unsloth/Kimi-K2.7-Code
21124
1{2 "architectures": [3 "KimiK25ForConditionalGeneration"4 ],5 "auto_map": {6 "AutoConfig": "configuration_kimi_k25.KimiK25Config",7 "AutoModel": "modeling_kimi_k25.KimiK25ForConditionalGeneration",8 "AutoModelForCausalLM": "modeling_kimi_k25.KimiK25ForConditionalGeneration"9 },10 "bos_token_id": 163584,11 "torch_dtype": "bfloat16",12 "eos_token_id": 163586,13 "ignore_index": -100,14 "media_placeholder_token_id": 163605,15 "model_type": "kimi_k25",16 "pad_token_id": 163839,17 "quantization_config": {18 "config_groups": {19 "group_0": {20 "input_activations": null,21 "output_activations": null,22 "targets": [23 "Linear"24 ],25 "weights": {26 "actorder": null,27 "block_structure": null,28 "dynamic": false,29 "group_size": 32,30 "num_bits": 4,31 "observer": "minmax",32 "observer_kwargs": {},33 "strategy": "group",34 "symmetric": true,35 "type": "int"36 }37 }38 },39 "format": "pack-quantized",40 "ignore": [41 "re:.*self_attn.*",42 "re:.*shared_experts.*",43 "re:.*mlp\\.(gate|up|gate_up|down)_proj.*",44 "re:.*lm_head.*",45 "re:.*vision_tower.*",46 "re:.*mm_projector.*"47 ],48 "kv_cache_scheme": null,49 "quant_method": "compressed-tensors",50 "quantization_status": "compressed"51 },52 "text_config": {53 "_name_or_path": "",54 "add_cross_attention": false,55 "architectures": [56 "DeepseekV3ForCausalLM"57 ],58 "attention_bias": false,59 "attention_dropout": 0.0,60 "auto_map": {61 "AutoConfig": "configuration_deepseek.DeepseekV3Config",62 "AutoModel": "modeling_deepseek.DeepseekV3Model",63 "AutoModelForCausalLM": "modeling_deepseek.DeepseekV3ForCausalLM"64 },65 "aux_loss_alpha": 0.001,66 "bos_token_id": 163584,67 "chunk_size_feed_forward": 0,68 "cross_attention_hidden_size": null,69 "decoder_start_token_id": null,70 "torch_dtype": "bfloat16",71 "eos_token_id": 163586,72 "ep_size": 1,73 "finetuning_task": null,74 "first_k_dense_replace": 1,75 "hidden_act": "silu",76 "hidden_size": 7168,77 "id2label": {78 "0": "LABEL_0",79 "1": "LABEL_1"80 },81 "initializer_range": 0.02,82 "intermediate_size": 18432,83 "is_decoder": false,84 "is_encoder_decoder": false,85 "kv_lora_rank": 512,86 "label2id": {87 "LABEL_0": 0,88 "LABEL_1": 189 },90 "max_position_embeddings": 262144,91 "model_type": "deepseek_v3",92 "moe_intermediate_size": 2048,93 "moe_layer_freq": 1,94 "n_group": 1,95 "n_routed_experts": 384,96 "n_shared_experts": 1,97 "norm_topk_prob": true,98 "num_attention_heads": 64,99 "num_experts_per_tok": 8,100 "num_hidden_layers": 61,101 "num_key_value_heads": 64,102 "num_nextn_predict_layers": 0,103 "output_attentions": false,104 "output_hidden_states": false,105 "pad_token_id": 163839,106 "prefix": null,107 "pretraining_tp": 1,108 "problem_type": null,109 "pruned_heads": {},110 "q_lora_rank": 1536,111 "qk_nope_head_dim": 128,112 "qk_rope_head_dim": 64,113 "quantization_config": {114 "config_groups": {115 "group_0": {116 "input_activations": null,117 "output_activations": null,118 "targets": [119 "Linear"120 ],121 "weights": {122 "actorder": null,123 "block_structure": null,124 "dynamic": false,125 "group_size": 32,126 "num_bits": 4,127 "observer": "minmax",128 "observer_kwargs": {},129 "strategy": "group",130 "symmetric": true,131 "type": "int"132 }133 }134 },135 "format": "pack-quantized",136 "ignore": [137 "re:.*self_attn.*",138 "re:.*shared_experts.*",139 "re:.*mlp\\.(gate|up|gate_up|down)_proj.*",140 "re:.*lm_head.*",141 "re:.*vision_tower.*",142 "re:.*mm_projector.*"143 ],144 "kv_cache_scheme": null,145 "quant_method": "compressed-tensors",146 "quantization_status": "compressed"147 },148 "return_dict": true,149 "rms_norm_eps": 1e-05,150 "rope_parameters": {151 "beta_fast": 32.0,152 "beta_slow": 1.0,153 "factor": 64.0,154 "mscale": 1.0,155 "mscale_all_dim": 1.0,156 "original_max_position_embeddings": 4096,157 "rope_theta": 50000.0,158 "rope_type": "yarn",159 "type": "yarn"160 },161 "rope_theta": 50000.0,162 "routed_scaling_factor": 2.827,163 "scoring_func": "sigmoid",164 "sep_token_id": null,165 "seq_aux": true,166 "task_specific_params": null,167 "tf_legacy_loss": false,168 "tie_encoder_decoder": false,169 "tie_word_embeddings": false,170 "tokenizer_class": null,171 "topk_group": 1,172 "topk_method": "noaux_tc",173 "torchscript": false,174 "use_bfloat16": false,175 "use_cache": true,176 "v_head_dim": 128,177 "vocab_size": 163840178 },179 "tie_word_embeddings": false,180 "transformers_version": "5.13.0.dev0",181 "unsloth_fixed": true,182 "use_unified_vision_chunk": true,183 "video_placeholder": "<|kimi_k25_video_placeholder|>",184 "vision_config": {185 "init_pos_emb_height": 64,186 "init_pos_emb_time": 4,187 "init_pos_emb_width": 64,188 "merge_kernel_size": [189 2,190 2191 ],192 "merge_type": "sd2_tpool",193 "mm_hidden_size": 1152,194 "mm_projector_type": "patchmerger",195 "model_type": "",196 "patch_size": 14,197 "pos_emb_type": "divided_fixed",198 "projector_hidden_act": "gelu",199 "projector_ln_eps": 1e-05,200 "text_hidden_size": 7168,201 "video_attn_type": "spatial_temporal",202 "vt_hidden_size": 1152,203 "vt_intermediate_size": 4304,204 "vt_num_attention_heads": 16,205 "vt_num_hidden_layers": 27206 }207}