
"model.layers.24.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.24.self_attn.q_proj": {
"stored_tensors": {
"model.layers.24.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.24.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.24.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.24.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.24.self_attn.k_proj": {
"stored_tensors": {
"model.layers.24.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.24.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.24.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.24.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.24.self_attn.v_proj": {
"stored_tensors": {
"model.layers.24.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.24.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.24.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.24.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.24.self_attn.o_proj": {
"stored_tensors": {
"model.layers.24.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.24.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.24.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.24.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.24.self_attn.q_norm": {
"stored_tensors": {
"model.layers.24.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.24.self_attn.k_norm": {
"stored_tensors": {
"model.layers.24.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.24.post_attention_layernorm": {
"stored_tensors": {
"model.layers.24.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.24.mlp.up_proj": {
"stored_tensors": {
"model.layers.24.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.24.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.24.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.24.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.24.mlp.gate_proj": {
"stored_tensors": {
"model.layers.24.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.24.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.24.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.24.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.24.mlp.down_proj": {
"stored_tensors": {
"model.layers.24.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.24.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.24.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.24.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.25.input_layernorm": {
"stored_tensors": {
"model.layers.25.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.25.self_attn.q_proj": {
"stored_tensors": {
"model.layers.25.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.25.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.25.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.25.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.25.self_attn.k_proj": {
"stored_tensors": {
"model.layers.25.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.25.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.25.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.25.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.25.self_attn.v_proj": {
"stored_tensors": {
"model.layers.25.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.25.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.25.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.25.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.25.self_attn.o_proj": {
"stored_tensors": {
"model.layers.25.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.25.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.25.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.25.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.25.self_attn.q_norm": {
"stored_tensors": {
"model.layers.25.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.25.self_attn.k_norm": {
"stored_tensors": {
"model.layers.25.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.25.post_attention_layernorm": {
"stored_tensors": {
"model.layers.25.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.25.mlp.up_proj": {
"stored_tensors": {
"model.layers.25.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.25.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.25.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.25.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.25.mlp.gate_proj": {
"stored_tensors": {
"model.layers.25.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.25.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.25.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.25.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.25.mlp.down_proj": {
"stored_tensors": {
"model.layers.25.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.25.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.25.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.25.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.26.input_layernorm": {
"stored_tensors": {
"model.layers.26.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.26.self_attn.q_proj": {
"stored_tensors": {
"model.layers.26.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.26.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.26.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.26.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.26.self_attn.k_proj": {
"stored_tensors": {
"model.layers.26.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.26.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.26.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.26.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.26.self_attn.v_proj": {
"stored_tensors": {
"model.layers.26.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.26.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.26.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.26.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.26.self_attn.o_proj": {
"stored_tensors": {
"model.layers.26.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.26.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.26.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.26.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.26.self_attn.q_norm": {
"stored_tensors": {
"model.layers.26.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.26.self_attn.k_norm": {
"stored_tensors": {
"model.layers.26.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.26.post_attention_layernorm": {
"stored_tensors": {
"model.layers.26.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.26.mlp.up_proj": {
"stored_tensors": {
"model.layers.26.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.26.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.26.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.26.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.26.mlp.gate_proj": {
"stored_tensors": {
"model.layers.26.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.26.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.26.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.26.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.26.mlp.down_proj": {
"stored_tensors": {
"model.layers.26.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.26.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.26.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.26.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.27.input_layernorm": {
"stored_tensors": {
"model.layers.27.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.27.self_attn.q_proj": {
"stored_tensors": {
"model.layers.27.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.27.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.27.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.27.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.27.self_attn.k_proj": {
"stored_tensors": {
"model.layers.27.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.27.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.27.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.27.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.27.self_attn.v_proj": {
"stored_tensors": {
"model.layers.27.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.27.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.27.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.27.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.27.self_attn.o_proj": {
"stored_tensors": {
"model.layers.27.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.27.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.27.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.27.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.27.self_attn.q_norm": {
"stored_tensors": {
"model.layers.27.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.27.self_attn.k_norm": {
"stored_tensors": {
"model.layers.27.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.27.post_attention_layernorm": {
"stored_tensors": {
"model.layers.27.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.27.mlp.up_proj": {
"stored_tensors": {
"model.layers.27.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.27.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.27.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.27.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.27.mlp.gate_proj": {
"stored_tensors": {
"model.layers.27.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.27.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.27.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.27.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.27.mlp.down_proj": {
"stored_tensors": {
"model.layers.27.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.27.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.27.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.27.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.28.input_layernorm": {
"stored_tensors": {
"model.layers.28.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.28.self_attn.q_proj": {
"stored_tensors": {
"model.layers.28.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.28.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.28.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.28.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.28.self_attn.k_proj": {
"stored_tensors": {
"model.layers.28.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.28.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.28.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.28.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.28.self_attn.v_proj": {
"stored_tensors": {
"model.layers.28.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.28.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.28.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.28.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.28.self_attn.o_proj": {
"stored_tensors": {
"model.layers.28.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.28.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.28.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.28.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.28.self_attn.q_norm": {
"stored_tensors": {
"model.layers.28.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.28.self_attn.k_norm": {
"stored_tensors": {
"model.layers.28.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.28.post_attention_layernorm": {
"stored_tensors": {
"model.layers.28.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.28.mlp.up_proj": {
"stored_tensors": {
"model.layers.28.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.28.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.28.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.28.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.28.mlp.gate_proj": {
"stored_tensors": {
"model.layers.28.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.28.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.28.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.28.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.28.mlp.down_proj": {
"stored_tensors": {
"model.layers.28.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.28.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.28.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.28.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.29.input_layernorm": {
"stored_tensors": {
"model.layers.29.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.29.self_attn.q_proj": {
"stored_tensors": {
"model.layers.29.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.29.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.29.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.29.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.29.self_attn.k_proj": {
"stored_tensors": {
"model.layers.29.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.29.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.29.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.29.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.29.self_attn.v_proj": {
"stored_tensors": {
"model.layers.29.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.29.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.29.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.29.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.29.self_attn.o_proj": {
"stored_tensors": {
"model.layers.29.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.29.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.29.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.29.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.29.self_attn.q_norm": {
"stored_tensors": {
"model.layers.29.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.29.self_attn.k_norm": {
"stored_tensors": {
"model.layers.29.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.29.post_attention_layernorm": {
"stored_tensors": {
"model.layers.29.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.29.mlp.up_proj": {
"stored_tensors": {
"model.layers.29.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.29.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.29.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.29.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.29.mlp.gate_proj": {
"stored_tensors": {
"model.layers.29.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.29.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.29.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.29.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.29.mlp.down_proj": {
"stored_tensors": {
"model.layers.29.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.29.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.29.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.29.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.30.input_layernorm": {
"stored_tensors": {
"model.layers.30.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.30.self_attn.q_proj": {
"stored_tensors": {
"model.layers.30.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.30.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.30.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.30.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.30.self_attn.k_proj": {
"stored_tensors": {
"model.layers.30.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.30.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.30.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.30.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.30.self_attn.v_proj": {
"stored_tensors": {
"model.layers.30.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.30.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.30.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.30.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.30.self_attn.o_proj": {
"stored_tensors": {
"model.layers.30.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.30.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.30.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.30.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.30.self_attn.q_norm": {
"stored_tensors": {
"model.layers.30.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.30.self_attn.k_norm": {
"stored_tensors": {
"model.layers.30.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.30.post_attention_layernorm": {
"stored_tensors": {
"model.layers.30.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.30.mlp.up_proj": {
"stored_tensors": {
"model.layers.30.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.30.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.30.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.30.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.30.mlp.gate_proj": {
"stored_tensors": {
"model.layers.30.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.30.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.30.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.30.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.30.mlp.down_proj": {
"stored_tensors": {
"model.layers.30.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.30.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.30.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.30.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.31.input_layernorm": {
"stored_tensors": {
"model.layers.31.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.31.self_attn.q_proj": {
"stored_tensors": {
"model.layers.31.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.31.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.31.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.31.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.31.self_attn.k_proj": {
"stored_tensors": {
"model.layers.31.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.31.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.31.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.31.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.31.self_attn.v_proj": {
"stored_tensors": {
"model.layers.31.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.31.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.31.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.31.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.31.self_attn.o_proj": {
"stored_tensors": {
"model.layers.31.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.31.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.31.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.31.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.31.self_attn.q_norm": {
"stored_tensors": {
"model.layers.31.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.31.self_attn.k_norm": {
"stored_tensors": {
"model.layers.31.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.31.post_attention_layernorm": {
"stored_tensors": {
"model.layers.31.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.31.mlp.up_proj": {
"stored_tensors": {
"model.layers.31.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.31.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.31.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.31.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.31.mlp.gate_proj": {
"stored_tensors": {
"model.layers.31.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.31.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.31.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.31.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.31.mlp.down_proj": {
"stored_tensors": {
"model.layers.31.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.31.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.31.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.31.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.32.input_layernorm": {
"stored_tensors": {
"model.layers.32.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.32.self_attn.q_proj": {
"stored_tensors": {
"model.layers.32.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.32.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.32.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.32.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.32.self_attn.k_proj": {
"stored_tensors": {
"model.layers.32.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.32.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.32.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.32.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.32.self_attn.v_proj": {
"stored_tensors": {
"model.layers.32.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.32.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.32.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.32.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.32.self_attn.o_proj": {
"stored_tensors": {
"model.layers.32.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.32.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.32.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.32.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.32.self_attn.q_norm": {
"stored_tensors": {
"model.layers.32.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.32.self_attn.k_norm": {
"stored_tensors": {
"model.layers.32.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.32.post_attention_layernorm": {
"stored_tensors": {
"model.layers.32.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.32.mlp.up_proj": {
"stored_tensors": {
"model.layers.32.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.32.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.32.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.32.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.32.mlp.gate_proj": {
"stored_tensors": {
"model.layers.32.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.32.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.32.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.32.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.32.mlp.down_proj": {
"stored_tensors": {
"model.layers.32.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.32.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.32.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.32.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.33.input_layernorm": {
"stored_tensors": {
"model.layers.33.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.33.self_attn.q_proj": {
"stored_tensors": {
"model.layers.33.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.33.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.33.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.33.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.33.self_attn.k_proj": {
"stored_tensors": {
"model.layers.33.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.33.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.33.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.33.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.33.self_attn.v_proj": {
"stored_tensors": {
"model.layers.33.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.33.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.33.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.33.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.33.self_attn.o_proj": {
"stored_tensors": {
"model.layers.33.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.33.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.33.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.33.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.33.self_attn.q_norm": {
"stored_tensors": {
"model.layers.33.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.33.self_attn.k_norm": {
"stored_tensors": {
"model.layers.33.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.33.post_attention_layernorm": {
"stored_tensors": {
"model.layers.33.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.33.mlp.up_proj": {
"stored_tensors": {
"model.layers.33.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.33.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.33.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.33.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.33.mlp.gate_proj": {
"stored_tensors": {
"model.layers.33.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.33.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.33.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.33.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.33.mlp.down_proj": {
"stored_tensors": {
"model.layers.33.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.33.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.33.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.33.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.34.input_layernorm": {
"stored_tensors": {
"model.layers.34.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.34.self_attn.q_proj": {
"stored_tensors": {
"model.layers.34.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.34.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.34.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.34.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.34.self_attn.k_proj": {
"stored_tensors": {
"model.layers.34.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.34.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.34.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.34.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.34.self_attn.v_proj": {
"stored_tensors": {
"model.layers.34.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.34.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.34.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.34.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.34.self_attn.o_proj": {
"stored_tensors": {
"model.layers.34.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.34.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.34.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.34.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.34.self_attn.q_norm": {
"stored_tensors": {
"model.layers.34.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.34.self_attn.k_norm": {
"stored_tensors": {
"model.layers.34.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.34.post_attention_layernorm": {
"stored_tensors": {
"model.layers.34.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.34.mlp.up_proj": {
"stored_tensors": {
"model.layers.34.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.34.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.34.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.34.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.34.mlp.gate_proj": {
"stored_tensors": {
"model.layers.34.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.34.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.34.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.34.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.34.mlp.down_proj": {
"stored_tensors": {
"model.layers.34.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.34.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.34.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.34.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.35.input_layernorm": {
"stored_tensors": {
"model.layers.35.input_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.35.self_attn.q_proj": {
"stored_tensors": {
"model.layers.35.self_attn.q_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.35.self_attn.q_proj.svh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.35.self_attn.q_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.35.self_attn.q_proj.trellis": {
"shape": [
160,
256,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.35.self_attn.k_proj": {
"stored_tensors": {
"model.layers.35.self_attn.k_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.35.self_attn.k_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.35.self_attn.k_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.35.self_attn.k_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.35.self_attn.v_proj": {
"stored_tensors": {
"model.layers.35.self_attn.v_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.35.self_attn.v_proj.svh": {
"shape": [
1024
],
"n_bytes": 2048,
"dtype": "torch.float16"
},
"model.layers.35.self_attn.v_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.35.self_attn.v_proj.trellis": {
"shape": [
160,
64,
64
],
"n_bytes": 1310720,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.35.self_attn.o_proj": {
"stored_tensors": {
"model.layers.35.self_attn.o_proj.suh": {
"shape": [
4096
],
"n_bytes": 8192,
"dtype": "torch.float16"
},
"model.layers.35.self_attn.o_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.35.self_attn.o_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.35.self_attn.o_proj.trellis": {
"shape": [
256,
160,
64
],
"n_bytes": 5242880,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.35.self_attn.q_norm": {
"stored_tensors": {
"model.layers.35.self_attn.q_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.35.self_attn.k_norm": {
"stored_tensors": {
"model.layers.35.self_attn.k_norm.weight": {
"shape": [
128
],
"n_bytes": 256,
"dtype": "torch.float16"
}
}
},
"model.layers.35.post_attention_layernorm": {
"stored_tensors": {
"model.layers.35.post_attention_layernorm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"model.layers.35.mlp.up_proj": {
"stored_tensors": {
"model.layers.35.mlp.up_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.35.mlp.up_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.35.mlp.up_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.35.mlp.up_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.35.mlp.gate_proj": {
"stored_tensors": {
"model.layers.35.mlp.gate_proj.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.35.mlp.gate_proj.svh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.35.mlp.gate_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.35.mlp.gate_proj.trellis": {
"shape": [
160,
608,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.layers.35.mlp.down_proj": {
"stored_tensors": {
"model.layers.35.mlp.down_proj.suh": {
"shape": [
9728
],
"n_bytes": 19456,
"dtype": "torch.float16"
},
"model.layers.35.mlp.down_proj.svh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"model.layers.35.mlp.down_proj.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"model.layers.35.mlp.down_proj.trellis": {
"shape": [
608,
160,
64
],
"n_bytes": 12451840,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 4,
"mcg_multiplier": 3417055213
},
"model.norm": {
"stored_tensors": {
"model.norm.weight": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
}
}
},
"lm_head": {
"stored_tensors": {
"lm_head.suh": {
"shape": [
2560
],
"n_bytes": 5120,
"dtype": "torch.float16"
},
"lm_head.svh": {
"shape": [
151936
],
"n_bytes": 303872,
"dtype": "torch.float16"
},
"lm_head.mcg": {
"shape": [],
"n_bytes": 4,
"dtype": "torch.int32"
},
"lm_head.trellis": {
"shape": [
160,
9496,
96
],
"n_bytes": 291717120,
"dtype": "torch.int16"
}
},
"quant_format": "exl3",
"bits_per_weight": 6,
"mcg_multiplier": 3417055213
}
}
}