hf-transformers-bot's picture
Update tiny models for Kimi_K25Model
a43f6cf verified
Raw
History Blame Contribute Delete
1.99 kB
{
"architectures": [
"Kimi_K25Model"
],
"dtype": "float32",
"image_token_id": 3,
"model_type": "kimi_k25",
"pad_token_id": 163606,
"projection_hidden_size": 32,
"projection_layer_norm_eps": 1e-05,
"text_config": {
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 163584,
"eos_token_id": 163585,
"first_k_dense_replace": 1,
"head_dim": 16,
"hidden_act": "silu",
"hidden_size": 32,
"initializer_range": 0.02,
"intermediate_size": 32,
"kv_lora_rank": 16,
"max_position_embeddings": 512,
"model_type": "deepseek_v3",
"moe_intermediate_size": 16,
"n_group": 2,
"n_routed_experts": 8,
"n_shared_experts": 1,
"norm_topk_prob": true,
"num_attention_heads": 2,
"num_experts_per_tok": 8,
"num_hidden_layers": 2,
"num_key_value_heads": 2,
"num_nextn_predict_layers": 1,
"pad_token_id": 163606,
"pretraining_tp": 1,
"q_lora_rank": 32,
"qk_head_dim": 48,
"qk_nope_head_dim": 32,
"qk_rope_head_dim": 16,
"rms_norm_eps": 1e-06,
"rope_interleave": true,
"rope_parameters": {
"rope_theta": 10000,
"rope_type": "default"
},
"routed_scaling_factor": 2.5,
"tie_word_embeddings": false,
"topk_group": 1,
"use_cache": true,
"v_head_dim": 32,
"vocab_size": 163607
},
"tie_word_embeddings": false,
"transformers_version": "5.16.0.dev0",
"video_token_id": 4,
"vision_config": {
"hidden_act": "silu",
"hidden_size": 32,
"intermediate_size": 32,
"max_position_embeddings": 512,
"merge_kernel_size": [
1,
1
],
"model_type": "kimi_k25_vision",
"num_attention_heads": 2,
"num_hidden_layers": 2,
"patch_size": 8,
"pos_emb_height": 4,
"pos_emb_time": 1,
"pos_emb_width": 4,
"rope_parameters": {
"rope_theta": 10000,
"rope_type": "default"
}
},
"vision_end_token_id": 163604,
"vision_start_token_id": 163602
}