{ "architectures": [ "MistralForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 1, "dtype": "bfloat16", "eos_token_id": 2, "head_dim": 128, "hidden_act": "silu", "hidden_size": 4096, "initializer_range": 0.02, "intermediate_size": 14336, "max_position_embeddings": 32768, "model_type": "mistral", "num_attention_heads": 32, "num_hidden_layers": 32, "num_key_value_heads": 8, "pad_token_id": null, "quantization_config": { "checkpoint_format": "dynquant-packed", "group_size": 128, "lm_head_quantized": true, "modules": { "lm_head": { "bits": 8, "out_features": 32768 }, "model.embed_tokens": { "bits": 4, "out_features": 32768 }, "model.layers.0.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.0.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.0.mlp.up_proj": { "bits": 3, "out_features": 14336 }, "model.layers.0.self_attn.k_proj": { "bits": 4, "out_features": 1024 }, "model.layers.0.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.0.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.0.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.1.mlp.down_proj": { "bits": 8, "out_features": 4096 }, "model.layers.1.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.1.mlp.up_proj": { "bits": 3, "out_features": 14336 }, "model.layers.1.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.1.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.1.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.1.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.10.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.10.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.10.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.10.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.10.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.10.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.10.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.11.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.11.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.11.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.11.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.11.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.11.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.11.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.12.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.12.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.12.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.12.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.12.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.12.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.12.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.13.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.13.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.13.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.13.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.13.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.13.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.13.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.14.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.14.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.14.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.14.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.14.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.14.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.14.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.15.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.15.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.15.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.15.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.15.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.15.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.15.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.16.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.16.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.16.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.16.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.16.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.16.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.16.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.17.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.17.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.17.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.17.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.17.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.17.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.17.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.18.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.18.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.18.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.18.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.18.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.18.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.18.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.19.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.19.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.19.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.19.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.19.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.19.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.19.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.2.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.2.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.2.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.2.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.2.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.2.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.2.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.20.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.20.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.20.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.20.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.20.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.20.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.20.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.21.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.21.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.21.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.21.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.21.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.21.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.21.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.22.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.22.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.22.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.22.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.22.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.22.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.22.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.23.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.23.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.23.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.23.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.23.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.23.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.23.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.24.mlp.down_proj": { "bits": 3, "out_features": 4096 }, "model.layers.24.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.24.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.24.self_attn.k_proj": { "bits": 4, "out_features": 1024 }, "model.layers.24.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.24.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.24.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.25.mlp.down_proj": { "bits": 3, "out_features": 4096 }, "model.layers.25.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.25.mlp.up_proj": { "bits": 3, "out_features": 14336 }, "model.layers.25.self_attn.k_proj": { "bits": 4, "out_features": 1024 }, "model.layers.25.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.25.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.25.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.26.mlp.down_proj": { "bits": 3, "out_features": 4096 }, "model.layers.26.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.26.mlp.up_proj": { "bits": 3, "out_features": 14336 }, "model.layers.26.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.26.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.26.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.26.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.27.mlp.down_proj": { "bits": 3, "out_features": 4096 }, "model.layers.27.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.27.mlp.up_proj": { "bits": 3, "out_features": 14336 }, "model.layers.27.self_attn.k_proj": { "bits": 4, "out_features": 1024 }, "model.layers.27.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.27.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.27.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.28.mlp.down_proj": { "bits": 3, "out_features": 4096 }, "model.layers.28.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.28.mlp.up_proj": { "bits": 3, "out_features": 14336 }, "model.layers.28.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.28.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.28.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.28.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.29.mlp.down_proj": { "bits": 3, "out_features": 4096 }, "model.layers.29.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.29.mlp.up_proj": { "bits": 3, "out_features": 14336 }, "model.layers.29.self_attn.k_proj": { "bits": 4, "out_features": 1024 }, "model.layers.29.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.29.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.29.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.3.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.3.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.3.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.3.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.3.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.3.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.3.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.30.mlp.down_proj": { "bits": 3, "out_features": 4096 }, "model.layers.30.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.30.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.30.self_attn.k_proj": { "bits": 4, "out_features": 1024 }, "model.layers.30.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.30.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.30.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.31.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.31.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.31.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.31.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.31.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.31.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.31.self_attn.v_proj": { "bits": 4, "out_features": 1024 }, "model.layers.4.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.4.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.4.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.4.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.4.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.4.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.4.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.5.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.5.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.5.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.5.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.5.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.5.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.5.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.6.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.6.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.6.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.6.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.6.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.6.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.6.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.7.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.7.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.7.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.7.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.7.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.7.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.7.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.8.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.8.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.8.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.8.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.8.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.8.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.8.self_attn.v_proj": { "bits": 8, "out_features": 1024 }, "model.layers.9.mlp.down_proj": { "bits": 4, "out_features": 4096 }, "model.layers.9.mlp.gate_proj": { "bits": 4, "out_features": 14336 }, "model.layers.9.mlp.up_proj": { "bits": 4, "out_features": 14336 }, "model.layers.9.self_attn.k_proj": { "bits": 8, "out_features": 1024 }, "model.layers.9.self_attn.o_proj": { "bits": 4, "out_features": 4096 }, "model.layers.9.self_attn.q_proj": { "bits": 4, "out_features": 4096 }, "model.layers.9.self_attn.v_proj": { "bits": 8, "out_features": 1024 } }, "modules_to_not_convert": [], "quant_method": "dynquant", "schema_version": 1, "symmetric": false, "version": "0.4.0" }, "rms_norm_eps": 1e-05, "rope_parameters": { "rope_theta": 1000000.0, "rope_type": "default" }, "sliding_window": null, "tie_word_embeddings": false, "transformers_version": "5.10.1", "use_cache": true, "vocab_size": 32768 }