Upload config.json with huggingface_hub
Browse files- config.json +109 -0
config.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3NextForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"decoder_sparse_step": 1,
|
| 9 |
+
"dtype": "float16",
|
| 10 |
+
"eos_token_id": 151645,
|
| 11 |
+
"full_attention_interval": 4,
|
| 12 |
+
"head_dim": 256,
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 2048,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 5120,
|
| 17 |
+
"layer_types": [
|
| 18 |
+
"linear_attention",
|
| 19 |
+
"linear_attention",
|
| 20 |
+
"linear_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"linear_attention",
|
| 23 |
+
"linear_attention",
|
| 24 |
+
"linear_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"linear_attention",
|
| 27 |
+
"linear_attention",
|
| 28 |
+
"linear_attention",
|
| 29 |
+
"full_attention",
|
| 30 |
+
"linear_attention",
|
| 31 |
+
"linear_attention",
|
| 32 |
+
"linear_attention",
|
| 33 |
+
"full_attention",
|
| 34 |
+
"linear_attention",
|
| 35 |
+
"linear_attention",
|
| 36 |
+
"linear_attention",
|
| 37 |
+
"full_attention",
|
| 38 |
+
"linear_attention",
|
| 39 |
+
"linear_attention",
|
| 40 |
+
"linear_attention",
|
| 41 |
+
"full_attention",
|
| 42 |
+
"linear_attention",
|
| 43 |
+
"linear_attention",
|
| 44 |
+
"linear_attention",
|
| 45 |
+
"full_attention",
|
| 46 |
+
"linear_attention",
|
| 47 |
+
"linear_attention",
|
| 48 |
+
"linear_attention",
|
| 49 |
+
"full_attention",
|
| 50 |
+
"linear_attention",
|
| 51 |
+
"linear_attention",
|
| 52 |
+
"linear_attention",
|
| 53 |
+
"full_attention",
|
| 54 |
+
"linear_attention",
|
| 55 |
+
"linear_attention",
|
| 56 |
+
"linear_attention",
|
| 57 |
+
"full_attention",
|
| 58 |
+
"linear_attention",
|
| 59 |
+
"linear_attention",
|
| 60 |
+
"linear_attention",
|
| 61 |
+
"full_attention",
|
| 62 |
+
"linear_attention",
|
| 63 |
+
"linear_attention",
|
| 64 |
+
"linear_attention",
|
| 65 |
+
"full_attention"
|
| 66 |
+
],
|
| 67 |
+
"linear_conv_kernel_dim": 4,
|
| 68 |
+
"linear_key_head_dim": 128,
|
| 69 |
+
"linear_num_key_heads": 16,
|
| 70 |
+
"linear_num_value_heads": 32,
|
| 71 |
+
"linear_value_head_dim": 128,
|
| 72 |
+
"max_position_embeddings": 262144,
|
| 73 |
+
"mlp_only_layers": [],
|
| 74 |
+
"model_type": "qwen3_next",
|
| 75 |
+
"moe_intermediate_size": 512,
|
| 76 |
+
"norm_topk_prob": true,
|
| 77 |
+
"num_attention_heads": 16,
|
| 78 |
+
"num_experts": 512,
|
| 79 |
+
"num_experts_per_tok": 10,
|
| 80 |
+
"num_hidden_layers": 48,
|
| 81 |
+
"num_key_value_heads": 2,
|
| 82 |
+
"output_router_logits": false,
|
| 83 |
+
"partial_rotary_factor": 0.25,
|
| 84 |
+
"quantization_config": {
|
| 85 |
+
"_load_in_4bit": false,
|
| 86 |
+
"_load_in_8bit": true,
|
| 87 |
+
"bnb_4bit_compute_dtype": "float32",
|
| 88 |
+
"bnb_4bit_quant_storage": "uint8",
|
| 89 |
+
"bnb_4bit_quant_type": "fp4",
|
| 90 |
+
"bnb_4bit_use_double_quant": false,
|
| 91 |
+
"llm_int8_enable_fp32_cpu_offload": true,
|
| 92 |
+
"llm_int8_has_fp16_weight": false,
|
| 93 |
+
"llm_int8_skip_modules": null,
|
| 94 |
+
"llm_int8_threshold": 6.0,
|
| 95 |
+
"load_in_4bit": false,
|
| 96 |
+
"load_in_8bit": true,
|
| 97 |
+
"quant_method": "bitsandbytes"
|
| 98 |
+
},
|
| 99 |
+
"rms_norm_eps": 1e-06,
|
| 100 |
+
"rope_scaling": null,
|
| 101 |
+
"rope_theta": 10000000,
|
| 102 |
+
"router_aux_loss_coef": 0.001,
|
| 103 |
+
"shared_expert_intermediate_size": 512,
|
| 104 |
+
"tie_word_embeddings": false,
|
| 105 |
+
"transformers_version": "4.57.0.dev0",
|
| 106 |
+
"use_cache": true,
|
| 107 |
+
"use_sliding_window": false,
|
| 108 |
+
"vocab_size": 151936
|
| 109 |
+
}
|