G.D.2. commited on
Commit
ee5cf17
·
verified ·
1 Parent(s): 63da969

Upload 2 files

Browse files
Files changed (2) hide show
  1. config.json +71 -0
  2. mergekit_moe_config.yml +14 -0
config.json ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3MoeForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": null,
8
+ "decoder_sparse_step": 1,
9
+ "dtype": "float16",
10
+ "eos_token_id": 151645,
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 1024,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention"
45
+ ],
46
+ "max_position_embeddings": 40960,
47
+ "max_window_layers": 28,
48
+ "mlp_only_layers": [],
49
+ "model_type": "qwen3_moe",
50
+ "moe_intermediate_size": 3072,
51
+ "norm_topk_prob": true,
52
+ "num_attention_heads": 16,
53
+ "num_experts_per_tok": 2,
54
+ "num_hidden_layers": 28,
55
+ "num_key_value_heads": 8,
56
+ "num_local_experts": 8,
57
+ "output_router_logits": false,
58
+ "pad_token_id": 151654,
59
+ "rms_norm_eps": 1e-06,
60
+ "rope_parameters": {
61
+ "rope_theta": 1000000,
62
+ "rope_type": "default"
63
+ },
64
+ "router_aux_loss_coef": 0.001,
65
+ "sliding_window": null,
66
+ "tie_word_embeddings": true,
67
+ "transformers_version": "5.1.0",
68
+ "use_cache": true,
69
+ "use_sliding_window": false,
70
+ "vocab_size": 151936
71
+ }
mergekit_moe_config.yml ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ base_model: C:\Users\GSS1147\Desktop\8\Qwen3-0.6B-rStar-Coder-IF-Expert
2
+ experts:
3
+ - source_model: C:\Users\GSS1147\Desktop\8\bigatuna-Qwen3-0.6B-Sushi-Coder
4
+ - source_model: C:\Users\GSS1147\Desktop\8\prithivMLmods-rStar-Coder-Qwen3-0.6B
5
+ - source_model: C:\Users\GSS1147\Desktop\8\Qwen3-0.6B-Qrazy-Qoder
6
+ - source_model: C:\Users\GSS1147\Desktop\8\Qwen3-0.6B-rStar-Coder-IF-Expert
7
+ - source_model: C:\Users\GSS1147\Desktop\8\Qwen3-0.6B-Sushi-Code-Expert
8
+ - source_model: C:\Users\GSS1147\Desktop\8\Qwen3-0.6B-Sushi-Math-Code-Expert
9
+ - source_model: C:\Users\GSS1147\Desktop\8\suayptalha-Qwen3-0.6B-Code-Expert
10
+ - source_model: C:\Users\GSS1147\Desktop\8\suayptalha-Qwen3-0.6B-IF-Expert
11
+ gate_mode: random
12
+ dtype: float16
13
+ experts_per_token: 2
14
+ architecture: Qwen3 MoE