{ "source": "/Users/arjun/Projects/Step Fun/models/Step-3.7-Flash", "mode": "affine", "method": "optiq_stream_frobenius", "target_bpw": 3.7, "achieved_bpw": 3.6930459517285583, "candidate_bits": [ 2, 3, 4, 8 ], "group_size": 64, "bit_counts": { "3": 66, "4": 62, "8": 574 }, "quantized_weight_count": 702, "passthrough_tensor_count": 769, "missing_allocations": [], "quantized": { "model.layers.1.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.1.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.1.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.1.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.1.mlp.gate_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.0.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.0.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.1.mlp.down_proj.weight": { "source_shape": [ 4096, 11264 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 46137344, "weight_shape": [ 4096, 2816 ], "scales_shape": [ 4096, 176 ], "biases_shape": [ 4096, 176 ] }, "model.layers.0.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.0.mlp.down_proj.weight": { "source_shape": [ 4096, 11264 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 46137344, "weight_shape": [ 4096, 2816 ], "scales_shape": [ 4096, 176 ], "biases_shape": [ 4096, 176 ] }, "model.layers.1.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.0.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.0.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.1.mlp.up_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.0.mlp.up_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.0.mlp.gate_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00001.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.3.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.3.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.3.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.3.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.3.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.3.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.3.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.2.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.2.mlp.down_proj.weight": { "source_shape": [ 4096, 11264 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 46137344, "weight_shape": [ 4096, 2816 ], "scales_shape": [ 4096, 176 ], "biases_shape": [ 4096, 176 ] }, "model.layers.3.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.3.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.2.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.2.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.2.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.3.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.2.mlp.up_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.2.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.3.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.2.mlp.gate_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.3.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00002.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.5.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.5.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.5.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.5.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.5.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.5.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.5.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.5.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.5.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.4.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.4.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.5.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.4.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.4.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.4.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.4.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.5.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.4.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.4.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.4.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.4.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.4.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.4.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.5.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00003.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.7.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.7.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.7.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.7.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.7.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.7.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.7.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.6.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.6.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.7.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.6.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.6.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.7.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.7.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.6.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.6.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.6.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.6.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.6.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.6.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.6.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.6.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.7.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.7.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00004.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.9.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.9.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.9.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.9.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.9.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.9.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.8.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.9.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.9.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.9.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.9.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.8.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.8.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.8.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.9.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.8.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.8.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.8.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.8.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.9.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.8.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.8.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.8.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.8.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00005.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.11.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.11.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.11.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.11.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.11.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.11.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.11.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.10.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.10.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.10.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.10.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.11.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.11.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.11.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.10.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.11.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.10.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.11.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.10.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.10.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.10.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.10.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.10.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.10.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00006.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.13.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.13.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.13.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.13.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.13.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.13.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.13.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.13.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.13.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.13.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.12.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.12.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.13.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.12.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.12.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.12.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.12.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.12.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.12.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.12.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.13.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.12.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.12.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.12.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00007.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.15.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.15.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.15.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.15.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.15.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.15.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.14.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.14.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.15.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.14.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.14.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.14.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.14.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.14.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.14.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.14.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.15.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.14.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.14.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.15.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.15.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.14.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.15.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.15.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00008.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.17.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.17.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.17.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.17.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.17.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.16.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.16.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.16.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.16.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.16.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.17.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.16.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.17.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.17.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.17.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.16.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.17.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.16.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.17.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.16.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.17.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.16.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.16.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.16.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00009.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.19.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.19.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.19.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.19.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.18.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.19.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.19.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.18.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.19.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.18.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.18.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.19.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.18.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.19.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.18.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.18.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.18.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.18.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.19.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.18.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.19.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.18.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.19.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.18.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00010.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.21.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.21.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.21.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.21.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.21.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.21.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.21.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.21.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.20.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.20.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.20.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.20.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.20.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.21.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.20.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.20.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.21.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.20.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.21.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.20.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.20.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.21.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.20.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.20.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00011.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.23.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.23.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.23.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.23.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.23.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.22.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.22.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.22.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.22.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.23.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.23.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.23.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.22.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.22.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.22.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.22.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.23.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.22.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.22.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.23.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.23.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.23.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.22.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.22.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00012.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.25.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.25.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.25.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.25.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.24.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.25.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.25.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.24.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.24.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.24.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.24.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.25.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.24.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.24.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.25.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.24.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.25.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.24.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.24.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.25.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.25.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.25.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.24.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.24.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00013.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.27.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.27.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.27.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.27.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.27.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.27.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.27.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.27.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.27.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.27.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.26.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.26.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.27.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.26.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.26.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.26.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.27.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.26.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.26.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.26.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.26.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.26.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.26.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.26.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00014.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.29.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.29.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.29.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.29.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.29.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.29.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.29.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.29.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.28.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.29.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.28.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.28.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.28.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.28.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.28.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.29.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.29.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.28.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.28.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.28.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.28.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.29.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.28.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.28.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00015.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.31.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.31.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.31.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.31.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.30.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.31.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.30.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.30.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.30.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.31.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.30.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.30.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.30.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.31.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.30.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.30.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.31.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.31.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.30.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 120 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.30.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.31.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.31.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.30.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.31.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00016.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.33.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.33.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.33.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.33.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.33.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.32.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.33.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.33.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.32.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.32.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.32.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.32.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.32.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.32.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.33.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.33.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.32.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.32.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.32.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.33.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.32.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.32.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.33.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.33.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00017.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.35.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.35.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.35.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.35.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.35.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.35.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.35.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.34.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.34.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.35.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.35.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.35.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.35.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.34.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.34.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.34.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.34.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.34.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.34.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.34.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.34.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.35.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.34.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.34.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00018.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.37.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.37.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.37.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.37.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.37.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.36.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.37.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.36.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.36.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.37.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.37.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.37.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.36.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.37.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.36.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.36.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.36.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.36.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.36.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.37.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.36.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.37.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.36.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.36.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00019.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.39.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.39.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.39.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.39.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.39.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.38.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.38.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.38.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.38.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.38.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.39.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.38.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.38.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.39.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.38.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.39.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.39.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.39.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.39.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.39.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.38.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.38.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 3, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 384 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.38.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.38.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00020.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.41.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.41.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.41.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.41.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.41.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.41.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.41.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.41.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.41.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.40.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.40.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.40.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.41.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.40.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.40.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.40.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.40.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.40.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.40.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.41.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.40.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.40.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.41.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.40.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00021.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.43.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.43.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.43.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.43.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.42.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.42.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.43.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.43.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.42.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.43.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.43.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.42.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.43.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.42.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.42.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.43.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.42.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.43.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.42.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.42.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.42.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.42.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.42.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.43.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00022.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.44.share_expert.down_proj.weight": { "source_shape": [ 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 5242880, "weight_shape": [ 4096, 320 ], "scales_shape": [ 4096, 20 ], "biases_shape": [ 4096, 20 ] }, "model.layers.44.share_expert.gate_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.44.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.44.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.44.share_expert.up_proj.weight": { "source_shape": [ 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 5242880, "weight_shape": [ 1280, 1024 ], "scales_shape": [ 1280, 64 ], "biases_shape": [ 1280, 64 ] }, "model.layers.44.moe.up_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.44.self_attn.o_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.44.moe.gate.weight": { "source_shape": [ 288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 1179648, "weight_shape": [ 288, 1024 ], "scales_shape": [ 288, 64 ], "biases_shape": [ 288, 64 ] }, "model.layers.44.moe.down_proj.weight": { "source_shape": [ 288, 4096, 1280 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 4096, 160 ], "scales_shape": [ 288, 4096, 20 ], "biases_shape": [ 288, 4096, 20 ] }, "model.layers.44.moe.gate_proj.weight": { "source_shape": [ 288, 1280, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 1509949440, "weight_shape": [ 288, 1280, 512 ], "scales_shape": [ 288, 1280, 64 ], "biases_shape": [ 288, 1280, 64 ] }, "model.layers.44.self_attn.g_proj.weight": { "source_shape": [ 64, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 262144, "weight_shape": [ 64, 1024 ], "scales_shape": [ 64, 64 ], "biases_shape": [ 64, 64 ] }, "model.layers.44.self_attn.q_proj.weight": { "source_shape": [ 8192, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00023.safetensors", "param_count": 33554432, "weight_shape": [ 8192, 1024 ], "scales_shape": [ 8192, 64 ], "biases_shape": [ 8192, 64 ] }, "model.layers.47.transformer.shared_head.output.weight": { "source_shape": [ 128896, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 527958016, "weight_shape": [ 128896, 1024 ], "scales_shape": [ 128896, 64 ], "biases_shape": [ 128896, 64 ] }, "model.layers.47.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.47.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.47.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "lm_head.weight": { "source_shape": [ 128896, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 527958016, "weight_shape": [ 128896, 1024 ], "scales_shape": [ 128896, 64 ], "biases_shape": [ 128896, 64 ] }, "model.layers.47.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.embed_tokens.weight": { "source_shape": [ 128896, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 527958016, "weight_shape": [ 128896, 1024 ], "scales_shape": [ 128896, 64 ], "biases_shape": [ 128896, 64 ] }, "model.layers.47.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.47.mlp.gate_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.45.transformer.shared_head.output.weight": { "source_shape": [ 128896, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 527958016, "weight_shape": [ 128896, 512 ], "scales_shape": [ 128896, 64 ], "biases_shape": [ 128896, 64 ] }, "model.layers.46.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.46.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.46.mlp.up_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.46.mlp.down_proj.weight": { "source_shape": [ 4096, 11264 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 46137344, "weight_shape": [ 4096, 2816 ], "scales_shape": [ 4096, 176 ], "biases_shape": [ 4096, 176 ] }, "model.layers.46.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.47.mlp.down_proj.weight": { "source_shape": [ 4096, 11264 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 46137344, "weight_shape": [ 4096, 2816 ], "scales_shape": [ 4096, 176 ], "biases_shape": [ 4096, 176 ] }, "model.layers.45.self_attn.v_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.45.mlp.up_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.47.mlp.up_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.46.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.45.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.47.eh_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.45.self_attn.o_proj.weight": { "source_shape": [ 4096, 12288 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 50331648, "weight_shape": [ 4096, 3072 ], "scales_shape": [ 4096, 192 ], "biases_shape": [ 4096, 192 ] }, "model.layers.45.mlp.gate_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.45.mlp.down_proj.weight": { "source_shape": [ 4096, 11264 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 46137344, "weight_shape": [ 4096, 2816 ], "scales_shape": [ 4096, 176 ], "biases_shape": [ 4096, 176 ] }, "model.layers.46.self_attn.q_proj.weight": { "source_shape": [ 12288, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 50331648, "weight_shape": [ 12288, 1024 ], "scales_shape": [ 12288, 64 ], "biases_shape": [ 12288, 64 ] }, "model.layers.46.transformer.shared_head.output.weight": { "source_shape": [ 128896, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 4, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 527958016, "weight_shape": [ 128896, 512 ], "scales_shape": [ 128896, 64 ], "biases_shape": [ 128896, 64 ] }, "model.layers.46.mlp.gate_proj.weight": { "source_shape": [ 11264, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 46137344, "weight_shape": [ 11264, 1024 ], "scales_shape": [ 11264, 64 ], "biases_shape": [ 11264, 64 ] }, "model.layers.45.self_attn.k_proj.weight": { "source_shape": [ 1024, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 4194304, "weight_shape": [ 1024, 1024 ], "scales_shape": [ 1024, 64 ], "biases_shape": [ 1024, 64 ] }, "model.layers.46.eh_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "model.layers.45.self_attn.g_proj.weight": { "source_shape": [ 96, 4096 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 393216, "weight_shape": [ 96, 1024 ], "scales_shape": [ 96, 64 ], "biases_shape": [ 96, 64 ] }, "model.layers.45.eh_proj.weight": { "source_shape": [ 4096, 8192 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-00024.safetensors", "param_count": 33554432, "weight_shape": [ 4096, 2048 ], "scales_shape": [ 4096, 128 ], "biases_shape": [ 4096, 128 ] }, "vision_model.transformer.resblocks.9.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.9.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.9.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.8.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.8.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.7.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.7.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.6.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.6.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.5.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.4.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.4.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.3.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.3.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.20.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.2.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.2.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.1.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.2.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.19.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vit_large_projector.weight": { "source_shape": [ 4096, 6144 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 25165824, "weight_shape": [ 4096, 1536 ], "scales_shape": [ 4096, 96 ], "biases_shape": [ 4096, 96 ] }, "vision_model.transformer.resblocks.16.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.11.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.19.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.18.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.17.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.5.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.10.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.16.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.6.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.14.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.15.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.5.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.15.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.19.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.8.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.11.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.16.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.13.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.7.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.10.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.13.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.12.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.14.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.12.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.12.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.11.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.18.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.20.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.17.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.0.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.15.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.3.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.0.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.1.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.17.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.20.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.10.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.18.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.4.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.1.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.13.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.14.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.0.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00001.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.46.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.46.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.44.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.43.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.42.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.41.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.41.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.41.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.40.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.40.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.40.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.39.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.38.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.38.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.37.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.36.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.43.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.42.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.35.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.35.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.34.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.43.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.44.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.34.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.22.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.42.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.26.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.23.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.37.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.32.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.37.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.32.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.32.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.31.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.33.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.27.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.31.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.30.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.45.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.23.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.34.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.30.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.33.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.29.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.29.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.39.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.45.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.28.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.28.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.29.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.35.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.22.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.28.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.22.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.27.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.24.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.44.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.24.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.26.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.36.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.27.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.31.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.25.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.21.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.25.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.24.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.23.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.25.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.21.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.33.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.45.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.30.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.36.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.26.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.21.mlp.c_proj.weight": { "source_shape": [ 1536, 8960 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 1536, 2240 ], "scales_shape": [ 1536, 140 ], "biases_shape": [ 1536, 140 ] }, "vision_model.transformer.resblocks.46.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] }, "vision_model.transformer.resblocks.39.mlp.c_fc.weight": { "source_shape": [ 8960, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 13762560, "weight_shape": [ 8960, 384 ], "scales_shape": [ 8960, 24 ], "biases_shape": [ 8960, 24 ] }, "vision_model.transformer.resblocks.38.attn.out_proj.weight": { "source_shape": [ 1536, 1536 ], "source_dtype": "mlx.core.bfloat16", "bits": 8, "group_size": 64, "shard": "model-vit-00002.safetensors", "param_count": 2359296, "weight_shape": [ 1536, 384 ], "scales_shape": [ 1536, 24 ], "biases_shape": [ 1536, 24 ] } } }