|
{ |
|
"architectures": [ |
|
"Gemma3nForConditionalGeneration" |
|
], |
|
"audio_config": { |
|
"conf_attention_chunk_size": 12, |
|
"conf_attention_context_left": 13, |
|
"conf_attention_context_right": 0, |
|
"conf_attention_logit_cap": 50.0, |
|
"conf_conv_kernel_size": 5, |
|
"conf_num_attention_heads": 8, |
|
"conf_num_hidden_layers": 12, |
|
"conf_positional_bias_size": 256, |
|
"conf_reduction_factor": 4, |
|
"conf_residual_weight": 0.5, |
|
"gradient_clipping": 10000000000.0, |
|
"hidden_size": 1536, |
|
"input_feat_size": 128, |
|
"model_type": "gemma3n_audio", |
|
"rms_norm_eps": 1e-06, |
|
"sscp_conv_channel_size": [ |
|
128, |
|
32 |
|
], |
|
"sscp_conv_eps": 0.001, |
|
"sscp_conv_kernel_size": [ |
|
[ |
|
3, |
|
3 |
|
], |
|
[ |
|
3, |
|
3 |
|
] |
|
], |
|
"sscp_conv_stride_size": [ |
|
[ |
|
2, |
|
2 |
|
], |
|
[ |
|
2, |
|
2 |
|
] |
|
], |
|
"torch_dtype": "bfloat16", |
|
"vocab_offset": 262272, |
|
"vocab_size": 128 |
|
}, |
|
"audio_soft_tokens_per_image": 188, |
|
"audio_token_id": 262273, |
|
"boa_token_id": 256000, |
|
"boi_token_id": 255999, |
|
"eoa_token_id": 262272, |
|
"eoi_token_id": 262144, |
|
"eos_token_id": [ |
|
1, |
|
106 |
|
], |
|
"image_token_id": 262145, |
|
"initializer_range": 0.02, |
|
"model_type": "gemma3n", |
|
"quantization": { |
|
"group_size": 64, |
|
"bits": 4, |
|
"model.language_model.embed_tokens": true, |
|
"model.language_model.layers.0.self_attn.q_proj": true, |
|
"model.language_model.layers.0.self_attn.k_proj": true, |
|
"model.language_model.layers.0.self_attn.v_proj": true, |
|
"model.language_model.layers.0.self_attn.o_proj": true, |
|
"model.language_model.layers.0.self_attn.q_norm": false, |
|
"model.language_model.layers.0.self_attn.k_norm": false, |
|
"model.language_model.layers.0.self_attn.v_norm": false, |
|
"model.language_model.layers.0.self_attn.rope": false, |
|
"model.language_model.layers.0.mlp.gate_proj": true, |
|
"model.language_model.layers.0.mlp.up_proj": true, |
|
"model.language_model.layers.0.mlp.down_proj": true, |
|
"model.language_model.layers.0.input_layernorm": false, |
|
"model.language_model.layers.0.post_attention_layernorm": false, |
|
"model.language_model.layers.0.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.0.post_feedforward_layernorm": false, |
|
"model.language_model.layers.0.altup.correction_coefs": false, |
|
"model.language_model.layers.0.altup.prediction_coefs": false, |
|
"model.language_model.layers.0.altup.modality_router": true, |
|
"model.language_model.layers.0.altup.router_norm": false, |
|
"model.language_model.layers.0.laurel.linear_left": true, |
|
"model.language_model.layers.0.laurel.linear_right": true, |
|
"model.language_model.layers.0.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.0.per_layer_input_gate": true, |
|
"model.language_model.layers.0.per_layer_projection": true, |
|
"model.language_model.layers.0.post_per_layer_input_norm": false, |
|
"model.language_model.layers.1.self_attn.q_proj": true, |
|
"model.language_model.layers.1.self_attn.k_proj": true, |
|
"model.language_model.layers.1.self_attn.v_proj": true, |
|
"model.language_model.layers.1.self_attn.o_proj": true, |
|
"model.language_model.layers.1.self_attn.q_norm": false, |
|
"model.language_model.layers.1.self_attn.k_norm": false, |
|
"model.language_model.layers.1.self_attn.v_norm": false, |
|
"model.language_model.layers.1.self_attn.rope": false, |
|
"model.language_model.layers.1.mlp.gate_proj": true, |
|
"model.language_model.layers.1.mlp.up_proj": true, |
|
"model.language_model.layers.1.mlp.down_proj": true, |
|
"model.language_model.layers.1.input_layernorm": false, |
|
"model.language_model.layers.1.post_attention_layernorm": false, |
|
"model.language_model.layers.1.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.1.post_feedforward_layernorm": false, |
|
"model.language_model.layers.1.altup.correction_coefs": false, |
|
"model.language_model.layers.1.altup.prediction_coefs": false, |
|
"model.language_model.layers.1.altup.modality_router": true, |
|
"model.language_model.layers.1.altup.router_norm": false, |
|
"model.language_model.layers.1.laurel.linear_left": true, |
|
"model.language_model.layers.1.laurel.linear_right": true, |
|
"model.language_model.layers.1.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.1.per_layer_input_gate": true, |
|
"model.language_model.layers.1.per_layer_projection": true, |
|
"model.language_model.layers.1.post_per_layer_input_norm": false, |
|
"model.language_model.layers.2.self_attn.q_proj": true, |
|
"model.language_model.layers.2.self_attn.k_proj": true, |
|
"model.language_model.layers.2.self_attn.v_proj": true, |
|
"model.language_model.layers.2.self_attn.o_proj": true, |
|
"model.language_model.layers.2.self_attn.q_norm": false, |
|
"model.language_model.layers.2.self_attn.k_norm": false, |
|
"model.language_model.layers.2.self_attn.v_norm": false, |
|
"model.language_model.layers.2.self_attn.rope": false, |
|
"model.language_model.layers.2.mlp.gate_proj": true, |
|
"model.language_model.layers.2.mlp.up_proj": true, |
|
"model.language_model.layers.2.mlp.down_proj": true, |
|
"model.language_model.layers.2.input_layernorm": false, |
|
"model.language_model.layers.2.post_attention_layernorm": false, |
|
"model.language_model.layers.2.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.2.post_feedforward_layernorm": false, |
|
"model.language_model.layers.2.altup.correction_coefs": false, |
|
"model.language_model.layers.2.altup.prediction_coefs": false, |
|
"model.language_model.layers.2.altup.modality_router": true, |
|
"model.language_model.layers.2.altup.router_norm": false, |
|
"model.language_model.layers.2.laurel.linear_left": true, |
|
"model.language_model.layers.2.laurel.linear_right": true, |
|
"model.language_model.layers.2.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.2.per_layer_input_gate": true, |
|
"model.language_model.layers.2.per_layer_projection": true, |
|
"model.language_model.layers.2.post_per_layer_input_norm": false, |
|
"model.language_model.layers.3.self_attn.q_proj": true, |
|
"model.language_model.layers.3.self_attn.k_proj": true, |
|
"model.language_model.layers.3.self_attn.v_proj": true, |
|
"model.language_model.layers.3.self_attn.o_proj": true, |
|
"model.language_model.layers.3.self_attn.q_norm": false, |
|
"model.language_model.layers.3.self_attn.k_norm": false, |
|
"model.language_model.layers.3.self_attn.v_norm": false, |
|
"model.language_model.layers.3.self_attn.rope": false, |
|
"model.language_model.layers.3.mlp.gate_proj": true, |
|
"model.language_model.layers.3.mlp.up_proj": true, |
|
"model.language_model.layers.3.mlp.down_proj": true, |
|
"model.language_model.layers.3.input_layernorm": false, |
|
"model.language_model.layers.3.post_attention_layernorm": false, |
|
"model.language_model.layers.3.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.3.post_feedforward_layernorm": false, |
|
"model.language_model.layers.3.altup.correction_coefs": false, |
|
"model.language_model.layers.3.altup.prediction_coefs": false, |
|
"model.language_model.layers.3.altup.modality_router": true, |
|
"model.language_model.layers.3.altup.router_norm": false, |
|
"model.language_model.layers.3.laurel.linear_left": true, |
|
"model.language_model.layers.3.laurel.linear_right": true, |
|
"model.language_model.layers.3.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.3.per_layer_input_gate": true, |
|
"model.language_model.layers.3.per_layer_projection": true, |
|
"model.language_model.layers.3.post_per_layer_input_norm": false, |
|
"model.language_model.layers.4.self_attn.q_proj": true, |
|
"model.language_model.layers.4.self_attn.k_proj": true, |
|
"model.language_model.layers.4.self_attn.v_proj": true, |
|
"model.language_model.layers.4.self_attn.o_proj": true, |
|
"model.language_model.layers.4.self_attn.q_norm": false, |
|
"model.language_model.layers.4.self_attn.k_norm": false, |
|
"model.language_model.layers.4.self_attn.v_norm": false, |
|
"model.language_model.layers.4.self_attn.rope": false, |
|
"model.language_model.layers.4.mlp.gate_proj": true, |
|
"model.language_model.layers.4.mlp.up_proj": true, |
|
"model.language_model.layers.4.mlp.down_proj": true, |
|
"model.language_model.layers.4.input_layernorm": false, |
|
"model.language_model.layers.4.post_attention_layernorm": false, |
|
"model.language_model.layers.4.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.4.post_feedforward_layernorm": false, |
|
"model.language_model.layers.4.altup.correction_coefs": false, |
|
"model.language_model.layers.4.altup.prediction_coefs": false, |
|
"model.language_model.layers.4.altup.modality_router": true, |
|
"model.language_model.layers.4.altup.router_norm": false, |
|
"model.language_model.layers.4.laurel.linear_left": true, |
|
"model.language_model.layers.4.laurel.linear_right": true, |
|
"model.language_model.layers.4.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.4.per_layer_input_gate": true, |
|
"model.language_model.layers.4.per_layer_projection": true, |
|
"model.language_model.layers.4.post_per_layer_input_norm": false, |
|
"model.language_model.layers.5.self_attn.q_proj": true, |
|
"model.language_model.layers.5.self_attn.k_proj": true, |
|
"model.language_model.layers.5.self_attn.v_proj": true, |
|
"model.language_model.layers.5.self_attn.o_proj": true, |
|
"model.language_model.layers.5.self_attn.q_norm": false, |
|
"model.language_model.layers.5.self_attn.k_norm": false, |
|
"model.language_model.layers.5.self_attn.v_norm": false, |
|
"model.language_model.layers.5.self_attn.rope": false, |
|
"model.language_model.layers.5.mlp.gate_proj": true, |
|
"model.language_model.layers.5.mlp.up_proj": true, |
|
"model.language_model.layers.5.mlp.down_proj": true, |
|
"model.language_model.layers.5.input_layernorm": false, |
|
"model.language_model.layers.5.post_attention_layernorm": false, |
|
"model.language_model.layers.5.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.5.post_feedforward_layernorm": false, |
|
"model.language_model.layers.5.altup.correction_coefs": false, |
|
"model.language_model.layers.5.altup.prediction_coefs": false, |
|
"model.language_model.layers.5.altup.modality_router": true, |
|
"model.language_model.layers.5.altup.router_norm": false, |
|
"model.language_model.layers.5.laurel.linear_left": true, |
|
"model.language_model.layers.5.laurel.linear_right": true, |
|
"model.language_model.layers.5.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.5.per_layer_input_gate": true, |
|
"model.language_model.layers.5.per_layer_projection": true, |
|
"model.language_model.layers.5.post_per_layer_input_norm": false, |
|
"model.language_model.layers.6.self_attn.q_proj": true, |
|
"model.language_model.layers.6.self_attn.k_proj": true, |
|
"model.language_model.layers.6.self_attn.v_proj": true, |
|
"model.language_model.layers.6.self_attn.o_proj": true, |
|
"model.language_model.layers.6.self_attn.q_norm": false, |
|
"model.language_model.layers.6.self_attn.k_norm": false, |
|
"model.language_model.layers.6.self_attn.v_norm": false, |
|
"model.language_model.layers.6.self_attn.rope": false, |
|
"model.language_model.layers.6.mlp.gate_proj": true, |
|
"model.language_model.layers.6.mlp.up_proj": true, |
|
"model.language_model.layers.6.mlp.down_proj": true, |
|
"model.language_model.layers.6.input_layernorm": false, |
|
"model.language_model.layers.6.post_attention_layernorm": false, |
|
"model.language_model.layers.6.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.6.post_feedforward_layernorm": false, |
|
"model.language_model.layers.6.altup.correction_coefs": false, |
|
"model.language_model.layers.6.altup.prediction_coefs": false, |
|
"model.language_model.layers.6.altup.modality_router": true, |
|
"model.language_model.layers.6.altup.router_norm": false, |
|
"model.language_model.layers.6.laurel.linear_left": true, |
|
"model.language_model.layers.6.laurel.linear_right": true, |
|
"model.language_model.layers.6.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.6.per_layer_input_gate": true, |
|
"model.language_model.layers.6.per_layer_projection": true, |
|
"model.language_model.layers.6.post_per_layer_input_norm": false, |
|
"model.language_model.layers.7.self_attn.q_proj": true, |
|
"model.language_model.layers.7.self_attn.k_proj": true, |
|
"model.language_model.layers.7.self_attn.v_proj": true, |
|
"model.language_model.layers.7.self_attn.o_proj": true, |
|
"model.language_model.layers.7.self_attn.q_norm": false, |
|
"model.language_model.layers.7.self_attn.k_norm": false, |
|
"model.language_model.layers.7.self_attn.v_norm": false, |
|
"model.language_model.layers.7.self_attn.rope": false, |
|
"model.language_model.layers.7.mlp.gate_proj": true, |
|
"model.language_model.layers.7.mlp.up_proj": true, |
|
"model.language_model.layers.7.mlp.down_proj": true, |
|
"model.language_model.layers.7.input_layernorm": false, |
|
"model.language_model.layers.7.post_attention_layernorm": false, |
|
"model.language_model.layers.7.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.7.post_feedforward_layernorm": false, |
|
"model.language_model.layers.7.altup.correction_coefs": false, |
|
"model.language_model.layers.7.altup.prediction_coefs": false, |
|
"model.language_model.layers.7.altup.modality_router": true, |
|
"model.language_model.layers.7.altup.router_norm": false, |
|
"model.language_model.layers.7.laurel.linear_left": true, |
|
"model.language_model.layers.7.laurel.linear_right": true, |
|
"model.language_model.layers.7.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.7.per_layer_input_gate": true, |
|
"model.language_model.layers.7.per_layer_projection": true, |
|
"model.language_model.layers.7.post_per_layer_input_norm": false, |
|
"model.language_model.layers.8.self_attn.q_proj": true, |
|
"model.language_model.layers.8.self_attn.k_proj": true, |
|
"model.language_model.layers.8.self_attn.v_proj": true, |
|
"model.language_model.layers.8.self_attn.o_proj": true, |
|
"model.language_model.layers.8.self_attn.q_norm": false, |
|
"model.language_model.layers.8.self_attn.k_norm": false, |
|
"model.language_model.layers.8.self_attn.v_norm": false, |
|
"model.language_model.layers.8.self_attn.rope": false, |
|
"model.language_model.layers.8.mlp.gate_proj": true, |
|
"model.language_model.layers.8.mlp.up_proj": true, |
|
"model.language_model.layers.8.mlp.down_proj": true, |
|
"model.language_model.layers.8.input_layernorm": false, |
|
"model.language_model.layers.8.post_attention_layernorm": false, |
|
"model.language_model.layers.8.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.8.post_feedforward_layernorm": false, |
|
"model.language_model.layers.8.altup.correction_coefs": false, |
|
"model.language_model.layers.8.altup.prediction_coefs": false, |
|
"model.language_model.layers.8.altup.modality_router": true, |
|
"model.language_model.layers.8.altup.router_norm": false, |
|
"model.language_model.layers.8.laurel.linear_left": true, |
|
"model.language_model.layers.8.laurel.linear_right": true, |
|
"model.language_model.layers.8.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.8.per_layer_input_gate": true, |
|
"model.language_model.layers.8.per_layer_projection": true, |
|
"model.language_model.layers.8.post_per_layer_input_norm": false, |
|
"model.language_model.layers.9.self_attn.q_proj": true, |
|
"model.language_model.layers.9.self_attn.k_proj": true, |
|
"model.language_model.layers.9.self_attn.v_proj": true, |
|
"model.language_model.layers.9.self_attn.o_proj": true, |
|
"model.language_model.layers.9.self_attn.q_norm": false, |
|
"model.language_model.layers.9.self_attn.k_norm": false, |
|
"model.language_model.layers.9.self_attn.v_norm": false, |
|
"model.language_model.layers.9.self_attn.rope": false, |
|
"model.language_model.layers.9.mlp.gate_proj": true, |
|
"model.language_model.layers.9.mlp.up_proj": true, |
|
"model.language_model.layers.9.mlp.down_proj": true, |
|
"model.language_model.layers.9.input_layernorm": false, |
|
"model.language_model.layers.9.post_attention_layernorm": false, |
|
"model.language_model.layers.9.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.9.post_feedforward_layernorm": false, |
|
"model.language_model.layers.9.altup.correction_coefs": false, |
|
"model.language_model.layers.9.altup.prediction_coefs": false, |
|
"model.language_model.layers.9.altup.modality_router": true, |
|
"model.language_model.layers.9.altup.router_norm": false, |
|
"model.language_model.layers.9.laurel.linear_left": true, |
|
"model.language_model.layers.9.laurel.linear_right": true, |
|
"model.language_model.layers.9.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.9.per_layer_input_gate": true, |
|
"model.language_model.layers.9.per_layer_projection": true, |
|
"model.language_model.layers.9.post_per_layer_input_norm": false, |
|
"model.language_model.layers.10.self_attn.q_proj": true, |
|
"model.language_model.layers.10.self_attn.k_proj": true, |
|
"model.language_model.layers.10.self_attn.v_proj": true, |
|
"model.language_model.layers.10.self_attn.o_proj": true, |
|
"model.language_model.layers.10.self_attn.q_norm": false, |
|
"model.language_model.layers.10.self_attn.k_norm": false, |
|
"model.language_model.layers.10.self_attn.v_norm": false, |
|
"model.language_model.layers.10.self_attn.rope": false, |
|
"model.language_model.layers.10.mlp.gate_proj": true, |
|
"model.language_model.layers.10.mlp.up_proj": true, |
|
"model.language_model.layers.10.mlp.down_proj": true, |
|
"model.language_model.layers.10.input_layernorm": false, |
|
"model.language_model.layers.10.post_attention_layernorm": false, |
|
"model.language_model.layers.10.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.10.post_feedforward_layernorm": false, |
|
"model.language_model.layers.10.altup.correction_coefs": false, |
|
"model.language_model.layers.10.altup.prediction_coefs": false, |
|
"model.language_model.layers.10.altup.modality_router": true, |
|
"model.language_model.layers.10.altup.router_norm": false, |
|
"model.language_model.layers.10.laurel.linear_left": true, |
|
"model.language_model.layers.10.laurel.linear_right": true, |
|
"model.language_model.layers.10.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.10.per_layer_input_gate": true, |
|
"model.language_model.layers.10.per_layer_projection": true, |
|
"model.language_model.layers.10.post_per_layer_input_norm": false, |
|
"model.language_model.layers.11.self_attn.q_proj": true, |
|
"model.language_model.layers.11.self_attn.k_proj": true, |
|
"model.language_model.layers.11.self_attn.v_proj": true, |
|
"model.language_model.layers.11.self_attn.o_proj": true, |
|
"model.language_model.layers.11.self_attn.q_norm": false, |
|
"model.language_model.layers.11.self_attn.k_norm": false, |
|
"model.language_model.layers.11.self_attn.v_norm": false, |
|
"model.language_model.layers.11.self_attn.rope": false, |
|
"model.language_model.layers.11.mlp.gate_proj": true, |
|
"model.language_model.layers.11.mlp.up_proj": true, |
|
"model.language_model.layers.11.mlp.down_proj": true, |
|
"model.language_model.layers.11.input_layernorm": false, |
|
"model.language_model.layers.11.post_attention_layernorm": false, |
|
"model.language_model.layers.11.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.11.post_feedforward_layernorm": false, |
|
"model.language_model.layers.11.altup.correction_coefs": false, |
|
"model.language_model.layers.11.altup.prediction_coefs": false, |
|
"model.language_model.layers.11.altup.modality_router": true, |
|
"model.language_model.layers.11.altup.router_norm": false, |
|
"model.language_model.layers.11.laurel.linear_left": true, |
|
"model.language_model.layers.11.laurel.linear_right": true, |
|
"model.language_model.layers.11.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.11.per_layer_input_gate": true, |
|
"model.language_model.layers.11.per_layer_projection": true, |
|
"model.language_model.layers.11.post_per_layer_input_norm": false, |
|
"model.language_model.layers.12.self_attn.q_proj": true, |
|
"model.language_model.layers.12.self_attn.k_proj": true, |
|
"model.language_model.layers.12.self_attn.v_proj": true, |
|
"model.language_model.layers.12.self_attn.o_proj": true, |
|
"model.language_model.layers.12.self_attn.q_norm": false, |
|
"model.language_model.layers.12.self_attn.k_norm": false, |
|
"model.language_model.layers.12.self_attn.v_norm": false, |
|
"model.language_model.layers.12.self_attn.rope": false, |
|
"model.language_model.layers.12.mlp.gate_proj": true, |
|
"model.language_model.layers.12.mlp.up_proj": true, |
|
"model.language_model.layers.12.mlp.down_proj": true, |
|
"model.language_model.layers.12.input_layernorm": false, |
|
"model.language_model.layers.12.post_attention_layernorm": false, |
|
"model.language_model.layers.12.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.12.post_feedforward_layernorm": false, |
|
"model.language_model.layers.12.altup.correction_coefs": false, |
|
"model.language_model.layers.12.altup.prediction_coefs": false, |
|
"model.language_model.layers.12.altup.modality_router": true, |
|
"model.language_model.layers.12.altup.router_norm": false, |
|
"model.language_model.layers.12.laurel.linear_left": true, |
|
"model.language_model.layers.12.laurel.linear_right": true, |
|
"model.language_model.layers.12.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.12.per_layer_input_gate": true, |
|
"model.language_model.layers.12.per_layer_projection": true, |
|
"model.language_model.layers.12.post_per_layer_input_norm": false, |
|
"model.language_model.layers.13.self_attn.q_proj": true, |
|
"model.language_model.layers.13.self_attn.k_proj": true, |
|
"model.language_model.layers.13.self_attn.v_proj": true, |
|
"model.language_model.layers.13.self_attn.o_proj": true, |
|
"model.language_model.layers.13.self_attn.q_norm": false, |
|
"model.language_model.layers.13.self_attn.k_norm": false, |
|
"model.language_model.layers.13.self_attn.v_norm": false, |
|
"model.language_model.layers.13.self_attn.rope": false, |
|
"model.language_model.layers.13.mlp.gate_proj": true, |
|
"model.language_model.layers.13.mlp.up_proj": true, |
|
"model.language_model.layers.13.mlp.down_proj": true, |
|
"model.language_model.layers.13.input_layernorm": false, |
|
"model.language_model.layers.13.post_attention_layernorm": false, |
|
"model.language_model.layers.13.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.13.post_feedforward_layernorm": false, |
|
"model.language_model.layers.13.altup.correction_coefs": false, |
|
"model.language_model.layers.13.altup.prediction_coefs": false, |
|
"model.language_model.layers.13.altup.modality_router": true, |
|
"model.language_model.layers.13.altup.router_norm": false, |
|
"model.language_model.layers.13.laurel.linear_left": true, |
|
"model.language_model.layers.13.laurel.linear_right": true, |
|
"model.language_model.layers.13.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.13.per_layer_input_gate": true, |
|
"model.language_model.layers.13.per_layer_projection": true, |
|
"model.language_model.layers.13.post_per_layer_input_norm": false, |
|
"model.language_model.layers.14.self_attn.q_proj": true, |
|
"model.language_model.layers.14.self_attn.k_proj": true, |
|
"model.language_model.layers.14.self_attn.v_proj": true, |
|
"model.language_model.layers.14.self_attn.o_proj": true, |
|
"model.language_model.layers.14.self_attn.q_norm": false, |
|
"model.language_model.layers.14.self_attn.k_norm": false, |
|
"model.language_model.layers.14.self_attn.v_norm": false, |
|
"model.language_model.layers.14.self_attn.rope": false, |
|
"model.language_model.layers.14.mlp.gate_proj": true, |
|
"model.language_model.layers.14.mlp.up_proj": true, |
|
"model.language_model.layers.14.mlp.down_proj": true, |
|
"model.language_model.layers.14.input_layernorm": false, |
|
"model.language_model.layers.14.post_attention_layernorm": false, |
|
"model.language_model.layers.14.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.14.post_feedforward_layernorm": false, |
|
"model.language_model.layers.14.altup.correction_coefs": false, |
|
"model.language_model.layers.14.altup.prediction_coefs": false, |
|
"model.language_model.layers.14.altup.modality_router": true, |
|
"model.language_model.layers.14.altup.router_norm": false, |
|
"model.language_model.layers.14.laurel.linear_left": true, |
|
"model.language_model.layers.14.laurel.linear_right": true, |
|
"model.language_model.layers.14.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.14.per_layer_input_gate": true, |
|
"model.language_model.layers.14.per_layer_projection": true, |
|
"model.language_model.layers.14.post_per_layer_input_norm": false, |
|
"model.language_model.layers.15.self_attn.q_proj": true, |
|
"model.language_model.layers.15.self_attn.k_proj": true, |
|
"model.language_model.layers.15.self_attn.v_proj": true, |
|
"model.language_model.layers.15.self_attn.o_proj": true, |
|
"model.language_model.layers.15.self_attn.q_norm": false, |
|
"model.language_model.layers.15.self_attn.k_norm": false, |
|
"model.language_model.layers.15.self_attn.v_norm": false, |
|
"model.language_model.layers.15.self_attn.rope": false, |
|
"model.language_model.layers.15.mlp.gate_proj": true, |
|
"model.language_model.layers.15.mlp.up_proj": true, |
|
"model.language_model.layers.15.mlp.down_proj": true, |
|
"model.language_model.layers.15.input_layernorm": false, |
|
"model.language_model.layers.15.post_attention_layernorm": false, |
|
"model.language_model.layers.15.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.15.post_feedforward_layernorm": false, |
|
"model.language_model.layers.15.altup.correction_coefs": false, |
|
"model.language_model.layers.15.altup.prediction_coefs": false, |
|
"model.language_model.layers.15.altup.modality_router": true, |
|
"model.language_model.layers.15.altup.router_norm": false, |
|
"model.language_model.layers.15.laurel.linear_left": true, |
|
"model.language_model.layers.15.laurel.linear_right": true, |
|
"model.language_model.layers.15.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.15.per_layer_input_gate": true, |
|
"model.language_model.layers.15.per_layer_projection": true, |
|
"model.language_model.layers.15.post_per_layer_input_norm": false, |
|
"model.language_model.layers.16.self_attn.q_proj": true, |
|
"model.language_model.layers.16.self_attn.k_proj": true, |
|
"model.language_model.layers.16.self_attn.v_proj": true, |
|
"model.language_model.layers.16.self_attn.o_proj": true, |
|
"model.language_model.layers.16.self_attn.q_norm": false, |
|
"model.language_model.layers.16.self_attn.k_norm": false, |
|
"model.language_model.layers.16.self_attn.v_norm": false, |
|
"model.language_model.layers.16.self_attn.rope": false, |
|
"model.language_model.layers.16.mlp.gate_proj": true, |
|
"model.language_model.layers.16.mlp.up_proj": true, |
|
"model.language_model.layers.16.mlp.down_proj": true, |
|
"model.language_model.layers.16.input_layernorm": false, |
|
"model.language_model.layers.16.post_attention_layernorm": false, |
|
"model.language_model.layers.16.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.16.post_feedforward_layernorm": false, |
|
"model.language_model.layers.16.altup.correction_coefs": false, |
|
"model.language_model.layers.16.altup.prediction_coefs": false, |
|
"model.language_model.layers.16.altup.modality_router": true, |
|
"model.language_model.layers.16.altup.router_norm": false, |
|
"model.language_model.layers.16.laurel.linear_left": true, |
|
"model.language_model.layers.16.laurel.linear_right": true, |
|
"model.language_model.layers.16.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.16.per_layer_input_gate": true, |
|
"model.language_model.layers.16.per_layer_projection": true, |
|
"model.language_model.layers.16.post_per_layer_input_norm": false, |
|
"model.language_model.layers.17.self_attn.q_proj": true, |
|
"model.language_model.layers.17.self_attn.k_proj": true, |
|
"model.language_model.layers.17.self_attn.v_proj": true, |
|
"model.language_model.layers.17.self_attn.o_proj": true, |
|
"model.language_model.layers.17.self_attn.q_norm": false, |
|
"model.language_model.layers.17.self_attn.k_norm": false, |
|
"model.language_model.layers.17.self_attn.v_norm": false, |
|
"model.language_model.layers.17.self_attn.rope": false, |
|
"model.language_model.layers.17.mlp.gate_proj": true, |
|
"model.language_model.layers.17.mlp.up_proj": true, |
|
"model.language_model.layers.17.mlp.down_proj": true, |
|
"model.language_model.layers.17.input_layernorm": false, |
|
"model.language_model.layers.17.post_attention_layernorm": false, |
|
"model.language_model.layers.17.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.17.post_feedforward_layernorm": false, |
|
"model.language_model.layers.17.altup.correction_coefs": false, |
|
"model.language_model.layers.17.altup.prediction_coefs": false, |
|
"model.language_model.layers.17.altup.modality_router": true, |
|
"model.language_model.layers.17.altup.router_norm": false, |
|
"model.language_model.layers.17.laurel.linear_left": true, |
|
"model.language_model.layers.17.laurel.linear_right": true, |
|
"model.language_model.layers.17.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.17.per_layer_input_gate": true, |
|
"model.language_model.layers.17.per_layer_projection": true, |
|
"model.language_model.layers.17.post_per_layer_input_norm": false, |
|
"model.language_model.layers.18.self_attn.q_proj": true, |
|
"model.language_model.layers.18.self_attn.k_proj": true, |
|
"model.language_model.layers.18.self_attn.v_proj": true, |
|
"model.language_model.layers.18.self_attn.o_proj": true, |
|
"model.language_model.layers.18.self_attn.q_norm": false, |
|
"model.language_model.layers.18.self_attn.k_norm": false, |
|
"model.language_model.layers.18.self_attn.v_norm": false, |
|
"model.language_model.layers.18.self_attn.rope": false, |
|
"model.language_model.layers.18.mlp.gate_proj": true, |
|
"model.language_model.layers.18.mlp.up_proj": true, |
|
"model.language_model.layers.18.mlp.down_proj": true, |
|
"model.language_model.layers.18.input_layernorm": false, |
|
"model.language_model.layers.18.post_attention_layernorm": false, |
|
"model.language_model.layers.18.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.18.post_feedforward_layernorm": false, |
|
"model.language_model.layers.18.altup.correction_coefs": false, |
|
"model.language_model.layers.18.altup.prediction_coefs": false, |
|
"model.language_model.layers.18.altup.modality_router": true, |
|
"model.language_model.layers.18.altup.router_norm": false, |
|
"model.language_model.layers.18.laurel.linear_left": true, |
|
"model.language_model.layers.18.laurel.linear_right": true, |
|
"model.language_model.layers.18.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.18.per_layer_input_gate": true, |
|
"model.language_model.layers.18.per_layer_projection": true, |
|
"model.language_model.layers.18.post_per_layer_input_norm": false, |
|
"model.language_model.layers.19.self_attn.q_proj": true, |
|
"model.language_model.layers.19.self_attn.k_proj": true, |
|
"model.language_model.layers.19.self_attn.v_proj": true, |
|
"model.language_model.layers.19.self_attn.o_proj": true, |
|
"model.language_model.layers.19.self_attn.q_norm": false, |
|
"model.language_model.layers.19.self_attn.k_norm": false, |
|
"model.language_model.layers.19.self_attn.v_norm": false, |
|
"model.language_model.layers.19.self_attn.rope": false, |
|
"model.language_model.layers.19.mlp.gate_proj": true, |
|
"model.language_model.layers.19.mlp.up_proj": true, |
|
"model.language_model.layers.19.mlp.down_proj": true, |
|
"model.language_model.layers.19.input_layernorm": false, |
|
"model.language_model.layers.19.post_attention_layernorm": false, |
|
"model.language_model.layers.19.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.19.post_feedforward_layernorm": false, |
|
"model.language_model.layers.19.altup.correction_coefs": false, |
|
"model.language_model.layers.19.altup.prediction_coefs": false, |
|
"model.language_model.layers.19.altup.modality_router": true, |
|
"model.language_model.layers.19.altup.router_norm": false, |
|
"model.language_model.layers.19.laurel.linear_left": true, |
|
"model.language_model.layers.19.laurel.linear_right": true, |
|
"model.language_model.layers.19.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.19.per_layer_input_gate": true, |
|
"model.language_model.layers.19.per_layer_projection": true, |
|
"model.language_model.layers.19.post_per_layer_input_norm": false, |
|
"model.language_model.layers.20.self_attn.q_proj": true, |
|
"model.language_model.layers.20.self_attn.k_proj": true, |
|
"model.language_model.layers.20.self_attn.v_proj": true, |
|
"model.language_model.layers.20.self_attn.o_proj": true, |
|
"model.language_model.layers.20.self_attn.q_norm": false, |
|
"model.language_model.layers.20.self_attn.k_norm": false, |
|
"model.language_model.layers.20.self_attn.v_norm": false, |
|
"model.language_model.layers.20.self_attn.rope": false, |
|
"model.language_model.layers.20.mlp.gate_proj": true, |
|
"model.language_model.layers.20.mlp.up_proj": true, |
|
"model.language_model.layers.20.mlp.down_proj": true, |
|
"model.language_model.layers.20.input_layernorm": false, |
|
"model.language_model.layers.20.post_attention_layernorm": false, |
|
"model.language_model.layers.20.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.20.post_feedforward_layernorm": false, |
|
"model.language_model.layers.20.altup.correction_coefs": false, |
|
"model.language_model.layers.20.altup.prediction_coefs": false, |
|
"model.language_model.layers.20.altup.modality_router": true, |
|
"model.language_model.layers.20.altup.router_norm": false, |
|
"model.language_model.layers.20.laurel.linear_left": true, |
|
"model.language_model.layers.20.laurel.linear_right": true, |
|
"model.language_model.layers.20.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.20.per_layer_input_gate": true, |
|
"model.language_model.layers.20.per_layer_projection": true, |
|
"model.language_model.layers.20.post_per_layer_input_norm": false, |
|
"model.language_model.layers.21.self_attn.q_proj": true, |
|
"model.language_model.layers.21.self_attn.k_proj": true, |
|
"model.language_model.layers.21.self_attn.v_proj": true, |
|
"model.language_model.layers.21.self_attn.o_proj": true, |
|
"model.language_model.layers.21.self_attn.q_norm": false, |
|
"model.language_model.layers.21.self_attn.k_norm": false, |
|
"model.language_model.layers.21.self_attn.v_norm": false, |
|
"model.language_model.layers.21.self_attn.rope": false, |
|
"model.language_model.layers.21.mlp.gate_proj": true, |
|
"model.language_model.layers.21.mlp.up_proj": true, |
|
"model.language_model.layers.21.mlp.down_proj": true, |
|
"model.language_model.layers.21.input_layernorm": false, |
|
"model.language_model.layers.21.post_attention_layernorm": false, |
|
"model.language_model.layers.21.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.21.post_feedforward_layernorm": false, |
|
"model.language_model.layers.21.altup.correction_coefs": false, |
|
"model.language_model.layers.21.altup.prediction_coefs": false, |
|
"model.language_model.layers.21.altup.modality_router": true, |
|
"model.language_model.layers.21.altup.router_norm": false, |
|
"model.language_model.layers.21.laurel.linear_left": true, |
|
"model.language_model.layers.21.laurel.linear_right": true, |
|
"model.language_model.layers.21.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.21.per_layer_input_gate": true, |
|
"model.language_model.layers.21.per_layer_projection": true, |
|
"model.language_model.layers.21.post_per_layer_input_norm": false, |
|
"model.language_model.layers.22.self_attn.q_proj": true, |
|
"model.language_model.layers.22.self_attn.k_proj": true, |
|
"model.language_model.layers.22.self_attn.v_proj": true, |
|
"model.language_model.layers.22.self_attn.o_proj": true, |
|
"model.language_model.layers.22.self_attn.q_norm": false, |
|
"model.language_model.layers.22.self_attn.k_norm": false, |
|
"model.language_model.layers.22.self_attn.v_norm": false, |
|
"model.language_model.layers.22.self_attn.rope": false, |
|
"model.language_model.layers.22.mlp.gate_proj": true, |
|
"model.language_model.layers.22.mlp.up_proj": true, |
|
"model.language_model.layers.22.mlp.down_proj": true, |
|
"model.language_model.layers.22.input_layernorm": false, |
|
"model.language_model.layers.22.post_attention_layernorm": false, |
|
"model.language_model.layers.22.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.22.post_feedforward_layernorm": false, |
|
"model.language_model.layers.22.altup.correction_coefs": false, |
|
"model.language_model.layers.22.altup.prediction_coefs": false, |
|
"model.language_model.layers.22.altup.modality_router": true, |
|
"model.language_model.layers.22.altup.router_norm": false, |
|
"model.language_model.layers.22.laurel.linear_left": true, |
|
"model.language_model.layers.22.laurel.linear_right": true, |
|
"model.language_model.layers.22.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.22.per_layer_input_gate": true, |
|
"model.language_model.layers.22.per_layer_projection": true, |
|
"model.language_model.layers.22.post_per_layer_input_norm": false, |
|
"model.language_model.layers.23.self_attn.q_proj": true, |
|
"model.language_model.layers.23.self_attn.k_proj": true, |
|
"model.language_model.layers.23.self_attn.v_proj": true, |
|
"model.language_model.layers.23.self_attn.o_proj": true, |
|
"model.language_model.layers.23.self_attn.q_norm": false, |
|
"model.language_model.layers.23.self_attn.k_norm": false, |
|
"model.language_model.layers.23.self_attn.v_norm": false, |
|
"model.language_model.layers.23.self_attn.rope": false, |
|
"model.language_model.layers.23.mlp.gate_proj": true, |
|
"model.language_model.layers.23.mlp.up_proj": true, |
|
"model.language_model.layers.23.mlp.down_proj": true, |
|
"model.language_model.layers.23.input_layernorm": false, |
|
"model.language_model.layers.23.post_attention_layernorm": false, |
|
"model.language_model.layers.23.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.23.post_feedforward_layernorm": false, |
|
"model.language_model.layers.23.altup.correction_coefs": false, |
|
"model.language_model.layers.23.altup.prediction_coefs": false, |
|
"model.language_model.layers.23.altup.modality_router": true, |
|
"model.language_model.layers.23.altup.router_norm": false, |
|
"model.language_model.layers.23.laurel.linear_left": true, |
|
"model.language_model.layers.23.laurel.linear_right": true, |
|
"model.language_model.layers.23.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.23.per_layer_input_gate": true, |
|
"model.language_model.layers.23.per_layer_projection": true, |
|
"model.language_model.layers.23.post_per_layer_input_norm": false, |
|
"model.language_model.layers.24.self_attn.q_proj": true, |
|
"model.language_model.layers.24.self_attn.k_proj": true, |
|
"model.language_model.layers.24.self_attn.v_proj": true, |
|
"model.language_model.layers.24.self_attn.o_proj": true, |
|
"model.language_model.layers.24.self_attn.q_norm": false, |
|
"model.language_model.layers.24.self_attn.k_norm": false, |
|
"model.language_model.layers.24.self_attn.v_norm": false, |
|
"model.language_model.layers.24.self_attn.rope": false, |
|
"model.language_model.layers.24.mlp.gate_proj": true, |
|
"model.language_model.layers.24.mlp.up_proj": true, |
|
"model.language_model.layers.24.mlp.down_proj": true, |
|
"model.language_model.layers.24.input_layernorm": false, |
|
"model.language_model.layers.24.post_attention_layernorm": false, |
|
"model.language_model.layers.24.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.24.post_feedforward_layernorm": false, |
|
"model.language_model.layers.24.altup.correction_coefs": false, |
|
"model.language_model.layers.24.altup.prediction_coefs": false, |
|
"model.language_model.layers.24.altup.modality_router": true, |
|
"model.language_model.layers.24.altup.router_norm": false, |
|
"model.language_model.layers.24.laurel.linear_left": true, |
|
"model.language_model.layers.24.laurel.linear_right": true, |
|
"model.language_model.layers.24.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.24.per_layer_input_gate": true, |
|
"model.language_model.layers.24.per_layer_projection": true, |
|
"model.language_model.layers.24.post_per_layer_input_norm": false, |
|
"model.language_model.layers.25.self_attn.q_proj": true, |
|
"model.language_model.layers.25.self_attn.k_proj": true, |
|
"model.language_model.layers.25.self_attn.v_proj": true, |
|
"model.language_model.layers.25.self_attn.o_proj": true, |
|
"model.language_model.layers.25.self_attn.q_norm": false, |
|
"model.language_model.layers.25.self_attn.k_norm": false, |
|
"model.language_model.layers.25.self_attn.v_norm": false, |
|
"model.language_model.layers.25.self_attn.rope": false, |
|
"model.language_model.layers.25.mlp.gate_proj": true, |
|
"model.language_model.layers.25.mlp.up_proj": true, |
|
"model.language_model.layers.25.mlp.down_proj": true, |
|
"model.language_model.layers.25.input_layernorm": false, |
|
"model.language_model.layers.25.post_attention_layernorm": false, |
|
"model.language_model.layers.25.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.25.post_feedforward_layernorm": false, |
|
"model.language_model.layers.25.altup.correction_coefs": false, |
|
"model.language_model.layers.25.altup.prediction_coefs": false, |
|
"model.language_model.layers.25.altup.modality_router": true, |
|
"model.language_model.layers.25.altup.router_norm": false, |
|
"model.language_model.layers.25.laurel.linear_left": true, |
|
"model.language_model.layers.25.laurel.linear_right": true, |
|
"model.language_model.layers.25.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.25.per_layer_input_gate": true, |
|
"model.language_model.layers.25.per_layer_projection": true, |
|
"model.language_model.layers.25.post_per_layer_input_norm": false, |
|
"model.language_model.layers.26.self_attn.q_proj": true, |
|
"model.language_model.layers.26.self_attn.k_proj": true, |
|
"model.language_model.layers.26.self_attn.v_proj": true, |
|
"model.language_model.layers.26.self_attn.o_proj": true, |
|
"model.language_model.layers.26.self_attn.q_norm": false, |
|
"model.language_model.layers.26.self_attn.k_norm": false, |
|
"model.language_model.layers.26.self_attn.v_norm": false, |
|
"model.language_model.layers.26.self_attn.rope": false, |
|
"model.language_model.layers.26.mlp.gate_proj": true, |
|
"model.language_model.layers.26.mlp.up_proj": true, |
|
"model.language_model.layers.26.mlp.down_proj": true, |
|
"model.language_model.layers.26.input_layernorm": false, |
|
"model.language_model.layers.26.post_attention_layernorm": false, |
|
"model.language_model.layers.26.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.26.post_feedforward_layernorm": false, |
|
"model.language_model.layers.26.altup.correction_coefs": false, |
|
"model.language_model.layers.26.altup.prediction_coefs": false, |
|
"model.language_model.layers.26.altup.modality_router": true, |
|
"model.language_model.layers.26.altup.router_norm": false, |
|
"model.language_model.layers.26.laurel.linear_left": true, |
|
"model.language_model.layers.26.laurel.linear_right": true, |
|
"model.language_model.layers.26.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.26.per_layer_input_gate": true, |
|
"model.language_model.layers.26.per_layer_projection": true, |
|
"model.language_model.layers.26.post_per_layer_input_norm": false, |
|
"model.language_model.layers.27.self_attn.q_proj": true, |
|
"model.language_model.layers.27.self_attn.k_proj": true, |
|
"model.language_model.layers.27.self_attn.v_proj": true, |
|
"model.language_model.layers.27.self_attn.o_proj": true, |
|
"model.language_model.layers.27.self_attn.q_norm": false, |
|
"model.language_model.layers.27.self_attn.k_norm": false, |
|
"model.language_model.layers.27.self_attn.v_norm": false, |
|
"model.language_model.layers.27.self_attn.rope": false, |
|
"model.language_model.layers.27.mlp.gate_proj": true, |
|
"model.language_model.layers.27.mlp.up_proj": true, |
|
"model.language_model.layers.27.mlp.down_proj": true, |
|
"model.language_model.layers.27.input_layernorm": false, |
|
"model.language_model.layers.27.post_attention_layernorm": false, |
|
"model.language_model.layers.27.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.27.post_feedforward_layernorm": false, |
|
"model.language_model.layers.27.altup.correction_coefs": false, |
|
"model.language_model.layers.27.altup.prediction_coefs": false, |
|
"model.language_model.layers.27.altup.modality_router": true, |
|
"model.language_model.layers.27.altup.router_norm": false, |
|
"model.language_model.layers.27.laurel.linear_left": true, |
|
"model.language_model.layers.27.laurel.linear_right": true, |
|
"model.language_model.layers.27.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.27.per_layer_input_gate": true, |
|
"model.language_model.layers.27.per_layer_projection": true, |
|
"model.language_model.layers.27.post_per_layer_input_norm": false, |
|
"model.language_model.layers.28.self_attn.q_proj": true, |
|
"model.language_model.layers.28.self_attn.k_proj": true, |
|
"model.language_model.layers.28.self_attn.v_proj": true, |
|
"model.language_model.layers.28.self_attn.o_proj": true, |
|
"model.language_model.layers.28.self_attn.q_norm": false, |
|
"model.language_model.layers.28.self_attn.k_norm": false, |
|
"model.language_model.layers.28.self_attn.v_norm": false, |
|
"model.language_model.layers.28.self_attn.rope": false, |
|
"model.language_model.layers.28.mlp.gate_proj": true, |
|
"model.language_model.layers.28.mlp.up_proj": true, |
|
"model.language_model.layers.28.mlp.down_proj": true, |
|
"model.language_model.layers.28.input_layernorm": false, |
|
"model.language_model.layers.28.post_attention_layernorm": false, |
|
"model.language_model.layers.28.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.28.post_feedforward_layernorm": false, |
|
"model.language_model.layers.28.altup.correction_coefs": false, |
|
"model.language_model.layers.28.altup.prediction_coefs": false, |
|
"model.language_model.layers.28.altup.modality_router": true, |
|
"model.language_model.layers.28.altup.router_norm": false, |
|
"model.language_model.layers.28.laurel.linear_left": true, |
|
"model.language_model.layers.28.laurel.linear_right": true, |
|
"model.language_model.layers.28.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.28.per_layer_input_gate": true, |
|
"model.language_model.layers.28.per_layer_projection": true, |
|
"model.language_model.layers.28.post_per_layer_input_norm": false, |
|
"model.language_model.layers.29.self_attn.q_proj": true, |
|
"model.language_model.layers.29.self_attn.k_proj": true, |
|
"model.language_model.layers.29.self_attn.v_proj": true, |
|
"model.language_model.layers.29.self_attn.o_proj": true, |
|
"model.language_model.layers.29.self_attn.q_norm": false, |
|
"model.language_model.layers.29.self_attn.k_norm": false, |
|
"model.language_model.layers.29.self_attn.v_norm": false, |
|
"model.language_model.layers.29.self_attn.rope": false, |
|
"model.language_model.layers.29.mlp.gate_proj": true, |
|
"model.language_model.layers.29.mlp.up_proj": true, |
|
"model.language_model.layers.29.mlp.down_proj": true, |
|
"model.language_model.layers.29.input_layernorm": false, |
|
"model.language_model.layers.29.post_attention_layernorm": false, |
|
"model.language_model.layers.29.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.29.post_feedforward_layernorm": false, |
|
"model.language_model.layers.29.altup.correction_coefs": false, |
|
"model.language_model.layers.29.altup.prediction_coefs": false, |
|
"model.language_model.layers.29.altup.modality_router": true, |
|
"model.language_model.layers.29.altup.router_norm": false, |
|
"model.language_model.layers.29.laurel.linear_left": true, |
|
"model.language_model.layers.29.laurel.linear_right": true, |
|
"model.language_model.layers.29.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.29.per_layer_input_gate": true, |
|
"model.language_model.layers.29.per_layer_projection": true, |
|
"model.language_model.layers.29.post_per_layer_input_norm": false, |
|
"model.language_model.layers.30.self_attn.q_proj": true, |
|
"model.language_model.layers.30.self_attn.k_proj": true, |
|
"model.language_model.layers.30.self_attn.v_proj": true, |
|
"model.language_model.layers.30.self_attn.o_proj": true, |
|
"model.language_model.layers.30.self_attn.q_norm": false, |
|
"model.language_model.layers.30.self_attn.k_norm": false, |
|
"model.language_model.layers.30.self_attn.v_norm": false, |
|
"model.language_model.layers.30.self_attn.rope": false, |
|
"model.language_model.layers.30.mlp.gate_proj": true, |
|
"model.language_model.layers.30.mlp.up_proj": true, |
|
"model.language_model.layers.30.mlp.down_proj": true, |
|
"model.language_model.layers.30.input_layernorm": false, |
|
"model.language_model.layers.30.post_attention_layernorm": false, |
|
"model.language_model.layers.30.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.30.post_feedforward_layernorm": false, |
|
"model.language_model.layers.30.altup.correction_coefs": false, |
|
"model.language_model.layers.30.altup.prediction_coefs": false, |
|
"model.language_model.layers.30.altup.modality_router": true, |
|
"model.language_model.layers.30.altup.router_norm": false, |
|
"model.language_model.layers.30.laurel.linear_left": true, |
|
"model.language_model.layers.30.laurel.linear_right": true, |
|
"model.language_model.layers.30.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.30.per_layer_input_gate": true, |
|
"model.language_model.layers.30.per_layer_projection": true, |
|
"model.language_model.layers.30.post_per_layer_input_norm": false, |
|
"model.language_model.layers.31.self_attn.q_proj": true, |
|
"model.language_model.layers.31.self_attn.k_proj": true, |
|
"model.language_model.layers.31.self_attn.v_proj": true, |
|
"model.language_model.layers.31.self_attn.o_proj": true, |
|
"model.language_model.layers.31.self_attn.q_norm": false, |
|
"model.language_model.layers.31.self_attn.k_norm": false, |
|
"model.language_model.layers.31.self_attn.v_norm": false, |
|
"model.language_model.layers.31.self_attn.rope": false, |
|
"model.language_model.layers.31.mlp.gate_proj": true, |
|
"model.language_model.layers.31.mlp.up_proj": true, |
|
"model.language_model.layers.31.mlp.down_proj": true, |
|
"model.language_model.layers.31.input_layernorm": false, |
|
"model.language_model.layers.31.post_attention_layernorm": false, |
|
"model.language_model.layers.31.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.31.post_feedforward_layernorm": false, |
|
"model.language_model.layers.31.altup.correction_coefs": false, |
|
"model.language_model.layers.31.altup.prediction_coefs": false, |
|
"model.language_model.layers.31.altup.modality_router": true, |
|
"model.language_model.layers.31.altup.router_norm": false, |
|
"model.language_model.layers.31.laurel.linear_left": true, |
|
"model.language_model.layers.31.laurel.linear_right": true, |
|
"model.language_model.layers.31.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.31.per_layer_input_gate": true, |
|
"model.language_model.layers.31.per_layer_projection": true, |
|
"model.language_model.layers.31.post_per_layer_input_norm": false, |
|
"model.language_model.layers.32.self_attn.q_proj": true, |
|
"model.language_model.layers.32.self_attn.k_proj": true, |
|
"model.language_model.layers.32.self_attn.v_proj": true, |
|
"model.language_model.layers.32.self_attn.o_proj": true, |
|
"model.language_model.layers.32.self_attn.q_norm": false, |
|
"model.language_model.layers.32.self_attn.k_norm": false, |
|
"model.language_model.layers.32.self_attn.v_norm": false, |
|
"model.language_model.layers.32.self_attn.rope": false, |
|
"model.language_model.layers.32.mlp.gate_proj": true, |
|
"model.language_model.layers.32.mlp.up_proj": true, |
|
"model.language_model.layers.32.mlp.down_proj": true, |
|
"model.language_model.layers.32.input_layernorm": false, |
|
"model.language_model.layers.32.post_attention_layernorm": false, |
|
"model.language_model.layers.32.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.32.post_feedforward_layernorm": false, |
|
"model.language_model.layers.32.altup.correction_coefs": false, |
|
"model.language_model.layers.32.altup.prediction_coefs": false, |
|
"model.language_model.layers.32.altup.modality_router": true, |
|
"model.language_model.layers.32.altup.router_norm": false, |
|
"model.language_model.layers.32.laurel.linear_left": true, |
|
"model.language_model.layers.32.laurel.linear_right": true, |
|
"model.language_model.layers.32.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.32.per_layer_input_gate": true, |
|
"model.language_model.layers.32.per_layer_projection": true, |
|
"model.language_model.layers.32.post_per_layer_input_norm": false, |
|
"model.language_model.layers.33.self_attn.q_proj": true, |
|
"model.language_model.layers.33.self_attn.k_proj": true, |
|
"model.language_model.layers.33.self_attn.v_proj": true, |
|
"model.language_model.layers.33.self_attn.o_proj": true, |
|
"model.language_model.layers.33.self_attn.q_norm": false, |
|
"model.language_model.layers.33.self_attn.k_norm": false, |
|
"model.language_model.layers.33.self_attn.v_norm": false, |
|
"model.language_model.layers.33.self_attn.rope": false, |
|
"model.language_model.layers.33.mlp.gate_proj": true, |
|
"model.language_model.layers.33.mlp.up_proj": true, |
|
"model.language_model.layers.33.mlp.down_proj": true, |
|
"model.language_model.layers.33.input_layernorm": false, |
|
"model.language_model.layers.33.post_attention_layernorm": false, |
|
"model.language_model.layers.33.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.33.post_feedforward_layernorm": false, |
|
"model.language_model.layers.33.altup.correction_coefs": false, |
|
"model.language_model.layers.33.altup.prediction_coefs": false, |
|
"model.language_model.layers.33.altup.modality_router": true, |
|
"model.language_model.layers.33.altup.router_norm": false, |
|
"model.language_model.layers.33.laurel.linear_left": true, |
|
"model.language_model.layers.33.laurel.linear_right": true, |
|
"model.language_model.layers.33.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.33.per_layer_input_gate": true, |
|
"model.language_model.layers.33.per_layer_projection": true, |
|
"model.language_model.layers.33.post_per_layer_input_norm": false, |
|
"model.language_model.layers.34.self_attn.q_proj": true, |
|
"model.language_model.layers.34.self_attn.k_proj": true, |
|
"model.language_model.layers.34.self_attn.v_proj": true, |
|
"model.language_model.layers.34.self_attn.o_proj": true, |
|
"model.language_model.layers.34.self_attn.q_norm": false, |
|
"model.language_model.layers.34.self_attn.k_norm": false, |
|
"model.language_model.layers.34.self_attn.v_norm": false, |
|
"model.language_model.layers.34.self_attn.rope": false, |
|
"model.language_model.layers.34.mlp.gate_proj": true, |
|
"model.language_model.layers.34.mlp.up_proj": true, |
|
"model.language_model.layers.34.mlp.down_proj": true, |
|
"model.language_model.layers.34.input_layernorm": false, |
|
"model.language_model.layers.34.post_attention_layernorm": false, |
|
"model.language_model.layers.34.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.34.post_feedforward_layernorm": false, |
|
"model.language_model.layers.34.altup.correction_coefs": false, |
|
"model.language_model.layers.34.altup.prediction_coefs": false, |
|
"model.language_model.layers.34.altup.modality_router": true, |
|
"model.language_model.layers.34.altup.router_norm": false, |
|
"model.language_model.layers.34.laurel.linear_left": true, |
|
"model.language_model.layers.34.laurel.linear_right": true, |
|
"model.language_model.layers.34.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.34.per_layer_input_gate": true, |
|
"model.language_model.layers.34.per_layer_projection": true, |
|
"model.language_model.layers.34.post_per_layer_input_norm": false, |
|
"model.language_model.embed_tokens_per_layer": true, |
|
"model.language_model.per_layer_model_projection": true, |
|
"model.language_model.per_layer_projection_norm": false, |
|
"model.language_model.altup_projections.0": true, |
|
"model.language_model.altup_projections.1": true, |
|
"model.language_model.altup_projections.2": true, |
|
"model.language_model.altup_unembed_projections.0": true, |
|
"model.language_model.altup_unembed_projections.1": true, |
|
"model.language_model.altup_unembed_projections.2": true, |
|
"model.language_model.norm": false |
|
}, |
|
"quantization_config": { |
|
"group_size": 64, |
|
"bits": 4, |
|
"model.language_model.embed_tokens": true, |
|
"model.language_model.layers.0.self_attn.q_proj": true, |
|
"model.language_model.layers.0.self_attn.k_proj": true, |
|
"model.language_model.layers.0.self_attn.v_proj": true, |
|
"model.language_model.layers.0.self_attn.o_proj": true, |
|
"model.language_model.layers.0.self_attn.q_norm": false, |
|
"model.language_model.layers.0.self_attn.k_norm": false, |
|
"model.language_model.layers.0.self_attn.v_norm": false, |
|
"model.language_model.layers.0.self_attn.rope": false, |
|
"model.language_model.layers.0.mlp.gate_proj": true, |
|
"model.language_model.layers.0.mlp.up_proj": true, |
|
"model.language_model.layers.0.mlp.down_proj": true, |
|
"model.language_model.layers.0.input_layernorm": false, |
|
"model.language_model.layers.0.post_attention_layernorm": false, |
|
"model.language_model.layers.0.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.0.post_feedforward_layernorm": false, |
|
"model.language_model.layers.0.altup.correction_coefs": false, |
|
"model.language_model.layers.0.altup.prediction_coefs": false, |
|
"model.language_model.layers.0.altup.modality_router": true, |
|
"model.language_model.layers.0.altup.router_norm": false, |
|
"model.language_model.layers.0.laurel.linear_left": true, |
|
"model.language_model.layers.0.laurel.linear_right": true, |
|
"model.language_model.layers.0.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.0.per_layer_input_gate": true, |
|
"model.language_model.layers.0.per_layer_projection": true, |
|
"model.language_model.layers.0.post_per_layer_input_norm": false, |
|
"model.language_model.layers.1.self_attn.q_proj": true, |
|
"model.language_model.layers.1.self_attn.k_proj": true, |
|
"model.language_model.layers.1.self_attn.v_proj": true, |
|
"model.language_model.layers.1.self_attn.o_proj": true, |
|
"model.language_model.layers.1.self_attn.q_norm": false, |
|
"model.language_model.layers.1.self_attn.k_norm": false, |
|
"model.language_model.layers.1.self_attn.v_norm": false, |
|
"model.language_model.layers.1.self_attn.rope": false, |
|
"model.language_model.layers.1.mlp.gate_proj": true, |
|
"model.language_model.layers.1.mlp.up_proj": true, |
|
"model.language_model.layers.1.mlp.down_proj": true, |
|
"model.language_model.layers.1.input_layernorm": false, |
|
"model.language_model.layers.1.post_attention_layernorm": false, |
|
"model.language_model.layers.1.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.1.post_feedforward_layernorm": false, |
|
"model.language_model.layers.1.altup.correction_coefs": false, |
|
"model.language_model.layers.1.altup.prediction_coefs": false, |
|
"model.language_model.layers.1.altup.modality_router": true, |
|
"model.language_model.layers.1.altup.router_norm": false, |
|
"model.language_model.layers.1.laurel.linear_left": true, |
|
"model.language_model.layers.1.laurel.linear_right": true, |
|
"model.language_model.layers.1.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.1.per_layer_input_gate": true, |
|
"model.language_model.layers.1.per_layer_projection": true, |
|
"model.language_model.layers.1.post_per_layer_input_norm": false, |
|
"model.language_model.layers.2.self_attn.q_proj": true, |
|
"model.language_model.layers.2.self_attn.k_proj": true, |
|
"model.language_model.layers.2.self_attn.v_proj": true, |
|
"model.language_model.layers.2.self_attn.o_proj": true, |
|
"model.language_model.layers.2.self_attn.q_norm": false, |
|
"model.language_model.layers.2.self_attn.k_norm": false, |
|
"model.language_model.layers.2.self_attn.v_norm": false, |
|
"model.language_model.layers.2.self_attn.rope": false, |
|
"model.language_model.layers.2.mlp.gate_proj": true, |
|
"model.language_model.layers.2.mlp.up_proj": true, |
|
"model.language_model.layers.2.mlp.down_proj": true, |
|
"model.language_model.layers.2.input_layernorm": false, |
|
"model.language_model.layers.2.post_attention_layernorm": false, |
|
"model.language_model.layers.2.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.2.post_feedforward_layernorm": false, |
|
"model.language_model.layers.2.altup.correction_coefs": false, |
|
"model.language_model.layers.2.altup.prediction_coefs": false, |
|
"model.language_model.layers.2.altup.modality_router": true, |
|
"model.language_model.layers.2.altup.router_norm": false, |
|
"model.language_model.layers.2.laurel.linear_left": true, |
|
"model.language_model.layers.2.laurel.linear_right": true, |
|
"model.language_model.layers.2.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.2.per_layer_input_gate": true, |
|
"model.language_model.layers.2.per_layer_projection": true, |
|
"model.language_model.layers.2.post_per_layer_input_norm": false, |
|
"model.language_model.layers.3.self_attn.q_proj": true, |
|
"model.language_model.layers.3.self_attn.k_proj": true, |
|
"model.language_model.layers.3.self_attn.v_proj": true, |
|
"model.language_model.layers.3.self_attn.o_proj": true, |
|
"model.language_model.layers.3.self_attn.q_norm": false, |
|
"model.language_model.layers.3.self_attn.k_norm": false, |
|
"model.language_model.layers.3.self_attn.v_norm": false, |
|
"model.language_model.layers.3.self_attn.rope": false, |
|
"model.language_model.layers.3.mlp.gate_proj": true, |
|
"model.language_model.layers.3.mlp.up_proj": true, |
|
"model.language_model.layers.3.mlp.down_proj": true, |
|
"model.language_model.layers.3.input_layernorm": false, |
|
"model.language_model.layers.3.post_attention_layernorm": false, |
|
"model.language_model.layers.3.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.3.post_feedforward_layernorm": false, |
|
"model.language_model.layers.3.altup.correction_coefs": false, |
|
"model.language_model.layers.3.altup.prediction_coefs": false, |
|
"model.language_model.layers.3.altup.modality_router": true, |
|
"model.language_model.layers.3.altup.router_norm": false, |
|
"model.language_model.layers.3.laurel.linear_left": true, |
|
"model.language_model.layers.3.laurel.linear_right": true, |
|
"model.language_model.layers.3.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.3.per_layer_input_gate": true, |
|
"model.language_model.layers.3.per_layer_projection": true, |
|
"model.language_model.layers.3.post_per_layer_input_norm": false, |
|
"model.language_model.layers.4.self_attn.q_proj": true, |
|
"model.language_model.layers.4.self_attn.k_proj": true, |
|
"model.language_model.layers.4.self_attn.v_proj": true, |
|
"model.language_model.layers.4.self_attn.o_proj": true, |
|
"model.language_model.layers.4.self_attn.q_norm": false, |
|
"model.language_model.layers.4.self_attn.k_norm": false, |
|
"model.language_model.layers.4.self_attn.v_norm": false, |
|
"model.language_model.layers.4.self_attn.rope": false, |
|
"model.language_model.layers.4.mlp.gate_proj": true, |
|
"model.language_model.layers.4.mlp.up_proj": true, |
|
"model.language_model.layers.4.mlp.down_proj": true, |
|
"model.language_model.layers.4.input_layernorm": false, |
|
"model.language_model.layers.4.post_attention_layernorm": false, |
|
"model.language_model.layers.4.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.4.post_feedforward_layernorm": false, |
|
"model.language_model.layers.4.altup.correction_coefs": false, |
|
"model.language_model.layers.4.altup.prediction_coefs": false, |
|
"model.language_model.layers.4.altup.modality_router": true, |
|
"model.language_model.layers.4.altup.router_norm": false, |
|
"model.language_model.layers.4.laurel.linear_left": true, |
|
"model.language_model.layers.4.laurel.linear_right": true, |
|
"model.language_model.layers.4.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.4.per_layer_input_gate": true, |
|
"model.language_model.layers.4.per_layer_projection": true, |
|
"model.language_model.layers.4.post_per_layer_input_norm": false, |
|
"model.language_model.layers.5.self_attn.q_proj": true, |
|
"model.language_model.layers.5.self_attn.k_proj": true, |
|
"model.language_model.layers.5.self_attn.v_proj": true, |
|
"model.language_model.layers.5.self_attn.o_proj": true, |
|
"model.language_model.layers.5.self_attn.q_norm": false, |
|
"model.language_model.layers.5.self_attn.k_norm": false, |
|
"model.language_model.layers.5.self_attn.v_norm": false, |
|
"model.language_model.layers.5.self_attn.rope": false, |
|
"model.language_model.layers.5.mlp.gate_proj": true, |
|
"model.language_model.layers.5.mlp.up_proj": true, |
|
"model.language_model.layers.5.mlp.down_proj": true, |
|
"model.language_model.layers.5.input_layernorm": false, |
|
"model.language_model.layers.5.post_attention_layernorm": false, |
|
"model.language_model.layers.5.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.5.post_feedforward_layernorm": false, |
|
"model.language_model.layers.5.altup.correction_coefs": false, |
|
"model.language_model.layers.5.altup.prediction_coefs": false, |
|
"model.language_model.layers.5.altup.modality_router": true, |
|
"model.language_model.layers.5.altup.router_norm": false, |
|
"model.language_model.layers.5.laurel.linear_left": true, |
|
"model.language_model.layers.5.laurel.linear_right": true, |
|
"model.language_model.layers.5.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.5.per_layer_input_gate": true, |
|
"model.language_model.layers.5.per_layer_projection": true, |
|
"model.language_model.layers.5.post_per_layer_input_norm": false, |
|
"model.language_model.layers.6.self_attn.q_proj": true, |
|
"model.language_model.layers.6.self_attn.k_proj": true, |
|
"model.language_model.layers.6.self_attn.v_proj": true, |
|
"model.language_model.layers.6.self_attn.o_proj": true, |
|
"model.language_model.layers.6.self_attn.q_norm": false, |
|
"model.language_model.layers.6.self_attn.k_norm": false, |
|
"model.language_model.layers.6.self_attn.v_norm": false, |
|
"model.language_model.layers.6.self_attn.rope": false, |
|
"model.language_model.layers.6.mlp.gate_proj": true, |
|
"model.language_model.layers.6.mlp.up_proj": true, |
|
"model.language_model.layers.6.mlp.down_proj": true, |
|
"model.language_model.layers.6.input_layernorm": false, |
|
"model.language_model.layers.6.post_attention_layernorm": false, |
|
"model.language_model.layers.6.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.6.post_feedforward_layernorm": false, |
|
"model.language_model.layers.6.altup.correction_coefs": false, |
|
"model.language_model.layers.6.altup.prediction_coefs": false, |
|
"model.language_model.layers.6.altup.modality_router": true, |
|
"model.language_model.layers.6.altup.router_norm": false, |
|
"model.language_model.layers.6.laurel.linear_left": true, |
|
"model.language_model.layers.6.laurel.linear_right": true, |
|
"model.language_model.layers.6.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.6.per_layer_input_gate": true, |
|
"model.language_model.layers.6.per_layer_projection": true, |
|
"model.language_model.layers.6.post_per_layer_input_norm": false, |
|
"model.language_model.layers.7.self_attn.q_proj": true, |
|
"model.language_model.layers.7.self_attn.k_proj": true, |
|
"model.language_model.layers.7.self_attn.v_proj": true, |
|
"model.language_model.layers.7.self_attn.o_proj": true, |
|
"model.language_model.layers.7.self_attn.q_norm": false, |
|
"model.language_model.layers.7.self_attn.k_norm": false, |
|
"model.language_model.layers.7.self_attn.v_norm": false, |
|
"model.language_model.layers.7.self_attn.rope": false, |
|
"model.language_model.layers.7.mlp.gate_proj": true, |
|
"model.language_model.layers.7.mlp.up_proj": true, |
|
"model.language_model.layers.7.mlp.down_proj": true, |
|
"model.language_model.layers.7.input_layernorm": false, |
|
"model.language_model.layers.7.post_attention_layernorm": false, |
|
"model.language_model.layers.7.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.7.post_feedforward_layernorm": false, |
|
"model.language_model.layers.7.altup.correction_coefs": false, |
|
"model.language_model.layers.7.altup.prediction_coefs": false, |
|
"model.language_model.layers.7.altup.modality_router": true, |
|
"model.language_model.layers.7.altup.router_norm": false, |
|
"model.language_model.layers.7.laurel.linear_left": true, |
|
"model.language_model.layers.7.laurel.linear_right": true, |
|
"model.language_model.layers.7.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.7.per_layer_input_gate": true, |
|
"model.language_model.layers.7.per_layer_projection": true, |
|
"model.language_model.layers.7.post_per_layer_input_norm": false, |
|
"model.language_model.layers.8.self_attn.q_proj": true, |
|
"model.language_model.layers.8.self_attn.k_proj": true, |
|
"model.language_model.layers.8.self_attn.v_proj": true, |
|
"model.language_model.layers.8.self_attn.o_proj": true, |
|
"model.language_model.layers.8.self_attn.q_norm": false, |
|
"model.language_model.layers.8.self_attn.k_norm": false, |
|
"model.language_model.layers.8.self_attn.v_norm": false, |
|
"model.language_model.layers.8.self_attn.rope": false, |
|
"model.language_model.layers.8.mlp.gate_proj": true, |
|
"model.language_model.layers.8.mlp.up_proj": true, |
|
"model.language_model.layers.8.mlp.down_proj": true, |
|
"model.language_model.layers.8.input_layernorm": false, |
|
"model.language_model.layers.8.post_attention_layernorm": false, |
|
"model.language_model.layers.8.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.8.post_feedforward_layernorm": false, |
|
"model.language_model.layers.8.altup.correction_coefs": false, |
|
"model.language_model.layers.8.altup.prediction_coefs": false, |
|
"model.language_model.layers.8.altup.modality_router": true, |
|
"model.language_model.layers.8.altup.router_norm": false, |
|
"model.language_model.layers.8.laurel.linear_left": true, |
|
"model.language_model.layers.8.laurel.linear_right": true, |
|
"model.language_model.layers.8.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.8.per_layer_input_gate": true, |
|
"model.language_model.layers.8.per_layer_projection": true, |
|
"model.language_model.layers.8.post_per_layer_input_norm": false, |
|
"model.language_model.layers.9.self_attn.q_proj": true, |
|
"model.language_model.layers.9.self_attn.k_proj": true, |
|
"model.language_model.layers.9.self_attn.v_proj": true, |
|
"model.language_model.layers.9.self_attn.o_proj": true, |
|
"model.language_model.layers.9.self_attn.q_norm": false, |
|
"model.language_model.layers.9.self_attn.k_norm": false, |
|
"model.language_model.layers.9.self_attn.v_norm": false, |
|
"model.language_model.layers.9.self_attn.rope": false, |
|
"model.language_model.layers.9.mlp.gate_proj": true, |
|
"model.language_model.layers.9.mlp.up_proj": true, |
|
"model.language_model.layers.9.mlp.down_proj": true, |
|
"model.language_model.layers.9.input_layernorm": false, |
|
"model.language_model.layers.9.post_attention_layernorm": false, |
|
"model.language_model.layers.9.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.9.post_feedforward_layernorm": false, |
|
"model.language_model.layers.9.altup.correction_coefs": false, |
|
"model.language_model.layers.9.altup.prediction_coefs": false, |
|
"model.language_model.layers.9.altup.modality_router": true, |
|
"model.language_model.layers.9.altup.router_norm": false, |
|
"model.language_model.layers.9.laurel.linear_left": true, |
|
"model.language_model.layers.9.laurel.linear_right": true, |
|
"model.language_model.layers.9.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.9.per_layer_input_gate": true, |
|
"model.language_model.layers.9.per_layer_projection": true, |
|
"model.language_model.layers.9.post_per_layer_input_norm": false, |
|
"model.language_model.layers.10.self_attn.q_proj": true, |
|
"model.language_model.layers.10.self_attn.k_proj": true, |
|
"model.language_model.layers.10.self_attn.v_proj": true, |
|
"model.language_model.layers.10.self_attn.o_proj": true, |
|
"model.language_model.layers.10.self_attn.q_norm": false, |
|
"model.language_model.layers.10.self_attn.k_norm": false, |
|
"model.language_model.layers.10.self_attn.v_norm": false, |
|
"model.language_model.layers.10.self_attn.rope": false, |
|
"model.language_model.layers.10.mlp.gate_proj": true, |
|
"model.language_model.layers.10.mlp.up_proj": true, |
|
"model.language_model.layers.10.mlp.down_proj": true, |
|
"model.language_model.layers.10.input_layernorm": false, |
|
"model.language_model.layers.10.post_attention_layernorm": false, |
|
"model.language_model.layers.10.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.10.post_feedforward_layernorm": false, |
|
"model.language_model.layers.10.altup.correction_coefs": false, |
|
"model.language_model.layers.10.altup.prediction_coefs": false, |
|
"model.language_model.layers.10.altup.modality_router": true, |
|
"model.language_model.layers.10.altup.router_norm": false, |
|
"model.language_model.layers.10.laurel.linear_left": true, |
|
"model.language_model.layers.10.laurel.linear_right": true, |
|
"model.language_model.layers.10.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.10.per_layer_input_gate": true, |
|
"model.language_model.layers.10.per_layer_projection": true, |
|
"model.language_model.layers.10.post_per_layer_input_norm": false, |
|
"model.language_model.layers.11.self_attn.q_proj": true, |
|
"model.language_model.layers.11.self_attn.k_proj": true, |
|
"model.language_model.layers.11.self_attn.v_proj": true, |
|
"model.language_model.layers.11.self_attn.o_proj": true, |
|
"model.language_model.layers.11.self_attn.q_norm": false, |
|
"model.language_model.layers.11.self_attn.k_norm": false, |
|
"model.language_model.layers.11.self_attn.v_norm": false, |
|
"model.language_model.layers.11.self_attn.rope": false, |
|
"model.language_model.layers.11.mlp.gate_proj": true, |
|
"model.language_model.layers.11.mlp.up_proj": true, |
|
"model.language_model.layers.11.mlp.down_proj": true, |
|
"model.language_model.layers.11.input_layernorm": false, |
|
"model.language_model.layers.11.post_attention_layernorm": false, |
|
"model.language_model.layers.11.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.11.post_feedforward_layernorm": false, |
|
"model.language_model.layers.11.altup.correction_coefs": false, |
|
"model.language_model.layers.11.altup.prediction_coefs": false, |
|
"model.language_model.layers.11.altup.modality_router": true, |
|
"model.language_model.layers.11.altup.router_norm": false, |
|
"model.language_model.layers.11.laurel.linear_left": true, |
|
"model.language_model.layers.11.laurel.linear_right": true, |
|
"model.language_model.layers.11.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.11.per_layer_input_gate": true, |
|
"model.language_model.layers.11.per_layer_projection": true, |
|
"model.language_model.layers.11.post_per_layer_input_norm": false, |
|
"model.language_model.layers.12.self_attn.q_proj": true, |
|
"model.language_model.layers.12.self_attn.k_proj": true, |
|
"model.language_model.layers.12.self_attn.v_proj": true, |
|
"model.language_model.layers.12.self_attn.o_proj": true, |
|
"model.language_model.layers.12.self_attn.q_norm": false, |
|
"model.language_model.layers.12.self_attn.k_norm": false, |
|
"model.language_model.layers.12.self_attn.v_norm": false, |
|
"model.language_model.layers.12.self_attn.rope": false, |
|
"model.language_model.layers.12.mlp.gate_proj": true, |
|
"model.language_model.layers.12.mlp.up_proj": true, |
|
"model.language_model.layers.12.mlp.down_proj": true, |
|
"model.language_model.layers.12.input_layernorm": false, |
|
"model.language_model.layers.12.post_attention_layernorm": false, |
|
"model.language_model.layers.12.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.12.post_feedforward_layernorm": false, |
|
"model.language_model.layers.12.altup.correction_coefs": false, |
|
"model.language_model.layers.12.altup.prediction_coefs": false, |
|
"model.language_model.layers.12.altup.modality_router": true, |
|
"model.language_model.layers.12.altup.router_norm": false, |
|
"model.language_model.layers.12.laurel.linear_left": true, |
|
"model.language_model.layers.12.laurel.linear_right": true, |
|
"model.language_model.layers.12.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.12.per_layer_input_gate": true, |
|
"model.language_model.layers.12.per_layer_projection": true, |
|
"model.language_model.layers.12.post_per_layer_input_norm": false, |
|
"model.language_model.layers.13.self_attn.q_proj": true, |
|
"model.language_model.layers.13.self_attn.k_proj": true, |
|
"model.language_model.layers.13.self_attn.v_proj": true, |
|
"model.language_model.layers.13.self_attn.o_proj": true, |
|
"model.language_model.layers.13.self_attn.q_norm": false, |
|
"model.language_model.layers.13.self_attn.k_norm": false, |
|
"model.language_model.layers.13.self_attn.v_norm": false, |
|
"model.language_model.layers.13.self_attn.rope": false, |
|
"model.language_model.layers.13.mlp.gate_proj": true, |
|
"model.language_model.layers.13.mlp.up_proj": true, |
|
"model.language_model.layers.13.mlp.down_proj": true, |
|
"model.language_model.layers.13.input_layernorm": false, |
|
"model.language_model.layers.13.post_attention_layernorm": false, |
|
"model.language_model.layers.13.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.13.post_feedforward_layernorm": false, |
|
"model.language_model.layers.13.altup.correction_coefs": false, |
|
"model.language_model.layers.13.altup.prediction_coefs": false, |
|
"model.language_model.layers.13.altup.modality_router": true, |
|
"model.language_model.layers.13.altup.router_norm": false, |
|
"model.language_model.layers.13.laurel.linear_left": true, |
|
"model.language_model.layers.13.laurel.linear_right": true, |
|
"model.language_model.layers.13.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.13.per_layer_input_gate": true, |
|
"model.language_model.layers.13.per_layer_projection": true, |
|
"model.language_model.layers.13.post_per_layer_input_norm": false, |
|
"model.language_model.layers.14.self_attn.q_proj": true, |
|
"model.language_model.layers.14.self_attn.k_proj": true, |
|
"model.language_model.layers.14.self_attn.v_proj": true, |
|
"model.language_model.layers.14.self_attn.o_proj": true, |
|
"model.language_model.layers.14.self_attn.q_norm": false, |
|
"model.language_model.layers.14.self_attn.k_norm": false, |
|
"model.language_model.layers.14.self_attn.v_norm": false, |
|
"model.language_model.layers.14.self_attn.rope": false, |
|
"model.language_model.layers.14.mlp.gate_proj": true, |
|
"model.language_model.layers.14.mlp.up_proj": true, |
|
"model.language_model.layers.14.mlp.down_proj": true, |
|
"model.language_model.layers.14.input_layernorm": false, |
|
"model.language_model.layers.14.post_attention_layernorm": false, |
|
"model.language_model.layers.14.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.14.post_feedforward_layernorm": false, |
|
"model.language_model.layers.14.altup.correction_coefs": false, |
|
"model.language_model.layers.14.altup.prediction_coefs": false, |
|
"model.language_model.layers.14.altup.modality_router": true, |
|
"model.language_model.layers.14.altup.router_norm": false, |
|
"model.language_model.layers.14.laurel.linear_left": true, |
|
"model.language_model.layers.14.laurel.linear_right": true, |
|
"model.language_model.layers.14.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.14.per_layer_input_gate": true, |
|
"model.language_model.layers.14.per_layer_projection": true, |
|
"model.language_model.layers.14.post_per_layer_input_norm": false, |
|
"model.language_model.layers.15.self_attn.q_proj": true, |
|
"model.language_model.layers.15.self_attn.k_proj": true, |
|
"model.language_model.layers.15.self_attn.v_proj": true, |
|
"model.language_model.layers.15.self_attn.o_proj": true, |
|
"model.language_model.layers.15.self_attn.q_norm": false, |
|
"model.language_model.layers.15.self_attn.k_norm": false, |
|
"model.language_model.layers.15.self_attn.v_norm": false, |
|
"model.language_model.layers.15.self_attn.rope": false, |
|
"model.language_model.layers.15.mlp.gate_proj": true, |
|
"model.language_model.layers.15.mlp.up_proj": true, |
|
"model.language_model.layers.15.mlp.down_proj": true, |
|
"model.language_model.layers.15.input_layernorm": false, |
|
"model.language_model.layers.15.post_attention_layernorm": false, |
|
"model.language_model.layers.15.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.15.post_feedforward_layernorm": false, |
|
"model.language_model.layers.15.altup.correction_coefs": false, |
|
"model.language_model.layers.15.altup.prediction_coefs": false, |
|
"model.language_model.layers.15.altup.modality_router": true, |
|
"model.language_model.layers.15.altup.router_norm": false, |
|
"model.language_model.layers.15.laurel.linear_left": true, |
|
"model.language_model.layers.15.laurel.linear_right": true, |
|
"model.language_model.layers.15.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.15.per_layer_input_gate": true, |
|
"model.language_model.layers.15.per_layer_projection": true, |
|
"model.language_model.layers.15.post_per_layer_input_norm": false, |
|
"model.language_model.layers.16.self_attn.q_proj": true, |
|
"model.language_model.layers.16.self_attn.k_proj": true, |
|
"model.language_model.layers.16.self_attn.v_proj": true, |
|
"model.language_model.layers.16.self_attn.o_proj": true, |
|
"model.language_model.layers.16.self_attn.q_norm": false, |
|
"model.language_model.layers.16.self_attn.k_norm": false, |
|
"model.language_model.layers.16.self_attn.v_norm": false, |
|
"model.language_model.layers.16.self_attn.rope": false, |
|
"model.language_model.layers.16.mlp.gate_proj": true, |
|
"model.language_model.layers.16.mlp.up_proj": true, |
|
"model.language_model.layers.16.mlp.down_proj": true, |
|
"model.language_model.layers.16.input_layernorm": false, |
|
"model.language_model.layers.16.post_attention_layernorm": false, |
|
"model.language_model.layers.16.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.16.post_feedforward_layernorm": false, |
|
"model.language_model.layers.16.altup.correction_coefs": false, |
|
"model.language_model.layers.16.altup.prediction_coefs": false, |
|
"model.language_model.layers.16.altup.modality_router": true, |
|
"model.language_model.layers.16.altup.router_norm": false, |
|
"model.language_model.layers.16.laurel.linear_left": true, |
|
"model.language_model.layers.16.laurel.linear_right": true, |
|
"model.language_model.layers.16.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.16.per_layer_input_gate": true, |
|
"model.language_model.layers.16.per_layer_projection": true, |
|
"model.language_model.layers.16.post_per_layer_input_norm": false, |
|
"model.language_model.layers.17.self_attn.q_proj": true, |
|
"model.language_model.layers.17.self_attn.k_proj": true, |
|
"model.language_model.layers.17.self_attn.v_proj": true, |
|
"model.language_model.layers.17.self_attn.o_proj": true, |
|
"model.language_model.layers.17.self_attn.q_norm": false, |
|
"model.language_model.layers.17.self_attn.k_norm": false, |
|
"model.language_model.layers.17.self_attn.v_norm": false, |
|
"model.language_model.layers.17.self_attn.rope": false, |
|
"model.language_model.layers.17.mlp.gate_proj": true, |
|
"model.language_model.layers.17.mlp.up_proj": true, |
|
"model.language_model.layers.17.mlp.down_proj": true, |
|
"model.language_model.layers.17.input_layernorm": false, |
|
"model.language_model.layers.17.post_attention_layernorm": false, |
|
"model.language_model.layers.17.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.17.post_feedforward_layernorm": false, |
|
"model.language_model.layers.17.altup.correction_coefs": false, |
|
"model.language_model.layers.17.altup.prediction_coefs": false, |
|
"model.language_model.layers.17.altup.modality_router": true, |
|
"model.language_model.layers.17.altup.router_norm": false, |
|
"model.language_model.layers.17.laurel.linear_left": true, |
|
"model.language_model.layers.17.laurel.linear_right": true, |
|
"model.language_model.layers.17.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.17.per_layer_input_gate": true, |
|
"model.language_model.layers.17.per_layer_projection": true, |
|
"model.language_model.layers.17.post_per_layer_input_norm": false, |
|
"model.language_model.layers.18.self_attn.q_proj": true, |
|
"model.language_model.layers.18.self_attn.k_proj": true, |
|
"model.language_model.layers.18.self_attn.v_proj": true, |
|
"model.language_model.layers.18.self_attn.o_proj": true, |
|
"model.language_model.layers.18.self_attn.q_norm": false, |
|
"model.language_model.layers.18.self_attn.k_norm": false, |
|
"model.language_model.layers.18.self_attn.v_norm": false, |
|
"model.language_model.layers.18.self_attn.rope": false, |
|
"model.language_model.layers.18.mlp.gate_proj": true, |
|
"model.language_model.layers.18.mlp.up_proj": true, |
|
"model.language_model.layers.18.mlp.down_proj": true, |
|
"model.language_model.layers.18.input_layernorm": false, |
|
"model.language_model.layers.18.post_attention_layernorm": false, |
|
"model.language_model.layers.18.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.18.post_feedforward_layernorm": false, |
|
"model.language_model.layers.18.altup.correction_coefs": false, |
|
"model.language_model.layers.18.altup.prediction_coefs": false, |
|
"model.language_model.layers.18.altup.modality_router": true, |
|
"model.language_model.layers.18.altup.router_norm": false, |
|
"model.language_model.layers.18.laurel.linear_left": true, |
|
"model.language_model.layers.18.laurel.linear_right": true, |
|
"model.language_model.layers.18.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.18.per_layer_input_gate": true, |
|
"model.language_model.layers.18.per_layer_projection": true, |
|
"model.language_model.layers.18.post_per_layer_input_norm": false, |
|
"model.language_model.layers.19.self_attn.q_proj": true, |
|
"model.language_model.layers.19.self_attn.k_proj": true, |
|
"model.language_model.layers.19.self_attn.v_proj": true, |
|
"model.language_model.layers.19.self_attn.o_proj": true, |
|
"model.language_model.layers.19.self_attn.q_norm": false, |
|
"model.language_model.layers.19.self_attn.k_norm": false, |
|
"model.language_model.layers.19.self_attn.v_norm": false, |
|
"model.language_model.layers.19.self_attn.rope": false, |
|
"model.language_model.layers.19.mlp.gate_proj": true, |
|
"model.language_model.layers.19.mlp.up_proj": true, |
|
"model.language_model.layers.19.mlp.down_proj": true, |
|
"model.language_model.layers.19.input_layernorm": false, |
|
"model.language_model.layers.19.post_attention_layernorm": false, |
|
"model.language_model.layers.19.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.19.post_feedforward_layernorm": false, |
|
"model.language_model.layers.19.altup.correction_coefs": false, |
|
"model.language_model.layers.19.altup.prediction_coefs": false, |
|
"model.language_model.layers.19.altup.modality_router": true, |
|
"model.language_model.layers.19.altup.router_norm": false, |
|
"model.language_model.layers.19.laurel.linear_left": true, |
|
"model.language_model.layers.19.laurel.linear_right": true, |
|
"model.language_model.layers.19.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.19.per_layer_input_gate": true, |
|
"model.language_model.layers.19.per_layer_projection": true, |
|
"model.language_model.layers.19.post_per_layer_input_norm": false, |
|
"model.language_model.layers.20.self_attn.q_proj": true, |
|
"model.language_model.layers.20.self_attn.k_proj": true, |
|
"model.language_model.layers.20.self_attn.v_proj": true, |
|
"model.language_model.layers.20.self_attn.o_proj": true, |
|
"model.language_model.layers.20.self_attn.q_norm": false, |
|
"model.language_model.layers.20.self_attn.k_norm": false, |
|
"model.language_model.layers.20.self_attn.v_norm": false, |
|
"model.language_model.layers.20.self_attn.rope": false, |
|
"model.language_model.layers.20.mlp.gate_proj": true, |
|
"model.language_model.layers.20.mlp.up_proj": true, |
|
"model.language_model.layers.20.mlp.down_proj": true, |
|
"model.language_model.layers.20.input_layernorm": false, |
|
"model.language_model.layers.20.post_attention_layernorm": false, |
|
"model.language_model.layers.20.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.20.post_feedforward_layernorm": false, |
|
"model.language_model.layers.20.altup.correction_coefs": false, |
|
"model.language_model.layers.20.altup.prediction_coefs": false, |
|
"model.language_model.layers.20.altup.modality_router": true, |
|
"model.language_model.layers.20.altup.router_norm": false, |
|
"model.language_model.layers.20.laurel.linear_left": true, |
|
"model.language_model.layers.20.laurel.linear_right": true, |
|
"model.language_model.layers.20.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.20.per_layer_input_gate": true, |
|
"model.language_model.layers.20.per_layer_projection": true, |
|
"model.language_model.layers.20.post_per_layer_input_norm": false, |
|
"model.language_model.layers.21.self_attn.q_proj": true, |
|
"model.language_model.layers.21.self_attn.k_proj": true, |
|
"model.language_model.layers.21.self_attn.v_proj": true, |
|
"model.language_model.layers.21.self_attn.o_proj": true, |
|
"model.language_model.layers.21.self_attn.q_norm": false, |
|
"model.language_model.layers.21.self_attn.k_norm": false, |
|
"model.language_model.layers.21.self_attn.v_norm": false, |
|
"model.language_model.layers.21.self_attn.rope": false, |
|
"model.language_model.layers.21.mlp.gate_proj": true, |
|
"model.language_model.layers.21.mlp.up_proj": true, |
|
"model.language_model.layers.21.mlp.down_proj": true, |
|
"model.language_model.layers.21.input_layernorm": false, |
|
"model.language_model.layers.21.post_attention_layernorm": false, |
|
"model.language_model.layers.21.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.21.post_feedforward_layernorm": false, |
|
"model.language_model.layers.21.altup.correction_coefs": false, |
|
"model.language_model.layers.21.altup.prediction_coefs": false, |
|
"model.language_model.layers.21.altup.modality_router": true, |
|
"model.language_model.layers.21.altup.router_norm": false, |
|
"model.language_model.layers.21.laurel.linear_left": true, |
|
"model.language_model.layers.21.laurel.linear_right": true, |
|
"model.language_model.layers.21.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.21.per_layer_input_gate": true, |
|
"model.language_model.layers.21.per_layer_projection": true, |
|
"model.language_model.layers.21.post_per_layer_input_norm": false, |
|
"model.language_model.layers.22.self_attn.q_proj": true, |
|
"model.language_model.layers.22.self_attn.k_proj": true, |
|
"model.language_model.layers.22.self_attn.v_proj": true, |
|
"model.language_model.layers.22.self_attn.o_proj": true, |
|
"model.language_model.layers.22.self_attn.q_norm": false, |
|
"model.language_model.layers.22.self_attn.k_norm": false, |
|
"model.language_model.layers.22.self_attn.v_norm": false, |
|
"model.language_model.layers.22.self_attn.rope": false, |
|
"model.language_model.layers.22.mlp.gate_proj": true, |
|
"model.language_model.layers.22.mlp.up_proj": true, |
|
"model.language_model.layers.22.mlp.down_proj": true, |
|
"model.language_model.layers.22.input_layernorm": false, |
|
"model.language_model.layers.22.post_attention_layernorm": false, |
|
"model.language_model.layers.22.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.22.post_feedforward_layernorm": false, |
|
"model.language_model.layers.22.altup.correction_coefs": false, |
|
"model.language_model.layers.22.altup.prediction_coefs": false, |
|
"model.language_model.layers.22.altup.modality_router": true, |
|
"model.language_model.layers.22.altup.router_norm": false, |
|
"model.language_model.layers.22.laurel.linear_left": true, |
|
"model.language_model.layers.22.laurel.linear_right": true, |
|
"model.language_model.layers.22.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.22.per_layer_input_gate": true, |
|
"model.language_model.layers.22.per_layer_projection": true, |
|
"model.language_model.layers.22.post_per_layer_input_norm": false, |
|
"model.language_model.layers.23.self_attn.q_proj": true, |
|
"model.language_model.layers.23.self_attn.k_proj": true, |
|
"model.language_model.layers.23.self_attn.v_proj": true, |
|
"model.language_model.layers.23.self_attn.o_proj": true, |
|
"model.language_model.layers.23.self_attn.q_norm": false, |
|
"model.language_model.layers.23.self_attn.k_norm": false, |
|
"model.language_model.layers.23.self_attn.v_norm": false, |
|
"model.language_model.layers.23.self_attn.rope": false, |
|
"model.language_model.layers.23.mlp.gate_proj": true, |
|
"model.language_model.layers.23.mlp.up_proj": true, |
|
"model.language_model.layers.23.mlp.down_proj": true, |
|
"model.language_model.layers.23.input_layernorm": false, |
|
"model.language_model.layers.23.post_attention_layernorm": false, |
|
"model.language_model.layers.23.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.23.post_feedforward_layernorm": false, |
|
"model.language_model.layers.23.altup.correction_coefs": false, |
|
"model.language_model.layers.23.altup.prediction_coefs": false, |
|
"model.language_model.layers.23.altup.modality_router": true, |
|
"model.language_model.layers.23.altup.router_norm": false, |
|
"model.language_model.layers.23.laurel.linear_left": true, |
|
"model.language_model.layers.23.laurel.linear_right": true, |
|
"model.language_model.layers.23.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.23.per_layer_input_gate": true, |
|
"model.language_model.layers.23.per_layer_projection": true, |
|
"model.language_model.layers.23.post_per_layer_input_norm": false, |
|
"model.language_model.layers.24.self_attn.q_proj": true, |
|
"model.language_model.layers.24.self_attn.k_proj": true, |
|
"model.language_model.layers.24.self_attn.v_proj": true, |
|
"model.language_model.layers.24.self_attn.o_proj": true, |
|
"model.language_model.layers.24.self_attn.q_norm": false, |
|
"model.language_model.layers.24.self_attn.k_norm": false, |
|
"model.language_model.layers.24.self_attn.v_norm": false, |
|
"model.language_model.layers.24.self_attn.rope": false, |
|
"model.language_model.layers.24.mlp.gate_proj": true, |
|
"model.language_model.layers.24.mlp.up_proj": true, |
|
"model.language_model.layers.24.mlp.down_proj": true, |
|
"model.language_model.layers.24.input_layernorm": false, |
|
"model.language_model.layers.24.post_attention_layernorm": false, |
|
"model.language_model.layers.24.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.24.post_feedforward_layernorm": false, |
|
"model.language_model.layers.24.altup.correction_coefs": false, |
|
"model.language_model.layers.24.altup.prediction_coefs": false, |
|
"model.language_model.layers.24.altup.modality_router": true, |
|
"model.language_model.layers.24.altup.router_norm": false, |
|
"model.language_model.layers.24.laurel.linear_left": true, |
|
"model.language_model.layers.24.laurel.linear_right": true, |
|
"model.language_model.layers.24.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.24.per_layer_input_gate": true, |
|
"model.language_model.layers.24.per_layer_projection": true, |
|
"model.language_model.layers.24.post_per_layer_input_norm": false, |
|
"model.language_model.layers.25.self_attn.q_proj": true, |
|
"model.language_model.layers.25.self_attn.k_proj": true, |
|
"model.language_model.layers.25.self_attn.v_proj": true, |
|
"model.language_model.layers.25.self_attn.o_proj": true, |
|
"model.language_model.layers.25.self_attn.q_norm": false, |
|
"model.language_model.layers.25.self_attn.k_norm": false, |
|
"model.language_model.layers.25.self_attn.v_norm": false, |
|
"model.language_model.layers.25.self_attn.rope": false, |
|
"model.language_model.layers.25.mlp.gate_proj": true, |
|
"model.language_model.layers.25.mlp.up_proj": true, |
|
"model.language_model.layers.25.mlp.down_proj": true, |
|
"model.language_model.layers.25.input_layernorm": false, |
|
"model.language_model.layers.25.post_attention_layernorm": false, |
|
"model.language_model.layers.25.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.25.post_feedforward_layernorm": false, |
|
"model.language_model.layers.25.altup.correction_coefs": false, |
|
"model.language_model.layers.25.altup.prediction_coefs": false, |
|
"model.language_model.layers.25.altup.modality_router": true, |
|
"model.language_model.layers.25.altup.router_norm": false, |
|
"model.language_model.layers.25.laurel.linear_left": true, |
|
"model.language_model.layers.25.laurel.linear_right": true, |
|
"model.language_model.layers.25.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.25.per_layer_input_gate": true, |
|
"model.language_model.layers.25.per_layer_projection": true, |
|
"model.language_model.layers.25.post_per_layer_input_norm": false, |
|
"model.language_model.layers.26.self_attn.q_proj": true, |
|
"model.language_model.layers.26.self_attn.k_proj": true, |
|
"model.language_model.layers.26.self_attn.v_proj": true, |
|
"model.language_model.layers.26.self_attn.o_proj": true, |
|
"model.language_model.layers.26.self_attn.q_norm": false, |
|
"model.language_model.layers.26.self_attn.k_norm": false, |
|
"model.language_model.layers.26.self_attn.v_norm": false, |
|
"model.language_model.layers.26.self_attn.rope": false, |
|
"model.language_model.layers.26.mlp.gate_proj": true, |
|
"model.language_model.layers.26.mlp.up_proj": true, |
|
"model.language_model.layers.26.mlp.down_proj": true, |
|
"model.language_model.layers.26.input_layernorm": false, |
|
"model.language_model.layers.26.post_attention_layernorm": false, |
|
"model.language_model.layers.26.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.26.post_feedforward_layernorm": false, |
|
"model.language_model.layers.26.altup.correction_coefs": false, |
|
"model.language_model.layers.26.altup.prediction_coefs": false, |
|
"model.language_model.layers.26.altup.modality_router": true, |
|
"model.language_model.layers.26.altup.router_norm": false, |
|
"model.language_model.layers.26.laurel.linear_left": true, |
|
"model.language_model.layers.26.laurel.linear_right": true, |
|
"model.language_model.layers.26.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.26.per_layer_input_gate": true, |
|
"model.language_model.layers.26.per_layer_projection": true, |
|
"model.language_model.layers.26.post_per_layer_input_norm": false, |
|
"model.language_model.layers.27.self_attn.q_proj": true, |
|
"model.language_model.layers.27.self_attn.k_proj": true, |
|
"model.language_model.layers.27.self_attn.v_proj": true, |
|
"model.language_model.layers.27.self_attn.o_proj": true, |
|
"model.language_model.layers.27.self_attn.q_norm": false, |
|
"model.language_model.layers.27.self_attn.k_norm": false, |
|
"model.language_model.layers.27.self_attn.v_norm": false, |
|
"model.language_model.layers.27.self_attn.rope": false, |
|
"model.language_model.layers.27.mlp.gate_proj": true, |
|
"model.language_model.layers.27.mlp.up_proj": true, |
|
"model.language_model.layers.27.mlp.down_proj": true, |
|
"model.language_model.layers.27.input_layernorm": false, |
|
"model.language_model.layers.27.post_attention_layernorm": false, |
|
"model.language_model.layers.27.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.27.post_feedforward_layernorm": false, |
|
"model.language_model.layers.27.altup.correction_coefs": false, |
|
"model.language_model.layers.27.altup.prediction_coefs": false, |
|
"model.language_model.layers.27.altup.modality_router": true, |
|
"model.language_model.layers.27.altup.router_norm": false, |
|
"model.language_model.layers.27.laurel.linear_left": true, |
|
"model.language_model.layers.27.laurel.linear_right": true, |
|
"model.language_model.layers.27.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.27.per_layer_input_gate": true, |
|
"model.language_model.layers.27.per_layer_projection": true, |
|
"model.language_model.layers.27.post_per_layer_input_norm": false, |
|
"model.language_model.layers.28.self_attn.q_proj": true, |
|
"model.language_model.layers.28.self_attn.k_proj": true, |
|
"model.language_model.layers.28.self_attn.v_proj": true, |
|
"model.language_model.layers.28.self_attn.o_proj": true, |
|
"model.language_model.layers.28.self_attn.q_norm": false, |
|
"model.language_model.layers.28.self_attn.k_norm": false, |
|
"model.language_model.layers.28.self_attn.v_norm": false, |
|
"model.language_model.layers.28.self_attn.rope": false, |
|
"model.language_model.layers.28.mlp.gate_proj": true, |
|
"model.language_model.layers.28.mlp.up_proj": true, |
|
"model.language_model.layers.28.mlp.down_proj": true, |
|
"model.language_model.layers.28.input_layernorm": false, |
|
"model.language_model.layers.28.post_attention_layernorm": false, |
|
"model.language_model.layers.28.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.28.post_feedforward_layernorm": false, |
|
"model.language_model.layers.28.altup.correction_coefs": false, |
|
"model.language_model.layers.28.altup.prediction_coefs": false, |
|
"model.language_model.layers.28.altup.modality_router": true, |
|
"model.language_model.layers.28.altup.router_norm": false, |
|
"model.language_model.layers.28.laurel.linear_left": true, |
|
"model.language_model.layers.28.laurel.linear_right": true, |
|
"model.language_model.layers.28.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.28.per_layer_input_gate": true, |
|
"model.language_model.layers.28.per_layer_projection": true, |
|
"model.language_model.layers.28.post_per_layer_input_norm": false, |
|
"model.language_model.layers.29.self_attn.q_proj": true, |
|
"model.language_model.layers.29.self_attn.k_proj": true, |
|
"model.language_model.layers.29.self_attn.v_proj": true, |
|
"model.language_model.layers.29.self_attn.o_proj": true, |
|
"model.language_model.layers.29.self_attn.q_norm": false, |
|
"model.language_model.layers.29.self_attn.k_norm": false, |
|
"model.language_model.layers.29.self_attn.v_norm": false, |
|
"model.language_model.layers.29.self_attn.rope": false, |
|
"model.language_model.layers.29.mlp.gate_proj": true, |
|
"model.language_model.layers.29.mlp.up_proj": true, |
|
"model.language_model.layers.29.mlp.down_proj": true, |
|
"model.language_model.layers.29.input_layernorm": false, |
|
"model.language_model.layers.29.post_attention_layernorm": false, |
|
"model.language_model.layers.29.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.29.post_feedforward_layernorm": false, |
|
"model.language_model.layers.29.altup.correction_coefs": false, |
|
"model.language_model.layers.29.altup.prediction_coefs": false, |
|
"model.language_model.layers.29.altup.modality_router": true, |
|
"model.language_model.layers.29.altup.router_norm": false, |
|
"model.language_model.layers.29.laurel.linear_left": true, |
|
"model.language_model.layers.29.laurel.linear_right": true, |
|
"model.language_model.layers.29.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.29.per_layer_input_gate": true, |
|
"model.language_model.layers.29.per_layer_projection": true, |
|
"model.language_model.layers.29.post_per_layer_input_norm": false, |
|
"model.language_model.layers.30.self_attn.q_proj": true, |
|
"model.language_model.layers.30.self_attn.k_proj": true, |
|
"model.language_model.layers.30.self_attn.v_proj": true, |
|
"model.language_model.layers.30.self_attn.o_proj": true, |
|
"model.language_model.layers.30.self_attn.q_norm": false, |
|
"model.language_model.layers.30.self_attn.k_norm": false, |
|
"model.language_model.layers.30.self_attn.v_norm": false, |
|
"model.language_model.layers.30.self_attn.rope": false, |
|
"model.language_model.layers.30.mlp.gate_proj": true, |
|
"model.language_model.layers.30.mlp.up_proj": true, |
|
"model.language_model.layers.30.mlp.down_proj": true, |
|
"model.language_model.layers.30.input_layernorm": false, |
|
"model.language_model.layers.30.post_attention_layernorm": false, |
|
"model.language_model.layers.30.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.30.post_feedforward_layernorm": false, |
|
"model.language_model.layers.30.altup.correction_coefs": false, |
|
"model.language_model.layers.30.altup.prediction_coefs": false, |
|
"model.language_model.layers.30.altup.modality_router": true, |
|
"model.language_model.layers.30.altup.router_norm": false, |
|
"model.language_model.layers.30.laurel.linear_left": true, |
|
"model.language_model.layers.30.laurel.linear_right": true, |
|
"model.language_model.layers.30.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.30.per_layer_input_gate": true, |
|
"model.language_model.layers.30.per_layer_projection": true, |
|
"model.language_model.layers.30.post_per_layer_input_norm": false, |
|
"model.language_model.layers.31.self_attn.q_proj": true, |
|
"model.language_model.layers.31.self_attn.k_proj": true, |
|
"model.language_model.layers.31.self_attn.v_proj": true, |
|
"model.language_model.layers.31.self_attn.o_proj": true, |
|
"model.language_model.layers.31.self_attn.q_norm": false, |
|
"model.language_model.layers.31.self_attn.k_norm": false, |
|
"model.language_model.layers.31.self_attn.v_norm": false, |
|
"model.language_model.layers.31.self_attn.rope": false, |
|
"model.language_model.layers.31.mlp.gate_proj": true, |
|
"model.language_model.layers.31.mlp.up_proj": true, |
|
"model.language_model.layers.31.mlp.down_proj": true, |
|
"model.language_model.layers.31.input_layernorm": false, |
|
"model.language_model.layers.31.post_attention_layernorm": false, |
|
"model.language_model.layers.31.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.31.post_feedforward_layernorm": false, |
|
"model.language_model.layers.31.altup.correction_coefs": false, |
|
"model.language_model.layers.31.altup.prediction_coefs": false, |
|
"model.language_model.layers.31.altup.modality_router": true, |
|
"model.language_model.layers.31.altup.router_norm": false, |
|
"model.language_model.layers.31.laurel.linear_left": true, |
|
"model.language_model.layers.31.laurel.linear_right": true, |
|
"model.language_model.layers.31.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.31.per_layer_input_gate": true, |
|
"model.language_model.layers.31.per_layer_projection": true, |
|
"model.language_model.layers.31.post_per_layer_input_norm": false, |
|
"model.language_model.layers.32.self_attn.q_proj": true, |
|
"model.language_model.layers.32.self_attn.k_proj": true, |
|
"model.language_model.layers.32.self_attn.v_proj": true, |
|
"model.language_model.layers.32.self_attn.o_proj": true, |
|
"model.language_model.layers.32.self_attn.q_norm": false, |
|
"model.language_model.layers.32.self_attn.k_norm": false, |
|
"model.language_model.layers.32.self_attn.v_norm": false, |
|
"model.language_model.layers.32.self_attn.rope": false, |
|
"model.language_model.layers.32.mlp.gate_proj": true, |
|
"model.language_model.layers.32.mlp.up_proj": true, |
|
"model.language_model.layers.32.mlp.down_proj": true, |
|
"model.language_model.layers.32.input_layernorm": false, |
|
"model.language_model.layers.32.post_attention_layernorm": false, |
|
"model.language_model.layers.32.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.32.post_feedforward_layernorm": false, |
|
"model.language_model.layers.32.altup.correction_coefs": false, |
|
"model.language_model.layers.32.altup.prediction_coefs": false, |
|
"model.language_model.layers.32.altup.modality_router": true, |
|
"model.language_model.layers.32.altup.router_norm": false, |
|
"model.language_model.layers.32.laurel.linear_left": true, |
|
"model.language_model.layers.32.laurel.linear_right": true, |
|
"model.language_model.layers.32.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.32.per_layer_input_gate": true, |
|
"model.language_model.layers.32.per_layer_projection": true, |
|
"model.language_model.layers.32.post_per_layer_input_norm": false, |
|
"model.language_model.layers.33.self_attn.q_proj": true, |
|
"model.language_model.layers.33.self_attn.k_proj": true, |
|
"model.language_model.layers.33.self_attn.v_proj": true, |
|
"model.language_model.layers.33.self_attn.o_proj": true, |
|
"model.language_model.layers.33.self_attn.q_norm": false, |
|
"model.language_model.layers.33.self_attn.k_norm": false, |
|
"model.language_model.layers.33.self_attn.v_norm": false, |
|
"model.language_model.layers.33.self_attn.rope": false, |
|
"model.language_model.layers.33.mlp.gate_proj": true, |
|
"model.language_model.layers.33.mlp.up_proj": true, |
|
"model.language_model.layers.33.mlp.down_proj": true, |
|
"model.language_model.layers.33.input_layernorm": false, |
|
"model.language_model.layers.33.post_attention_layernorm": false, |
|
"model.language_model.layers.33.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.33.post_feedforward_layernorm": false, |
|
"model.language_model.layers.33.altup.correction_coefs": false, |
|
"model.language_model.layers.33.altup.prediction_coefs": false, |
|
"model.language_model.layers.33.altup.modality_router": true, |
|
"model.language_model.layers.33.altup.router_norm": false, |
|
"model.language_model.layers.33.laurel.linear_left": true, |
|
"model.language_model.layers.33.laurel.linear_right": true, |
|
"model.language_model.layers.33.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.33.per_layer_input_gate": true, |
|
"model.language_model.layers.33.per_layer_projection": true, |
|
"model.language_model.layers.33.post_per_layer_input_norm": false, |
|
"model.language_model.layers.34.self_attn.q_proj": true, |
|
"model.language_model.layers.34.self_attn.k_proj": true, |
|
"model.language_model.layers.34.self_attn.v_proj": true, |
|
"model.language_model.layers.34.self_attn.o_proj": true, |
|
"model.language_model.layers.34.self_attn.q_norm": false, |
|
"model.language_model.layers.34.self_attn.k_norm": false, |
|
"model.language_model.layers.34.self_attn.v_norm": false, |
|
"model.language_model.layers.34.self_attn.rope": false, |
|
"model.language_model.layers.34.mlp.gate_proj": true, |
|
"model.language_model.layers.34.mlp.up_proj": true, |
|
"model.language_model.layers.34.mlp.down_proj": true, |
|
"model.language_model.layers.34.input_layernorm": false, |
|
"model.language_model.layers.34.post_attention_layernorm": false, |
|
"model.language_model.layers.34.pre_feedforward_layernorm": false, |
|
"model.language_model.layers.34.post_feedforward_layernorm": false, |
|
"model.language_model.layers.34.altup.correction_coefs": false, |
|
"model.language_model.layers.34.altup.prediction_coefs": false, |
|
"model.language_model.layers.34.altup.modality_router": true, |
|
"model.language_model.layers.34.altup.router_norm": false, |
|
"model.language_model.layers.34.laurel.linear_left": true, |
|
"model.language_model.layers.34.laurel.linear_right": true, |
|
"model.language_model.layers.34.laurel.post_laurel_norm": false, |
|
"model.language_model.layers.34.per_layer_input_gate": true, |
|
"model.language_model.layers.34.per_layer_projection": true, |
|
"model.language_model.layers.34.post_per_layer_input_norm": false, |
|
"model.language_model.embed_tokens_per_layer": true, |
|
"model.language_model.per_layer_model_projection": true, |
|
"model.language_model.per_layer_projection_norm": false, |
|
"model.language_model.altup_projections.0": true, |
|
"model.language_model.altup_projections.1": true, |
|
"model.language_model.altup_projections.2": true, |
|
"model.language_model.altup_unembed_projections.0": true, |
|
"model.language_model.altup_unembed_projections.1": true, |
|
"model.language_model.altup_unembed_projections.2": true, |
|
"model.language_model.norm": false |
|
}, |
|
"text_config": { |
|
"activation_sparsity_pattern": [ |
|
0.95, |
|
0.95, |
|
0.95, |
|
0.95, |
|
0.95, |
|
0.95, |
|
0.95, |
|
0.95, |
|
0.95, |
|
0.95, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0, |
|
0.0 |
|
], |
|
"altup_active_idx": 0, |
|
"altup_coef_clip": 120.0, |
|
"altup_correct_scale": true, |
|
"altup_lr_multiplier": 1.0, |
|
"altup_num_inputs": 4, |
|
"attention_bias": false, |
|
"attention_dropout": 0.0, |
|
"final_logit_softcapping": 30.0, |
|
"head_dim": 256, |
|
"hidden_activation": "gelu_pytorch_tanh", |
|
"hidden_size": 2048, |
|
"hidden_size_per_layer_input": 256, |
|
"initializer_range": 0.02, |
|
"intermediate_size": 16384, |
|
"laurel_rank": 64, |
|
"layer_types": [ |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"full_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"full_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"full_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"full_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"full_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"full_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"sliding_attention", |
|
"full_attention" |
|
], |
|
"max_position_embeddings": 32768, |
|
"model_type": "gemma3n_text", |
|
"num_attention_heads": 8, |
|
"num_hidden_layers": 35, |
|
"num_key_value_heads": 2, |
|
"num_kv_shared_layers": 15, |
|
"query_pre_attn_scalar": 256, |
|
"rms_norm_eps": 1e-06, |
|
"rope_local_base_freq": 10000.0, |
|
"rope_scaling": null, |
|
"rope_theta": 1000000.0, |
|
"sliding_window": 512, |
|
"torch_dtype": "bfloat16", |
|
"use_cache": true, |
|
"vocab_size": 262400, |
|
"vocab_size_per_layer_input": 262144 |
|
}, |
|
"torch_dtype": "bfloat16", |
|
"transformers_version": "4.53.0.dev0", |
|
"vision_soft_tokens_per_image": 256 |
|
} |