state_dict_shape.json

{"model.word_embeddings.weight": [[151936, 2048], "torch.bfloat16"], "model.layers.0.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.0.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.0.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.0.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.0.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.0.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.0.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.0.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.0.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.0.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.1.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.1.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.1.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.1.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.1.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.1.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.1.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.1.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.1.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.1.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.2.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.2.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.2.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.2.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.2.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.2.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.2.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.2.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.2.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.2.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.3.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.3.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.3.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.3.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.3.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.3.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.3.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.3.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.3.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.3.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.4.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.4.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.4.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.4.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.4.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.4.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.4.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.4.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.4.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.4.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.5.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.5.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.5.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.5.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.5.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.5.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.5.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.5.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.5.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.5.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.6.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.6.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.6.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.6.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.6.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.6.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.6.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.6.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.6.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.6.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.7.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.7.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.7.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.7.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.7.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.7.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.7.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.7.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.7.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.7.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.8.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.8.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.8.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.8.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.8.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.8.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.8.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.8.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.8.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.8.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.9.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.9.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.9.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.9.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.9.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.9.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.9.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.9.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.9.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.9.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.10.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.10.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.10.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.10.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.10.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.10.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.10.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.10.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.10.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.10.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.11.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.11.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.11.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.11.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.11.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.11.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.11.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.11.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.11.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.11.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.12.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.12.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.12.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.12.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.12.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.12.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.12.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.12.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.12.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.12.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.13.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.13.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.13.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.13.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.13.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.13.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.13.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.13.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.13.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.13.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.14.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.14.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.14.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.14.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.14.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.14.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.14.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.14.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.14.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.14.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.15.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.15.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.15.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.15.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.15.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.15.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.15.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.15.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.15.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.15.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.16.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.16.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.16.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.16.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.16.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.16.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.16.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.16.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.16.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.16.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.17.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.17.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.17.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.17.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.17.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.17.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.17.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.17.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.17.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.17.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.18.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.18.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.18.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.18.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.18.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.18.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.18.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.18.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.18.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.18.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.19.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.19.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.19.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.19.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.19.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.19.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.19.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.19.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.19.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.19.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.20.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.20.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.20.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.20.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.20.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.20.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.20.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.20.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.20.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.20.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.21.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.21.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.21.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.21.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.21.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.21.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.21.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.21.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.21.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.21.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.22.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.22.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.22.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.22.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.22.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.22.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.22.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.22.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.22.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.22.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.23.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.23.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.23.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.23.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.23.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.23.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.23.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.23.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.23.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.23.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.24.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.24.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.24.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.24.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.24.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.24.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.24.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.24.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.24.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.24.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.25.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.25.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.25.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.25.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.25.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.25.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.25.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.25.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.25.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.25.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.26.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.26.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.26.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.26.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.26.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.26.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.26.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.26.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.26.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.26.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.27.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.27.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.27.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.27.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.27.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.27.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.27.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.27.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.27.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.27.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.28.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.28.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.28.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.28.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.28.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.28.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.28.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.28.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.28.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.28.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.29.attention.query_key_value.weight": [[3072, 2048], "torch.bfloat16"], "model.layers.29.attention.query_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.29.attention.key_layernorm.weight": [[128], "torch.bfloat16"], "model.layers.29.attention.dense.weight": [[2048, 2048], "torch.bfloat16"], "model.layers.29.mlp.gate.weight": [[224, 2048], "torch.bfloat16"], "model.layers.29.mlp.gate.expert_bias": [[224], "torch.bfloat16"], "model.layers.29.mlp.experts.gate_up_proj": [[224, 1024, 2048], "torch.bfloat16"], "model.layers.29.mlp.experts.down_proj": [[224, 2048, 512], "torch.bfloat16"], "model.layers.29.input_layernorm.weight": [[2048], "torch.bfloat16"], "model.layers.29.post_attention_layernorm.weight": [[2048], "torch.bfloat16"], "model.norm.weight": [[2048], "torch.bfloat16"], "lm_head.weight": [[151936, 2048], "torch.bfloat16"]}