{ "algorithm": "dspark", "base_model": "CohereLabs/North-Mini-Code-1.0-w4a16", "checkpoint_epoch": 4, "default_speculative_tokens": 4, "largest_tensors": [ { "dtype": "BF16", "name": "embed_tokens.weight", "parameters": 536870912, "shape": [ 262144, 2048 ] }, { "dtype": "BF16", "name": "markov_head.markov_w1.weight", "parameters": 67108864, "shape": [ 262144, 256 ] }, { "dtype": "BF16", "name": "lm_head.weight", "parameters": 65536000, "shape": [ 32000, 2048 ] }, { "dtype": "BF16", "name": "fc.weight", "parameters": 12582912, "shape": [ 2048, 6144 ] }, { "dtype": "BF16", "name": "layers.0.self_attn.o_proj.weight", "parameters": 8388608, "shape": [ 2048, 4096 ] }, { "dtype": "BF16", "name": "layers.0.self_attn.q_proj.weight", "parameters": 8388608, "shape": [ 4096, 2048 ] }, { "dtype": "BF16", "name": "layers.1.self_attn.o_proj.weight", "parameters": 8388608, "shape": [ 2048, 4096 ] }, { "dtype": "BF16", "name": "layers.1.self_attn.q_proj.weight", "parameters": 8388608, "shape": [ 4096, 2048 ] }, { "dtype": "BF16", "name": "layers.2.self_attn.o_proj.weight", "parameters": 8388608, "shape": [ 2048, 4096 ] }, { "dtype": "BF16", "name": "layers.2.self_attn.q_proj.weight", "parameters": 8388608, "shape": [ 4096, 2048 ] }, { "dtype": "BF16", "name": "layers.3.self_attn.o_proj.weight", "parameters": 8388608, "shape": [ 2048, 4096 ] }, { "dtype": "BF16", "name": "layers.3.self_attn.q_proj.weight", "parameters": 8388608, "shape": [ 4096, 2048 ] }, { "dtype": "BF16", "name": "layers.4.self_attn.o_proj.weight", "parameters": 8388608, "shape": [ 2048, 4096 ] }, { "dtype": "BF16", "name": "layers.4.self_attn.q_proj.weight", "parameters": 8388608, "shape": [ 4096, 2048 ] }, { "dtype": "BF16", "name": "markov_head.markov_w2.weight", "parameters": 8192000, "shape": [ 32000, 256 ] }, { "dtype": "BF16", "name": "layers.0.mlp.down_proj.weight", "parameters": 1572864, "shape": [ 2048, 768 ] }, { "dtype": "BF16", "name": "layers.0.mlp.gate_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.0.mlp.up_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.1.mlp.down_proj.weight", "parameters": 1572864, "shape": [ 2048, 768 ] }, { "dtype": "BF16", "name": "layers.1.mlp.gate_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] } ], "model_family": "North Mini Code speculative draft", "model_file": { "name": "model.safetensors", "safetensors_metadata": { "format": "pt" }, "sha256": "64fd7d60c1c5a91d030dae0b7300bf8593b536b0a5569f53db7209e4a5f5b7f6", "size_bytes": 1617092434 }, "parameter_count": 808577793, "parameters_by_dtype": { "BF16": 808283649, "BOOL": 262144, "I64": 32000 }, "schema_version": 1, "tensor_count": 66, "tensors": [ { "dtype": "BF16", "name": "confidence_head.proj.bias", "parameters": 1, "shape": [ 1 ] }, { "dtype": "BF16", "name": "confidence_head.proj.weight", "parameters": 2304, "shape": [ 1, 2304 ] }, { "dtype": "I64", "name": "d2t", "parameters": 32000, "shape": [ 32000 ] }, { "dtype": "BF16", "name": "embed_tokens.weight", "parameters": 536870912, "shape": [ 262144, 2048 ] }, { "dtype": "BF16", "name": "fc.weight", "parameters": 12582912, "shape": [ 2048, 6144 ] }, { "dtype": "BF16", "name": "hidden_norm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.0.input_layernorm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.0.mlp.down_proj.weight", "parameters": 1572864, "shape": [ 2048, 768 ] }, { "dtype": "BF16", "name": "layers.0.mlp.gate_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.0.mlp.up_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.0.post_attention_layernorm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.0.self_attn.k_norm.weight", "parameters": 128, "shape": [ 128 ] }, { "dtype": "BF16", "name": "layers.0.self_attn.k_proj.weight", "parameters": 1048576, "shape": [ 512, 2048 ] }, { "dtype": "BF16", "name": "layers.0.self_attn.o_proj.weight", "parameters": 8388608, "shape": [ 2048, 4096 ] }, { "dtype": "BF16", "name": "layers.0.self_attn.q_norm.weight", "parameters": 128, "shape": [ 128 ] }, { "dtype": "BF16", "name": "layers.0.self_attn.q_proj.weight", "parameters": 8388608, "shape": [ 4096, 2048 ] }, { "dtype": "BF16", "name": "layers.0.self_attn.v_proj.weight", "parameters": 1048576, "shape": [ 512, 2048 ] }, { "dtype": "BF16", "name": "layers.1.input_layernorm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.1.mlp.down_proj.weight", "parameters": 1572864, "shape": [ 2048, 768 ] }, { "dtype": "BF16", "name": "layers.1.mlp.gate_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.1.mlp.up_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.1.post_attention_layernorm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.1.self_attn.k_norm.weight", "parameters": 128, "shape": [ 128 ] }, { "dtype": "BF16", "name": "layers.1.self_attn.k_proj.weight", "parameters": 1048576, "shape": [ 512, 2048 ] }, { "dtype": "BF16", "name": "layers.1.self_attn.o_proj.weight", "parameters": 8388608, "shape": [ 2048, 4096 ] }, { "dtype": "BF16", "name": "layers.1.self_attn.q_norm.weight", "parameters": 128, "shape": [ 128 ] }, { "dtype": "BF16", "name": "layers.1.self_attn.q_proj.weight", "parameters": 8388608, "shape": [ 4096, 2048 ] }, { "dtype": "BF16", "name": "layers.1.self_attn.v_proj.weight", "parameters": 1048576, "shape": [ 512, 2048 ] }, { "dtype": "BF16", "name": "layers.2.input_layernorm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.2.mlp.down_proj.weight", "parameters": 1572864, "shape": [ 2048, 768 ] }, { "dtype": "BF16", "name": "layers.2.mlp.gate_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.2.mlp.up_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.2.post_attention_layernorm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.2.self_attn.k_norm.weight", "parameters": 128, "shape": [ 128 ] }, { "dtype": "BF16", "name": "layers.2.self_attn.k_proj.weight", "parameters": 1048576, "shape": [ 512, 2048 ] }, { "dtype": "BF16", "name": "layers.2.self_attn.o_proj.weight", "parameters": 8388608, "shape": [ 2048, 4096 ] }, { "dtype": "BF16", "name": "layers.2.self_attn.q_norm.weight", "parameters": 128, "shape": [ 128 ] }, { "dtype": "BF16", "name": "layers.2.self_attn.q_proj.weight", "parameters": 8388608, "shape": [ 4096, 2048 ] }, { "dtype": "BF16", "name": "layers.2.self_attn.v_proj.weight", "parameters": 1048576, "shape": [ 512, 2048 ] }, { "dtype": "BF16", "name": "layers.3.input_layernorm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.3.mlp.down_proj.weight", "parameters": 1572864, "shape": [ 2048, 768 ] }, { "dtype": "BF16", "name": "layers.3.mlp.gate_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.3.mlp.up_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.3.post_attention_layernorm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.3.self_attn.k_norm.weight", "parameters": 128, "shape": [ 128 ] }, { "dtype": "BF16", "name": "layers.3.self_attn.k_proj.weight", "parameters": 1048576, "shape": [ 512, 2048 ] }, { "dtype": "BF16", "name": "layers.3.self_attn.o_proj.weight", "parameters": 8388608, "shape": [ 2048, 4096 ] }, { "dtype": "BF16", "name": "layers.3.self_attn.q_norm.weight", "parameters": 128, "shape": [ 128 ] }, { "dtype": "BF16", "name": "layers.3.self_attn.q_proj.weight", "parameters": 8388608, "shape": [ 4096, 2048 ] }, { "dtype": "BF16", "name": "layers.3.self_attn.v_proj.weight", "parameters": 1048576, "shape": [ 512, 2048 ] }, { "dtype": "BF16", "name": "layers.4.input_layernorm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.4.mlp.down_proj.weight", "parameters": 1572864, "shape": [ 2048, 768 ] }, { "dtype": "BF16", "name": "layers.4.mlp.gate_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.4.mlp.up_proj.weight", "parameters": 1572864, "shape": [ 768, 2048 ] }, { "dtype": "BF16", "name": "layers.4.post_attention_layernorm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BF16", "name": "layers.4.self_attn.k_norm.weight", "parameters": 128, "shape": [ 128 ] }, { "dtype": "BF16", "name": "layers.4.self_attn.k_proj.weight", "parameters": 1048576, "shape": [ 512, 2048 ] }, { "dtype": "BF16", "name": "layers.4.self_attn.o_proj.weight", "parameters": 8388608, "shape": [ 2048, 4096 ] }, { "dtype": "BF16", "name": "layers.4.self_attn.q_norm.weight", "parameters": 128, "shape": [ 128 ] }, { "dtype": "BF16", "name": "layers.4.self_attn.q_proj.weight", "parameters": 8388608, "shape": [ 4096, 2048 ] }, { "dtype": "BF16", "name": "layers.4.self_attn.v_proj.weight", "parameters": 1048576, "shape": [ 512, 2048 ] }, { "dtype": "BF16", "name": "lm_head.weight", "parameters": 65536000, "shape": [ 32000, 2048 ] }, { "dtype": "BF16", "name": "markov_head.markov_w1.weight", "parameters": 67108864, "shape": [ 262144, 256 ] }, { "dtype": "BF16", "name": "markov_head.markov_w2.weight", "parameters": 8192000, "shape": [ 32000, 256 ] }, { "dtype": "BF16", "name": "norm.weight", "parameters": 2048, "shape": [ 2048 ] }, { "dtype": "BOOL", "name": "t2d", "parameters": 262144, "shape": [ 262144 ] } ] }