{ "_name_or_path": "checkpoints/stabilityai/stablelm-zephyr-3b", "anyprec": { "arch_config": { "layers_name": "layers", "model_name": "model", "module_names": [ "self_attn.q_proj", "self_attn.k_proj", "self_attn.v_proj", "self_attn.o_proj", "mlp.gate_proj", "mlp.up_proj", "mlp.down_proj" ] }, "group_count": 1, "parent_precision": 4, "seed_precision": 2, "sparse_numvals": { "model.layers.0.mlp.down_proj": 361261, "model.layers.0.mlp.gate_proj": 409239, "model.layers.0.mlp.up_proj": 382725, "model.layers.0.self_attn.k_proj": 1249422, "model.layers.0.self_attn.o_proj": 493934, "model.layers.0.self_attn.q_proj": 1383514, "model.layers.0.self_attn.v_proj": 582091, "model.layers.1.mlp.down_proj": 371941, "model.layers.1.mlp.gate_proj": 312127, "model.layers.1.mlp.up_proj": 350541, "model.layers.1.self_attn.k_proj": 739147, "model.layers.1.self_attn.o_proj": 870440, "model.layers.1.self_attn.q_proj": 754729, "model.layers.1.self_attn.v_proj": 790019, "model.layers.10.mlp.down_proj": 400494, "model.layers.10.mlp.gate_proj": 437481, "model.layers.10.mlp.up_proj": 389717, "model.layers.10.self_attn.k_proj": 217707, "model.layers.10.self_attn.o_proj": 159623, "model.layers.10.self_attn.q_proj": 204767, "model.layers.10.self_attn.v_proj": 168686, "model.layers.11.mlp.down_proj": 398251, "model.layers.11.mlp.gate_proj": 468437, "model.layers.11.mlp.up_proj": 399514, "model.layers.11.self_attn.k_proj": 223137, "model.layers.11.self_attn.o_proj": 153385, "model.layers.11.self_attn.q_proj": 207469, "model.layers.11.self_attn.v_proj": 157690, "model.layers.12.mlp.down_proj": 392789, "model.layers.12.mlp.gate_proj": 511766, "model.layers.12.mlp.up_proj": 388221, "model.layers.12.self_attn.k_proj": 195398, "model.layers.12.self_attn.o_proj": 142264, "model.layers.12.self_attn.q_proj": 185287, "model.layers.12.self_attn.v_proj": 152926, "model.layers.13.mlp.down_proj": 395950, "model.layers.13.mlp.gate_proj": 495319, "model.layers.13.mlp.up_proj": 404143, "model.layers.13.self_attn.k_proj": 218597, "model.layers.13.self_attn.o_proj": 159306, "model.layers.13.self_attn.q_proj": 206129, "model.layers.13.self_attn.v_proj": 168994, "model.layers.14.mlp.down_proj": 395967, "model.layers.14.mlp.gate_proj": 562906, "model.layers.14.mlp.up_proj": 390940, "model.layers.14.self_attn.k_proj": 194047, "model.layers.14.self_attn.o_proj": 145953, "model.layers.14.self_attn.q_proj": 182838, "model.layers.14.self_attn.v_proj": 153758, "model.layers.15.mlp.down_proj": 386374, "model.layers.15.mlp.gate_proj": 522112, "model.layers.15.mlp.up_proj": 395074, "model.layers.15.self_attn.k_proj": 196997, "model.layers.15.self_attn.o_proj": 150452, "model.layers.15.self_attn.q_proj": 184345, "model.layers.15.self_attn.v_proj": 156440, "model.layers.16.mlp.down_proj": 384573, "model.layers.16.mlp.gate_proj": 477452, "model.layers.16.mlp.up_proj": 386693, "model.layers.16.self_attn.k_proj": 215088, "model.layers.16.self_attn.o_proj": 145176, "model.layers.16.self_attn.q_proj": 193489, "model.layers.16.self_attn.v_proj": 157381, "model.layers.17.mlp.down_proj": 376292, "model.layers.17.mlp.gate_proj": 440742, "model.layers.17.mlp.up_proj": 372740, "model.layers.17.self_attn.k_proj": 198626, "model.layers.17.self_attn.o_proj": 138659, "model.layers.17.self_attn.q_proj": 184019, "model.layers.17.self_attn.v_proj": 146408, "model.layers.18.mlp.down_proj": 372954, "model.layers.18.mlp.gate_proj": 407526, "model.layers.18.mlp.up_proj": 359072, "model.layers.18.self_attn.k_proj": 195759, "model.layers.18.self_attn.o_proj": 143198, "model.layers.18.self_attn.q_proj": 178705, "model.layers.18.self_attn.v_proj": 154099, "model.layers.19.mlp.down_proj": 372170, "model.layers.19.mlp.gate_proj": 404678, "model.layers.19.mlp.up_proj": 362693, "model.layers.19.self_attn.k_proj": 181548, "model.layers.19.self_attn.o_proj": 141818, "model.layers.19.self_attn.q_proj": 167249, "model.layers.19.self_attn.v_proj": 144673, "model.layers.2.mlp.down_proj": 343350, "model.layers.2.mlp.gate_proj": 318733, "model.layers.2.mlp.up_proj": 347181, "model.layers.2.self_attn.k_proj": 727364, "model.layers.2.self_attn.o_proj": 206116, "model.layers.2.self_attn.q_proj": 632838, "model.layers.2.self_attn.v_proj": 178305, "model.layers.20.mlp.down_proj": 358950, "model.layers.20.mlp.gate_proj": 407599, "model.layers.20.mlp.up_proj": 359506, "model.layers.20.self_attn.k_proj": 174602, "model.layers.20.self_attn.o_proj": 139503, "model.layers.20.self_attn.q_proj": 169296, "model.layers.20.self_attn.v_proj": 136204, "model.layers.21.mlp.down_proj": 357960, "model.layers.21.mlp.gate_proj": 385992, "model.layers.21.mlp.up_proj": 348288, "model.layers.21.self_attn.k_proj": 185211, "model.layers.21.self_attn.o_proj": 142190, "model.layers.21.self_attn.q_proj": 176282, "model.layers.21.self_attn.v_proj": 142371, "model.layers.22.mlp.down_proj": 361293, "model.layers.22.mlp.gate_proj": 377918, "model.layers.22.mlp.up_proj": 353747, "model.layers.22.self_attn.k_proj": 179724, "model.layers.22.self_attn.o_proj": 143106, "model.layers.22.self_attn.q_proj": 167648, "model.layers.22.self_attn.v_proj": 145138, "model.layers.23.mlp.down_proj": 358986, "model.layers.23.mlp.gate_proj": 378110, "model.layers.23.mlp.up_proj": 356957, "model.layers.23.self_attn.k_proj": 178196, "model.layers.23.self_attn.o_proj": 142376, "model.layers.23.self_attn.q_proj": 168788, "model.layers.23.self_attn.v_proj": 143664, "model.layers.24.mlp.down_proj": 354332, "model.layers.24.mlp.gate_proj": 366565, "model.layers.24.mlp.up_proj": 346526, "model.layers.24.self_attn.k_proj": 185428, "model.layers.24.self_attn.o_proj": 140813, "model.layers.24.self_attn.q_proj": 176689, "model.layers.24.self_attn.v_proj": 139483, "model.layers.25.mlp.down_proj": 353602, "model.layers.25.mlp.gate_proj": 343542, "model.layers.25.mlp.up_proj": 345700, "model.layers.25.self_attn.k_proj": 178173, "model.layers.25.self_attn.o_proj": 146206, "model.layers.25.self_attn.q_proj": 166860, "model.layers.25.self_attn.v_proj": 144246, "model.layers.26.mlp.down_proj": 356005, "model.layers.26.mlp.gate_proj": 339690, "model.layers.26.mlp.up_proj": 347487, "model.layers.26.self_attn.k_proj": 196789, "model.layers.26.self_attn.o_proj": 140962, "model.layers.26.self_attn.q_proj": 179600, "model.layers.26.self_attn.v_proj": 145076, "model.layers.27.mlp.down_proj": 354147, "model.layers.27.mlp.gate_proj": 333874, "model.layers.27.mlp.up_proj": 342862, "model.layers.27.self_attn.k_proj": 191073, "model.layers.27.self_attn.o_proj": 138949, "model.layers.27.self_attn.q_proj": 174841, "model.layers.27.self_attn.v_proj": 139074, "model.layers.28.mlp.down_proj": 350013, "model.layers.28.mlp.gate_proj": 343786, "model.layers.28.mlp.up_proj": 341398, "model.layers.28.self_attn.k_proj": 193596, "model.layers.28.self_attn.o_proj": 141809, "model.layers.28.self_attn.q_proj": 174614, "model.layers.28.self_attn.v_proj": 143142, "model.layers.29.mlp.down_proj": 356645, "model.layers.29.mlp.gate_proj": 382614, "model.layers.29.mlp.up_proj": 351816, "model.layers.29.self_attn.k_proj": 192295, "model.layers.29.self_attn.o_proj": 147619, "model.layers.29.self_attn.q_proj": 174798, "model.layers.29.self_attn.v_proj": 148455, "model.layers.3.mlp.down_proj": 353085, "model.layers.3.mlp.gate_proj": 310395, "model.layers.3.mlp.up_proj": 352469, "model.layers.3.self_attn.k_proj": 329393, "model.layers.3.self_attn.o_proj": 179470, "model.layers.3.self_attn.q_proj": 307652, "model.layers.3.self_attn.v_proj": 163581, "model.layers.30.mlp.down_proj": 380293, "model.layers.30.mlp.gate_proj": 435280, "model.layers.30.mlp.up_proj": 352518, "model.layers.30.self_attn.k_proj": 178469, "model.layers.30.self_attn.o_proj": 156350, "model.layers.30.self_attn.q_proj": 165420, "model.layers.30.self_attn.v_proj": 157109, "model.layers.31.mlp.down_proj": 407998, "model.layers.31.mlp.gate_proj": 478129, "model.layers.31.mlp.up_proj": 360125, "model.layers.31.self_attn.k_proj": 162466, "model.layers.31.self_attn.o_proj": 180364, "model.layers.31.self_attn.q_proj": 160350, "model.layers.31.self_attn.v_proj": 166443, "model.layers.4.mlp.down_proj": 351602, "model.layers.4.mlp.gate_proj": 303771, "model.layers.4.mlp.up_proj": 351399, "model.layers.4.self_attn.k_proj": 269033, "model.layers.4.self_attn.o_proj": 159082, "model.layers.4.self_attn.q_proj": 257637, "model.layers.4.self_attn.v_proj": 156983, "model.layers.5.mlp.down_proj": 350684, "model.layers.5.mlp.gate_proj": 325463, "model.layers.5.mlp.up_proj": 347872, "model.layers.5.self_attn.k_proj": 237619, "model.layers.5.self_attn.o_proj": 179358, "model.layers.5.self_attn.q_proj": 227821, "model.layers.5.self_attn.v_proj": 191027, "model.layers.6.mlp.down_proj": 364690, "model.layers.6.mlp.gate_proj": 341908, "model.layers.6.mlp.up_proj": 356264, "model.layers.6.self_attn.k_proj": 200776, "model.layers.6.self_attn.o_proj": 144559, "model.layers.6.self_attn.q_proj": 189252, "model.layers.6.self_attn.v_proj": 163452, "model.layers.7.mlp.down_proj": 367623, "model.layers.7.mlp.gate_proj": 355908, "model.layers.7.mlp.up_proj": 361007, "model.layers.7.self_attn.k_proj": 192816, "model.layers.7.self_attn.o_proj": 141675, "model.layers.7.self_attn.q_proj": 182592, "model.layers.7.self_attn.v_proj": 147459, "model.layers.8.mlp.down_proj": 381346, "model.layers.8.mlp.gate_proj": 392814, "model.layers.8.mlp.up_proj": 385174, "model.layers.8.self_attn.k_proj": 214820, "model.layers.8.self_attn.o_proj": 148998, "model.layers.8.self_attn.q_proj": 205237, "model.layers.8.self_attn.v_proj": 154614, "model.layers.9.mlp.down_proj": 384446, "model.layers.9.mlp.gate_proj": 408833, "model.layers.9.mlp.up_proj": 386970, "model.layers.9.self_attn.k_proj": 203316, "model.layers.9.self_attn.o_proj": 150194, "model.layers.9.self_attn.q_proj": 197655, "model.layers.9.self_attn.v_proj": 157709 } }, "architectures": [ "StableLmForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 0, "eos_token_id": 0, "hidden_act": "silu", "hidden_dropout": 0.0, "hidden_size": 2560, "initializer_range": 0.02, "intermediate_size": 6912, "layer_norm_eps": 1e-05, "max_position_embeddings": 4096, "model_type": "stablelm", "num_attention_heads": 32, "num_hidden_layers": 32, "num_key_value_heads": 32, "partial_rotary_factor": 0.25, "rope_scaling": null, "rope_theta": 10000, "tie_word_embeddings": false, "torch_dtype": "float16", "transformers_version": "4.39.3", "use_cache": true, "use_qkv_bias": false, "vocab_size": 50304 }