{ "architectures": [ "LagunaForCausalLM" ], "auto_map": { "AutoConfig": "configuration_laguna.LagunaConfig", "AutoModelForCausalLM": "modeling_laguna.LagunaForCausalLM" }, "model_type": "laguna", "vocab_size": 100352, "hidden_size": 4096, "intermediate_size": 16384, "num_hidden_layers": 70, "num_attention_heads": 64, "num_key_value_heads": 8, "head_dim": 128, "max_position_embeddings": 262144, "attention_bias": false, "attention_dropout": 0.0, "rms_norm_eps": 1e-06, "num_experts": 256, "num_experts_per_tok": 16, "moe_intermediate_size": 1024, "shared_expert_intermediate_size": 1024, "router_aux_loss_coef": 0.0, "bos_token_id": 2, "eos_token_id": [ 2, 24 ], "pad_token_id": 9, "tie_word_embeddings": false, "use_cache": true, "torch_dtype": "bfloat16", "gating": "per-element", "rope_parameters": { "full_attention": { "rope_theta": 500000.0, "rope_type": "yarn", "factor": 64.0, "original_max_position_embeddings": 4096, "beta_slow": 1.0, "beta_fast": 64.0, "attention_factor": 1.0, "partial_rotary_factor": 1.0 } }, "moe_apply_router_weight_on_input": false, "partial_rotary_factor": 1.0, "mlp_layer_types": [ "dense", "dense", "dense", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse", "sparse" ], "use_bidirectional_attention": false, "moe_routed_scaling_factor": 1.0, "compression_config": { "config_groups": { "group_0": { "format": "float-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": 128, "num_bits": 8, "observer": null, "observer_kwargs": {}, "strategy": "group", "symmetric": true, "type": "float" }, "output_activations": null, "targets": [ "Linear" ], "weights": { "actorder": null, "block_structure": [ 128, 128 ], "dynamic": false, "group_size": null, "num_bits": 8, "observer": "minmax", "observer_kwargs": {}, "strategy": "block", "symmetric": true, "type": "float" } } }, "format": "float-quantized", "global_compression_ratio": null, "ignore": [ "model.layers.0.self_attn.o_proj", "model.layers.1.self_attn.o_proj", "model.layers.2.self_attn.o_proj", "model.layers.3.self_attn.o_proj", "model.layers.3.mlp.gate", "model.layers.4.self_attn.o_proj", "model.layers.4.mlp.gate", "model.layers.5.self_attn.o_proj", "model.layers.5.mlp.gate", "model.layers.6.self_attn.o_proj", "model.layers.6.mlp.gate", "model.layers.7.self_attn.o_proj", "model.layers.7.mlp.gate", "model.layers.8.self_attn.o_proj", "model.layers.8.mlp.gate", "model.layers.9.self_attn.o_proj", "model.layers.9.mlp.gate", "model.layers.10.self_attn.o_proj", "model.layers.10.mlp.gate", "model.layers.11.self_attn.o_proj", "model.layers.11.mlp.gate", "model.layers.12.self_attn.o_proj", "model.layers.12.mlp.gate", "model.layers.13.self_attn.o_proj", "model.layers.13.mlp.gate", "model.layers.14.self_attn.o_proj", "model.layers.14.mlp.gate", "model.layers.15.self_attn.o_proj", "model.layers.15.mlp.gate", "model.layers.16.self_attn.o_proj", "model.layers.16.mlp.gate", "model.layers.17.self_attn.o_proj", "model.layers.17.mlp.gate", "model.layers.18.self_attn.o_proj", "model.layers.18.mlp.gate", "model.layers.19.self_attn.o_proj", "model.layers.19.mlp.gate", "model.layers.20.self_attn.o_proj", "model.layers.20.mlp.gate", "model.layers.21.self_attn.o_proj", "model.layers.21.mlp.gate", "model.layers.22.self_attn.o_proj", "model.layers.22.mlp.gate", "model.layers.23.self_attn.o_proj", "model.layers.23.mlp.gate", "model.layers.24.self_attn.o_proj", "model.layers.24.mlp.gate", "model.layers.25.self_attn.o_proj", "model.layers.25.mlp.gate", "model.layers.26.self_attn.o_proj", "model.layers.26.mlp.gate", "model.layers.27.self_attn.o_proj", "model.layers.27.mlp.gate", "model.layers.28.self_attn.o_proj", "model.layers.28.mlp.gate", "model.layers.29.self_attn.o_proj", "model.layers.29.mlp.gate", "model.layers.30.self_attn.o_proj", "model.layers.30.mlp.gate", "model.layers.31.self_attn.o_proj", "model.layers.31.mlp.gate", "model.layers.32.self_attn.o_proj", "model.layers.32.mlp.gate", "model.layers.33.self_attn.o_proj", "model.layers.33.mlp.gate", "model.layers.34.self_attn.o_proj", "model.layers.34.mlp.gate", "model.layers.35.self_attn.o_proj", "model.layers.35.mlp.gate", "model.layers.36.self_attn.o_proj", "model.layers.36.mlp.gate", "model.layers.37.self_attn.o_proj", "model.layers.37.mlp.gate", "model.layers.38.self_attn.o_proj", "model.layers.38.mlp.gate", "model.layers.39.self_attn.o_proj", "model.layers.39.mlp.gate", "model.layers.40.self_attn.o_proj", "model.layers.40.mlp.gate", "model.layers.41.self_attn.o_proj", "model.layers.41.mlp.gate", "model.layers.42.self_attn.o_proj", "model.layers.42.mlp.gate", "model.layers.43.self_attn.o_proj", "model.layers.43.mlp.gate", "model.layers.44.self_attn.o_proj", "model.layers.44.mlp.gate", "model.layers.45.self_attn.o_proj", "model.layers.45.mlp.gate", "model.layers.46.self_attn.o_proj", "model.layers.46.mlp.gate", "model.layers.47.self_attn.o_proj", "model.layers.47.mlp.gate", "model.layers.48.self_attn.o_proj", "model.layers.48.mlp.gate", "model.layers.49.self_attn.o_proj", "model.layers.49.mlp.gate", "model.layers.50.self_attn.o_proj", "model.layers.50.mlp.gate", "model.layers.51.self_attn.o_proj", "model.layers.51.mlp.gate", "model.layers.52.self_attn.o_proj", "model.layers.52.mlp.gate", "model.layers.53.self_attn.o_proj", "model.layers.53.mlp.gate", "model.layers.54.self_attn.o_proj", "model.layers.54.mlp.gate", "model.layers.55.self_attn.o_proj", "model.layers.55.mlp.gate", "model.layers.56.self_attn.o_proj", "model.layers.56.mlp.gate", "model.layers.57.self_attn.o_proj", "model.layers.57.mlp.gate", "model.layers.58.self_attn.o_proj", "model.layers.58.mlp.gate", "model.layers.59.self_attn.o_proj", "model.layers.59.mlp.gate", "model.layers.60.self_attn.o_proj", "model.layers.60.mlp.gate", "model.layers.61.self_attn.o_proj", "model.layers.61.mlp.gate", "model.layers.62.self_attn.o_proj", "model.layers.62.mlp.gate", "model.layers.63.self_attn.o_proj", "model.layers.63.mlp.gate", "model.layers.64.self_attn.o_proj", "model.layers.64.mlp.gate", "model.layers.65.self_attn.o_proj", "model.layers.65.mlp.gate", "model.layers.66.self_attn.o_proj", "model.layers.66.mlp.gate", "model.layers.67.self_attn.o_proj", "model.layers.67.mlp.gate", "model.layers.68.self_attn.o_proj", "model.layers.68.mlp.gate", "model.layers.69.self_attn.o_proj", "model.layers.69.mlp.gate", "lm_head" ], "kv_cache_scheme": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 8, "observer": "minmax", "observer_kwargs": {}, "strategy": "tensor", "symmetric": true, "type": "float" }, "quant_method": "compressed-tensors", "quantization_status": "compressed", "sparsity_config": {}, "transform_config": {}, "version": "0.11.0" }, "quantization_config": { "config_groups": { "group_0": { "format": "float-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": 128, "num_bits": 8, "observer": null, "observer_kwargs": {}, "strategy": "group", "symmetric": true, "type": "float" }, "output_activations": null, "targets": [ "Linear" ], "weights": { "actorder": null, "block_structure": [ 128, 128 ], "dynamic": false, "group_size": null, "num_bits": 8, "observer": "minmax", "observer_kwargs": {}, "strategy": "block", "symmetric": true, "type": "float" } } }, "format": "float-quantized", "global_compression_ratio": null, "ignore": [ "model.layers.0.self_attn.o_proj", "model.layers.1.self_attn.o_proj", "model.layers.2.self_attn.o_proj", "model.layers.3.self_attn.o_proj", "model.layers.3.mlp.gate", "model.layers.4.self_attn.o_proj", "model.layers.4.mlp.gate", "model.layers.5.self_attn.o_proj", "model.layers.5.mlp.gate", "model.layers.6.self_attn.o_proj", "model.layers.6.mlp.gate", "model.layers.7.self_attn.o_proj", "model.layers.7.mlp.gate", "model.layers.8.self_attn.o_proj", "model.layers.8.mlp.gate", "model.layers.9.self_attn.o_proj", "model.layers.9.mlp.gate", "model.layers.10.self_attn.o_proj", "model.layers.10.mlp.gate", "model.layers.11.self_attn.o_proj", "model.layers.11.mlp.gate", "model.layers.12.self_attn.o_proj", "model.layers.12.mlp.gate", "model.layers.13.self_attn.o_proj", "model.layers.13.mlp.gate", "model.layers.14.self_attn.o_proj", "model.layers.14.mlp.gate", "model.layers.15.self_attn.o_proj", "model.layers.15.mlp.gate", "model.layers.16.self_attn.o_proj", "model.layers.16.mlp.gate", "model.layers.17.self_attn.o_proj", "model.layers.17.mlp.gate", "model.layers.18.self_attn.o_proj", "model.layers.18.mlp.gate", "model.layers.19.self_attn.o_proj", "model.layers.19.mlp.gate", "model.layers.20.self_attn.o_proj", "model.layers.20.mlp.gate", "model.layers.21.self_attn.o_proj", "model.layers.21.mlp.gate", "model.layers.22.self_attn.o_proj", "model.layers.22.mlp.gate", "model.layers.23.self_attn.o_proj", "model.layers.23.mlp.gate", "model.layers.24.self_attn.o_proj", "model.layers.24.mlp.gate", "model.layers.25.self_attn.o_proj", "model.layers.25.mlp.gate", "model.layers.26.self_attn.o_proj", "model.layers.26.mlp.gate", "model.layers.27.self_attn.o_proj", "model.layers.27.mlp.gate", "model.layers.28.self_attn.o_proj", "model.layers.28.mlp.gate", "model.layers.29.self_attn.o_proj", "model.layers.29.mlp.gate", "model.layers.30.self_attn.o_proj", "model.layers.30.mlp.gate", "model.layers.31.self_attn.o_proj", "model.layers.31.mlp.gate", "model.layers.32.self_attn.o_proj", "model.layers.32.mlp.gate", "model.layers.33.self_attn.o_proj", "model.layers.33.mlp.gate", "model.layers.34.self_attn.o_proj", "model.layers.34.mlp.gate", "model.layers.35.self_attn.o_proj", "model.layers.35.mlp.gate", "model.layers.36.self_attn.o_proj", "model.layers.36.mlp.gate", "model.layers.37.self_attn.o_proj", "model.layers.37.mlp.gate", "model.layers.38.self_attn.o_proj", "model.layers.38.mlp.gate", "model.layers.39.self_attn.o_proj", "model.layers.39.mlp.gate", "model.layers.40.self_attn.o_proj", "model.layers.40.mlp.gate", "model.layers.41.self_attn.o_proj", "model.layers.41.mlp.gate", "model.layers.42.self_attn.o_proj", "model.layers.42.mlp.gate", "model.layers.43.self_attn.o_proj", "model.layers.43.mlp.gate", "model.layers.44.self_attn.o_proj", "model.layers.44.mlp.gate", "model.layers.45.self_attn.o_proj", "model.layers.45.mlp.gate", "model.layers.46.self_attn.o_proj", "model.layers.46.mlp.gate", "model.layers.47.self_attn.o_proj", "model.layers.47.mlp.gate", "model.layers.48.self_attn.o_proj", "model.layers.48.mlp.gate", "model.layers.49.self_attn.o_proj", "model.layers.49.mlp.gate", "model.layers.50.self_attn.o_proj", "model.layers.50.mlp.gate", "model.layers.51.self_attn.o_proj", "model.layers.51.mlp.gate", "model.layers.52.self_attn.o_proj", "model.layers.52.mlp.gate", "model.layers.53.self_attn.o_proj", "model.layers.53.mlp.gate", "model.layers.54.self_attn.o_proj", "model.layers.54.mlp.gate", "model.layers.55.self_attn.o_proj", "model.layers.55.mlp.gate", "model.layers.56.self_attn.o_proj", "model.layers.56.mlp.gate", "model.layers.57.self_attn.o_proj", "model.layers.57.mlp.gate", "model.layers.58.self_attn.o_proj", "model.layers.58.mlp.gate", "model.layers.59.self_attn.o_proj", "model.layers.59.mlp.gate", "model.layers.60.self_attn.o_proj", "model.layers.60.mlp.gate", "model.layers.61.self_attn.o_proj", "model.layers.61.mlp.gate", "model.layers.62.self_attn.o_proj", "model.layers.62.mlp.gate", "model.layers.63.self_attn.o_proj", "model.layers.63.mlp.gate", "model.layers.64.self_attn.o_proj", "model.layers.64.mlp.gate", "model.layers.65.self_attn.o_proj", "model.layers.65.mlp.gate", "model.layers.66.self_attn.o_proj", "model.layers.66.mlp.gate", "model.layers.67.self_attn.o_proj", "model.layers.67.mlp.gate", "model.layers.68.self_attn.o_proj", "model.layers.68.mlp.gate", "model.layers.69.self_attn.o_proj", "model.layers.69.mlp.gate", "lm_head" ], "kv_cache_scheme": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 8, "observer": "minmax", "observer_kwargs": {}, "strategy": "tensor", "symmetric": true, "type": "float" }, "quant_method": "compressed-tensors", "quantization_status": "compressed", "sparsity_config": {}, "transform_config": {}, "version": "0.11.0" }, "norm_topk_prob": true, "mlp_only_layers": [ 0, 1, 2 ], "decoder_sparse_step": 1, "sliding_window": 0 }