Hugging Face's logo

fklska
/
school-answer

Safetensors
GGUF
qwen3
llama.cpp
unsloth
8-bit precision
Model card Files
xet
 Community
fklska commited on
Commit
cca0006
·
verified ·
1 Parent(s): 8be1884

Trained with Unsloth - config

Browse files
Files changed (1) hide show
  1. config.json +67 -115
config.json CHANGED
@@ -1,117 +1,69 @@
1
  {
2
- "architectures": [
3
- "Qwen3ForCausalLM"
4
- ],
5
- "attention_bias": false,
6
- "attention_dropout": 0.0,
7
- "torch_dtype": "float16",
8
- "eos_token_id": 151645,
9
- "head_dim": 128,
10
- "hidden_act": "silu",
11
- "hidden_size": 2560,
12
- "initializer_range": 0.02,
13
- "intermediate_size": 9728,
14
- "layer_types": [
15
- "full_attention",
16
- "full_attention",
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
- "full_attention",
46
- "full_attention",
47
- "full_attention",
48
- "full_attention",
49
- "full_attention",
50
- "full_attention"
51
- ],
52
- "max_position_embeddings": 262144,
53
- "max_window_layers": 36,
54
- "model_type": "qwen3",
55
- "num_attention_heads": 32,
56
- "num_hidden_layers": 36,
57
- "num_key_value_heads": 8,
58
- "pad_token_id": 151654,
59
- "quantization_config": {
60
- "bnb_4bit_compute_dtype": "float16",
61
- "bnb_4bit_quant_type": "nf4",
62
- "bnb_4bit_use_double_quant": true,
63
- "llm_int8_enable_fp32_cpu_offload": false,
64
- "llm_int8_has_fp16_weight": false,
65
- "llm_int8_skip_modules": [
66
- "model.layers.0.self_attn.q_proj",
67
- "model.layers.0.self_attn.k_proj",
68
- "model.layers.0.self_attn.v_proj",
69
- "model.layers.0.self_attn.o_proj",
70
- "model.layers.1.mlp.gate_proj",
71
- "model.layers.1.mlp.up_proj",
72
- "model.layers.1.mlp.down_proj",
73
- "model.layers.2.mlp.gate_proj",
74
- "model.layers.2.mlp.up_proj",
75
- "model.layers.2.mlp.down_proj",
76
- "model.layers.3.self_attn.q_proj",
77
- "model.layers.3.self_attn.k_proj",
78
- "model.layers.3.self_attn.v_proj",
79
- "model.layers.3.self_attn.o_proj",
80
- "model.layers.3.mlp.gate_proj",
81
- "model.layers.3.mlp.up_proj",
82
- "model.layers.3.mlp.down_proj",
83
- "model.layers.5.mlp.gate_proj",
84
- "model.layers.5.mlp.up_proj",
85
- "model.layers.5.mlp.down_proj",
86
- "model.layers.6.self_attn.q_proj",
87
- "model.layers.6.self_attn.k_proj",
88
- "model.layers.6.self_attn.v_proj",
89
- "model.layers.6.self_attn.o_proj",
90
- "model.layers.6.mlp.gate_proj",
91
- "model.layers.6.mlp.up_proj",
92
- "model.layers.6.mlp.down_proj",
93
- "model.layers.34.mlp.gate_proj",
94
- "model.layers.34.mlp.up_proj",
95
- "model.layers.34.mlp.down_proj",
96
- "model.layers.35.mlp.gate_proj",
97
- "model.layers.35.mlp.up_proj",
98
- "model.layers.35.mlp.down_proj",
99
- "lm_head"
100
  ],
101
- "llm_int8_threshold": 6.0,
102
- "load_in_4bit": true,
103
- "load_in_8bit": false,
104
- "quant_method": "bitsandbytes"
105
- },
106
- "rms_norm_eps": 1e-06,
107
- "rope_scaling": null,
108
- "rope_theta": 5000000,
109
- "sliding_window": null,
110
- "tie_word_embeddings": true,
111
- "transformers_version": "4.56.2",
112
- "unsloth_fixed": true,
113
- "unsloth_version": "2026.5.8",
114
- "use_cache": true,
115
- "use_sliding_window": false,
116
- "vocab_size": 151936
117
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  {
2
+ "architectures": [
3
+ "Qwen3ForCausalLM"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4
  ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "torch_dtype": "float16",
8
+ "eos_token_id": 151645,
9
+ "head_dim": 128,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 2560,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 9728,
14
+ "layer_types": [
15
+ "full_attention",
16
+ "full_attention",
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
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention"
51
+ ],
52
+ "max_position_embeddings": 262144,
53
+ "max_window_layers": 36,
54
+ "model_type": "qwen3",
55
+ "num_attention_heads": 32,
56
+ "num_hidden_layers": 36,
57
+ "num_key_value_heads": 8,
58
+ "pad_token_id": 151654,
59
+ "rms_norm_eps": 1e-06,
60
+ "rope_scaling": null,
61
+ "rope_theta": 5000000,
62
+ "sliding_window": null,
63
+ "tie_word_embeddings": true,
64
+ "unsloth_fixed": true,
65
+ "unsloth_version": "2026.5.8",
66
+ "use_cache": true,
67
+ "use_sliding_window": false,
68
+ "vocab_size": 151936
69
+ }