{ "_name_or_path": "/home/eldar/.cache/huggingface/hub/models--mistralai--Mixtral-8x7B-v0.1/snapshots/ffe1a706bacbd5abddc5ff99432ee38f7e0662fb", "architectures": [ "MixtralForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 1, "eos_token_id": 2, "hidden_act": "silu", "hidden_size": 4096, "initializer_range": 0.02, "intermediate_size": 14336, "max_position_embeddings": 32768, "model_type": "mixtral", "num_attention_heads": 32, "num_experts_per_tok": 2, "num_hidden_layers": 32, "num_key_value_heads": 8, "num_local_experts": 8, "output_router_logits": false, "quantization_config": { "config_groups": { "group_0": { "input_activations": null, "output_activations": null, "targets": [ "Linear" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": 128, "num_bits": 4, "observer": "mse", "observer_kwargs": {}, "strategy": "group", "symmetric": true, "type": "int" } } }, "format": "pack-quantized", "global_compression_ratio": 2.0968847636762966, "ignore": [ "model.layers.0.block_sparse_moe.gate", "model.layers.1.block_sparse_moe.gate", "model.layers.2.block_sparse_moe.gate", "model.layers.3.block_sparse_moe.gate", "model.layers.4.block_sparse_moe.gate", "model.layers.5.block_sparse_moe.gate", "model.layers.6.block_sparse_moe.gate", "model.layers.7.block_sparse_moe.gate", "model.layers.8.block_sparse_moe.gate", "model.layers.9.block_sparse_moe.gate", "model.layers.10.block_sparse_moe.gate", "model.layers.11.block_sparse_moe.gate", "model.layers.12.block_sparse_moe.gate", "model.layers.13.block_sparse_moe.gate", "model.layers.14.block_sparse_moe.gate", "model.layers.15.block_sparse_moe.gate", "model.layers.16.block_sparse_moe.gate", "model.layers.17.block_sparse_moe.gate", "model.layers.18.block_sparse_moe.gate", "model.layers.19.block_sparse_moe.gate", "model.layers.20.block_sparse_moe.gate", "model.layers.21.block_sparse_moe.gate", "model.layers.22.block_sparse_moe.gate", "model.layers.23.block_sparse_moe.gate", "model.layers.24.block_sparse_moe.gate", "model.layers.25.block_sparse_moe.gate", "model.layers.26.block_sparse_moe.gate", "model.layers.27.block_sparse_moe.gate", "model.layers.28.block_sparse_moe.gate", "model.layers.29.block_sparse_moe.gate", "model.layers.30.block_sparse_moe.gate", "model.layers.31.block_sparse_moe.gate", "lm_head" ], "kv_cache_scheme": null, "quant_method": "compressed-tensors", "quantization_status": "compressed", "sparsity_config": { "format": "dense", "global_sparsity": 0.14580331375755926, "ignore": [ "model.layers.0.self_attn.v_proj", "model.layers.0.self_attn.o_proj", "model.layers.0.block_sparse_moe.gate", "model.layers.0.block_sparse_moe.experts.0.w1", "model.layers.0.block_sparse_moe.experts.0.w2", "model.layers.0.block_sparse_moe.experts.0.w3", "model.layers.0.block_sparse_moe.experts.1.w1", "model.layers.0.block_sparse_moe.experts.1.w2", "model.layers.0.block_sparse_moe.experts.1.w3", "model.layers.0.block_sparse_moe.experts.2.w1", "model.layers.0.block_sparse_moe.experts.2.w2", "model.layers.0.block_sparse_moe.experts.2.w3", "model.layers.0.block_sparse_moe.experts.3.w1", "model.layers.0.block_sparse_moe.experts.3.w2", "model.layers.0.block_sparse_moe.experts.3.w3", "model.layers.0.block_sparse_moe.experts.4.w1", "model.layers.0.block_sparse_moe.experts.4.w2", "model.layers.0.block_sparse_moe.experts.4.w3", "model.layers.0.block_sparse_moe.experts.5.w1", "model.layers.0.block_sparse_moe.experts.5.w2", "model.layers.0.block_sparse_moe.experts.5.w3", "model.layers.0.block_sparse_moe.experts.6.w1", "model.layers.0.block_sparse_moe.experts.6.w2", "model.layers.0.block_sparse_moe.experts.6.w3", "model.layers.0.block_sparse_moe.experts.7.w1", "model.layers.0.block_sparse_moe.experts.7.w2", "model.layers.0.block_sparse_moe.experts.7.w3", "model.layers.1.self_attn.q_proj", "model.layers.1.self_attn.k_proj", "model.layers.1.self_attn.v_proj", "model.layers.1.self_attn.o_proj", "model.layers.1.block_sparse_moe.gate", "model.layers.1.block_sparse_moe.experts.0.w1", "model.layers.1.block_sparse_moe.experts.0.w2", "model.layers.1.block_sparse_moe.experts.0.w3", "model.layers.1.block_sparse_moe.experts.1.w1", "model.layers.1.block_sparse_moe.experts.1.w2", "model.layers.1.block_sparse_moe.experts.1.w3", "model.layers.1.block_sparse_moe.experts.2.w1", "model.layers.1.block_sparse_moe.experts.2.w2", "model.layers.1.block_sparse_moe.experts.2.w3", "model.layers.1.block_sparse_moe.experts.3.w1", "model.layers.1.block_sparse_moe.experts.3.w2", "model.layers.1.block_sparse_moe.experts.3.w3", "model.layers.1.block_sparse_moe.experts.4.w1", "model.layers.1.block_sparse_moe.experts.4.w2", "model.layers.1.block_sparse_moe.experts.4.w3", "model.layers.1.block_sparse_moe.experts.5.w1", "model.layers.1.block_sparse_moe.experts.5.w2", "model.layers.1.block_sparse_moe.experts.5.w3", "model.layers.1.block_sparse_moe.experts.6.w1", "model.layers.1.block_sparse_moe.experts.6.w2", "model.layers.1.block_sparse_moe.experts.6.w3", "model.layers.1.block_sparse_moe.experts.7.w1", "model.layers.1.block_sparse_moe.experts.7.w2", "model.layers.1.block_sparse_moe.experts.7.w3", "model.layers.2.self_attn.q_proj", "model.layers.2.self_attn.k_proj", "model.layers.2.self_attn.v_proj", "model.layers.2.self_attn.o_proj", "model.layers.2.block_sparse_moe.gate", "model.layers.2.block_sparse_moe.experts.0.w1", "model.layers.2.block_sparse_moe.experts.0.w2", "model.layers.2.block_sparse_moe.experts.0.w3", "model.layers.2.block_sparse_moe.experts.1.w1", "model.layers.2.block_sparse_moe.experts.1.w2", "model.layers.2.block_sparse_moe.experts.1.w3", "model.layers.2.block_sparse_moe.experts.2.w1", "model.layers.2.block_sparse_moe.experts.2.w2", "model.layers.2.block_sparse_moe.experts.2.w3", "model.layers.2.block_sparse_moe.experts.3.w1", "model.layers.2.block_sparse_moe.experts.3.w2", "model.layers.2.block_sparse_moe.experts.3.w3", "model.layers.2.block_sparse_moe.experts.4.w1", "model.layers.2.block_sparse_moe.experts.4.w2", "model.layers.2.block_sparse_moe.experts.4.w3", "model.layers.2.block_sparse_moe.experts.5.w1", "model.layers.2.block_sparse_moe.experts.5.w2", "model.layers.2.block_sparse_moe.experts.5.w3", "model.layers.2.block_sparse_moe.experts.6.w1", "model.layers.2.block_sparse_moe.experts.6.w2", "model.layers.2.block_sparse_moe.experts.6.w3", "model.layers.2.block_sparse_moe.experts.7.w1", "model.layers.2.block_sparse_moe.experts.7.w2", "model.layers.2.block_sparse_moe.experts.7.w3", "model.layers.3.self_attn.q_proj", "model.layers.3.self_attn.k_proj", "model.layers.3.self_attn.v_proj", "model.layers.3.self_attn.o_proj", "model.layers.3.block_sparse_moe.gate", "model.layers.3.block_sparse_moe.experts.0.w1", "model.layers.3.block_sparse_moe.experts.0.w2", "model.layers.3.block_sparse_moe.experts.0.w3", "model.layers.3.block_sparse_moe.experts.1.w1", "model.layers.3.block_sparse_moe.experts.1.w2", "model.layers.3.block_sparse_moe.experts.1.w3", "model.layers.3.block_sparse_moe.experts.2.w1", "model.layers.3.block_sparse_moe.experts.2.w2", "model.layers.3.block_sparse_moe.experts.2.w3", "model.layers.3.block_sparse_moe.experts.3.w1", "model.layers.3.block_sparse_moe.experts.3.w2", "model.layers.3.block_sparse_moe.experts.3.w3", "model.layers.3.block_sparse_moe.experts.4.w1", "model.layers.3.block_sparse_moe.experts.4.w2", "model.layers.3.block_sparse_moe.experts.4.w3", "model.layers.3.block_sparse_moe.experts.5.w1", "model.layers.3.block_sparse_moe.experts.5.w2", "model.layers.3.block_sparse_moe.experts.5.w3", "model.layers.3.block_sparse_moe.experts.6.w1", "model.layers.3.block_sparse_moe.experts.6.w2", "model.layers.3.block_sparse_moe.experts.6.w3", "model.layers.3.block_sparse_moe.experts.7.w1", "model.layers.3.block_sparse_moe.experts.7.w2", "model.layers.3.block_sparse_moe.experts.7.w3", "model.layers.4.self_attn.q_proj", "model.layers.4.self_attn.k_proj", "model.layers.4.self_attn.v_proj", "model.layers.4.self_attn.o_proj", "model.layers.4.block_sparse_moe.gate", "model.layers.4.block_sparse_moe.experts.0.w1", "model.layers.4.block_sparse_moe.experts.0.w2", "model.layers.4.block_sparse_moe.experts.0.w3", "model.layers.4.block_sparse_moe.experts.1.w1", "model.layers.4.block_sparse_moe.experts.1.w2", "model.layers.4.block_sparse_moe.experts.1.w3", "model.layers.4.block_sparse_moe.experts.2.w1", "model.layers.4.block_sparse_moe.experts.2.w2", "model.layers.4.block_sparse_moe.experts.2.w3", "model.layers.4.block_sparse_moe.experts.3.w1", "model.layers.4.block_sparse_moe.experts.3.w2", "model.layers.4.block_sparse_moe.experts.3.w3", "model.layers.4.block_sparse_moe.experts.4.w1", "model.layers.4.block_sparse_moe.experts.4.w2", "model.layers.4.block_sparse_moe.experts.4.w3", "model.layers.4.block_sparse_moe.experts.5.w1", "model.layers.4.block_sparse_moe.experts.5.w2", "model.layers.4.block_sparse_moe.experts.5.w3", "model.layers.4.block_sparse_moe.experts.6.w1", "model.layers.4.block_sparse_moe.experts.6.w2", "model.layers.4.block_sparse_moe.experts.6.w3", "model.layers.4.block_sparse_moe.experts.7.w1", "model.layers.4.block_sparse_moe.experts.7.w2", "model.layers.4.block_sparse_moe.experts.7.w3", "model.layers.5.self_attn.q_proj", "model.layers.5.self_attn.k_proj", "model.layers.5.self_attn.v_proj", "model.layers.5.self_attn.o_proj", "model.layers.5.block_sparse_moe.gate", "model.layers.5.block_sparse_moe.experts.0.w1", "model.layers.5.block_sparse_moe.experts.0.w2", "model.layers.5.block_sparse_moe.experts.0.w3", "model.layers.5.block_sparse_moe.experts.1.w1", "model.layers.5.block_sparse_moe.experts.1.w2", "model.layers.5.block_sparse_moe.experts.1.w3", "model.layers.5.block_sparse_moe.experts.2.w1", "model.layers.5.block_sparse_moe.experts.2.w2", "model.layers.5.block_sparse_moe.experts.2.w3", "model.layers.5.block_sparse_moe.experts.3.w1", "model.layers.5.block_sparse_moe.experts.3.w2", "model.layers.5.block_sparse_moe.experts.3.w3", "model.layers.5.block_sparse_moe.experts.4.w1", "model.layers.5.block_sparse_moe.experts.4.w2", "model.layers.5.block_sparse_moe.experts.4.w3", "model.layers.5.block_sparse_moe.experts.5.w1", "model.layers.5.block_sparse_moe.experts.5.w2", "model.layers.5.block_sparse_moe.experts.5.w3", "model.layers.5.block_sparse_moe.experts.6.w1", "model.layers.5.block_sparse_moe.experts.6.w2", "model.layers.5.block_sparse_moe.experts.6.w3", "model.layers.5.block_sparse_moe.experts.7.w1", "model.layers.5.block_sparse_moe.experts.7.w2", "model.layers.5.block_sparse_moe.experts.7.w3", "model.layers.6.self_attn.q_proj", "model.layers.6.self_attn.k_proj", "model.layers.6.self_attn.v_proj", "model.layers.6.self_attn.o_proj", "model.layers.6.block_sparse_moe.gate", "model.layers.6.block_sparse_moe.experts.0.w1", "model.layers.6.block_sparse_moe.experts.0.w2", "model.layers.6.block_sparse_moe.experts.0.w3", "model.layers.6.block_sparse_moe.experts.1.w1", "model.layers.6.block_sparse_moe.experts.1.w2", "model.layers.6.block_sparse_moe.experts.1.w3", "model.layers.6.block_sparse_moe.experts.2.w1", "model.layers.6.block_sparse_moe.experts.2.w2", "model.layers.6.block_sparse_moe.experts.2.w3", "model.layers.6.block_sparse_moe.experts.3.w1", "model.layers.6.block_sparse_moe.experts.3.w2", "model.layers.6.block_sparse_moe.experts.3.w3", "model.layers.6.block_sparse_moe.experts.4.w1", "model.layers.6.block_sparse_moe.experts.4.w2", "model.layers.6.block_sparse_moe.experts.4.w3", "model.layers.6.block_sparse_moe.experts.5.w1", "model.layers.6.block_sparse_moe.experts.5.w2", "model.layers.6.block_sparse_moe.experts.5.w3", "model.layers.6.block_sparse_moe.experts.6.w1", "model.layers.6.block_sparse_moe.experts.6.w2", "model.layers.6.block_sparse_moe.experts.6.w3", "model.layers.6.block_sparse_moe.experts.7.w1", "model.layers.6.block_sparse_moe.experts.7.w2", "model.layers.6.block_sparse_moe.experts.7.w3", "model.layers.7.self_attn.q_proj", "model.layers.7.self_attn.k_proj", "model.layers.7.self_attn.v_proj", "model.layers.7.self_attn.o_proj", "model.layers.7.block_sparse_moe.gate", "model.layers.7.block_sparse_moe.experts.0.w1", "model.layers.7.block_sparse_moe.experts.0.w2", "model.layers.7.block_sparse_moe.experts.0.w3", "model.layers.7.block_sparse_moe.experts.1.w1", "model.layers.7.block_sparse_moe.experts.1.w2", "model.layers.7.block_sparse_moe.experts.1.w3", "model.layers.7.block_sparse_moe.experts.2.w1", "model.layers.7.block_sparse_moe.experts.2.w2", "model.layers.7.block_sparse_moe.experts.2.w3", "model.layers.7.block_sparse_moe.experts.3.w1", "model.layers.7.block_sparse_moe.experts.3.w2", "model.layers.7.block_sparse_moe.experts.3.w3", "model.layers.7.block_sparse_moe.experts.4.w1", "model.layers.7.block_sparse_moe.experts.4.w2", "model.layers.7.block_sparse_moe.experts.4.w3", "model.layers.7.block_sparse_moe.experts.5.w1", "model.layers.7.block_sparse_moe.experts.5.w2", "model.layers.7.block_sparse_moe.experts.5.w3", "model.layers.7.block_sparse_moe.experts.6.w1", "model.layers.7.block_sparse_moe.experts.6.w2", "model.layers.7.block_sparse_moe.experts.6.w3", "model.layers.7.block_sparse_moe.experts.7.w1", "model.layers.7.block_sparse_moe.experts.7.w2", "model.layers.7.block_sparse_moe.experts.7.w3", "model.layers.8.self_attn.q_proj", "model.layers.8.self_attn.k_proj", "model.layers.8.self_attn.v_proj", "model.layers.8.self_attn.o_proj", "model.layers.8.block_sparse_moe.gate", "model.layers.8.block_sparse_moe.experts.0.w1", "model.layers.8.block_sparse_moe.experts.0.w2", "model.layers.8.block_sparse_moe.experts.0.w3", "model.layers.8.block_sparse_moe.experts.1.w1", "model.layers.8.block_sparse_moe.experts.1.w2", "model.layers.8.block_sparse_moe.experts.1.w3", "model.layers.8.block_sparse_moe.experts.2.w1", "model.layers.8.block_sparse_moe.experts.2.w2", "model.layers.8.block_sparse_moe.experts.2.w3", "model.layers.8.block_sparse_moe.experts.3.w1", "model.layers.8.block_sparse_moe.experts.3.w2", "model.layers.8.block_sparse_moe.experts.3.w3", "model.layers.8.block_sparse_moe.experts.4.w1", "model.layers.8.block_sparse_moe.experts.4.w2", "model.layers.8.block_sparse_moe.experts.4.w3", "model.layers.8.block_sparse_moe.experts.5.w1", "model.layers.8.block_sparse_moe.experts.5.w2", "model.layers.8.block_sparse_moe.experts.5.w3", "model.layers.8.block_sparse_moe.experts.6.w1", "model.layers.8.block_sparse_moe.experts.6.w2", "model.layers.8.block_sparse_moe.experts.6.w3", "model.layers.8.block_sparse_moe.experts.7.w1", "model.layers.8.block_sparse_moe.experts.7.w2", "model.layers.8.block_sparse_moe.experts.7.w3", "model.layers.9.self_attn.q_proj", "model.layers.9.self_attn.k_proj", "model.layers.9.self_attn.v_proj", "model.layers.9.self_attn.o_proj", "model.layers.9.block_sparse_moe.gate", "model.layers.9.block_sparse_moe.experts.0.w1", "model.layers.9.block_sparse_moe.experts.0.w2", "model.layers.9.block_sparse_moe.experts.0.w3", "model.layers.9.block_sparse_moe.experts.1.w1", "model.layers.9.block_sparse_moe.experts.1.w2", "model.layers.9.block_sparse_moe.experts.1.w3", "model.layers.9.block_sparse_moe.experts.2.w1", "model.layers.9.block_sparse_moe.experts.2.w2", "model.layers.9.block_sparse_moe.experts.2.w3", "model.layers.9.block_sparse_moe.experts.3.w1", "model.layers.9.block_sparse_moe.experts.3.w2", "model.layers.9.block_sparse_moe.experts.3.w3", "model.layers.9.block_sparse_moe.experts.4.w1", "model.layers.9.block_sparse_moe.experts.4.w2", "model.layers.9.block_sparse_moe.experts.4.w3", "model.layers.9.block_sparse_moe.experts.5.w1", "model.layers.9.block_sparse_moe.experts.5.w2", "model.layers.9.block_sparse_moe.experts.5.w3", "model.layers.9.block_sparse_moe.experts.6.w1", "model.layers.9.block_sparse_moe.experts.6.w2", "model.layers.9.block_sparse_moe.experts.6.w3", "model.layers.9.block_sparse_moe.experts.7.w1", "model.layers.9.block_sparse_moe.experts.7.w2", "model.layers.9.block_sparse_moe.experts.7.w3", "model.layers.10.self_attn.q_proj", "model.layers.10.self_attn.k_proj", "model.layers.10.self_attn.v_proj", "model.layers.10.self_attn.o_proj", "model.layers.10.block_sparse_moe.gate", "model.layers.10.block_sparse_moe.experts.0.w1", "model.layers.10.block_sparse_moe.experts.0.w2", "model.layers.10.block_sparse_moe.experts.0.w3", "model.layers.10.block_sparse_moe.experts.1.w1", "model.layers.10.block_sparse_moe.experts.1.w2", "model.layers.10.block_sparse_moe.experts.1.w3", "model.layers.10.block_sparse_moe.experts.2.w1", "model.layers.10.block_sparse_moe.experts.2.w2", "model.layers.10.block_sparse_moe.experts.2.w3", "model.layers.10.block_sparse_moe.experts.3.w1", "model.layers.10.block_sparse_moe.experts.3.w2", "model.layers.10.block_sparse_moe.experts.3.w3", "model.layers.10.block_sparse_moe.experts.4.w1", "model.layers.10.block_sparse_moe.experts.4.w2", "model.layers.10.block_sparse_moe.experts.4.w3", "model.layers.10.block_sparse_moe.experts.5.w1", "model.layers.10.block_sparse_moe.experts.5.w2", "model.layers.10.block_sparse_moe.experts.5.w3", "model.layers.10.block_sparse_moe.experts.6.w1", "model.layers.10.block_sparse_moe.experts.6.w2", "model.layers.10.block_sparse_moe.experts.6.w3", "model.layers.10.block_sparse_moe.experts.7.w1", "model.layers.10.block_sparse_moe.experts.7.w2", "model.layers.10.block_sparse_moe.experts.7.w3", "model.layers.11.self_attn.q_proj", "model.layers.11.self_attn.k_proj", "model.layers.11.self_attn.v_proj", "model.layers.11.self_attn.o_proj", "model.layers.11.block_sparse_moe.gate", "model.layers.11.block_sparse_moe.experts.0.w1", "model.layers.11.block_sparse_moe.experts.0.w2", "model.layers.11.block_sparse_moe.experts.0.w3", "model.layers.11.block_sparse_moe.experts.1.w1", "model.layers.11.block_sparse_moe.experts.1.w2", "model.layers.11.block_sparse_moe.experts.1.w3", "model.layers.11.block_sparse_moe.experts.2.w1", "model.layers.11.block_sparse_moe.experts.2.w2", "model.layers.11.block_sparse_moe.experts.2.w3", "model.layers.11.block_sparse_moe.experts.3.w1", "model.layers.11.block_sparse_moe.experts.3.w2", "model.layers.11.block_sparse_moe.experts.3.w3", "model.layers.11.block_sparse_moe.experts.4.w1", "model.layers.11.block_sparse_moe.experts.4.w2", "model.layers.11.block_sparse_moe.experts.4.w3", "model.layers.11.block_sparse_moe.experts.5.w1", "model.layers.11.block_sparse_moe.experts.5.w2", "model.layers.11.block_sparse_moe.experts.5.w3", "model.layers.11.block_sparse_moe.experts.6.w1", "model.layers.11.block_sparse_moe.experts.6.w2", "model.layers.11.block_sparse_moe.experts.6.w3", "model.layers.11.block_sparse_moe.experts.7.w1", "model.layers.11.block_sparse_moe.experts.7.w2", "model.layers.11.block_sparse_moe.experts.7.w3", "model.layers.12.self_attn.q_proj", "model.layers.12.self_attn.k_proj", "model.layers.12.self_attn.v_proj", "model.layers.12.self_attn.o_proj", "model.layers.12.block_sparse_moe.gate", "model.layers.12.block_sparse_moe.experts.0.w1", "model.layers.12.block_sparse_moe.experts.0.w2", "model.layers.12.block_sparse_moe.experts.0.w3", "model.layers.12.block_sparse_moe.experts.1.w1", "model.layers.12.block_sparse_moe.experts.1.w2", "model.layers.12.block_sparse_moe.experts.1.w3", "model.layers.12.block_sparse_moe.experts.2.w1", "model.layers.12.block_sparse_moe.experts.2.w2", "model.layers.12.block_sparse_moe.experts.2.w3", "model.layers.12.block_sparse_moe.experts.3.w1", "model.layers.12.block_sparse_moe.experts.3.w2", "model.layers.12.block_sparse_moe.experts.3.w3", "model.layers.12.block_sparse_moe.experts.4.w1", "model.layers.12.block_sparse_moe.experts.4.w2", "model.layers.12.block_sparse_moe.experts.4.w3", "model.layers.12.block_sparse_moe.experts.5.w1", "model.layers.12.block_sparse_moe.experts.5.w2", "model.layers.12.block_sparse_moe.experts.5.w3", "model.layers.12.block_sparse_moe.experts.6.w1", "model.layers.12.block_sparse_moe.experts.6.w2", "model.layers.12.block_sparse_moe.experts.6.w3", "model.layers.12.block_sparse_moe.experts.7.w1", "model.layers.12.block_sparse_moe.experts.7.w2", "model.layers.12.block_sparse_moe.experts.7.w3", "model.layers.13.self_attn.q_proj", "model.layers.13.self_attn.k_proj", "model.layers.13.self_attn.v_proj", "model.layers.13.self_attn.o_proj", "model.layers.13.block_sparse_moe.gate", "model.layers.13.block_sparse_moe.experts.0.w1", "model.layers.13.block_sparse_moe.experts.0.w2", "model.layers.13.block_sparse_moe.experts.0.w3", "model.layers.13.block_sparse_moe.experts.1.w1", "model.layers.13.block_sparse_moe.experts.1.w2", "model.layers.13.block_sparse_moe.experts.1.w3", "model.layers.13.block_sparse_moe.experts.2.w1", "model.layers.13.block_sparse_moe.experts.2.w2", "model.layers.13.block_sparse_moe.experts.2.w3", "model.layers.13.block_sparse_moe.experts.3.w1", "model.layers.13.block_sparse_moe.experts.3.w2", "model.layers.13.block_sparse_moe.experts.3.w3", "model.layers.13.block_sparse_moe.experts.4.w1", "model.layers.13.block_sparse_moe.experts.4.w2", "model.layers.13.block_sparse_moe.experts.4.w3", "model.layers.13.block_sparse_moe.experts.5.w1", "model.layers.13.block_sparse_moe.experts.5.w2", "model.layers.13.block_sparse_moe.experts.5.w3", "model.layers.13.block_sparse_moe.experts.6.w1", "model.layers.13.block_sparse_moe.experts.6.w2", "model.layers.13.block_sparse_moe.experts.6.w3", "model.layers.13.block_sparse_moe.experts.7.w1", "model.layers.13.block_sparse_moe.experts.7.w2", "model.layers.13.block_sparse_moe.experts.7.w3", "model.layers.14.self_attn.q_proj", "model.layers.14.self_attn.k_proj", "model.layers.14.self_attn.v_proj", "model.layers.14.self_attn.o_proj", "model.layers.14.block_sparse_moe.gate", "model.layers.14.block_sparse_moe.experts.0.w1", "model.layers.14.block_sparse_moe.experts.0.w2", "model.layers.14.block_sparse_moe.experts.0.w3", "model.layers.14.block_sparse_moe.experts.1.w1", "model.layers.14.block_sparse_moe.experts.1.w2", "model.layers.14.block_sparse_moe.experts.1.w3", "model.layers.14.block_sparse_moe.experts.2.w1", "model.layers.14.block_sparse_moe.experts.2.w2", "model.layers.14.block_sparse_moe.experts.2.w3", "model.layers.14.block_sparse_moe.experts.3.w1", "model.layers.14.block_sparse_moe.experts.3.w2", "model.layers.14.block_sparse_moe.experts.3.w3", "model.layers.14.block_sparse_moe.experts.4.w1", "model.layers.14.block_sparse_moe.experts.4.w2", "model.layers.14.block_sparse_moe.experts.4.w3", "model.layers.14.block_sparse_moe.experts.5.w1", "model.layers.14.block_sparse_moe.experts.5.w2", "model.layers.14.block_sparse_moe.experts.5.w3", "model.layers.14.block_sparse_moe.experts.6.w1", "model.layers.14.block_sparse_moe.experts.6.w2", "model.layers.14.block_sparse_moe.experts.6.w3", "model.layers.14.block_sparse_moe.experts.7.w1", "model.layers.14.block_sparse_moe.experts.7.w2", "model.layers.14.block_sparse_moe.experts.7.w3", "model.layers.15.self_attn.q_proj", "model.layers.15.self_attn.k_proj", "model.layers.15.self_attn.v_proj", "model.layers.15.self_attn.o_proj", "model.layers.15.block_sparse_moe.gate", "model.layers.15.block_sparse_moe.experts.0.w1", "model.layers.15.block_sparse_moe.experts.0.w2", "model.layers.15.block_sparse_moe.experts.0.w3", "model.layers.15.block_sparse_moe.experts.1.w1", "model.layers.15.block_sparse_moe.experts.1.w2", "model.layers.15.block_sparse_moe.experts.1.w3", "model.layers.15.block_sparse_moe.experts.2.w1", "model.layers.15.block_sparse_moe.experts.2.w2", "model.layers.15.block_sparse_moe.experts.2.w3", "model.layers.15.block_sparse_moe.experts.3.w1", "model.layers.15.block_sparse_moe.experts.3.w2", "model.layers.15.block_sparse_moe.experts.3.w3", "model.layers.15.block_sparse_moe.experts.4.w1", "model.layers.15.block_sparse_moe.experts.4.w2", "model.layers.15.block_sparse_moe.experts.4.w3", "model.layers.15.block_sparse_moe.experts.5.w1", "model.layers.15.block_sparse_moe.experts.5.w2", "model.layers.15.block_sparse_moe.experts.5.w3", "model.layers.15.block_sparse_moe.experts.6.w1", "model.layers.15.block_sparse_moe.experts.6.w2", "model.layers.15.block_sparse_moe.experts.6.w3", "model.layers.15.block_sparse_moe.experts.7.w1", "model.layers.15.block_sparse_moe.experts.7.w2", "model.layers.15.block_sparse_moe.experts.7.w3", "model.layers.16.self_attn.q_proj", "model.layers.16.self_attn.k_proj", "model.layers.16.self_attn.v_proj", "model.layers.16.self_attn.o_proj", "model.layers.16.block_sparse_moe.gate", "model.layers.16.block_sparse_moe.experts.0.w1", "model.layers.16.block_sparse_moe.experts.0.w2", "model.layers.16.block_sparse_moe.experts.0.w3", "model.layers.16.block_sparse_moe.experts.1.w1", "model.layers.16.block_sparse_moe.experts.1.w2", "model.layers.16.block_sparse_moe.experts.1.w3", "model.layers.16.block_sparse_moe.experts.2.w1", "model.layers.16.block_sparse_moe.experts.2.w2", "model.layers.16.block_sparse_moe.experts.2.w3", "model.layers.16.block_sparse_moe.experts.3.w1", "model.layers.16.block_sparse_moe.experts.3.w2", "model.layers.16.block_sparse_moe.experts.3.w3", "model.layers.16.block_sparse_moe.experts.4.w1", "model.layers.16.block_sparse_moe.experts.4.w2", "model.layers.16.block_sparse_moe.experts.4.w3", "model.layers.16.block_sparse_moe.experts.5.w1", "model.layers.16.block_sparse_moe.experts.5.w2", "model.layers.16.block_sparse_moe.experts.5.w3", "model.layers.16.block_sparse_moe.experts.6.w1", "model.layers.16.block_sparse_moe.experts.6.w2", "model.layers.16.block_sparse_moe.experts.6.w3", "model.layers.16.block_sparse_moe.experts.7.w1", "model.layers.16.block_sparse_moe.experts.7.w2", "model.layers.16.block_sparse_moe.experts.7.w3", "model.layers.17.self_attn.q_proj", "model.layers.17.self_attn.k_proj", "model.layers.17.self_attn.v_proj", "model.layers.17.self_attn.o_proj", "model.layers.17.block_sparse_moe.gate", "model.layers.17.block_sparse_moe.experts.0.w1", "model.layers.17.block_sparse_moe.experts.0.w2", "model.layers.17.block_sparse_moe.experts.0.w3", "model.layers.17.block_sparse_moe.experts.1.w1", "model.layers.17.block_sparse_moe.experts.1.w2", "model.layers.17.block_sparse_moe.experts.1.w3", "model.layers.17.block_sparse_moe.experts.2.w1", "model.layers.17.block_sparse_moe.experts.2.w2", "model.layers.17.block_sparse_moe.experts.2.w3", "model.layers.17.block_sparse_moe.experts.3.w1", "model.layers.17.block_sparse_moe.experts.3.w2", "model.layers.17.block_sparse_moe.experts.3.w3", "model.layers.17.block_sparse_moe.experts.4.w1", "model.layers.17.block_sparse_moe.experts.4.w2", "model.layers.17.block_sparse_moe.experts.4.w3", "model.layers.17.block_sparse_moe.experts.5.w1", "model.layers.17.block_sparse_moe.experts.5.w2", "model.layers.17.block_sparse_moe.experts.5.w3", "model.layers.17.block_sparse_moe.experts.6.w1", "model.layers.17.block_sparse_moe.experts.6.w2", "model.layers.17.block_sparse_moe.experts.6.w3", "model.layers.17.block_sparse_moe.experts.7.w1", "model.layers.17.block_sparse_moe.experts.7.w2", "model.layers.17.block_sparse_moe.experts.7.w3", "model.layers.18.self_attn.q_proj", "model.layers.18.self_attn.k_proj", "model.layers.18.self_attn.v_proj", "model.layers.18.self_attn.o_proj", "model.layers.18.block_sparse_moe.gate", "model.layers.18.block_sparse_moe.experts.0.w1", "model.layers.18.block_sparse_moe.experts.0.w2", "model.layers.18.block_sparse_moe.experts.0.w3", "model.layers.18.block_sparse_moe.experts.1.w1", "model.layers.18.block_sparse_moe.experts.1.w2", "model.layers.18.block_sparse_moe.experts.1.w3", "model.layers.18.block_sparse_moe.experts.2.w1", "model.layers.18.block_sparse_moe.experts.2.w2", "model.layers.18.block_sparse_moe.experts.2.w3", "model.layers.18.block_sparse_moe.experts.3.w1", "model.layers.18.block_sparse_moe.experts.3.w2", "model.layers.18.block_sparse_moe.experts.3.w3", "model.layers.18.block_sparse_moe.experts.4.w1", "model.layers.18.block_sparse_moe.experts.4.w2", "model.layers.18.block_sparse_moe.experts.4.w3", "model.layers.18.block_sparse_moe.experts.5.w1", "model.layers.18.block_sparse_moe.experts.5.w2", "model.layers.18.block_sparse_moe.experts.5.w3", "model.layers.18.block_sparse_moe.experts.6.w1", "model.layers.18.block_sparse_moe.experts.6.w2", "model.layers.18.block_sparse_moe.experts.6.w3", "model.layers.18.block_sparse_moe.experts.7.w1", "model.layers.18.block_sparse_moe.experts.7.w2", "model.layers.18.block_sparse_moe.experts.7.w3", "model.layers.19.self_attn.q_proj", "model.layers.19.self_attn.k_proj", "model.layers.19.self_attn.v_proj", "model.layers.19.self_attn.o_proj", "model.layers.19.block_sparse_moe.gate", "model.layers.19.block_sparse_moe.experts.0.w1", "model.layers.19.block_sparse_moe.experts.0.w2", "model.layers.19.block_sparse_moe.experts.0.w3", "model.layers.19.block_sparse_moe.experts.1.w1", "model.layers.19.block_sparse_moe.experts.1.w2", "model.layers.19.block_sparse_moe.experts.1.w3", "model.layers.19.block_sparse_moe.experts.2.w1", "model.layers.19.block_sparse_moe.experts.2.w2", "model.layers.19.block_sparse_moe.experts.2.w3", "model.layers.19.block_sparse_moe.experts.3.w1", "model.layers.19.block_sparse_moe.experts.3.w2", "model.layers.19.block_sparse_moe.experts.3.w3", "model.layers.19.block_sparse_moe.experts.4.w1", "model.layers.19.block_sparse_moe.experts.4.w2", "model.layers.19.block_sparse_moe.experts.4.w3", "model.layers.19.block_sparse_moe.experts.5.w1", "model.layers.19.block_sparse_moe.experts.5.w2", "model.layers.19.block_sparse_moe.experts.5.w3", "model.layers.19.block_sparse_moe.experts.6.w1", "model.layers.19.block_sparse_moe.experts.6.w2", "model.layers.19.block_sparse_moe.experts.6.w3", "model.layers.19.block_sparse_moe.experts.7.w1", "model.layers.19.block_sparse_moe.experts.7.w2", "model.layers.19.block_sparse_moe.experts.7.w3", "model.layers.20.self_attn.q_proj", "model.layers.20.self_attn.k_proj", "model.layers.20.self_attn.v_proj", "model.layers.20.self_attn.o_proj", "model.layers.20.block_sparse_moe.gate", "model.layers.20.block_sparse_moe.experts.0.w1", "model.layers.20.block_sparse_moe.experts.0.w2", "model.layers.20.block_sparse_moe.experts.0.w3", "model.layers.20.block_sparse_moe.experts.1.w1", "model.layers.20.block_sparse_moe.experts.1.w2", "model.layers.20.block_sparse_moe.experts.1.w3", "model.layers.20.block_sparse_moe.experts.2.w1", "model.layers.20.block_sparse_moe.experts.2.w2", "model.layers.20.block_sparse_moe.experts.2.w3", "model.layers.20.block_sparse_moe.experts.3.w1", "model.layers.20.block_sparse_moe.experts.3.w2", "model.layers.20.block_sparse_moe.experts.3.w3", "model.layers.20.block_sparse_moe.experts.4.w1", "model.layers.20.block_sparse_moe.experts.4.w2", "model.layers.20.block_sparse_moe.experts.4.w3", "model.layers.20.block_sparse_moe.experts.5.w1", "model.layers.20.block_sparse_moe.experts.5.w2", "model.layers.20.block_sparse_moe.experts.5.w3", "model.layers.20.block_sparse_moe.experts.6.w1", "model.layers.20.block_sparse_moe.experts.6.w2", "model.layers.20.block_sparse_moe.experts.6.w3", "model.layers.20.block_sparse_moe.experts.7.w1", "model.layers.20.block_sparse_moe.experts.7.w2", "model.layers.20.block_sparse_moe.experts.7.w3", "model.layers.21.self_attn.q_proj", "model.layers.21.self_attn.k_proj", "model.layers.21.self_attn.v_proj", "model.layers.21.self_attn.o_proj", "model.layers.21.block_sparse_moe.gate", "model.layers.21.block_sparse_moe.experts.0.w1", "model.layers.21.block_sparse_moe.experts.0.w2", "model.layers.21.block_sparse_moe.experts.0.w3", "model.layers.21.block_sparse_moe.experts.1.w1", "model.layers.21.block_sparse_moe.experts.1.w2", "model.layers.21.block_sparse_moe.experts.1.w3", "model.layers.21.block_sparse_moe.experts.2.w1", "model.layers.21.block_sparse_moe.experts.2.w2", "model.layers.21.block_sparse_moe.experts.2.w3", "model.layers.21.block_sparse_moe.experts.3.w1", "model.layers.21.block_sparse_moe.experts.3.w2", "model.layers.21.block_sparse_moe.experts.3.w3", "model.layers.21.block_sparse_moe.experts.4.w1", "model.layers.21.block_sparse_moe.experts.4.w2", "model.layers.21.block_sparse_moe.experts.4.w3", "model.layers.21.block_sparse_moe.experts.5.w1", "model.layers.21.block_sparse_moe.experts.5.w2", "model.layers.21.block_sparse_moe.experts.5.w3", "model.layers.21.block_sparse_moe.experts.6.w1", "model.layers.21.block_sparse_moe.experts.6.w2", "model.layers.21.block_sparse_moe.experts.6.w3", "model.layers.21.block_sparse_moe.experts.7.w1", "model.layers.21.block_sparse_moe.experts.7.w2", "model.layers.21.block_sparse_moe.experts.7.w3", "model.layers.22.self_attn.q_proj", "model.layers.22.self_attn.k_proj", "model.layers.22.self_attn.v_proj", "model.layers.22.self_attn.o_proj", "model.layers.22.block_sparse_moe.gate", "model.layers.22.block_sparse_moe.experts.0.w1", "model.layers.22.block_sparse_moe.experts.0.w2", "model.layers.22.block_sparse_moe.experts.0.w3", "model.layers.22.block_sparse_moe.experts.1.w1", "model.layers.22.block_sparse_moe.experts.1.w2", "model.layers.22.block_sparse_moe.experts.1.w3", "model.layers.22.block_sparse_moe.experts.2.w1", "model.layers.22.block_sparse_moe.experts.2.w2", "model.layers.22.block_sparse_moe.experts.2.w3", "model.layers.22.block_sparse_moe.experts.3.w1", "model.layers.22.block_sparse_moe.experts.3.w2", "model.layers.22.block_sparse_moe.experts.3.w3", "model.layers.22.block_sparse_moe.experts.4.w1", "model.layers.22.block_sparse_moe.experts.4.w2", "model.layers.22.block_sparse_moe.experts.4.w3", "model.layers.22.block_sparse_moe.experts.5.w1", "model.layers.22.block_sparse_moe.experts.5.w2", "model.layers.22.block_sparse_moe.experts.5.w3", "model.layers.22.block_sparse_moe.experts.6.w1", "model.layers.22.block_sparse_moe.experts.6.w2", "model.layers.22.block_sparse_moe.experts.6.w3", "model.layers.22.block_sparse_moe.experts.7.w1", "model.layers.22.block_sparse_moe.experts.7.w2", "model.layers.22.block_sparse_moe.experts.7.w3", "model.layers.23.self_attn.q_proj", "model.layers.23.self_attn.k_proj", "model.layers.23.self_attn.v_proj", "model.layers.23.self_attn.o_proj", "model.layers.23.block_sparse_moe.gate", "model.layers.23.block_sparse_moe.experts.0.w1", "model.layers.23.block_sparse_moe.experts.0.w2", "model.layers.23.block_sparse_moe.experts.0.w3", "model.layers.23.block_sparse_moe.experts.1.w1", "model.layers.23.block_sparse_moe.experts.1.w2", "model.layers.23.block_sparse_moe.experts.1.w3", "model.layers.23.block_sparse_moe.experts.2.w1", "model.layers.23.block_sparse_moe.experts.2.w2", "model.layers.23.block_sparse_moe.experts.2.w3", "model.layers.23.block_sparse_moe.experts.3.w1", "model.layers.23.block_sparse_moe.experts.3.w2", "model.layers.23.block_sparse_moe.experts.3.w3", "model.layers.23.block_sparse_moe.experts.4.w1", "model.layers.23.block_sparse_moe.experts.4.w2", "model.layers.23.block_sparse_moe.experts.4.w3", "model.layers.23.block_sparse_moe.experts.5.w1", "model.layers.23.block_sparse_moe.experts.5.w2", "model.layers.23.block_sparse_moe.experts.5.w3", "model.layers.23.block_sparse_moe.experts.6.w1", "model.layers.23.block_sparse_moe.experts.6.w2", "model.layers.23.block_sparse_moe.experts.6.w3", "model.layers.23.block_sparse_moe.experts.7.w1", "model.layers.23.block_sparse_moe.experts.7.w2", "model.layers.23.block_sparse_moe.experts.7.w3", "model.layers.24.self_attn.q_proj", "model.layers.24.self_attn.k_proj", "model.layers.24.self_attn.v_proj", "model.layers.24.self_attn.o_proj", "model.layers.24.block_sparse_moe.gate", "model.layers.24.block_sparse_moe.experts.0.w1", "model.layers.24.block_sparse_moe.experts.0.w2", "model.layers.24.block_sparse_moe.experts.0.w3", "model.layers.24.block_sparse_moe.experts.1.w1", "model.layers.24.block_sparse_moe.experts.1.w2", "model.layers.24.block_sparse_moe.experts.1.w3", "model.layers.24.block_sparse_moe.experts.2.w1", "model.layers.24.block_sparse_moe.experts.2.w2", "model.layers.24.block_sparse_moe.experts.2.w3", "model.layers.24.block_sparse_moe.experts.3.w1", "model.layers.24.block_sparse_moe.experts.3.w2", "model.layers.24.block_sparse_moe.experts.3.w3", "model.layers.24.block_sparse_moe.experts.4.w1", "model.layers.24.block_sparse_moe.experts.4.w2", "model.layers.24.block_sparse_moe.experts.4.w3", "model.layers.24.block_sparse_moe.experts.5.w1", "model.layers.24.block_sparse_moe.experts.5.w2", "model.layers.24.block_sparse_moe.experts.5.w3", "model.layers.24.block_sparse_moe.experts.6.w1", "model.layers.24.block_sparse_moe.experts.6.w2", "model.layers.24.block_sparse_moe.experts.6.w3", "model.layers.24.block_sparse_moe.experts.7.w1", "model.layers.24.block_sparse_moe.experts.7.w2", "model.layers.24.block_sparse_moe.experts.7.w3", "model.layers.25.self_attn.q_proj", "model.layers.25.self_attn.k_proj", "model.layers.25.self_attn.v_proj", "model.layers.25.self_attn.o_proj", "model.layers.25.block_sparse_moe.gate", "model.layers.25.block_sparse_moe.experts.0.w1", "model.layers.25.block_sparse_moe.experts.0.w2", "model.layers.25.block_sparse_moe.experts.0.w3", "model.layers.25.block_sparse_moe.experts.1.w1", "model.layers.25.block_sparse_moe.experts.1.w2", "model.layers.25.block_sparse_moe.experts.1.w3", "model.layers.25.block_sparse_moe.experts.2.w1", "model.layers.25.block_sparse_moe.experts.2.w2", "model.layers.25.block_sparse_moe.experts.2.w3", "model.layers.25.block_sparse_moe.experts.3.w1", "model.layers.25.block_sparse_moe.experts.3.w2", "model.layers.25.block_sparse_moe.experts.3.w3", "model.layers.25.block_sparse_moe.experts.4.w1", "model.layers.25.block_sparse_moe.experts.4.w2", "model.layers.25.block_sparse_moe.experts.4.w3", "model.layers.25.block_sparse_moe.experts.5.w1", "model.layers.25.block_sparse_moe.experts.5.w2", "model.layers.25.block_sparse_moe.experts.5.w3", "model.layers.25.block_sparse_moe.experts.6.w1", "model.layers.25.block_sparse_moe.experts.6.w2", "model.layers.25.block_sparse_moe.experts.6.w3", "model.layers.25.block_sparse_moe.experts.7.w1", "model.layers.25.block_sparse_moe.experts.7.w2", "model.layers.25.block_sparse_moe.experts.7.w3", "model.layers.26.self_attn.q_proj", "model.layers.26.self_attn.k_proj", "model.layers.26.self_attn.v_proj", "model.layers.26.self_attn.o_proj", "model.layers.26.block_sparse_moe.gate", "model.layers.26.block_sparse_moe.experts.0.w1", "model.layers.26.block_sparse_moe.experts.0.w2", "model.layers.26.block_sparse_moe.experts.0.w3", "model.layers.26.block_sparse_moe.experts.1.w1", "model.layers.26.block_sparse_moe.experts.1.w2", "model.layers.26.block_sparse_moe.experts.1.w3", "model.layers.26.block_sparse_moe.experts.2.w1", "model.layers.26.block_sparse_moe.experts.2.w2", "model.layers.26.block_sparse_moe.experts.2.w3", "model.layers.26.block_sparse_moe.experts.3.w1", "model.layers.26.block_sparse_moe.experts.3.w2", "model.layers.26.block_sparse_moe.experts.3.w3", "model.layers.26.block_sparse_moe.experts.4.w1", "model.layers.26.block_sparse_moe.experts.4.w2", "model.layers.26.block_sparse_moe.experts.4.w3", "model.layers.26.block_sparse_moe.experts.5.w1", "model.layers.26.block_sparse_moe.experts.5.w2", "model.layers.26.block_sparse_moe.experts.5.w3", "model.layers.26.block_sparse_moe.experts.6.w1", "model.layers.26.block_sparse_moe.experts.6.w2", "model.layers.26.block_sparse_moe.experts.6.w3", "model.layers.26.block_sparse_moe.experts.7.w1", "model.layers.26.block_sparse_moe.experts.7.w2", "model.layers.26.block_sparse_moe.experts.7.w3", "model.layers.27.self_attn.q_proj", "model.layers.27.self_attn.k_proj", "model.layers.27.self_attn.v_proj", "model.layers.27.self_attn.o_proj", "model.layers.27.block_sparse_moe.gate", "model.layers.27.block_sparse_moe.experts.0.w1", "model.layers.27.block_sparse_moe.experts.0.w2", "model.layers.27.block_sparse_moe.experts.0.w3", "model.layers.27.block_sparse_moe.experts.1.w1", "model.layers.27.block_sparse_moe.experts.1.w2", "model.layers.27.block_sparse_moe.experts.1.w3", "model.layers.27.block_sparse_moe.experts.2.w1", "model.layers.27.block_sparse_moe.experts.2.w2", "model.layers.27.block_sparse_moe.experts.2.w3", "model.layers.27.block_sparse_moe.experts.3.w1", "model.layers.27.block_sparse_moe.experts.3.w2", "model.layers.27.block_sparse_moe.experts.3.w3", "model.layers.27.block_sparse_moe.experts.4.w1", "model.layers.27.block_sparse_moe.experts.4.w2", "model.layers.27.block_sparse_moe.experts.4.w3", "model.layers.27.block_sparse_moe.experts.5.w1", "model.layers.27.block_sparse_moe.experts.5.w2", "model.layers.27.block_sparse_moe.experts.5.w3", "model.layers.27.block_sparse_moe.experts.6.w1", "model.layers.27.block_sparse_moe.experts.6.w2", "model.layers.27.block_sparse_moe.experts.6.w3", "model.layers.27.block_sparse_moe.experts.7.w1", "model.layers.27.block_sparse_moe.experts.7.w2", "model.layers.27.block_sparse_moe.experts.7.w3", "model.layers.28.self_attn.q_proj", "model.layers.28.self_attn.k_proj", "model.layers.28.self_attn.v_proj", "model.layers.28.self_attn.o_proj", "model.layers.28.block_sparse_moe.gate", "model.layers.28.block_sparse_moe.experts.0.w1", "model.layers.28.block_sparse_moe.experts.0.w2", "model.layers.28.block_sparse_moe.experts.0.w3", "model.layers.28.block_sparse_moe.experts.1.w1", "model.layers.28.block_sparse_moe.experts.1.w2", "model.layers.28.block_sparse_moe.experts.1.w3", "model.layers.28.block_sparse_moe.experts.2.w1", "model.layers.28.block_sparse_moe.experts.2.w2", "model.layers.28.block_sparse_moe.experts.2.w3", "model.layers.28.block_sparse_moe.experts.3.w1", "model.layers.28.block_sparse_moe.experts.3.w2", "model.layers.28.block_sparse_moe.experts.3.w3", "model.layers.28.block_sparse_moe.experts.4.w1", "model.layers.28.block_sparse_moe.experts.4.w2", "model.layers.28.block_sparse_moe.experts.4.w3", "model.layers.28.block_sparse_moe.experts.5.w1", "model.layers.28.block_sparse_moe.experts.5.w2", "model.layers.28.block_sparse_moe.experts.5.w3", "model.layers.28.block_sparse_moe.experts.6.w1", "model.layers.28.block_sparse_moe.experts.6.w2", "model.layers.28.block_sparse_moe.experts.6.w3", "model.layers.28.block_sparse_moe.experts.7.w1", "model.layers.28.block_sparse_moe.experts.7.w2", "model.layers.28.block_sparse_moe.experts.7.w3", "model.layers.29.self_attn.q_proj", "model.layers.29.self_attn.k_proj", "model.layers.29.self_attn.v_proj", "model.layers.29.self_attn.o_proj", "model.layers.29.block_sparse_moe.gate", "model.layers.29.block_sparse_moe.experts.0.w1", "model.layers.29.block_sparse_moe.experts.0.w2", "model.layers.29.block_sparse_moe.experts.0.w3", "model.layers.29.block_sparse_moe.experts.1.w1", "model.layers.29.block_sparse_moe.experts.1.w2", "model.layers.29.block_sparse_moe.experts.1.w3", "model.layers.29.block_sparse_moe.experts.2.w1", "model.layers.29.block_sparse_moe.experts.2.w2", "model.layers.29.block_sparse_moe.experts.2.w3", "model.layers.29.block_sparse_moe.experts.3.w1", "model.layers.29.block_sparse_moe.experts.3.w2", "model.layers.29.block_sparse_moe.experts.3.w3", "model.layers.29.block_sparse_moe.experts.4.w1", "model.layers.29.block_sparse_moe.experts.4.w2", "model.layers.29.block_sparse_moe.experts.4.w3", "model.layers.29.block_sparse_moe.experts.5.w1", "model.layers.29.block_sparse_moe.experts.5.w2", "model.layers.29.block_sparse_moe.experts.5.w3", "model.layers.29.block_sparse_moe.experts.6.w1", "model.layers.29.block_sparse_moe.experts.6.w2", "model.layers.29.block_sparse_moe.experts.6.w3", "model.layers.29.block_sparse_moe.experts.7.w1", "model.layers.29.block_sparse_moe.experts.7.w2", "model.layers.29.block_sparse_moe.experts.7.w3", "model.layers.30.self_attn.q_proj", "model.layers.30.self_attn.k_proj", "model.layers.30.self_attn.v_proj", "model.layers.30.self_attn.o_proj", "model.layers.30.block_sparse_moe.gate", "model.layers.30.block_sparse_moe.experts.0.w1", "model.layers.30.block_sparse_moe.experts.0.w2", "model.layers.30.block_sparse_moe.experts.0.w3", "model.layers.30.block_sparse_moe.experts.1.w1", "model.layers.30.block_sparse_moe.experts.1.w2", "model.layers.30.block_sparse_moe.experts.1.w3", "model.layers.30.block_sparse_moe.experts.2.w1", "model.layers.30.block_sparse_moe.experts.2.w2", "model.layers.30.block_sparse_moe.experts.2.w3", "model.layers.30.block_sparse_moe.experts.3.w1", "model.layers.30.block_sparse_moe.experts.3.w2", "model.layers.30.block_sparse_moe.experts.3.w3", "model.layers.30.block_sparse_moe.experts.4.w1", "model.layers.30.block_sparse_moe.experts.4.w2", "model.layers.30.block_sparse_moe.experts.4.w3", "model.layers.30.block_sparse_moe.experts.5.w1", "model.layers.30.block_sparse_moe.experts.5.w2", "model.layers.30.block_sparse_moe.experts.5.w3", "model.layers.30.block_sparse_moe.experts.6.w1", "model.layers.30.block_sparse_moe.experts.6.w2", "model.layers.30.block_sparse_moe.experts.6.w3", "model.layers.30.block_sparse_moe.experts.7.w1", "model.layers.30.block_sparse_moe.experts.7.w2", "model.layers.30.block_sparse_moe.experts.7.w3", "model.layers.31.self_attn.q_proj", "model.layers.31.self_attn.k_proj", "model.layers.31.self_attn.v_proj", "model.layers.31.self_attn.o_proj", "model.layers.31.block_sparse_moe.gate", "model.layers.31.block_sparse_moe.experts.0.w1", "model.layers.31.block_sparse_moe.experts.0.w2", "model.layers.31.block_sparse_moe.experts.0.w3", "model.layers.31.block_sparse_moe.experts.1.w1", "model.layers.31.block_sparse_moe.experts.1.w2", "model.layers.31.block_sparse_moe.experts.1.w3", "model.layers.31.block_sparse_moe.experts.2.w1", "model.layers.31.block_sparse_moe.experts.2.w2", "model.layers.31.block_sparse_moe.experts.2.w3", "model.layers.31.block_sparse_moe.experts.3.w1", "model.layers.31.block_sparse_moe.experts.3.w2", "model.layers.31.block_sparse_moe.experts.3.w3", "model.layers.31.block_sparse_moe.experts.4.w1", "model.layers.31.block_sparse_moe.experts.4.w2", "model.layers.31.block_sparse_moe.experts.4.w3", "model.layers.31.block_sparse_moe.experts.5.w1", "model.layers.31.block_sparse_moe.experts.5.w2", "model.layers.31.block_sparse_moe.experts.5.w3", "model.layers.31.block_sparse_moe.experts.6.w1", "model.layers.31.block_sparse_moe.experts.6.w2", "model.layers.31.block_sparse_moe.experts.6.w3", "model.layers.31.block_sparse_moe.experts.7.w1", "model.layers.31.block_sparse_moe.experts.7.w2", "model.layers.31.block_sparse_moe.experts.7.w3", "lm_head" ], "registry_requires_subclass": false, "sparsity_structure": "unstructured", "targets": [ "model.layers.0.self_attn.q_proj", "model.layers.0.self_attn.k_proj" ] } }, "rms_norm_eps": 1e-05, "rope_theta": 1000000.0, "router_aux_loss_coef": 0.02, "router_jitter_noise": 0.0, "sliding_window": null, "tie_word_embeddings": false, "torch_dtype": "bfloat16", "transformers_version": "4.46.2", "use_cache": false, "vocab_size": 32000 }