{ "metadata": { "total_size": 3870532216 }, "weight_map": { "model.audio_tower.conformer.0.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.0.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.1.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.10.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.11.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.2.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.3.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.4.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.5.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.6.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.7.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.8.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.attention.post.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.attention.post_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.conformer.9.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.subsample_conv_projection.conv_0.conv.weight": "model-00001-of-00001.safetensors", "model.audio_tower.subsample_conv_projection.conv_0.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.subsample_conv_projection.conv_1.conv.weight": "model-00001-of-00001.safetensors", "model.audio_tower.subsample_conv_projection.conv_1.norm.weight": "model-00001-of-00001.safetensors", "model.audio_tower.subsample_conv_projection.input_proj_linear.weight": "model-00001-of-00001.safetensors", "model.embed_audio.embedding.weight": "model-00001-of-00001.safetensors", "model.embed_vision.embedding.weight": "model-00001-of-00001.safetensors", "model.language_model.altup_projections.0.weight": "model-00001-of-00001.safetensors", "model.language_model.altup_projections.1.weight": "model-00001-of-00001.safetensors", "model.language_model.altup_projections.2.weight": "model-00001-of-00001.safetensors", "model.language_model.altup_unembed_projections.0.weight": "model-00001-of-00001.safetensors", "model.language_model.altup_unembed_projections.1.weight": "model-00001-of-00001.safetensors", "model.language_model.altup_unembed_projections.2.weight": "model-00001-of-00001.safetensors", "model.language_model.embed_tokens.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.0.bn1.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.0.bn2.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.0.conv_exp.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.0.conv_pwl.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.1.bn1.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.1.bn2.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.1.conv_exp.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.1.conv_pwl.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.2.bn1.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.2.bn2.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.2.conv_exp.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.0.2.conv_pwl.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.0.dw_mid.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.0.dw_mid.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.0.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.0.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.0.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.0.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.0.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.0.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.0.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.1.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.1.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.1.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.1.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.1.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.1.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.1.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.2.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.2.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.2.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.2.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.2.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.2.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.2.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.3.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.3.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.3.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.3.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.3.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.3.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.3.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.4.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.4.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.4.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.4.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.4.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.4.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.1.4.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.0.dw_mid.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.0.dw_mid.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.0.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.0.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.0.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.0.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.0.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.0.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.0.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.1.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.1.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.1.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.1.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.1.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.1.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.1.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.10.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.10.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.10.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.10.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.10.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.11.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.11.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.11.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.11.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.11.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.11.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.11.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.11.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.11.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.11.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.12.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.12.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.12.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.12.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.12.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.13.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.13.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.13.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.13.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.13.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.13.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.13.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.13.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.13.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.13.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.14.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.14.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.14.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.14.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.14.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.15.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.15.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.15.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.15.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.15.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.15.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.15.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.15.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.15.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.15.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.16.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.16.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.16.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.16.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.16.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.17.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.17.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.17.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.17.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.17.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.17.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.17.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.17.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.17.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.17.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.18.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.18.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.18.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.18.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.18.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.19.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.19.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.19.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.19.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.19.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.19.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.19.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.19.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.19.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.19.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.2.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.2.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.2.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.2.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.2.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.2.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.2.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.20.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.20.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.20.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.20.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.20.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.21.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.21.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.21.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.21.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.21.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.21.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.21.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.21.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.21.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.21.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.22.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.22.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.22.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.22.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.22.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.23.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.23.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.23.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.23.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.23.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.23.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.23.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.23.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.23.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.23.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.24.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.24.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.24.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.24.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.24.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.25.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.25.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.25.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.25.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.25.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.25.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.25.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.25.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.25.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.25.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.26.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.26.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.26.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.26.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.26.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.27.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.27.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.27.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.27.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.27.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.27.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.27.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.27.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.27.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.27.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.28.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.28.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.28.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.28.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.28.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.29.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.29.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.29.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.29.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.29.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.29.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.29.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.29.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.29.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.29.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.3.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.3.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.3.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.3.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.3.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.3.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.3.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.30.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.30.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.30.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.30.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.30.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.31.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.31.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.31.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.31.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.31.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.31.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.31.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.31.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.31.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.31.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.32.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.32.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.32.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.32.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.32.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.33.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.33.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.33.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.33.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.33.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.33.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.33.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.33.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.33.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.33.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.34.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.34.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.34.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.34.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.34.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.35.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.35.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.35.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.35.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.35.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.35.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.35.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.35.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.35.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.35.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.36.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.36.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.36.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.36.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.36.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.4.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.4.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.4.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.4.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.4.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.4.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.4.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.5.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.5.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.5.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.5.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.5.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.5.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.5.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.6.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.6.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.6.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.6.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.6.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.6.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.6.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.7.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.7.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.7.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.7.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.7.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.7.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.7.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.8.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.8.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.8.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.8.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.8.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.9.attn.key.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.9.attn.key.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.9.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.9.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.9.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.9.attn.value.down_conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.9.attn.value.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.9.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.9.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.2.9.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.0.dw_mid.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.0.dw_mid.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.0.dw_start.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.0.dw_start.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.0.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.0.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.0.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.0.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.0.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.1.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.1.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.1.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.1.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.1.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.1.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.10.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.10.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.10.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.10.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.10.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.11.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.11.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.11.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.11.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.11.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.11.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.12.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.12.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.12.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.12.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.12.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.13.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.13.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.13.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.13.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.13.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.13.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.14.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.14.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.14.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.14.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.14.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.15.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.15.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.15.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.15.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.15.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.15.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.16.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.16.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.16.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.16.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.16.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.17.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.17.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.17.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.17.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.17.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.17.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.18.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.18.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.18.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.18.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.18.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.19.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.19.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.19.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.19.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.19.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.19.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.2.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.2.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.2.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.2.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.2.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.20.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.20.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.20.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.20.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.20.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.21.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.21.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.21.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.21.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.21.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.21.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.22.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.22.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.22.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.22.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.22.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.23.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.23.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.23.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.23.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.23.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.23.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.24.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.24.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.24.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.24.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.24.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.25.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.25.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.25.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.25.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.25.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.25.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.26.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.26.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.26.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.26.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.26.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.27.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.27.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.27.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.27.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.27.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.27.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.28.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.28.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.28.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.28.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.28.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.29.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.29.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.29.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.29.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.29.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.29.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.3.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.3.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.3.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.3.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.3.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.3.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.30.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.30.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.30.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.30.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.30.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.31.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.31.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.31.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.31.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.31.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.31.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.32.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.32.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.32.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.32.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.32.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.33.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.33.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.33.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.33.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.33.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.33.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.34.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.34.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.34.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.34.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.34.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.35.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.35.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.35.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.35.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.35.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.35.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.36.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.36.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.36.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.36.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.36.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.37.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.37.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.37.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.37.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.37.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.37.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.38.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.38.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.38.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.38.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.38.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.4.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.4.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.4.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.4.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.4.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.5.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.5.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.5.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.5.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.5.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.5.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.6.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.6.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.6.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.6.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.6.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.7.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.7.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.7.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.7.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.7.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.7.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.8.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.8.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.8.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.8.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.8.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.9.attn.key.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.9.attn.output.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.9.attn.query.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.9.attn.value.proj.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.9.layer_scale.gamma": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.blocks.3.9.norm.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.conv_stem.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.conv_stem.conv.bias": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.conv_stem.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.msfa.ffn.pw_exp.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.msfa.ffn.pw_exp.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.msfa.ffn.pw_proj.bn.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.msfa.ffn.pw_proj.conv.weight": "model-00001-of-00001.safetensors", "model.vision_tower.timm_model.msfa.norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.altup.correct_output_scale": "model-00001-of-00001.safetensors", "model.language_model.layers.1.altup.correction_coefs.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.altup.modality_router.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.altup.prediction_coefs.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.altup.router_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.laurel.linear_left.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.laurel.linear_right.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.laurel.post_laurel_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.per_layer_input_gate.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.per_layer_projection.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.post_feedforward_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.post_per_layer_input_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.pre_feedforward_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.self_attn.k_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.altup.correct_output_scale": "model-00001-of-00001.safetensors", "model.language_model.layers.2.altup.correction_coefs.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.altup.modality_router.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.altup.prediction_coefs.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.altup.router_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.laurel.linear_left.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.laurel.linear_right.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.laurel.post_laurel_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.per_layer_input_gate.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.per_layer_projection.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.post_feedforward_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.post_per_layer_input_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.pre_feedforward_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.self_attn.k_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.self_attn.q_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.embed_audio.embedding_projection.weight": "model-00001-of-00001.safetensors", "model.embed_audio.hard_embedding_norm.weight": "model-00001-of-00001.safetensors", "model.embed_audio.soft_embedding_norm.weight": "model-00001-of-00001.safetensors", "model.embed_vision.embedding_projection.weight": "model-00001-of-00001.safetensors", "model.embed_vision.hard_embedding_norm.weight": "model-00001-of-00001.safetensors", "model.embed_vision.soft_embedding_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.embed_tokens_per_layer.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.altup.correct_output_scale": "model-00001-of-00001.safetensors", "model.language_model.layers.0.altup.correction_coefs.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.altup.modality_router.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.altup.prediction_coefs.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.altup.router_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.laurel.linear_left.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.laurel.linear_right.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.laurel.post_laurel_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.per_layer_input_gate.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.per_layer_projection.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.post_feedforward_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.post_per_layer_input_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.pre_feedforward_layernorm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.self_attn.k_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.layers.0.self_attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.language_model.norm.weight": "model-00001-of-00001.safetensors", "model.language_model.per_layer_model_projection.weight": "model-00001-of-00001.safetensors", "model.language_model.per_layer_projection_norm.weight": "model-00001-of-00001.safetensors" } }