{ "converted_bf16_to_fp16": 668, "converted_integer_or_quantized_tensors": 0, "created_at": "2026-06-07T02:01:52+00:00", "files_copied": [ { "kind": "file-copy", "name": "chat_template.jinja", "size_bytes": 7756 }, { "kind": "file-copy", "name": "model.safetensors.index.json", "size_bytes": 95099 }, { "kind": "file-copy", "name": "tokenizer.json", "size_bytes": 19989325 }, { "kind": "file-copy", "name": "tokenizer_config.json", "size_bytes": 1161 } ], "intended_default_for": [ "m1", "m2" ], "local_paths_required": false, "policy": "convert bf16 floating tensors to fp16; preserve packed integer/quantized tensors", "precision_variant": "fp16", "preserve_mtp_compatibility": true, "preserved_integer_or_quantized_tensors": 274, "repo_id": "Youssofal/Qwen3.5-9B-MTPLX-Optimized-Speed-FP16", "safetensors": [ { "converted_bf16_to_fp16": 528, "name": "model-00001-of-00002.safetensors", "output_sha256": "314432c10e8546e442a72852e8109bfcafa26ac5d8c2a2dd21469255cf27a573", "output_size_bytes": 5358171972, "preserved": 222, "source_sha256": "f9738b8b326e67e5213e46f1ebd59c558f60146ae50f4af9bfebe1ec7a42bc9d", "source_size_bytes": 5358172442, "tensor_count": 750, "tensors": [ { "converted": true, "name": "language_model.model.embed_tokens.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 248320, 64 ] }, { "converted": true, "name": "language_model.model.embed_tokens.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 248320, 64 ] }, { "converted": false, "name": "language_model.model.embed_tokens.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 248320, 768 ] }, { "converted": true, "name": "language_model.model.layers.0.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.0.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.0.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.0.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.0.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.0.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.0.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.0.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.0.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.0.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.0.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.0.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.0.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.0.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.0.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.0.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.0.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.0.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.1.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.1.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.1.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.1.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.1.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.1.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.1.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.1.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.1.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.1.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.1.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.1.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.1.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.1.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.1.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.1.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.1.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.1.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.10.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.10.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.10.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.10.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.10.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.10.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.10.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.10.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.10.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.10.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.10.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.10.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.10.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.10.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.10.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.10.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.10.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.10.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.11.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.11.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.11.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.11.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.11.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.11.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.11.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.11.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.11.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.11.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.11.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.11.self_attn.k_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.11.self_attn.k_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.11.self_attn.k_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.11.self_attn.k_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.11.self_attn.o_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.11.self_attn.o_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.11.self_attn.o_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.11.self_attn.q_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.11.self_attn.q_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.11.self_attn.q_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.11.self_attn.q_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.11.self_attn.v_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.11.self_attn.v_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.11.self_attn.v_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.12.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.12.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.12.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.12.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.12.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.12.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.12.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.12.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.12.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.12.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.12.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.12.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.12.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.12.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.12.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.12.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.12.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.12.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.13.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.13.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.13.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.13.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.13.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.13.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.13.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.13.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.13.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.13.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.13.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.13.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.13.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.13.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.13.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.13.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.13.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.13.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.14.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.14.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.14.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.14.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.14.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.14.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.14.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.14.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.14.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.14.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.14.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.14.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.14.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.14.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.14.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.14.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.14.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.14.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.15.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.15.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.15.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.15.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.15.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.15.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.15.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.15.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.15.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.15.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.15.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.15.self_attn.k_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.15.self_attn.k_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.15.self_attn.k_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.15.self_attn.k_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.15.self_attn.o_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.15.self_attn.o_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.15.self_attn.o_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.15.self_attn.q_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.15.self_attn.q_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.15.self_attn.q_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.15.self_attn.q_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.15.self_attn.v_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.15.self_attn.v_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.15.self_attn.v_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.16.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.16.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.16.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.16.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.16.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.16.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.16.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.16.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.16.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.16.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.16.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.16.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.16.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.16.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.16.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.16.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.16.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.16.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.17.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.17.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.17.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.17.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.17.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.17.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.17.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.17.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.17.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.17.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.17.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.17.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.17.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.17.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.17.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.17.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.17.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.17.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.18.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.18.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.18.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.18.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.18.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.18.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.18.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.18.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.18.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.18.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.18.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.18.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.18.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.18.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.18.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.18.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.18.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.18.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.19.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.19.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.19.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.19.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.19.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.19.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.19.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.19.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.19.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.19.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.19.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.19.self_attn.k_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.19.self_attn.k_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.19.self_attn.k_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.19.self_attn.k_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.19.self_attn.o_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.19.self_attn.o_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.19.self_attn.o_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.19.self_attn.q_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.19.self_attn.q_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.19.self_attn.q_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.19.self_attn.q_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.19.self_attn.v_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.19.self_attn.v_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.19.self_attn.v_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.2.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.2.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.2.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.2.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.2.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.2.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.2.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.2.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.2.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.2.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.2.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.2.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.2.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.2.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.2.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.2.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.2.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.2.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.20.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.20.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.20.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.20.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.20.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.20.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.20.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.20.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.20.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.20.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.20.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.20.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.20.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.20.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.20.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.20.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.20.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.20.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.21.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.21.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.21.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.21.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.21.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.21.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.21.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.21.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.21.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.21.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.21.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.21.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.21.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.21.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.21.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.21.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.21.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.21.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.22.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.22.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.22.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.22.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.22.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.22.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.22.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.22.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.22.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.22.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.22.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.22.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.22.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.22.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.22.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.22.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.22.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.22.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.23.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.23.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.23.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.23.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.23.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.23.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.23.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.23.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.23.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.23.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.23.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.23.self_attn.k_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.23.self_attn.k_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.23.self_attn.k_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.23.self_attn.k_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.23.self_attn.o_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.23.self_attn.o_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.23.self_attn.o_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.23.self_attn.q_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.23.self_attn.q_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.23.self_attn.q_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.23.self_attn.q_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.23.self_attn.v_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.23.self_attn.v_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.23.self_attn.v_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.24.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.24.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.24.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.24.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.24.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.24.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.24.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.24.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.24.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.24.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.24.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.24.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.24.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.24.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.24.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.24.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.24.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.24.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.25.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.25.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.25.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.25.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.25.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.25.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.25.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.25.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.25.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.25.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.25.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.25.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.25.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.25.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.25.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.3.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.3.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.3.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.3.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.3.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.3.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.3.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.3.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.3.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.3.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.3.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.3.self_attn.k_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.3.self_attn.k_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.3.self_attn.k_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.3.self_attn.k_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.3.self_attn.o_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.3.self_attn.o_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.3.self_attn.o_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.3.self_attn.q_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.3.self_attn.q_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.3.self_attn.q_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.3.self_attn.q_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.3.self_attn.v_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.3.self_attn.v_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.3.self_attn.v_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.4.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.4.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.4.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.4.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.4.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.4.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.4.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.4.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.4.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.4.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.4.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.4.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.4.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.4.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.4.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.4.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.4.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.4.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.5.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.5.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.5.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.5.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.5.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.5.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.5.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.5.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.5.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.5.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.5.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.5.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.5.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.5.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.5.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.5.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.5.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.5.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.6.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.6.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.6.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.6.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.6.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.6.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.6.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.6.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.6.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.6.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.6.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.6.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.6.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.6.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.6.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.6.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.6.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.6.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.7.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.7.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.7.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.7.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.7.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.7.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.7.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.7.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.7.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.7.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.7.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.7.self_attn.k_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.7.self_attn.k_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.7.self_attn.k_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.7.self_attn.k_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.7.self_attn.o_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.7.self_attn.o_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.7.self_attn.o_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.7.self_attn.q_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.7.self_attn.q_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.7.self_attn.q_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.7.self_attn.q_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.7.self_attn.v_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.7.self_attn.v_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.7.self_attn.v_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.8.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.8.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.8.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.8.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.8.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.8.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.8.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.8.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.8.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.8.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.8.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.8.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.8.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.8.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.8.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.8.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.8.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.8.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.9.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.9.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.9.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.9.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.9.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.9.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.9.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.9.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.9.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.9.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.9.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.9.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.9.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.9.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.9.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.9.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.9.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.9.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] } ] }, { "converted_bf16_to_fp16": 125, "name": "model-00002-of-00002.safetensors", "output_sha256": "a9bbca7e94392b6ff83acc5cdd35a797454ee29a1b601f81ec7315212d753900", "output_size_bytes": 1918175118, "preserved": 52, "source_sha256": "57afa8da33ded0f5ac1801cf4d69b4dcae5318933beb48c86f284a908f4341c5", "source_size_bytes": 1918175495, "tensor_count": 177, "tensors": [ { "converted": true, "name": "language_model.lm_head.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 248320, 64 ] }, { "converted": true, "name": "language_model.lm_head.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 248320, 64 ] }, { "converted": false, "name": "language_model.lm_head.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 248320, 768 ] }, { "converted": true, "name": "language_model.model.layers.25.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.25.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.25.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.26.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.26.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.26.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.26.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.26.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.26.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.26.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.26.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.26.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.26.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.26.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.26.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.26.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.26.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.26.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.26.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.26.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.26.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.27.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.27.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.27.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.27.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.27.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.27.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.27.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.27.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.27.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.27.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.27.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.27.self_attn.k_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.27.self_attn.k_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.27.self_attn.k_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.27.self_attn.k_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.27.self_attn.o_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.27.self_attn.o_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.27.self_attn.o_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.27.self_attn.q_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.27.self_attn.q_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.27.self_attn.q_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.27.self_attn.q_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.27.self_attn.v_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.27.self_attn.v_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.27.self_attn.v_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.28.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.28.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.28.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.28.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.28.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.28.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.28.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.28.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.28.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.28.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.28.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.28.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.28.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.28.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.28.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.28.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.28.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.28.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.29.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.29.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.29.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.29.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.29.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.29.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.29.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.29.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.29.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.29.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.29.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.29.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.29.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.29.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.29.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.29.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.29.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.29.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.30.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": false, "name": "language_model.model.layers.30.linear_attn.A_log", "new_dtype": "mlx.core.float32", "old_dtype": "mlx.core.float32", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.conv1d.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4, 1 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.dt_bias", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.in_proj_a.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.in_proj_a.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.30.linear_attn.in_proj_a.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.in_proj_b.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.in_proj_b.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 32, 64 ] }, { "converted": false, "name": "language_model.model.layers.30.linear_attn.in_proj_b.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 32, 768 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.in_proj_qkv.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.in_proj_qkv.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.30.linear_attn.in_proj_qkv.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.in_proj_z.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.in_proj_z.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.30.linear_attn.in_proj_z.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 128 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.out_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.30.linear_attn.out_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.30.linear_attn.out_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.30.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.30.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.30.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.30.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.30.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.30.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.30.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.30.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.30.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.30.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.31.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.31.mlp.down_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": true, "name": "language_model.model.layers.31.mlp.down_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 192 ] }, { "converted": false, "name": "language_model.model.layers.31.mlp.down_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 2304 ] }, { "converted": true, "name": "language_model.model.layers.31.mlp.gate_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.31.mlp.gate_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.31.mlp.gate_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.31.mlp.up_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": true, "name": "language_model.model.layers.31.mlp.up_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 64 ] }, { "converted": false, "name": "language_model.model.layers.31.mlp.up_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 12288, 768 ] }, { "converted": true, "name": "language_model.model.layers.31.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "language_model.model.layers.31.self_attn.k_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.31.self_attn.k_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.31.self_attn.k_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.31.self_attn.k_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.layers.31.self_attn.o_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": true, "name": "language_model.model.layers.31.self_attn.o_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 64 ] }, { "converted": false, "name": "language_model.model.layers.31.self_attn.o_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 4096, 768 ] }, { "converted": true, "name": "language_model.model.layers.31.self_attn.q_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "language_model.model.layers.31.self_attn.q_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": true, "name": "language_model.model.layers.31.self_attn.q_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 64 ] }, { "converted": false, "name": "language_model.model.layers.31.self_attn.q_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 8192, 768 ] }, { "converted": true, "name": "language_model.model.layers.31.self_attn.v_proj.biases", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": true, "name": "language_model.model.layers.31.self_attn.v_proj.scales", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 64 ] }, { "converted": false, "name": "language_model.model.layers.31.self_attn.v_proj.weight", "new_dtype": "mlx.core.uint32", "old_dtype": "mlx.core.uint32", "shape": [ 1024, 768 ] }, { "converted": true, "name": "language_model.model.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] } ] }, { "converted_bf16_to_fp16": 15, "name": "mtp.safetensors", "output_sha256": "8461334aec5d4c29515aa18cb0476b79660ba2916ede72f1943237ed847d13fe", "output_size_bytes": 486582844, "preserved": 0, "source_sha256": "b480f630a574a19aa2d1c0efd92bb18de8cb6fa7c05c7cb043eecb07201ce7d2", "source_size_bytes": 486582837, "tensor_count": 15, "tensors": [ { "converted": true, "name": "mtp.fc.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 8192 ] }, { "converted": true, "name": "mtp.layers.0.input_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "mtp.layers.0.mlp.down_proj.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 12288 ] }, { "converted": true, "name": "mtp.layers.0.mlp.gate_proj.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 4096 ] }, { "converted": true, "name": "mtp.layers.0.mlp.up_proj.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 12288, 4096 ] }, { "converted": true, "name": "mtp.layers.0.post_attention_layernorm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "mtp.layers.0.self_attn.k_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "mtp.layers.0.self_attn.k_proj.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 4096 ] }, { "converted": true, "name": "mtp.layers.0.self_attn.o_proj.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096, 4096 ] }, { "converted": true, "name": "mtp.layers.0.self_attn.q_norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 256 ] }, { "converted": true, "name": "mtp.layers.0.self_attn.q_proj.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 8192, 4096 ] }, { "converted": true, "name": "mtp.layers.0.self_attn.v_proj.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 1024, 4096 ] }, { "converted": true, "name": "mtp.norm.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "mtp.pre_fc_norm_embedding.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] }, { "converted": true, "name": "mtp.pre_fc_norm_hidden.weight", "new_dtype": "mlx.core.float16", "old_dtype": "mlx.core.bfloat16", "shape": [ 4096 ] } ] } ], "source_repo": "Youssofal/Qwen3.5-9B-MTPLX-Optimized-Speed" }