diff --git "a/model.safetensors.index.json" "b/model.safetensors.index.json" new file mode 100644--- /dev/null +++ "b/model.safetensors.index.json" @@ -0,0 +1,1258 @@ +{ + "metadata": { + "total_size": 206281293312 + }, + "weight_map": { + "lm_head.weight": "model-00043-of-00043.safetensors", + "model.embed_tokens.weight": "model-00001-of-00043.safetensors", + "model.final_layernorm.weight": "model-00042-of-00043.safetensors", + "model.layers.0.feed_forward.down_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.0.feed_forward.gate_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.0.feed_forward.up_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.A_log": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.D": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.b_layernorm.weight": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.c_layernorm.weight": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.conv1d.bias": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.conv1d.weight": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.dt_layernorm.weight": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.dt_proj.bias": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.dt_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.in_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.out_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.0.mamba.x_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.0.pre_ff_layernorm.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.0.down_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.0.gate_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.0.up_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.1.down_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.1.gate_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.1.up_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.10.down_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.10.gate_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.10.up_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.11.down_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.11.gate_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.11.up_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.12.down_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.12.gate_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.12.up_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.13.down_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.13.gate_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.13.up_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.14.down_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.14.gate_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.14.up_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.15.down_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.15.gate_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.15.up_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.1.feed_forward.experts.2.down_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.2.gate_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.2.up_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.3.down_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.3.gate_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.feed_forward.experts.3.up_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.4.down_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.4.gate_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.4.up_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.5.down_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.5.gate_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.5.up_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.6.down_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.6.gate_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.6.up_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.7.down_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.7.gate_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.7.up_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.8.down_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.8.gate_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.8.up_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.9.down_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.9.gate_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.experts.9.up_proj.weight": "model-00002-of-00043.safetensors", + "model.layers.1.feed_forward.router.weight": "model-00001-of-00043.safetensors", + "model.layers.1.input_layernorm.weight": "model-00003-of-00043.safetensors", + "model.layers.1.mamba.A_log": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.D": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.b_layernorm.weight": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.c_layernorm.weight": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.conv1d.bias": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.conv1d.weight": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.dt_layernorm.weight": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.dt_proj.bias": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.dt_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.in_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.out_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.mamba.x_proj.weight": "model-00001-of-00043.safetensors", + "model.layers.1.pre_ff_layernorm.weight": "model-00003-of-00043.safetensors", + "model.layers.10.feed_forward.down_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.10.feed_forward.gate_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.10.feed_forward.up_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.10.input_layernorm.weight": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.A_log": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.D": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.b_layernorm.weight": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.c_layernorm.weight": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.conv1d.bias": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.conv1d.weight": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.dt_layernorm.weight": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.dt_proj.bias": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.dt_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.in_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.out_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.10.mamba.x_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.10.pre_ff_layernorm.weight": "model-00014-of-00043.safetensors", + "model.layers.11.feed_forward.experts.0.down_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.feed_forward.experts.0.gate_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.feed_forward.experts.0.up_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.feed_forward.experts.1.down_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.feed_forward.experts.1.gate_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.feed_forward.experts.1.up_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.feed_forward.experts.10.down_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.10.gate_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.10.up_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.11.down_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.11.gate_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.11.up_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.12.down_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.12.gate_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.12.up_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.13.down_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.13.gate_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.13.up_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.14.down_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.14.gate_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.14.up_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.15.down_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.15.gate_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.15.up_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.11.feed_forward.experts.2.down_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.feed_forward.experts.2.gate_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.feed_forward.experts.2.up_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.feed_forward.experts.3.down_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.3.gate_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.3.up_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.4.down_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.4.gate_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.4.up_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.5.down_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.5.gate_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.5.up_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.6.down_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.6.gate_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.6.up_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.7.down_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.7.gate_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.7.up_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.8.down_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.8.gate_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.8.up_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.9.down_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.9.gate_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.experts.9.up_proj.weight": "model-00015-of-00043.safetensors", + "model.layers.11.feed_forward.router.weight": "model-00014-of-00043.safetensors", + "model.layers.11.input_layernorm.weight": "model-00016-of-00043.safetensors", + "model.layers.11.mamba.A_log": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.D": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.b_layernorm.weight": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.c_layernorm.weight": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.conv1d.bias": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.conv1d.weight": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.dt_layernorm.weight": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.dt_proj.bias": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.dt_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.in_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.out_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.mamba.x_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.11.pre_ff_layernorm.weight": "model-00016-of-00043.safetensors", + "model.layers.12.feed_forward.down_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.12.feed_forward.gate_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.12.feed_forward.up_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.12.input_layernorm.weight": "model-00017-of-00043.safetensors", + "model.layers.12.pre_ff_layernorm.weight": "model-00017-of-00043.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00016-of-00043.safetensors", + "model.layers.13.feed_forward.experts.0.down_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.0.gate_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.0.up_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.1.down_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.1.gate_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.1.up_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.10.down_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.10.gate_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.10.up_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.11.down_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.11.gate_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.11.up_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.12.down_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.12.gate_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.12.up_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.13.down_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.13.feed_forward.experts.13.gate_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.13.feed_forward.experts.13.up_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.13.feed_forward.experts.14.down_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.13.feed_forward.experts.14.gate_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.13.feed_forward.experts.14.up_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.13.feed_forward.experts.15.down_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.13.feed_forward.experts.15.gate_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.13.feed_forward.experts.15.up_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.13.feed_forward.experts.2.down_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.2.gate_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.2.up_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.3.down_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.3.gate_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.3.up_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.4.down_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.4.gate_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.4.up_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.5.down_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.5.gate_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.5.up_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.feed_forward.experts.6.down_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.6.gate_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.6.up_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.7.down_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.7.gate_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.7.up_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.8.down_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.8.gate_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.8.up_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.9.down_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.9.gate_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.experts.9.up_proj.weight": "model-00018-of-00043.safetensors", + "model.layers.13.feed_forward.router.weight": "model-00017-of-00043.safetensors", + "model.layers.13.input_layernorm.weight": "model-00019-of-00043.safetensors", + "model.layers.13.mamba.A_log": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.D": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.b_layernorm.weight": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.c_layernorm.weight": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.conv1d.bias": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.conv1d.weight": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.dt_layernorm.weight": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.dt_proj.bias": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.dt_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.in_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.out_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.mamba.x_proj.weight": "model-00017-of-00043.safetensors", + "model.layers.13.pre_ff_layernorm.weight": "model-00019-of-00043.safetensors", + "model.layers.14.feed_forward.down_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.14.feed_forward.gate_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.14.feed_forward.up_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.14.input_layernorm.weight": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.A_log": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.D": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.b_layernorm.weight": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.c_layernorm.weight": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.conv1d.bias": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.conv1d.weight": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.dt_layernorm.weight": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.dt_proj.bias": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.dt_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.in_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.out_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.14.mamba.x_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.14.pre_ff_layernorm.weight": "model-00019-of-00043.safetensors", + "model.layers.15.feed_forward.experts.0.down_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.15.feed_forward.experts.0.gate_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.15.feed_forward.experts.0.up_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.15.feed_forward.experts.1.down_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.1.gate_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.15.feed_forward.experts.1.up_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.15.feed_forward.experts.10.down_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.10.gate_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.10.up_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.11.down_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.11.gate_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.11.up_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.12.down_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.12.gate_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.12.up_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.13.down_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.13.gate_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.13.up_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.14.down_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.14.gate_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.14.up_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.15.down_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.15.feed_forward.experts.15.gate_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.15.up_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.2.down_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.2.gate_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.2.up_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.3.down_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.3.gate_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.3.up_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.4.down_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.4.gate_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.4.up_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.5.down_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.5.gate_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.5.up_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.6.down_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.6.gate_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.6.up_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.7.down_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.7.gate_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.7.up_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.8.down_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.8.gate_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.8.up_proj.weight": "model-00020-of-00043.safetensors", + "model.layers.15.feed_forward.experts.9.down_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.9.gate_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.experts.9.up_proj.weight": "model-00021-of-00043.safetensors", + "model.layers.15.feed_forward.router.weight": "model-00019-of-00043.safetensors", + "model.layers.15.input_layernorm.weight": "model-00022-of-00043.safetensors", + "model.layers.15.mamba.A_log": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.D": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.b_layernorm.weight": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.c_layernorm.weight": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.conv1d.bias": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.conv1d.weight": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.dt_layernorm.weight": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.dt_proj.bias": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.dt_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.in_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.out_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.15.mamba.x_proj.weight": "model-00019-of-00043.safetensors", + "model.layers.15.pre_ff_layernorm.weight": "model-00022-of-00043.safetensors", + "model.layers.16.feed_forward.down_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.16.feed_forward.gate_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.16.feed_forward.up_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.16.input_layernorm.weight": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.A_log": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.D": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.b_layernorm.weight": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.c_layernorm.weight": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.conv1d.bias": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.conv1d.weight": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.dt_layernorm.weight": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.dt_proj.bias": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.dt_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.in_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.out_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.16.mamba.x_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.16.pre_ff_layernorm.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.0.down_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.0.gate_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.0.up_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.1.down_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.1.gate_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.1.up_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.10.down_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.10.gate_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.10.up_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.11.down_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.11.gate_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.11.up_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.12.down_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.12.gate_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.12.up_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.13.down_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.13.gate_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.13.up_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.14.down_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.14.gate_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.14.up_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.15.down_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.15.gate_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.15.up_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.17.feed_forward.experts.2.down_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.2.gate_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.2.up_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.3.down_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.3.gate_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.3.up_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.4.down_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.4.gate_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.feed_forward.experts.4.up_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.5.down_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.5.gate_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.5.up_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.6.down_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.6.gate_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.6.up_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.7.down_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.7.gate_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.7.up_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.8.down_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.8.gate_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.8.up_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.9.down_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.9.gate_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.experts.9.up_proj.weight": "model-00023-of-00043.safetensors", + "model.layers.17.feed_forward.router.weight": "model-00022-of-00043.safetensors", + "model.layers.17.input_layernorm.weight": "model-00024-of-00043.safetensors", + "model.layers.17.mamba.A_log": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.D": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.b_layernorm.weight": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.c_layernorm.weight": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.conv1d.bias": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.conv1d.weight": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.dt_layernorm.weight": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.dt_proj.bias": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.dt_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.in_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.out_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.mamba.x_proj.weight": "model-00022-of-00043.safetensors", + "model.layers.17.pre_ff_layernorm.weight": "model-00024-of-00043.safetensors", + "model.layers.18.feed_forward.down_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.18.feed_forward.gate_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.18.feed_forward.up_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.18.input_layernorm.weight": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.A_log": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.D": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.b_layernorm.weight": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.c_layernorm.weight": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.conv1d.bias": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.conv1d.weight": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.dt_layernorm.weight": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.dt_proj.bias": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.dt_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.in_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.out_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.18.mamba.x_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.18.pre_ff_layernorm.weight": "model-00024-of-00043.safetensors", + "model.layers.19.feed_forward.experts.0.down_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.0.gate_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.0.up_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.1.down_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.1.gate_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.1.up_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.10.down_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.10.gate_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.10.up_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.11.down_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.11.gate_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.11.up_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.12.down_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.12.gate_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.12.up_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.13.down_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.13.gate_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.13.up_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.14.down_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.19.feed_forward.experts.14.gate_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.19.feed_forward.experts.14.up_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.19.feed_forward.experts.15.down_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.19.feed_forward.experts.15.gate_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.19.feed_forward.experts.15.up_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.19.feed_forward.experts.2.down_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.2.gate_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.2.up_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.3.down_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.3.gate_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.3.up_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.4.down_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.4.gate_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.4.up_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.5.down_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.5.gate_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.5.up_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.6.down_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.6.gate_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.6.up_proj.weight": "model-00025-of-00043.safetensors", + "model.layers.19.feed_forward.experts.7.down_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.7.gate_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.7.up_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.8.down_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.8.gate_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.8.up_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.9.down_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.9.gate_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.experts.9.up_proj.weight": "model-00026-of-00043.safetensors", + "model.layers.19.feed_forward.router.weight": "model-00024-of-00043.safetensors", + "model.layers.19.input_layernorm.weight": "model-00027-of-00043.safetensors", + "model.layers.19.mamba.A_log": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.D": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.b_layernorm.weight": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.c_layernorm.weight": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.conv1d.bias": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.conv1d.weight": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.dt_layernorm.weight": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.dt_proj.bias": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.dt_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.in_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.out_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.19.mamba.x_proj.weight": "model-00024-of-00043.safetensors", + "model.layers.19.pre_ff_layernorm.weight": "model-00027-of-00043.safetensors", + "model.layers.2.feed_forward.down_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.2.feed_forward.gate_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.2.feed_forward.up_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.2.input_layernorm.weight": "model-00004-of-00043.safetensors", + "model.layers.2.mamba.A_log": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.D": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.b_layernorm.weight": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.c_layernorm.weight": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.conv1d.bias": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.conv1d.weight": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.dt_layernorm.weight": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.dt_proj.bias": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.dt_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.in_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.out_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.2.mamba.x_proj.weight": "model-00003-of-00043.safetensors", + "model.layers.2.pre_ff_layernorm.weight": "model-00004-of-00043.safetensors", + "model.layers.20.feed_forward.down_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.20.feed_forward.gate_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.20.feed_forward.up_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.20.input_layernorm.weight": "model-00027-of-00043.safetensors", + "model.layers.20.pre_ff_layernorm.weight": "model-00027-of-00043.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.feed_forward.experts.0.down_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.feed_forward.experts.0.gate_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.feed_forward.experts.0.up_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.feed_forward.experts.1.down_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.feed_forward.experts.1.gate_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.feed_forward.experts.1.up_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.feed_forward.experts.10.down_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.10.gate_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.10.up_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.11.down_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.11.gate_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.11.up_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.12.down_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.12.gate_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.12.up_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.13.down_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.13.gate_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.13.up_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.14.down_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.14.gate_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.14.up_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.15.down_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.15.gate_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.15.up_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.21.feed_forward.experts.2.down_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.feed_forward.experts.2.gate_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.feed_forward.experts.2.up_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.feed_forward.experts.3.down_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.3.gate_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.3.up_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.4.down_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.4.gate_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.4.up_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.5.down_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.5.gate_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.5.up_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.6.down_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.6.gate_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.6.up_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.7.down_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.7.gate_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.7.up_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.8.down_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.8.gate_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.8.up_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.9.down_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.9.gate_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.experts.9.up_proj.weight": "model-00028-of-00043.safetensors", + "model.layers.21.feed_forward.router.weight": "model-00027-of-00043.safetensors", + "model.layers.21.input_layernorm.weight": "model-00029-of-00043.safetensors", + "model.layers.21.mamba.A_log": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.D": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.b_layernorm.weight": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.c_layernorm.weight": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.conv1d.bias": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.conv1d.weight": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.dt_layernorm.weight": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.dt_proj.bias": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.dt_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.in_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.out_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.mamba.x_proj.weight": "model-00027-of-00043.safetensors", + "model.layers.21.pre_ff_layernorm.weight": "model-00029-of-00043.safetensors", + "model.layers.22.feed_forward.down_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.22.feed_forward.gate_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.22.feed_forward.up_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.22.input_layernorm.weight": "model-00030-of-00043.safetensors", + "model.layers.22.mamba.A_log": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.D": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.b_layernorm.weight": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.c_layernorm.weight": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.conv1d.bias": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.conv1d.weight": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.dt_layernorm.weight": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.dt_proj.bias": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.dt_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.in_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.out_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.22.mamba.x_proj.weight": "model-00029-of-00043.safetensors", + "model.layers.22.pre_ff_layernorm.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.0.down_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.0.gate_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.0.up_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.1.down_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.1.gate_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.1.up_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.10.down_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.10.gate_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.10.up_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.11.down_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.11.gate_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.11.up_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.12.down_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.23.feed_forward.experts.12.gate_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.12.up_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.13.down_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.23.feed_forward.experts.13.gate_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.23.feed_forward.experts.13.up_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.23.feed_forward.experts.14.down_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.23.feed_forward.experts.14.gate_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.23.feed_forward.experts.14.up_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.23.feed_forward.experts.15.down_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.23.feed_forward.experts.15.gate_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.23.feed_forward.experts.15.up_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.23.feed_forward.experts.2.down_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.2.gate_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.2.up_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.3.down_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.3.gate_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.3.up_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.4.down_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.4.gate_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.4.up_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.5.down_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.5.gate_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.5.up_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.feed_forward.experts.6.down_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.6.gate_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.6.up_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.7.down_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.7.gate_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.7.up_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.8.down_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.8.gate_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.8.up_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.9.down_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.9.gate_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.experts.9.up_proj.weight": "model-00031-of-00043.safetensors", + "model.layers.23.feed_forward.router.weight": "model-00030-of-00043.safetensors", + "model.layers.23.input_layernorm.weight": "model-00032-of-00043.safetensors", + "model.layers.23.mamba.A_log": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.D": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.b_layernorm.weight": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.c_layernorm.weight": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.conv1d.bias": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.conv1d.weight": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.dt_layernorm.weight": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.dt_proj.bias": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.dt_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.in_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.out_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.mamba.x_proj.weight": "model-00030-of-00043.safetensors", + "model.layers.23.pre_ff_layernorm.weight": "model-00032-of-00043.safetensors", + "model.layers.24.feed_forward.down_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.24.feed_forward.gate_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.24.feed_forward.up_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.24.input_layernorm.weight": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.A_log": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.D": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.b_layernorm.weight": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.c_layernorm.weight": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.conv1d.bias": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.conv1d.weight": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.dt_layernorm.weight": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.dt_proj.bias": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.dt_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.in_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.out_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.24.mamba.x_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.24.pre_ff_layernorm.weight": "model-00032-of-00043.safetensors", + "model.layers.25.feed_forward.experts.0.down_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.25.feed_forward.experts.0.gate_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.25.feed_forward.experts.0.up_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.25.feed_forward.experts.1.down_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.1.gate_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.25.feed_forward.experts.1.up_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.10.down_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.10.gate_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.10.up_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.11.down_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.11.gate_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.11.up_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.12.down_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.12.gate_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.12.up_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.13.down_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.13.gate_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.13.up_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.14.down_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.14.gate_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.14.up_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.15.down_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.25.feed_forward.experts.15.gate_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.15.up_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.25.feed_forward.experts.2.down_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.2.gate_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.2.up_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.3.down_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.3.gate_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.3.up_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.4.down_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.4.gate_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.4.up_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.5.down_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.5.gate_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.5.up_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.6.down_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.6.gate_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.6.up_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.7.down_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.7.gate_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.7.up_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.8.down_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.8.gate_proj.weight": "model-00033-of-00043.safetensors", + "model.layers.25.feed_forward.experts.8.up_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.9.down_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.9.gate_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.experts.9.up_proj.weight": "model-00034-of-00043.safetensors", + "model.layers.25.feed_forward.router.weight": "model-00032-of-00043.safetensors", + "model.layers.25.input_layernorm.weight": "model-00035-of-00043.safetensors", + "model.layers.25.mamba.A_log": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.D": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.b_layernorm.weight": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.c_layernorm.weight": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.conv1d.bias": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.conv1d.weight": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.dt_layernorm.weight": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.dt_proj.bias": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.dt_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.in_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.out_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.25.mamba.x_proj.weight": "model-00032-of-00043.safetensors", + "model.layers.25.pre_ff_layernorm.weight": "model-00035-of-00043.safetensors", + "model.layers.26.feed_forward.down_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.26.feed_forward.gate_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.26.feed_forward.up_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.26.input_layernorm.weight": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.A_log": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.D": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.b_layernorm.weight": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.c_layernorm.weight": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.conv1d.bias": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.conv1d.weight": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.dt_layernorm.weight": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.dt_proj.bias": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.dt_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.in_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.out_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.26.mamba.x_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.26.pre_ff_layernorm.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.0.down_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.0.gate_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.0.up_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.1.down_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.1.gate_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.1.up_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.10.down_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.10.gate_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.10.up_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.11.down_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.11.gate_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.11.up_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.12.down_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.12.gate_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.12.up_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.13.down_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.13.gate_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.13.up_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.14.down_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.14.gate_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.14.up_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.15.down_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.15.gate_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.15.up_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.27.feed_forward.experts.2.down_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.2.gate_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.2.up_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.3.down_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.3.gate_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.3.up_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.feed_forward.experts.4.down_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.4.gate_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.4.up_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.5.down_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.5.gate_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.5.up_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.6.down_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.6.gate_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.6.up_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.7.down_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.7.gate_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.7.up_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.8.down_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.8.gate_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.8.up_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.9.down_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.9.gate_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.experts.9.up_proj.weight": "model-00036-of-00043.safetensors", + "model.layers.27.feed_forward.router.weight": "model-00035-of-00043.safetensors", + "model.layers.27.input_layernorm.weight": "model-00037-of-00043.safetensors", + "model.layers.27.mamba.A_log": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.D": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.b_layernorm.weight": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.c_layernorm.weight": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.conv1d.bias": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.conv1d.weight": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.dt_layernorm.weight": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.dt_proj.bias": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.dt_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.in_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.out_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.mamba.x_proj.weight": "model-00035-of-00043.safetensors", + "model.layers.27.pre_ff_layernorm.weight": "model-00037-of-00043.safetensors", + "model.layers.28.feed_forward.down_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.28.feed_forward.gate_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.28.feed_forward.up_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.28.input_layernorm.weight": "model-00037-of-00043.safetensors", + "model.layers.28.pre_ff_layernorm.weight": "model-00037-of-00043.safetensors", + "model.layers.28.self_attn.k_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.28.self_attn.o_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.28.self_attn.q_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.28.self_attn.v_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.29.feed_forward.experts.0.down_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.0.gate_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.0.up_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.1.down_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.1.gate_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.1.up_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.10.down_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.10.gate_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.10.up_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.11.down_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.11.gate_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.11.up_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.12.down_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.12.gate_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.12.up_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.13.down_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.13.gate_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.13.up_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.14.down_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.29.feed_forward.experts.14.gate_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.29.feed_forward.experts.14.up_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.29.feed_forward.experts.15.down_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.29.feed_forward.experts.15.gate_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.29.feed_forward.experts.15.up_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.29.feed_forward.experts.2.down_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.2.gate_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.2.up_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.3.down_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.3.gate_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.3.up_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.4.down_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.4.gate_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.4.up_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.5.down_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.5.gate_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.5.up_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.6.down_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.6.gate_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.6.up_proj.weight": "model-00038-of-00043.safetensors", + "model.layers.29.feed_forward.experts.7.down_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.7.gate_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.7.up_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.8.down_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.8.gate_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.8.up_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.9.down_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.9.gate_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.experts.9.up_proj.weight": "model-00039-of-00043.safetensors", + "model.layers.29.feed_forward.router.weight": "model-00037-of-00043.safetensors", + "model.layers.29.input_layernorm.weight": "model-00040-of-00043.safetensors", + "model.layers.29.mamba.A_log": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.D": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.b_layernorm.weight": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.c_layernorm.weight": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.conv1d.bias": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.conv1d.weight": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.dt_layernorm.weight": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.dt_proj.bias": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.dt_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.in_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.out_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.29.mamba.x_proj.weight": "model-00037-of-00043.safetensors", + "model.layers.29.pre_ff_layernorm.weight": "model-00040-of-00043.safetensors", + "model.layers.3.feed_forward.experts.0.down_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.0.gate_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.0.up_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.1.down_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.1.gate_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.1.up_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.10.down_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.10.gate_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.10.up_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.11.down_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.11.gate_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.11.up_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.12.down_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.12.gate_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.12.up_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.13.down_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.3.feed_forward.experts.13.gate_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.3.feed_forward.experts.13.up_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.3.feed_forward.experts.14.down_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.3.feed_forward.experts.14.gate_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.3.feed_forward.experts.14.up_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.3.feed_forward.experts.15.down_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.3.feed_forward.experts.15.gate_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.3.feed_forward.experts.15.up_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.3.feed_forward.experts.2.down_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.2.gate_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.2.up_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.3.down_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.3.gate_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.3.up_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.4.down_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.4.gate_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.4.up_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.5.down_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.5.gate_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.5.up_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.feed_forward.experts.6.down_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.6.gate_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.6.up_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.7.down_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.7.gate_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.7.up_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.8.down_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.8.gate_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.8.up_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.9.down_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.9.gate_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.experts.9.up_proj.weight": "model-00005-of-00043.safetensors", + "model.layers.3.feed_forward.router.weight": "model-00004-of-00043.safetensors", + "model.layers.3.input_layernorm.weight": "model-00006-of-00043.safetensors", + "model.layers.3.mamba.A_log": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.D": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.b_layernorm.weight": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.c_layernorm.weight": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.conv1d.bias": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.conv1d.weight": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.dt_layernorm.weight": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.dt_proj.bias": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.dt_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.in_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.out_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.mamba.x_proj.weight": "model-00004-of-00043.safetensors", + "model.layers.3.pre_ff_layernorm.weight": "model-00006-of-00043.safetensors", + "model.layers.30.feed_forward.down_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.30.feed_forward.gate_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.30.feed_forward.up_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.30.input_layernorm.weight": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.A_log": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.D": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.b_layernorm.weight": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.c_layernorm.weight": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.conv1d.bias": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.conv1d.weight": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.dt_layernorm.weight": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.dt_proj.bias": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.dt_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.in_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.out_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.30.mamba.x_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.30.pre_ff_layernorm.weight": "model-00040-of-00043.safetensors", + "model.layers.31.feed_forward.experts.0.down_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.feed_forward.experts.0.gate_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.feed_forward.experts.0.up_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.feed_forward.experts.1.down_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.feed_forward.experts.1.gate_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.feed_forward.experts.1.up_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.feed_forward.experts.10.down_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.10.gate_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.10.up_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.11.down_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.11.gate_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.11.up_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.12.down_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.12.gate_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.12.up_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.13.down_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.13.gate_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.13.up_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.14.down_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.14.gate_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.14.up_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.15.down_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.15.gate_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.15.up_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.2.down_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.2.gate_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.feed_forward.experts.2.up_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.feed_forward.experts.3.down_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.3.gate_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.3.up_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.4.down_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.4.gate_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.4.up_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.5.down_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.5.gate_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.5.up_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.6.down_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.6.gate_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.6.up_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.7.down_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.7.gate_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.7.up_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.8.down_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.8.gate_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.8.up_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.9.down_proj.weight": "model-00042-of-00043.safetensors", + "model.layers.31.feed_forward.experts.9.gate_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.experts.9.up_proj.weight": "model-00041-of-00043.safetensors", + "model.layers.31.feed_forward.router.weight": "model-00040-of-00043.safetensors", + "model.layers.31.input_layernorm.weight": "model-00042-of-00043.safetensors", + "model.layers.31.mamba.A_log": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.D": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.b_layernorm.weight": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.c_layernorm.weight": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.conv1d.bias": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.conv1d.weight": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.dt_layernorm.weight": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.dt_proj.bias": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.dt_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.in_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.out_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.mamba.x_proj.weight": "model-00040-of-00043.safetensors", + "model.layers.31.pre_ff_layernorm.weight": "model-00042-of-00043.safetensors", + "model.layers.4.feed_forward.down_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.4.feed_forward.gate_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.4.feed_forward.up_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.4.input_layernorm.weight": "model-00006-of-00043.safetensors", + "model.layers.4.pre_ff_layernorm.weight": "model-00006-of-00043.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.feed_forward.experts.0.down_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.feed_forward.experts.0.gate_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.feed_forward.experts.0.up_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.feed_forward.experts.1.down_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.feed_forward.experts.1.gate_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.feed_forward.experts.1.up_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.feed_forward.experts.10.down_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.10.gate_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.10.up_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.11.down_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.11.gate_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.11.up_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.12.down_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.12.gate_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.12.up_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.13.down_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.13.gate_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.13.up_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.14.down_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.14.gate_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.14.up_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.15.down_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.15.gate_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.15.up_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.2.down_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.2.gate_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.2.up_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.3.down_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.3.gate_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.3.up_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.4.down_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.4.gate_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.4.up_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.5.down_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.5.gate_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.5.up_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.6.down_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.6.gate_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.6.up_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.7.down_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.7.gate_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.7.up_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.8.down_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.8.gate_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.8.up_proj.weight": "model-00007-of-00043.safetensors", + "model.layers.5.feed_forward.experts.9.down_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.9.gate_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.experts.9.up_proj.weight": "model-00008-of-00043.safetensors", + "model.layers.5.feed_forward.router.weight": "model-00006-of-00043.safetensors", + "model.layers.5.input_layernorm.weight": "model-00008-of-00043.safetensors", + "model.layers.5.mamba.A_log": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.D": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.b_layernorm.weight": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.c_layernorm.weight": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.conv1d.bias": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.conv1d.weight": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.dt_layernorm.weight": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.dt_proj.bias": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.dt_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.in_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.out_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.mamba.x_proj.weight": "model-00006-of-00043.safetensors", + "model.layers.5.pre_ff_layernorm.weight": "model-00008-of-00043.safetensors", + "model.layers.6.feed_forward.down_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.6.feed_forward.gate_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.6.feed_forward.up_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.6.input_layernorm.weight": "model-00009-of-00043.safetensors", + "model.layers.6.mamba.A_log": "model-00008-of-00043.safetensors", + "model.layers.6.mamba.D": "model-00008-of-00043.safetensors", + "model.layers.6.mamba.b_layernorm.weight": "model-00009-of-00043.safetensors", + "model.layers.6.mamba.c_layernorm.weight": "model-00009-of-00043.safetensors", + "model.layers.6.mamba.conv1d.bias": "model-00008-of-00043.safetensors", + "model.layers.6.mamba.conv1d.weight": "model-00008-of-00043.safetensors", + "model.layers.6.mamba.dt_layernorm.weight": "model-00009-of-00043.safetensors", + "model.layers.6.mamba.dt_proj.bias": "model-00009-of-00043.safetensors", + "model.layers.6.mamba.dt_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.6.mamba.in_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.6.mamba.out_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.6.mamba.x_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.6.pre_ff_layernorm.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.0.down_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.0.gate_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.0.up_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.1.down_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.1.gate_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.1.up_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.10.down_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.10.gate_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.10.up_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.11.down_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.11.gate_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.11.up_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.12.down_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.12.gate_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.12.up_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.13.down_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.13.gate_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.13.up_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.14.down_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.14.gate_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.14.up_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.15.down_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.15.gate_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.15.up_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.7.feed_forward.experts.2.down_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.2.gate_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.2.up_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.3.down_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.3.gate_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.3.up_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.4.down_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.4.gate_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.4.up_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.feed_forward.experts.5.down_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.5.gate_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.5.up_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.6.down_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.6.gate_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.6.up_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.7.down_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.7.gate_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.7.up_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.8.down_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.8.gate_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.8.up_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.9.down_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.9.gate_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.experts.9.up_proj.weight": "model-00010-of-00043.safetensors", + "model.layers.7.feed_forward.router.weight": "model-00009-of-00043.safetensors", + "model.layers.7.input_layernorm.weight": "model-00011-of-00043.safetensors", + "model.layers.7.mamba.A_log": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.D": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.b_layernorm.weight": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.c_layernorm.weight": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.conv1d.bias": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.conv1d.weight": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.dt_layernorm.weight": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.dt_proj.bias": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.dt_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.in_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.out_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.mamba.x_proj.weight": "model-00009-of-00043.safetensors", + "model.layers.7.pre_ff_layernorm.weight": "model-00011-of-00043.safetensors", + "model.layers.8.feed_forward.down_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.8.feed_forward.gate_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.8.feed_forward.up_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.8.input_layernorm.weight": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.A_log": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.D": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.b_layernorm.weight": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.c_layernorm.weight": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.conv1d.bias": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.conv1d.weight": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.dt_layernorm.weight": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.dt_proj.bias": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.dt_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.in_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.out_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.8.mamba.x_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.8.pre_ff_layernorm.weight": "model-00011-of-00043.safetensors", + "model.layers.9.feed_forward.experts.0.down_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.0.gate_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.9.feed_forward.experts.0.up_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.1.down_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.1.gate_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.1.up_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.10.down_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.10.gate_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.10.up_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.11.down_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.11.gate_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.11.up_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.12.down_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.12.gate_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.12.up_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.13.down_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.13.gate_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.13.up_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.14.down_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.9.feed_forward.experts.14.gate_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.14.up_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.9.feed_forward.experts.15.down_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.9.feed_forward.experts.15.gate_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.9.feed_forward.experts.15.up_proj.weight": "model-00014-of-00043.safetensors", + "model.layers.9.feed_forward.experts.2.down_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.2.gate_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.2.up_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.3.down_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.3.gate_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.3.up_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.4.down_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.4.gate_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.4.up_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.5.down_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.5.gate_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.5.up_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.6.down_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.6.gate_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.6.up_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.7.down_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.7.gate_proj.weight": "model-00012-of-00043.safetensors", + "model.layers.9.feed_forward.experts.7.up_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.8.down_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.8.gate_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.8.up_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.9.down_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.9.gate_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.experts.9.up_proj.weight": "model-00013-of-00043.safetensors", + "model.layers.9.feed_forward.router.weight": "model-00011-of-00043.safetensors", + "model.layers.9.input_layernorm.weight": "model-00014-of-00043.safetensors", + "model.layers.9.mamba.A_log": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.D": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.b_layernorm.weight": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.c_layernorm.weight": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.conv1d.bias": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.conv1d.weight": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.dt_layernorm.weight": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.dt_proj.bias": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.dt_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.in_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.out_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.9.mamba.x_proj.weight": "model-00011-of-00043.safetensors", + "model.layers.9.pre_ff_layernorm.weight": "model-00014-of-00043.safetensors" + } +}