AutoDecompiler commited on
Commit
dab0ade
·
verified ·
1 Parent(s): a89cf7b

Delete config.json

Browse files
Files changed (1) hide show
  1. config.json +0 -41
config.json DELETED
@@ -1,41 +0,0 @@
1
- {
2
- "architectures": [
3
- "Qwen3MoeForCausalLM"
4
- ],
5
- "attention_bias": false,
6
- "attention_dropout": 0.0,
7
- "decoder_sparse_step": 1,
8
- "dtype": "bfloat16",
9
- "eos_token_id": 151645,
10
- "head_dim": 128,
11
- "hidden_act": "silu",
12
- "hidden_size": 2048,
13
- "initializer_range": 0.02,
14
- "intermediate_size": 5472,
15
- "max_position_embeddings": 262144,
16
- "max_window_layers": 28,
17
- "mlp_only_layers": [],
18
- "model_type": "qwen3_moe",
19
- "moe_intermediate_size": 768,
20
- "norm_topk_prob": true,
21
- "num_attention_heads": 32,
22
- "num_experts": 128,
23
- "num_experts_per_tok": 8,
24
- "num_hidden_layers": 48,
25
- "num_key_value_heads": 4,
26
- "output_router_logits": false,
27
- "pad_token_id": 151643,
28
- "qkv_bias": false,
29
- "rms_norm_eps": 1e-06,
30
- "rope_scaling": null,
31
- "rope_theta": 10000000,
32
- "router_aux_loss_coef": 0.0,
33
- "shared_expert_intermediate_size": 0,
34
- "sliding_window": null,
35
- "tie_word_embeddings": false,
36
- "transformers_version": "4.57.3",
37
- "use_cache": false,
38
- "use_qk_norm": true,
39
- "use_sliding_window": false,
40
- "vocab_size": 151936
41
- }