Hugging Face's logo

Hack337
/
BuildGemma4-GGUF

GGUF
gemma4
llama.cpp
unsloth
vision-language-model
conversational
Model card Files
xet
 Community
Hack337 commited on
Commit
21ca3fd
·
verified ·
1 Parent(s): 9be2206

Trained with Unsloth - config

Browse files
Files changed (1) hide show
  1. config.json +146 -0
config.json ADDED
@@ -0,0 +1,146 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma4ForConditionalGeneration"
4
+ ],
5
+ "audio_config": null,
6
+ "audio_token_id": 258881,
7
+ "boa_token_id": 256000,
8
+ "boi_token_id": 255999,
9
+ "torch_dtype": "bfloat16",
10
+ "eoa_token_id": 258883,
11
+ "eoa_token_index": 258883,
12
+ "eoi_token_id": 258882,
13
+ "eos_token_id": 106,
14
+ "image_token_id": 258880,
15
+ "initializer_range": 0.02,
16
+ "model_name": "unsloth/gemma-4-26b-a4b-it",
17
+ "model_type": "gemma4",
18
+ "pad_token_id": 0,
19
+ "text_config": {
20
+ "attention_bias": false,
21
+ "attention_dropout": 0.0,
22
+ "attention_k_eq_v": true,
23
+ "bos_token_id": 2,
24
+ "torch_dtype": "bfloat16",
25
+ "enable_moe_block": true,
26
+ "eos_token_id": 1,
27
+ "final_logit_softcapping": 30.0,
28
+ "global_head_dim": 512,
29
+ "head_dim": 256,
30
+ "hidden_activation": "gelu_pytorch_tanh",
31
+ "hidden_size": 2816,
32
+ "hidden_size_per_layer_input": 0,
33
+ "initializer_range": 0.02,
34
+ "intermediate_size": 2112,
35
+ "layer_types": [
36
+ "sliding_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "sliding_attention",
40
+ "sliding_attention",
41
+ "full_attention",
42
+ "sliding_attention",
43
+ "sliding_attention",
44
+ "sliding_attention",
45
+ "sliding_attention",
46
+ "sliding_attention",
47
+ "full_attention",
48
+ "sliding_attention",
49
+ "sliding_attention",
50
+ "sliding_attention",
51
+ "sliding_attention",
52
+ "sliding_attention",
53
+ "full_attention",
54
+ "sliding_attention",
55
+ "sliding_attention",
56
+ "sliding_attention",
57
+ "sliding_attention",
58
+ "sliding_attention",
59
+ "full_attention",
60
+ "sliding_attention",
61
+ "sliding_attention",
62
+ "sliding_attention",
63
+ "sliding_attention",
64
+ "sliding_attention",
65
+ "full_attention"
66
+ ],
67
+ "max_position_embeddings": 262144,
68
+ "model_type": "gemma4_text",
69
+ "moe_intermediate_size": 704,
70
+ "num_attention_heads": 16,
71
+ "num_experts": 128,
72
+ "num_global_key_value_heads": 2,
73
+ "num_hidden_layers": 30,
74
+ "num_key_value_heads": 8,
75
+ "num_kv_shared_layers": 0,
76
+ "pad_token_id": 0,
77
+ "rms_norm_eps": 1e-06,
78
+ "rope_parameters": {
79
+ "full_attention": {
80
+ "partial_rotary_factor": 0.25,
81
+ "rope_theta": 1000000.0,
82
+ "rope_type": "proportional"
83
+ },
84
+ "sliding_attention": {
85
+ "rope_theta": 10000.0,
86
+ "rope_type": "default"
87
+ }
88
+ },
89
+ "sliding_window": 1024,
90
+ "tie_word_embeddings": true,
91
+ "top_k_experts": 8,
92
+ "use_bidirectional_attention": "vision",
93
+ "use_cache": true,
94
+ "use_double_wide_mlp": false,
95
+ "vocab_size": 262144,
96
+ "vocab_size_per_layer_input": 262144
97
+ },
98
+ "tie_word_embeddings": true,
99
+ "unsloth_fixed": true,
100
+ "unsloth_version": "2026.5.9",
101
+ "video_token_id": 258884,
102
+ "vision_config": {
103
+ "_name_or_path": "",
104
+ "architectures": null,
105
+ "attention_bias": false,
106
+ "attention_dropout": 0.0,
107
+ "chunk_size_feed_forward": 0,
108
+ "default_output_length": 280,
109
+ "torch_dtype": "bfloat16",
110
+ "global_head_dim": 72,
111
+ "head_dim": 72,
112
+ "hidden_activation": "gelu_pytorch_tanh",
113
+ "hidden_size": 1152,
114
+ "id2label": {
115
+ "0": "LABEL_0",
116
+ "1": "LABEL_1"
117
+ },
118
+ "initializer_range": 0.02,
119
+ "intermediate_size": 4304,
120
+ "is_encoder_decoder": false,
121
+ "label2id": {
122
+ "LABEL_0": 0,
123
+ "LABEL_1": 1
124
+ },
125
+ "max_position_embeddings": 131072,
126
+ "model_type": "gemma4_vision",
127
+ "num_attention_heads": 16,
128
+ "num_hidden_layers": 27,
129
+ "num_key_value_heads": 16,
130
+ "output_attentions": false,
131
+ "output_hidden_states": false,
132
+ "patch_size": 16,
133
+ "pooling_kernel_size": 3,
134
+ "position_embedding_size": 10240,
135
+ "problem_type": null,
136
+ "return_dict": true,
137
+ "rms_norm_eps": 1e-06,
138
+ "rope_parameters": {
139
+ "rope_theta": 100.0,
140
+ "rope_type": "default"
141
+ },
142
+ "standardize": true,
143
+ "use_clipped_linears": false
144
+ },
145
+ "vision_soft_tokens_per_image": 280
146
+ }