NullSense/Nanbeige4.2-3B-FP8-Dynamic

1{2  "architectures": [3    "NanbeigeForCausalLM"4  ],5  "attention_bias": false,6  "attention_dropout": 0.0,7  "auto_map": {8    "AutoConfig": "configuration_nanbeige.NanbeigeConfig",9    "AutoModel": "modeling_nanbeige.NanbeigeModel",10    "AutoModelForCausalLM": "modeling_nanbeige.NanbeigeForCausalLM"11  },12  "bos_token_id": 166100,13  "depth_attention_recent_window": 0,14  "depth_attention_static_anchor_once": true,15  "depth_attention_stride": null,16  "dtype": "bfloat16",17  "emb_neighbor_num": null,18  "emb_split_num": null,19  "emb_tp_num": null,20  "enable_depth_attention": false,21  "enable_double_loop_split": false,22  "enable_h_res_identity": false,23  "enable_hyper_connection": false,24  "enable_mhc": false,25  "eos_token_id": 166101,26  "head_dim": 128,27  "hidden_act": "silu",28  "hidden_size": 3072,29  "initializer_range": 0.02,30  "insert_ngram_layer_idx": [],31  "intermediate_size": 10752,32  "kv_channels": 128,33  "loop_loss_weights": [],34  "loop_middle_layers": null,35  "loop_share_kv": false,36  "max_position_embeddings": 262144,37  "mhc_diff_for_loop": false,38  "mhc_double_stream_position_for_loop": null,39  "mhc_identity_nohresparam": false,40  "mhc_init_gating_factor": 0.01,41  "mhc_sinkhorn_iterations": 20,42  "mlp_bias": false,43  "model_type": "nanbeige",44  "ngram_compressed_tokenizer": false,45  "ngram_embedding_hidden_size": null,46  "ngram_fused_mode": "average",47  "ngram_insert_all_layers": false,48  "ngram_layer_downproject_size": null,49  "ngram_mod_force_prime": false,50  "ngram_vocab_size_ratio": null,51  "num_attention_heads": 48,52  "num_hidden_layers": 22,53  "num_key_value_heads": 8,54  "num_loops": 2,55  "num_residual_streams": 4,56  "pad_token_id": 0,57  "pretraining_tp": 1,58  "qk_layernorm": false,59  "quantization_config": {60    "config_groups": {61      "group_0": {62        "format": "float-quantized",63        "input_activations": {64          "actorder": null,65          "block_structure": null,66          "dynamic": true,67          "group_size": null,68          "num_bits": 8,69          "observer": null,70          "observer_kwargs": {},71          "scale_dtype": null,72          "strategy": "token",73          "symmetric": true,74          "type": "float",75          "zp_dtype": null76        },77        "output_activations": null,78        "targets": [79          "Linear"80        ],81        "weights": {82          "actorder": null,83          "block_structure": null,84          "dynamic": false,85          "group_size": null,86          "num_bits": 8,87          "observer": "memoryless_minmax",88          "observer_kwargs": {},89          "scale_dtype": null,90          "strategy": "channel",91          "symmetric": true,92          "type": "float",93          "zp_dtype": null94        }95      }96    },97    "format": "float-quantized",98    "global_compression_ratio": null,99    "ignore": [100      "lm_head"101    ],102    "kv_cache_scheme": null,103    "quant_method": "compressed-tensors",104    "quantization_status": "compressed",105    "sparsity_config": {},106    "transform_config": {},107    "version": "0.17.1"108  },109  "rms_norm_eps": 1e-05,110  "rope_parameters": null,111  "rope_theta": 70000000,112  "skip_loop_final_norm": false,113  "skip_ngram_for_input": false,114  "tie_word_embeddings": false,115  "transformers_version": "5.10.1",116  "use_cache": true,117  "vocab_size": 166144118}