Hugging Face's logo

kvfs
/
NFA-Panel-AI

GGUF
qwen2_5_vl
llama.cpp
unsloth
vision-language-model
conversational
Model card Files
xet
 Community
kvfs commited on
Commit
60b176c
·
verified ·
1 Parent(s): 3ec5fc3

Trained with Unsloth - config

Browse files
Files changed (1) hide show
  1. config.json +115 -0
config.json ADDED
@@ -0,0 +1,115 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2_5_VLForConditionalGeneration"
4
+ ],
5
+ "bos_token_id": null,
6
+ "torch_dtype": "float16",
7
+ "eos_token_id": 151645,
8
+ "image_token_id": 151655,
9
+ "model_name": "unsloth/Qwen2.5-VL-7B-Instruct-bnb-4bit",
10
+ "model_type": "qwen2_5_vl",
11
+ "pad_token_id": 151654,
12
+ "text_config": {
13
+ "architectures": [
14
+ "Qwen2_5_VLForConditionalGeneration"
15
+ ],
16
+ "attention_dropout": 0.0,
17
+ "bos_token_id": 151643,
18
+ "torch_dtype": "float16",
19
+ "eos_token_id": 151645,
20
+ "hidden_act": "silu",
21
+ "hidden_size": 3584,
22
+ "image_token_id": null,
23
+ "initializer_range": 0.02,
24
+ "intermediate_size": 18944,
25
+ "layer_types": [
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
+ "full_attention",
52
+ "full_attention",
53
+ "full_attention"
54
+ ],
55
+ "max_position_embeddings": 128000,
56
+ "max_window_layers": 28,
57
+ "model_type": "qwen2_5_vl_text",
58
+ "num_attention_heads": 28,
59
+ "num_hidden_layers": 28,
60
+ "num_key_value_heads": 4,
61
+ "pad_token_id": null,
62
+ "rms_norm_eps": 1e-06,
63
+ "rope_parameters": {
64
+ "mrope_section": [
65
+ 16,
66
+ 24,
67
+ 24
68
+ ],
69
+ "rope_theta": 1000000.0,
70
+ "rope_type": "default",
71
+ "type": "default"
72
+ },
73
+ "sliding_window": null,
74
+ "use_cache": true,
75
+ "use_sliding_window": false,
76
+ "video_token_id": null,
77
+ "vision_end_token_id": 151653,
78
+ "vision_start_token_id": 151652,
79
+ "vision_token_id": 151654,
80
+ "vocab_size": 152064
81
+ },
82
+ "tie_word_embeddings": false,
83
+ "unsloth_fixed": true,
84
+ "unsloth_version": "2026.6.9",
85
+ "use_cache": false,
86
+ "video_token_id": 151656,
87
+ "vision_config": {
88
+ "depth": 32,
89
+ "torch_dtype": "float16",
90
+ "fullatt_block_indexes": [
91
+ 7,
92
+ 15,
93
+ 23,
94
+ 31
95
+ ],
96
+ "hidden_act": "silu",
97
+ "hidden_size": 1280,
98
+ "in_channels": 3,
99
+ "in_chans": 3,
100
+ "initializer_range": 0.02,
101
+ "intermediate_size": 3420,
102
+ "model_type": "qwen2_5_vl_vision",
103
+ "num_heads": 16,
104
+ "out_hidden_size": 3584,
105
+ "patch_size": 14,
106
+ "spatial_merge_size": 2,
107
+ "spatial_patch_size": 14,
108
+ "temporal_patch_size": 2,
109
+ "tokens_per_second": 2,
110
+ "window_size": 112
111
+ },
112
+ "vision_end_token_id": 151653,
113
+ "vision_start_token_id": 151652,
114
+ "vision_token_id": 151654
115
+ }