Hugging Face's logo

humpf
/
LightOnOCR-4bit

Image-Text-to-Text
MLX
Safetensors
mistral3
ocr
document-understanding
vision-language
pdf
tables
forms
conversational
4-bit precision
Model card Files
xet
 Community
LightOnOCR-4bit
File size: 3,011 Bytes
1835135
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
{
    "architectures": [
        "LightOnOCRForConditionalGeneration"
    ],
    "dtype": "bfloat16",
    "eos_token_id": [
        151645,
        151643
    ],
    "image_token_id": 151655,
    "model_type": "mistral3",
    "multimodal_projector_bias": false,
    "pad_token_id": 151643,
    "projector_hidden_act": "gelu",
    "quantization": {
        "group_size": 64,
        "bits": 4,
        "mode": "affine"
    },
    "quantization_config": {
        "group_size": 64,
        "bits": 4,
        "mode": "affine"
    },
    "spatial_merge_size": 2,
    "text_config": {
        "architectures": [
            "Qwen3ForCausalLM"
        ],
        "attention_bias": false,
        "attention_dropout": 0,
        "dtype": "bfloat16",
        "head_dim": 128,
        "hidden_act": "silu",
        "hidden_size": 1024,
        "initializer_range": 0.02,
        "intermediate_size": 3072,
        "layer_types": [
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention",
            "full_attention"
        ],
        "max_position_embeddings": 16384,
        "max_window_layers": 28,
        "model_type": "qwen3",
        "num_attention_heads": 16,
        "num_hidden_layers": 28,
        "num_key_value_heads": 8,
        "rms_norm_eps": 1e-06,
        "rope_parameters": {
            "rope_theta": 1000000,
            "rope_type": "default"
        },
        "rope_theta": 1000000,
        "sliding_window": null,
        "use_cache": true,
        "use_sliding_window": false,
        "tie_word_embeddings": true,
        "vocab_size": 151936,
        "use_qk_norm": true
    },
    "transformers_version": "5.0.0.dev0",
    "use_cache": false,
    "vision_config": {
        "attention_dropout": 0,
        "dtype": "bfloat16",
        "head_dim": 64,
        "hidden_act": "silu",
        "hidden_size": 1024,
        "image_size": 1540,
        "initializer_range": 0.02,
        "intermediate_size": 4096,
        "model_type": "pixtral",
        "num_attention_heads": 16,
        "num_channels": 3,
        "num_hidden_layers": 24,
        "patch_size": 14,
        "rope_parameters": {
            "rope_theta": 10000,
            "rope_type": "default"
        },
        "rope_theta": 10000
    },
    "vision_feature_layer": -1
}