Hugging Face's logo

LiquidAI
/
LFM2.5-VL-1.6B-ONNX

Image-Text-to-Text
ONNX
lfm2_vl
liquid
edge
lfm2.5-vl
lfm2.5
onnxruntime
webgpu
conversational
Model card Files
xet
 Community
2
ykhrustalev commited on
Commit
832f5ca
·
verified ·
1 Parent(s): 855e09f

Upload folder using huggingface_hub

Browse files
Files changed (26) hide show
  1. .gitattributes +9 -0
  2. chat_template.jinja +4 -0
  3. config.json +97 -0
  4. generation_config.json +7 -0
  5. onnx/decoder.onnx +3 -0
  6. onnx/decoder.onnx_data +3 -0
  7. onnx/decoder_fp16.onnx +3 -0
  8. onnx/decoder_fp16.onnx_data +3 -0
  9. onnx/decoder_q4.onnx +3 -0
  10. onnx/decoder_q4.onnx_data +3 -0
  11. onnx/decoder_q8.onnx +3 -0
  12. onnx/decoder_q8.onnx_data +3 -0
  13. onnx/embed_images.onnx +3 -0
  14. onnx/embed_images.onnx_data +3 -0
  15. onnx/embed_images_fp16.onnx +3 -0
  16. onnx/embed_images_fp16.onnx_data +3 -0
  17. onnx/embed_images_q4.onnx +3 -0
  18. onnx/embed_images_q4.onnx_data +3 -0
  19. onnx/embed_images_q8.onnx +3 -0
  20. onnx/embed_images_q8.onnx_data +3 -0
  21. onnx/embed_tokens.onnx +3 -0
  22. onnx/embed_tokens_fp16.onnx +3 -0
  23. onnx/embed_tokens_fp16.onnx_data +3 -0
  24. processor_config.json +49 -0
  25. tokenizer.json +0 -0
  26. tokenizer_config.json +22 -0
.gitattributes CHANGED
@@ -33,3 +33,12 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ onnx/decoder.onnx_data filter=lfs diff=lfs merge=lfs -text
37
+ onnx/decoder_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text
38
+ onnx/decoder_q4.onnx_data filter=lfs diff=lfs merge=lfs -text
39
+ onnx/decoder_q8.onnx_data filter=lfs diff=lfs merge=lfs -text
40
+ onnx/embed_images.onnx_data filter=lfs diff=lfs merge=lfs -text
41
+ onnx/embed_images_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text
42
+ onnx/embed_images_q4.onnx_data filter=lfs diff=lfs merge=lfs -text
43
+ onnx/embed_images_q8.onnx_data filter=lfs diff=lfs merge=lfs -text
44
+ onnx/embed_tokens_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text
chat_template.jinja ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {{bos_token}}{% for message in messages %}{{'<|im_start|>' + message['role'] + '
2
+ '}}{% if message['content'] is string %}{{ message['content'] }}{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' %}{{ '<image>' }}{% elif content['type'] == 'text' %}{{ content['text'] }}{% endif %}{% endfor %}{% endif %}{{'<|im_end|>
3
+ '}}{% endfor %}{% if add_generation_prompt %}{{'<|im_start|>assistant
4
+ ' }}{% endif %}
config.json ADDED
@@ -0,0 +1,97 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Lfm2VlForConditionalGeneration"
4
+ ],
5
+ "do_image_splitting": true,
6
+ "downsample_factor": 2,
7
+ "dtype": "bfloat16",
8
+ "encoder_patch_size": 16,
9
+ "image_token_id": 396,
10
+ "max_image_tokens": 256,
11
+ "max_pixels_tolerance": 2.0,
12
+ "max_tiles": 10,
13
+ "min_image_tokens": 64,
14
+ "min_tiles": 2,
15
+ "model_type": "lfm2_vl",
16
+ "projector_bias": true,
17
+ "projector_hidden_act": "gelu",
18
+ "projector_hidden_size": 2048,
19
+ "projector_use_layernorm": false,
20
+ "text_config": {
21
+ "_name_or_path": "LiquidAI/LFM2-1.2B",
22
+ "architectures": [
23
+ "Lfm2ForCausalLM"
24
+ ],
25
+ "block_auto_adjust_ff_dim": true,
26
+ "block_dim": 2048,
27
+ "block_ff_dim": 12288,
28
+ "block_ffn_dim_multiplier": 1.0,
29
+ "block_mlp_init_scale": 1.0,
30
+ "block_multiple_of": 256,
31
+ "block_norm_eps": 1e-05,
32
+ "block_out_init_scale": 1.0,
33
+ "block_use_swiglu": true,
34
+ "block_use_xavier_init": true,
35
+ "conv_L_cache": 3,
36
+ "conv_bias": false,
37
+ "conv_dim": 2048,
38
+ "conv_dim_out": 2048,
39
+ "conv_use_xavier_init": true,
40
+ "dtype": "bfloat16",
41
+ "eos_token_id": 7,
42
+ "hidden_size": 2048,
43
+ "initializer_range": 0.02,
44
+ "intermediate_size": 12288,
45
+ "layer_types": [
46
+ "conv",
47
+ "conv",
48
+ "full_attention",
49
+ "conv",
50
+ "conv",
51
+ "full_attention",
52
+ "conv",
53
+ "conv",
54
+ "full_attention",
55
+ "conv",
56
+ "full_attention",
57
+ "conv",
58
+ "full_attention",
59
+ "conv",
60
+ "full_attention",
61
+ "conv"
62
+ ],
63
+ "max_position_embeddings": 128000,
64
+ "model_type": "lfm2",
65
+ "norm_eps": 1e-05,
66
+ "num_attention_heads": 32,
67
+ "num_heads": 32,
68
+ "num_hidden_layers": 16,
69
+ "num_key_value_heads": 8,
70
+ "rope_parameters": {
71
+ "rope_theta": 1000000.0,
72
+ "rope_type": "default"
73
+ },
74
+ "use_cache": true,
75
+ "use_pos_enc": true,
76
+ "vocab_size": 65536
77
+ },
78
+ "tile_size": 512,
79
+ "transformers_version": "5.0.0.dev0",
80
+ "use_image_special_tokens": true,
81
+ "use_thumbnail": true,
82
+ "vision_config": {
83
+ "attention_dropout": 0.0,
84
+ "dtype": "bfloat16",
85
+ "hidden_act": "gelu_pytorch_tanh",
86
+ "hidden_size": 1152,
87
+ "intermediate_size": 4304,
88
+ "layer_norm_eps": 1e-06,
89
+ "model_type": "siglip2_vision_model",
90
+ "num_attention_heads": 16,
91
+ "num_channels": 3,
92
+ "num_hidden_layers": 27,
93
+ "num_patches": 256,
94
+ "patch_size": 16,
95
+ "vision_use_head": false
96
+ }
97
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 7,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.57.0"
7
+ }
onnx/decoder.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e17579934e9937b66151e0685bcd547a67516dd86884c378e3aa6bb9e5d10c4f
3
+ size 94129
onnx/decoder.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63acf182b47b3f03469489eb4c97a656dccedb756027b50243a1d0f20e7efa9b
3
+ size 4714127360
onnx/decoder_fp16.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:396dabb6865ff0b86c3e375c588d899b60db44bcd1c6fbb4cdd981d9b18d5c4d
3
+ size 100129
onnx/decoder_fp16.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:784e258e9b24f23b2a73147112fabe4fabf8592239dfb43dbcee5310b05793c0
3
+ size 2357063680
onnx/decoder_q4.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e1c94b89a28544c0d0d39d2ae3854696368e8cea13e2e77c898a7be9b28b6f30
3
+ size 121865
onnx/decoder_q4.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:144e30048e0de1af8b9524fc4bdca99286090ff92c0f12792ccedd31804f2d73
3
+ size 1217650688
onnx/decoder_q8.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31854b0d992594073bbcc572f0509ba35b9570498c25b300eeb91059bac4770a
3
+ size 137849
onnx/decoder_q8.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f6c8f6e2b088703d65a7a90727cc2d04fbcb343eaa1868c8e3ddec2f5e0d0ec
3
+ size 1768022016
onnx/embed_images.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edd03fbb991c739f93dece0cad07c9262bda11e87125d61ac14e7a04ee171cb3
3
+ size 174037
onnx/embed_images.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a29f768340b9d9755ebca6ac928058d2f7dcda4c688dfa214f5eea3cca7e687
3
+ size 1705141184
onnx/embed_images_fp16.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:759d9d72127c2426c2c1468a7b0a15effffe5c92e531b0d5b7b88fc68b48e6ba
3
+ size 176398
onnx/embed_images_fp16.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8054413051a026379524506513c30489dea5ba8830a8d5f7a1bb6fd3d6221041
3
+ size 852570592
onnx/embed_images_q4.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b9666bbdebd0705a643e6b24600c3e8587c292642f37a66262e4e91cfe6dd6b8
3
+ size 223748
onnx/embed_images_q4.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78ad0694364f0b9c0d3b3a5b0d2f57c402dc676e75aacdf79f67e02b1d458b25
3
+ size 269097152
onnx/embed_images_q8.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:56882806a6c70e4e44198f4188c1c73caa4b33f3dfd86199e437aa39d6826a06
3
+ size 252424
onnx/embed_images_q8.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8aaeea172f0fdab2b1ba893814d4c5f4ced62d4c3f15ec05b013526cbcd21ce2
3
+ size 495454464
onnx/embed_tokens.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ebeefbe5d4b82537a8f1d11a570579763a97f4e8b98156d6bb4e24856cf194d8
3
+ size 536871177
onnx/embed_tokens_fp16.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9447ff7bd0cd971d7bc7f5243b92cd557b1864c1acdb186f1bd4d600e71acff0
3
+ size 479
onnx/embed_tokens_fp16.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f7fe1eac2569a8be47a56bb721fdb8d98bc34ce5edc5fce073b5db7c0b1785d
3
+ size 268435456
processor_config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_processor": {
3
+ "crop_size": null,
4
+ "data_format": "channels_first",
5
+ "device": null,
6
+ "disable_grouping": null,
7
+ "do_center_crop": null,
8
+ "do_convert_rgb": null,
9
+ "do_image_splitting": true,
10
+ "do_normalize": true,
11
+ "do_pad": true,
12
+ "do_rescale": true,
13
+ "do_resize": true,
14
+ "downsample_factor": 2,
15
+ "encoder_patch_size": 16,
16
+ "image_mean": [
17
+ 0.5,
18
+ 0.5,
19
+ 0.5
20
+ ],
21
+ "image_processor_type": "Lfm2VlImageProcessorFast",
22
+ "image_seq_length": null,
23
+ "image_std": [
24
+ 0.5,
25
+ 0.5,
26
+ 0.5
27
+ ],
28
+ "input_data_format": null,
29
+ "max_image_tokens": 256,
30
+ "max_num_patches": 1024,
31
+ "max_pixels_tolerance": 2.0,
32
+ "max_tiles": 10,
33
+ "min_image_tokens": 64,
34
+ "min_tiles": 2,
35
+ "pad_size": null,
36
+ "processor_class": "Lfm2VlProcessor",
37
+ "resample": 2,
38
+ "rescale_factor": 0.00392156862745098,
39
+ "return_row_col_info": true,
40
+ "return_tensors": null,
41
+ "size": {
42
+ "height": 512,
43
+ "width": 512
44
+ },
45
+ "tile_size": 512,
46
+ "use_thumbnail": true
47
+ },
48
+ "processor_class": "Lfm2VlProcessor"
49
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": null,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|startoftext|>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "is_local": true,
8
+ "legacy": false,
9
+ "model_input_names": [
10
+ "input_ids",
11
+ "attention_mask"
12
+ ],
13
+ "model_max_length": 1000000000000000019884624838656,
14
+ "model_specific_special_tokens": {},
15
+ "pad_token": "<|pad|>",
16
+ "processor_class": "Lfm2VlProcessor",
17
+ "sp_model_kwargs": {},
18
+ "spaces_between_special_tokens": false,
19
+ "tokenizer_class": "TokenizersBackend",
20
+ "use_default_system_prompt": false,
21
+ "use_fast": true
22
+ }