Hugging Face's logo

mlx-community
/
LFM2.5-VL-1.6B-bf16

Image-Text-to-Text
Transformers
Safetensors
MLX
lfm2_vl
liquid
lfm2
lfm2-vl
edge
lfm2.5-vl
lfm2.5
conversational
Model card Files
xet
 Community
prince-canuma commited on
Commit
16a710c
·
verified ·
1 Parent(s): 5acdd7d

Upload folder using huggingface_hub

Browse files
Files changed (9) hide show
  1. README.md +38 -0
  2. chat_template.jinja +55 -0
  3. config.json +98 -0
  4. generation_config.json +7 -0
  5. model.safetensors +3 -0
  6. model.safetensors.index.json +596 -0
  7. processor_config.json +49 -0
  8. tokenizer.json +0 -0
  9. tokenizer_config.json +29 -0
README.md ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: other
4
+ license_name: lfm1.0
5
+ license_link: LICENSE
6
+ language:
7
+ - en
8
+ - ja
9
+ - ko
10
+ - fr
11
+ - es
12
+ - de
13
+ - ar
14
+ - zh
15
+ pipeline_tag: image-text-to-text
16
+ tags:
17
+ - liquid
18
+ - lfm2
19
+ - lfm2-vl
20
+ - edge
21
+ - lfm2.5-vl
22
+ - lfm2.5
23
+ - mlx
24
+ base_model: LiquidAI/LFM2.5-1.2B-Base
25
+ ---
26
+
27
+ # mlx-community/LFM2.5-VL-1.6B-bf16
28
+ This model was converted to MLX format from [`LiquidAI/LFM2.5-VL-1.6B`]() using mlx-vlm version **0.3.10**.
29
+ Refer to the [original model card](https://huggingface.co/LiquidAI/LFM2.5-VL-1.6B) for more details on the model.
30
+ ## Use with mlx
31
+
32
+ ```bash
33
+ pip install -U mlx-vlm
34
+ ```
35
+
36
+ ```bash
37
+ python -m mlx_vlm.generate --model mlx-community/LFM2.5-VL-1.6B-bf16 --max-tokens 100 --temperature 0.0 --prompt "Describe this image." --image <path_to_image>
38
+ ```
chat_template.jinja ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {{- bos_token -}}
2
+ {%- set keep_past_thinking = keep_past_thinking | default(false) -%}
3
+ {%- set ns = namespace(system_prompt="") -%}
4
+ {%- if messages[0]["role"] == "system" -%}
5
+ {%- set ns.system_prompt = messages[0]["content"] -%}
6
+ {%- set messages = messages[1:] -%}
7
+ {%- endif -%}
8
+ {%- if tools -%}
9
+ {%- set ns.system_prompt = ns.system_prompt + ("\n" if ns.system_prompt else "") + "List of tools: [" -%}
10
+ {%- for tool in tools -%}
11
+ {%- if tool is not string -%}
12
+ {%- set tool = tool | tojson -%}
13
+ {%- endif -%}
14
+ {%- set ns.system_prompt = ns.system_prompt + tool -%}
15
+ {%- if not loop.last -%}
16
+ {%- set ns.system_prompt = ns.system_prompt + ", " -%}
17
+ {%- endif -%}
18
+ {%- endfor -%}
19
+ {%- set ns.system_prompt = ns.system_prompt + "]" -%}
20
+ {%- endif -%}
21
+ {%- if ns.system_prompt -%}
22
+ {{- "<|im_start|>system\n" + ns.system_prompt + "<|im_end|>\n" -}}
23
+ {%- endif -%}
24
+ {%- set ns.last_assistant_index = -1 -%}
25
+ {%- for message in messages -%}
26
+ {%- if message["role"] == "assistant" -%}
27
+ {%- set ns.last_assistant_index = loop.index0 -%}
28
+ {%- endif -%}
29
+ {%- endfor -%}
30
+ {%- for message in messages -%}
31
+ {{- "<|im_start|>" + message["role"] + "\n" -}}
32
+ {%- set content = message["content"] -%}
33
+ {%- if content is not string -%}
34
+ {%- set ns.content = "" -%}
35
+ {%- for item in content -%}
36
+ {%- if item["type"] == "image" -%}
37
+ {%- set ns.content = ns.content + "<image>" -%}
38
+ {%- elif item["type"] == "text" -%}
39
+ {%- set ns.content = ns.content + item["text"] -%}
40
+ {%- else -%}
41
+ {%- set ns.content = ns.content + item | tojson -%}
42
+ {%- endif -%}
43
+ {%- endfor -%}
44
+ {%- set content = ns.content -%}
45
+ {%- endif -%}
46
+ {%- if message["role"] == "assistant" and not keep_past_thinking and loop.index0 != ns.last_assistant_index -%}
47
+ {%- if "</think>" in content -%}
48
+ {%- set content = content.split("</think>")[-1] | trim -%}
49
+ {%- endif -%}
50
+ {%- endif -%}
51
+ {{- content + "<|im_end|>\n" -}}
52
+ {%- endfor -%}
53
+ {%- if add_generation_prompt -%}
54
+ {{- "<|im_start|>assistant\n" -}}
55
+ {%- endif -%}
config.json ADDED
@@ -0,0 +1,98 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Lfm2VlForConditionalGeneration"
4
+ ],
5
+ "do_image_splitting": true,
6
+ "downsample_factor": 2,
7
+ "dtype": "bfloat16",
8
+ "encoder_patch_size": 16,
9
+ "eos_token_id": 7,
10
+ "image_token_id": 396,
11
+ "max_image_tokens": 256,
12
+ "max_pixels_tolerance": 2.0,
13
+ "max_tiles": 10,
14
+ "min_image_tokens": 64,
15
+ "min_tiles": 2,
16
+ "model_type": "lfm2_vl",
17
+ "projector_bias": true,
18
+ "projector_hidden_act": "gelu",
19
+ "projector_hidden_size": 2048,
20
+ "projector_use_layernorm": false,
21
+ "text_config": {
22
+ "_name_or_path": "LiquidAI/LFM2-1.2B",
23
+ "architectures": [
24
+ "Lfm2ForCausalLM"
25
+ ],
26
+ "block_auto_adjust_ff_dim": true,
27
+ "block_dim": 2048,
28
+ "block_ff_dim": 12288,
29
+ "block_ffn_dim_multiplier": 1.0,
30
+ "block_mlp_init_scale": 1.0,
31
+ "block_multiple_of": 256,
32
+ "block_norm_eps": 1e-05,
33
+ "block_out_init_scale": 1.0,
34
+ "block_use_swiglu": true,
35
+ "block_use_xavier_init": true,
36
+ "conv_L_cache": 3,
37
+ "conv_bias": false,
38
+ "conv_dim": 2048,
39
+ "conv_dim_out": 2048,
40
+ "conv_use_xavier_init": true,
41
+ "dtype": "bfloat16",
42
+ "eos_token_id": 7,
43
+ "hidden_size": 2048,
44
+ "initializer_range": 0.02,
45
+ "intermediate_size": 12288,
46
+ "layer_types": [
47
+ "conv",
48
+ "conv",
49
+ "full_attention",
50
+ "conv",
51
+ "conv",
52
+ "full_attention",
53
+ "conv",
54
+ "conv",
55
+ "full_attention",
56
+ "conv",
57
+ "full_attention",
58
+ "conv",
59
+ "full_attention",
60
+ "conv",
61
+ "full_attention",
62
+ "conv"
63
+ ],
64
+ "max_position_embeddings": 128000,
65
+ "model_type": "lfm2",
66
+ "norm_eps": 1e-05,
67
+ "num_attention_heads": 32,
68
+ "num_heads": 32,
69
+ "num_hidden_layers": 16,
70
+ "num_key_value_heads": 8,
71
+ "rope_parameters": {
72
+ "rope_theta": 1000000.0,
73
+ "rope_type": "default"
74
+ },
75
+ "use_cache": true,
76
+ "use_pos_enc": true,
77
+ "vocab_size": 65536
78
+ },
79
+ "tile_size": 512,
80
+ "transformers_version": "5.0.0.dev0",
81
+ "use_image_special_tokens": true,
82
+ "use_thumbnail": true,
83
+ "vision_config": {
84
+ "attention_dropout": 0.0,
85
+ "dtype": "bfloat16",
86
+ "hidden_act": "gelu_pytorch_tanh",
87
+ "hidden_size": 1152,
88
+ "intermediate_size": 4304,
89
+ "layer_norm_eps": 1e-06,
90
+ "model_type": "siglip2_vision_model",
91
+ "num_attention_heads": 16,
92
+ "num_channels": 3,
93
+ "num_hidden_layers": 27,
94
+ "num_patches": 256,
95
+ "patch_size": 16,
96
+ "vision_use_head": false
97
+ }
98
+ }
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": "5.0.0.dev0"
7
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08c72d8d6a8543f0462f65526dd5a2ae5f673f71953b51d9c84cd73b0d04f653
3
+ size 3193325814
model.safetensors.index.json ADDED
@@ -0,0 +1,596 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 3193251808
4
+ },
5
+ "weight_map": {
6
+ "language_model.model.embed_tokens.weight": "model.safetensors",
7
+ "language_model.model.embedding_norm.weight": "model.safetensors",
8
+ "language_model.model.layers.0.conv.conv.weight": "model.safetensors",
9
+ "language_model.model.layers.0.conv.in_proj.weight": "model.safetensors",
10
+ "language_model.model.layers.0.conv.out_proj.weight": "model.safetensors",
11
+ "language_model.model.layers.0.feed_forward.w1.weight": "model.safetensors",
12
+ "language_model.model.layers.0.feed_forward.w2.weight": "model.safetensors",
13
+ "language_model.model.layers.0.feed_forward.w3.weight": "model.safetensors",
14
+ "language_model.model.layers.0.ffn_norm.weight": "model.safetensors",
15
+ "language_model.model.layers.0.operator_norm.weight": "model.safetensors",
16
+ "language_model.model.layers.1.conv.conv.weight": "model.safetensors",
17
+ "language_model.model.layers.1.conv.in_proj.weight": "model.safetensors",
18
+ "language_model.model.layers.1.conv.out_proj.weight": "model.safetensors",
19
+ "language_model.model.layers.1.feed_forward.w1.weight": "model.safetensors",
20
+ "language_model.model.layers.1.feed_forward.w2.weight": "model.safetensors",
21
+ "language_model.model.layers.1.feed_forward.w3.weight": "model.safetensors",
22
+ "language_model.model.layers.1.ffn_norm.weight": "model.safetensors",
23
+ "language_model.model.layers.1.operator_norm.weight": "model.safetensors",
24
+ "language_model.model.layers.10.feed_forward.w1.weight": "model.safetensors",
25
+ "language_model.model.layers.10.feed_forward.w2.weight": "model.safetensors",
26
+ "language_model.model.layers.10.feed_forward.w3.weight": "model.safetensors",
27
+ "language_model.model.layers.10.ffn_norm.weight": "model.safetensors",
28
+ "language_model.model.layers.10.operator_norm.weight": "model.safetensors",
29
+ "language_model.model.layers.10.self_attn.k_layernorm.weight": "model.safetensors",
30
+ "language_model.model.layers.10.self_attn.k_proj.weight": "model.safetensors",
31
+ "language_model.model.layers.10.self_attn.out_proj.weight": "model.safetensors",
32
+ "language_model.model.layers.10.self_attn.q_layernorm.weight": "model.safetensors",
33
+ "language_model.model.layers.10.self_attn.q_proj.weight": "model.safetensors",
34
+ "language_model.model.layers.10.self_attn.v_proj.weight": "model.safetensors",
35
+ "language_model.model.layers.11.conv.conv.weight": "model.safetensors",
36
+ "language_model.model.layers.11.conv.in_proj.weight": "model.safetensors",
37
+ "language_model.model.layers.11.conv.out_proj.weight": "model.safetensors",
38
+ "language_model.model.layers.11.feed_forward.w1.weight": "model.safetensors",
39
+ "language_model.model.layers.11.feed_forward.w2.weight": "model.safetensors",
40
+ "language_model.model.layers.11.feed_forward.w3.weight": "model.safetensors",
41
+ "language_model.model.layers.11.ffn_norm.weight": "model.safetensors",
42
+ "language_model.model.layers.11.operator_norm.weight": "model.safetensors",
43
+ "language_model.model.layers.12.feed_forward.w1.weight": "model.safetensors",
44
+ "language_model.model.layers.12.feed_forward.w2.weight": "model.safetensors",
45
+ "language_model.model.layers.12.feed_forward.w3.weight": "model.safetensors",
46
+ "language_model.model.layers.12.ffn_norm.weight": "model.safetensors",
47
+ "language_model.model.layers.12.operator_norm.weight": "model.safetensors",
48
+ "language_model.model.layers.12.self_attn.k_layernorm.weight": "model.safetensors",
49
+ "language_model.model.layers.12.self_attn.k_proj.weight": "model.safetensors",
50
+ "language_model.model.layers.12.self_attn.out_proj.weight": "model.safetensors",
51
+ "language_model.model.layers.12.self_attn.q_layernorm.weight": "model.safetensors",
52
+ "language_model.model.layers.12.self_attn.q_proj.weight": "model.safetensors",
53
+ "language_model.model.layers.12.self_attn.v_proj.weight": "model.safetensors",
54
+ "language_model.model.layers.13.conv.conv.weight": "model.safetensors",
55
+ "language_model.model.layers.13.conv.in_proj.weight": "model.safetensors",
56
+ "language_model.model.layers.13.conv.out_proj.weight": "model.safetensors",
57
+ "language_model.model.layers.13.feed_forward.w1.weight": "model.safetensors",
58
+ "language_model.model.layers.13.feed_forward.w2.weight": "model.safetensors",
59
+ "language_model.model.layers.13.feed_forward.w3.weight": "model.safetensors",
60
+ "language_model.model.layers.13.ffn_norm.weight": "model.safetensors",
61
+ "language_model.model.layers.13.operator_norm.weight": "model.safetensors",
62
+ "language_model.model.layers.14.feed_forward.w1.weight": "model.safetensors",
63
+ "language_model.model.layers.14.feed_forward.w2.weight": "model.safetensors",
64
+ "language_model.model.layers.14.feed_forward.w3.weight": "model.safetensors",
65
+ "language_model.model.layers.14.ffn_norm.weight": "model.safetensors",
66
+ "language_model.model.layers.14.operator_norm.weight": "model.safetensors",
67
+ "language_model.model.layers.14.self_attn.k_layernorm.weight": "model.safetensors",
68
+ "language_model.model.layers.14.self_attn.k_proj.weight": "model.safetensors",
69
+ "language_model.model.layers.14.self_attn.out_proj.weight": "model.safetensors",
70
+ "language_model.model.layers.14.self_attn.q_layernorm.weight": "model.safetensors",
71
+ "language_model.model.layers.14.self_attn.q_proj.weight": "model.safetensors",
72
+ "language_model.model.layers.14.self_attn.v_proj.weight": "model.safetensors",
73
+ "language_model.model.layers.15.conv.conv.weight": "model.safetensors",
74
+ "language_model.model.layers.15.conv.in_proj.weight": "model.safetensors",
75
+ "language_model.model.layers.15.conv.out_proj.weight": "model.safetensors",
76
+ "language_model.model.layers.15.feed_forward.w1.weight": "model.safetensors",
77
+ "language_model.model.layers.15.feed_forward.w2.weight": "model.safetensors",
78
+ "language_model.model.layers.15.feed_forward.w3.weight": "model.safetensors",
79
+ "language_model.model.layers.15.ffn_norm.weight": "model.safetensors",
80
+ "language_model.model.layers.15.operator_norm.weight": "model.safetensors",
81
+ "language_model.model.layers.2.feed_forward.w1.weight": "model.safetensors",
82
+ "language_model.model.layers.2.feed_forward.w2.weight": "model.safetensors",
83
+ "language_model.model.layers.2.feed_forward.w3.weight": "model.safetensors",
84
+ "language_model.model.layers.2.ffn_norm.weight": "model.safetensors",
85
+ "language_model.model.layers.2.operator_norm.weight": "model.safetensors",
86
+ "language_model.model.layers.2.self_attn.k_layernorm.weight": "model.safetensors",
87
+ "language_model.model.layers.2.self_attn.k_proj.weight": "model.safetensors",
88
+ "language_model.model.layers.2.self_attn.out_proj.weight": "model.safetensors",
89
+ "language_model.model.layers.2.self_attn.q_layernorm.weight": "model.safetensors",
90
+ "language_model.model.layers.2.self_attn.q_proj.weight": "model.safetensors",
91
+ "language_model.model.layers.2.self_attn.v_proj.weight": "model.safetensors",
92
+ "language_model.model.layers.3.conv.conv.weight": "model.safetensors",
93
+ "language_model.model.layers.3.conv.in_proj.weight": "model.safetensors",
94
+ "language_model.model.layers.3.conv.out_proj.weight": "model.safetensors",
95
+ "language_model.model.layers.3.feed_forward.w1.weight": "model.safetensors",
96
+ "language_model.model.layers.3.feed_forward.w2.weight": "model.safetensors",
97
+ "language_model.model.layers.3.feed_forward.w3.weight": "model.safetensors",
98
+ "language_model.model.layers.3.ffn_norm.weight": "model.safetensors",
99
+ "language_model.model.layers.3.operator_norm.weight": "model.safetensors",
100
+ "language_model.model.layers.4.conv.conv.weight": "model.safetensors",
101
+ "language_model.model.layers.4.conv.in_proj.weight": "model.safetensors",
102
+ "language_model.model.layers.4.conv.out_proj.weight": "model.safetensors",
103
+ "language_model.model.layers.4.feed_forward.w1.weight": "model.safetensors",
104
+ "language_model.model.layers.4.feed_forward.w2.weight": "model.safetensors",
105
+ "language_model.model.layers.4.feed_forward.w3.weight": "model.safetensors",
106
+ "language_model.model.layers.4.ffn_norm.weight": "model.safetensors",
107
+ "language_model.model.layers.4.operator_norm.weight": "model.safetensors",
108
+ "language_model.model.layers.5.feed_forward.w1.weight": "model.safetensors",
109
+ "language_model.model.layers.5.feed_forward.w2.weight": "model.safetensors",
110
+ "language_model.model.layers.5.feed_forward.w3.weight": "model.safetensors",
111
+ "language_model.model.layers.5.ffn_norm.weight": "model.safetensors",
112
+ "language_model.model.layers.5.operator_norm.weight": "model.safetensors",
113
+ "language_model.model.layers.5.self_attn.k_layernorm.weight": "model.safetensors",
114
+ "language_model.model.layers.5.self_attn.k_proj.weight": "model.safetensors",
115
+ "language_model.model.layers.5.self_attn.out_proj.weight": "model.safetensors",
116
+ "language_model.model.layers.5.self_attn.q_layernorm.weight": "model.safetensors",
117
+ "language_model.model.layers.5.self_attn.q_proj.weight": "model.safetensors",
118
+ "language_model.model.layers.5.self_attn.v_proj.weight": "model.safetensors",
119
+ "language_model.model.layers.6.conv.conv.weight": "model.safetensors",
120
+ "language_model.model.layers.6.conv.in_proj.weight": "model.safetensors",
121
+ "language_model.model.layers.6.conv.out_proj.weight": "model.safetensors",
122
+ "language_model.model.layers.6.feed_forward.w1.weight": "model.safetensors",
123
+ "language_model.model.layers.6.feed_forward.w2.weight": "model.safetensors",
124
+ "language_model.model.layers.6.feed_forward.w3.weight": "model.safetensors",
125
+ "language_model.model.layers.6.ffn_norm.weight": "model.safetensors",
126
+ "language_model.model.layers.6.operator_norm.weight": "model.safetensors",
127
+ "language_model.model.layers.7.conv.conv.weight": "model.safetensors",
128
+ "language_model.model.layers.7.conv.in_proj.weight": "model.safetensors",
129
+ "language_model.model.layers.7.conv.out_proj.weight": "model.safetensors",
130
+ "language_model.model.layers.7.feed_forward.w1.weight": "model.safetensors",
131
+ "language_model.model.layers.7.feed_forward.w2.weight": "model.safetensors",
132
+ "language_model.model.layers.7.feed_forward.w3.weight": "model.safetensors",
133
+ "language_model.model.layers.7.ffn_norm.weight": "model.safetensors",
134
+ "language_model.model.layers.7.operator_norm.weight": "model.safetensors",
135
+ "language_model.model.layers.8.feed_forward.w1.weight": "model.safetensors",
136
+ "language_model.model.layers.8.feed_forward.w2.weight": "model.safetensors",
137
+ "language_model.model.layers.8.feed_forward.w3.weight": "model.safetensors",
138
+ "language_model.model.layers.8.ffn_norm.weight": "model.safetensors",
139
+ "language_model.model.layers.8.operator_norm.weight": "model.safetensors",
140
+ "language_model.model.layers.8.self_attn.k_layernorm.weight": "model.safetensors",
141
+ "language_model.model.layers.8.self_attn.k_proj.weight": "model.safetensors",
142
+ "language_model.model.layers.8.self_attn.out_proj.weight": "model.safetensors",
143
+ "language_model.model.layers.8.self_attn.q_layernorm.weight": "model.safetensors",
144
+ "language_model.model.layers.8.self_attn.q_proj.weight": "model.safetensors",
145
+ "language_model.model.layers.8.self_attn.v_proj.weight": "model.safetensors",
146
+ "language_model.model.layers.9.conv.conv.weight": "model.safetensors",
147
+ "language_model.model.layers.9.conv.in_proj.weight": "model.safetensors",
148
+ "language_model.model.layers.9.conv.out_proj.weight": "model.safetensors",
149
+ "language_model.model.layers.9.feed_forward.w1.weight": "model.safetensors",
150
+ "language_model.model.layers.9.feed_forward.w2.weight": "model.safetensors",
151
+ "language_model.model.layers.9.feed_forward.w3.weight": "model.safetensors",
152
+ "language_model.model.layers.9.ffn_norm.weight": "model.safetensors",
153
+ "language_model.model.layers.9.operator_norm.weight": "model.safetensors",
154
+ "multi_modal_projector.linear_1.bias": "model.safetensors",
155
+ "multi_modal_projector.linear_1.weight": "model.safetensors",
156
+ "multi_modal_projector.linear_2.bias": "model.safetensors",
157
+ "multi_modal_projector.linear_2.weight": "model.safetensors",
158
+ "vision_tower.embeddings.patch_embedding.bias": "model.safetensors",
159
+ "vision_tower.embeddings.patch_embedding.weight": "model.safetensors",
160
+ "vision_tower.embeddings.position_embedding.weight": "model.safetensors",
161
+ "vision_tower.encoder.layers.0.layer_norm1.bias": "model.safetensors",
162
+ "vision_tower.encoder.layers.0.layer_norm1.weight": "model.safetensors",
163
+ "vision_tower.encoder.layers.0.layer_norm2.bias": "model.safetensors",
164
+ "vision_tower.encoder.layers.0.layer_norm2.weight": "model.safetensors",
165
+ "vision_tower.encoder.layers.0.mlp.fc1.bias": "model.safetensors",
166
+ "vision_tower.encoder.layers.0.mlp.fc1.weight": "model.safetensors",
167
+ "vision_tower.encoder.layers.0.mlp.fc2.bias": "model.safetensors",
168
+ "vision_tower.encoder.layers.0.mlp.fc2.weight": "model.safetensors",
169
+ "vision_tower.encoder.layers.0.self_attn.k_proj.bias": "model.safetensors",
170
+ "vision_tower.encoder.layers.0.self_attn.k_proj.weight": "model.safetensors",
171
+ "vision_tower.encoder.layers.0.self_attn.out_proj.bias": "model.safetensors",
172
+ "vision_tower.encoder.layers.0.self_attn.out_proj.weight": "model.safetensors",
173
+ "vision_tower.encoder.layers.0.self_attn.q_proj.bias": "model.safetensors",
174
+ "vision_tower.encoder.layers.0.self_attn.q_proj.weight": "model.safetensors",
175
+ "vision_tower.encoder.layers.0.self_attn.v_proj.bias": "model.safetensors",
176
+ "vision_tower.encoder.layers.0.self_attn.v_proj.weight": "model.safetensors",
177
+ "vision_tower.encoder.layers.1.layer_norm1.bias": "model.safetensors",
178
+ "vision_tower.encoder.layers.1.layer_norm1.weight": "model.safetensors",
179
+ "vision_tower.encoder.layers.1.layer_norm2.bias": "model.safetensors",
180
+ "vision_tower.encoder.layers.1.layer_norm2.weight": "model.safetensors",
181
+ "vision_tower.encoder.layers.1.mlp.fc1.bias": "model.safetensors",
182
+ "vision_tower.encoder.layers.1.mlp.fc1.weight": "model.safetensors",
183
+ "vision_tower.encoder.layers.1.mlp.fc2.bias": "model.safetensors",
184
+ "vision_tower.encoder.layers.1.mlp.fc2.weight": "model.safetensors",
185
+ "vision_tower.encoder.layers.1.self_attn.k_proj.bias": "model.safetensors",
186
+ "vision_tower.encoder.layers.1.self_attn.k_proj.weight": "model.safetensors",
187
+ "vision_tower.encoder.layers.1.self_attn.out_proj.bias": "model.safetensors",
188
+ "vision_tower.encoder.layers.1.self_attn.out_proj.weight": "model.safetensors",
189
+ "vision_tower.encoder.layers.1.self_attn.q_proj.bias": "model.safetensors",
190
+ "vision_tower.encoder.layers.1.self_attn.q_proj.weight": "model.safetensors",
191
+ "vision_tower.encoder.layers.1.self_attn.v_proj.bias": "model.safetensors",
192
+ "vision_tower.encoder.layers.1.self_attn.v_proj.weight": "model.safetensors",
193
+ "vision_tower.encoder.layers.10.layer_norm1.bias": "model.safetensors",
194
+ "vision_tower.encoder.layers.10.layer_norm1.weight": "model.safetensors",
195
+ "vision_tower.encoder.layers.10.layer_norm2.bias": "model.safetensors",
196
+ "vision_tower.encoder.layers.10.layer_norm2.weight": "model.safetensors",
197
+ "vision_tower.encoder.layers.10.mlp.fc1.bias": "model.safetensors",
198
+ "vision_tower.encoder.layers.10.mlp.fc1.weight": "model.safetensors",
199
+ "vision_tower.encoder.layers.10.mlp.fc2.bias": "model.safetensors",
200
+ "vision_tower.encoder.layers.10.mlp.fc2.weight": "model.safetensors",
201
+ "vision_tower.encoder.layers.10.self_attn.k_proj.bias": "model.safetensors",
202
+ "vision_tower.encoder.layers.10.self_attn.k_proj.weight": "model.safetensors",
203
+ "vision_tower.encoder.layers.10.self_attn.out_proj.bias": "model.safetensors",
204
+ "vision_tower.encoder.layers.10.self_attn.out_proj.weight": "model.safetensors",
205
+ "vision_tower.encoder.layers.10.self_attn.q_proj.bias": "model.safetensors",
206
+ "vision_tower.encoder.layers.10.self_attn.q_proj.weight": "model.safetensors",
207
+ "vision_tower.encoder.layers.10.self_attn.v_proj.bias": "model.safetensors",
208
+ "vision_tower.encoder.layers.10.self_attn.v_proj.weight": "model.safetensors",
209
+ "vision_tower.encoder.layers.11.layer_norm1.bias": "model.safetensors",
210
+ "vision_tower.encoder.layers.11.layer_norm1.weight": "model.safetensors",
211
+ "vision_tower.encoder.layers.11.layer_norm2.bias": "model.safetensors",
212
+ "vision_tower.encoder.layers.11.layer_norm2.weight": "model.safetensors",
213
+ "vision_tower.encoder.layers.11.mlp.fc1.bias": "model.safetensors",
214
+ "vision_tower.encoder.layers.11.mlp.fc1.weight": "model.safetensors",
215
+ "vision_tower.encoder.layers.11.mlp.fc2.bias": "model.safetensors",
216
+ "vision_tower.encoder.layers.11.mlp.fc2.weight": "model.safetensors",
217
+ "vision_tower.encoder.layers.11.self_attn.k_proj.bias": "model.safetensors",
218
+ "vision_tower.encoder.layers.11.self_attn.k_proj.weight": "model.safetensors",
219
+ "vision_tower.encoder.layers.11.self_attn.out_proj.bias": "model.safetensors",
220
+ "vision_tower.encoder.layers.11.self_attn.out_proj.weight": "model.safetensors",
221
+ "vision_tower.encoder.layers.11.self_attn.q_proj.bias": "model.safetensors",
222
+ "vision_tower.encoder.layers.11.self_attn.q_proj.weight": "model.safetensors",
223
+ "vision_tower.encoder.layers.11.self_attn.v_proj.bias": "model.safetensors",
224
+ "vision_tower.encoder.layers.11.self_attn.v_proj.weight": "model.safetensors",
225
+ "vision_tower.encoder.layers.12.layer_norm1.bias": "model.safetensors",
226
+ "vision_tower.encoder.layers.12.layer_norm1.weight": "model.safetensors",
227
+ "vision_tower.encoder.layers.12.layer_norm2.bias": "model.safetensors",
228
+ "vision_tower.encoder.layers.12.layer_norm2.weight": "model.safetensors",
229
+ "vision_tower.encoder.layers.12.mlp.fc1.bias": "model.safetensors",
230
+ "vision_tower.encoder.layers.12.mlp.fc1.weight": "model.safetensors",
231
+ "vision_tower.encoder.layers.12.mlp.fc2.bias": "model.safetensors",
232
+ "vision_tower.encoder.layers.12.mlp.fc2.weight": "model.safetensors",
233
+ "vision_tower.encoder.layers.12.self_attn.k_proj.bias": "model.safetensors",
234
+ "vision_tower.encoder.layers.12.self_attn.k_proj.weight": "model.safetensors",
235
+ "vision_tower.encoder.layers.12.self_attn.out_proj.bias": "model.safetensors",
236
+ "vision_tower.encoder.layers.12.self_attn.out_proj.weight": "model.safetensors",
237
+ "vision_tower.encoder.layers.12.self_attn.q_proj.bias": "model.safetensors",
238
+ "vision_tower.encoder.layers.12.self_attn.q_proj.weight": "model.safetensors",
239
+ "vision_tower.encoder.layers.12.self_attn.v_proj.bias": "model.safetensors",
240
+ "vision_tower.encoder.layers.12.self_attn.v_proj.weight": "model.safetensors",
241
+ "vision_tower.encoder.layers.13.layer_norm1.bias": "model.safetensors",
242
+ "vision_tower.encoder.layers.13.layer_norm1.weight": "model.safetensors",
243
+ "vision_tower.encoder.layers.13.layer_norm2.bias": "model.safetensors",
244
+ "vision_tower.encoder.layers.13.layer_norm2.weight": "model.safetensors",
245
+ "vision_tower.encoder.layers.13.mlp.fc1.bias": "model.safetensors",
246
+ "vision_tower.encoder.layers.13.mlp.fc1.weight": "model.safetensors",
247
+ "vision_tower.encoder.layers.13.mlp.fc2.bias": "model.safetensors",
248
+ "vision_tower.encoder.layers.13.mlp.fc2.weight": "model.safetensors",
249
+ "vision_tower.encoder.layers.13.self_attn.k_proj.bias": "model.safetensors",
250
+ "vision_tower.encoder.layers.13.self_attn.k_proj.weight": "model.safetensors",
251
+ "vision_tower.encoder.layers.13.self_attn.out_proj.bias": "model.safetensors",
252
+ "vision_tower.encoder.layers.13.self_attn.out_proj.weight": "model.safetensors",
253
+ "vision_tower.encoder.layers.13.self_attn.q_proj.bias": "model.safetensors",
254
+ "vision_tower.encoder.layers.13.self_attn.q_proj.weight": "model.safetensors",
255
+ "vision_tower.encoder.layers.13.self_attn.v_proj.bias": "model.safetensors",
256
+ "vision_tower.encoder.layers.13.self_attn.v_proj.weight": "model.safetensors",
257
+ "vision_tower.encoder.layers.14.layer_norm1.bias": "model.safetensors",
258
+ "vision_tower.encoder.layers.14.layer_norm1.weight": "model.safetensors",
259
+ "vision_tower.encoder.layers.14.layer_norm2.bias": "model.safetensors",
260
+ "vision_tower.encoder.layers.14.layer_norm2.weight": "model.safetensors",
261
+ "vision_tower.encoder.layers.14.mlp.fc1.bias": "model.safetensors",
262
+ "vision_tower.encoder.layers.14.mlp.fc1.weight": "model.safetensors",
263
+ "vision_tower.encoder.layers.14.mlp.fc2.bias": "model.safetensors",
264
+ "vision_tower.encoder.layers.14.mlp.fc2.weight": "model.safetensors",
265
+ "vision_tower.encoder.layers.14.self_attn.k_proj.bias": "model.safetensors",
266
+ "vision_tower.encoder.layers.14.self_attn.k_proj.weight": "model.safetensors",
267
+ "vision_tower.encoder.layers.14.self_attn.out_proj.bias": "model.safetensors",
268
+ "vision_tower.encoder.layers.14.self_attn.out_proj.weight": "model.safetensors",
269
+ "vision_tower.encoder.layers.14.self_attn.q_proj.bias": "model.safetensors",
270
+ "vision_tower.encoder.layers.14.self_attn.q_proj.weight": "model.safetensors",
271
+ "vision_tower.encoder.layers.14.self_attn.v_proj.bias": "model.safetensors",
272
+ "vision_tower.encoder.layers.14.self_attn.v_proj.weight": "model.safetensors",
273
+ "vision_tower.encoder.layers.15.layer_norm1.bias": "model.safetensors",
274
+ "vision_tower.encoder.layers.15.layer_norm1.weight": "model.safetensors",
275
+ "vision_tower.encoder.layers.15.layer_norm2.bias": "model.safetensors",
276
+ "vision_tower.encoder.layers.15.layer_norm2.weight": "model.safetensors",
277
+ "vision_tower.encoder.layers.15.mlp.fc1.bias": "model.safetensors",
278
+ "vision_tower.encoder.layers.15.mlp.fc1.weight": "model.safetensors",
279
+ "vision_tower.encoder.layers.15.mlp.fc2.bias": "model.safetensors",
280
+ "vision_tower.encoder.layers.15.mlp.fc2.weight": "model.safetensors",
281
+ "vision_tower.encoder.layers.15.self_attn.k_proj.bias": "model.safetensors",
282
+ "vision_tower.encoder.layers.15.self_attn.k_proj.weight": "model.safetensors",
283
+ "vision_tower.encoder.layers.15.self_attn.out_proj.bias": "model.safetensors",
284
+ "vision_tower.encoder.layers.15.self_attn.out_proj.weight": "model.safetensors",
285
+ "vision_tower.encoder.layers.15.self_attn.q_proj.bias": "model.safetensors",
286
+ "vision_tower.encoder.layers.15.self_attn.q_proj.weight": "model.safetensors",
287
+ "vision_tower.encoder.layers.15.self_attn.v_proj.bias": "model.safetensors",
288
+ "vision_tower.encoder.layers.15.self_attn.v_proj.weight": "model.safetensors",
289
+ "vision_tower.encoder.layers.16.layer_norm1.bias": "model.safetensors",
290
+ "vision_tower.encoder.layers.16.layer_norm1.weight": "model.safetensors",
291
+ "vision_tower.encoder.layers.16.layer_norm2.bias": "model.safetensors",
292
+ "vision_tower.encoder.layers.16.layer_norm2.weight": "model.safetensors",
293
+ "vision_tower.encoder.layers.16.mlp.fc1.bias": "model.safetensors",
294
+ "vision_tower.encoder.layers.16.mlp.fc1.weight": "model.safetensors",
295
+ "vision_tower.encoder.layers.16.mlp.fc2.bias": "model.safetensors",
296
+ "vision_tower.encoder.layers.16.mlp.fc2.weight": "model.safetensors",
297
+ "vision_tower.encoder.layers.16.self_attn.k_proj.bias": "model.safetensors",
298
+ "vision_tower.encoder.layers.16.self_attn.k_proj.weight": "model.safetensors",
299
+ "vision_tower.encoder.layers.16.self_attn.out_proj.bias": "model.safetensors",
300
+ "vision_tower.encoder.layers.16.self_attn.out_proj.weight": "model.safetensors",
301
+ "vision_tower.encoder.layers.16.self_attn.q_proj.bias": "model.safetensors",
302
+ "vision_tower.encoder.layers.16.self_attn.q_proj.weight": "model.safetensors",
303
+ "vision_tower.encoder.layers.16.self_attn.v_proj.bias": "model.safetensors",
304
+ "vision_tower.encoder.layers.16.self_attn.v_proj.weight": "model.safetensors",
305
+ "vision_tower.encoder.layers.17.layer_norm1.bias": "model.safetensors",
306
+ "vision_tower.encoder.layers.17.layer_norm1.weight": "model.safetensors",
307
+ "vision_tower.encoder.layers.17.layer_norm2.bias": "model.safetensors",
308
+ "vision_tower.encoder.layers.17.layer_norm2.weight": "model.safetensors",
309
+ "vision_tower.encoder.layers.17.mlp.fc1.bias": "model.safetensors",
310
+ "vision_tower.encoder.layers.17.mlp.fc1.weight": "model.safetensors",
311
+ "vision_tower.encoder.layers.17.mlp.fc2.bias": "model.safetensors",
312
+ "vision_tower.encoder.layers.17.mlp.fc2.weight": "model.safetensors",
313
+ "vision_tower.encoder.layers.17.self_attn.k_proj.bias": "model.safetensors",
314
+ "vision_tower.encoder.layers.17.self_attn.k_proj.weight": "model.safetensors",
315
+ "vision_tower.encoder.layers.17.self_attn.out_proj.bias": "model.safetensors",
316
+ "vision_tower.encoder.layers.17.self_attn.out_proj.weight": "model.safetensors",
317
+ "vision_tower.encoder.layers.17.self_attn.q_proj.bias": "model.safetensors",
318
+ "vision_tower.encoder.layers.17.self_attn.q_proj.weight": "model.safetensors",
319
+ "vision_tower.encoder.layers.17.self_attn.v_proj.bias": "model.safetensors",
320
+ "vision_tower.encoder.layers.17.self_attn.v_proj.weight": "model.safetensors",
321
+ "vision_tower.encoder.layers.18.layer_norm1.bias": "model.safetensors",
322
+ "vision_tower.encoder.layers.18.layer_norm1.weight": "model.safetensors",
323
+ "vision_tower.encoder.layers.18.layer_norm2.bias": "model.safetensors",
324
+ "vision_tower.encoder.layers.18.layer_norm2.weight": "model.safetensors",
325
+ "vision_tower.encoder.layers.18.mlp.fc1.bias": "model.safetensors",
326
+ "vision_tower.encoder.layers.18.mlp.fc1.weight": "model.safetensors",
327
+ "vision_tower.encoder.layers.18.mlp.fc2.bias": "model.safetensors",
328
+ "vision_tower.encoder.layers.18.mlp.fc2.weight": "model.safetensors",
329
+ "vision_tower.encoder.layers.18.self_attn.k_proj.bias": "model.safetensors",
330
+ "vision_tower.encoder.layers.18.self_attn.k_proj.weight": "model.safetensors",
331
+ "vision_tower.encoder.layers.18.self_attn.out_proj.bias": "model.safetensors",
332
+ "vision_tower.encoder.layers.18.self_attn.out_proj.weight": "model.safetensors",
333
+ "vision_tower.encoder.layers.18.self_attn.q_proj.bias": "model.safetensors",
334
+ "vision_tower.encoder.layers.18.self_attn.q_proj.weight": "model.safetensors",
335
+ "vision_tower.encoder.layers.18.self_attn.v_proj.bias": "model.safetensors",
336
+ "vision_tower.encoder.layers.18.self_attn.v_proj.weight": "model.safetensors",
337
+ "vision_tower.encoder.layers.19.layer_norm1.bias": "model.safetensors",
338
+ "vision_tower.encoder.layers.19.layer_norm1.weight": "model.safetensors",
339
+ "vision_tower.encoder.layers.19.layer_norm2.bias": "model.safetensors",
340
+ "vision_tower.encoder.layers.19.layer_norm2.weight": "model.safetensors",
341
+ "vision_tower.encoder.layers.19.mlp.fc1.bias": "model.safetensors",
342
+ "vision_tower.encoder.layers.19.mlp.fc1.weight": "model.safetensors",
343
+ "vision_tower.encoder.layers.19.mlp.fc2.bias": "model.safetensors",
344
+ "vision_tower.encoder.layers.19.mlp.fc2.weight": "model.safetensors",
345
+ "vision_tower.encoder.layers.19.self_attn.k_proj.bias": "model.safetensors",
346
+ "vision_tower.encoder.layers.19.self_attn.k_proj.weight": "model.safetensors",
347
+ "vision_tower.encoder.layers.19.self_attn.out_proj.bias": "model.safetensors",
348
+ "vision_tower.encoder.layers.19.self_attn.out_proj.weight": "model.safetensors",
349
+ "vision_tower.encoder.layers.19.self_attn.q_proj.bias": "model.safetensors",
350
+ "vision_tower.encoder.layers.19.self_attn.q_proj.weight": "model.safetensors",
351
+ "vision_tower.encoder.layers.19.self_attn.v_proj.bias": "model.safetensors",
352
+ "vision_tower.encoder.layers.19.self_attn.v_proj.weight": "model.safetensors",
353
+ "vision_tower.encoder.layers.2.layer_norm1.bias": "model.safetensors",
354
+ "vision_tower.encoder.layers.2.layer_norm1.weight": "model.safetensors",
355
+ "vision_tower.encoder.layers.2.layer_norm2.bias": "model.safetensors",
356
+ "vision_tower.encoder.layers.2.layer_norm2.weight": "model.safetensors",
357
+ "vision_tower.encoder.layers.2.mlp.fc1.bias": "model.safetensors",
358
+ "vision_tower.encoder.layers.2.mlp.fc1.weight": "model.safetensors",
359
+ "vision_tower.encoder.layers.2.mlp.fc2.bias": "model.safetensors",
360
+ "vision_tower.encoder.layers.2.mlp.fc2.weight": "model.safetensors",
361
+ "vision_tower.encoder.layers.2.self_attn.k_proj.bias": "model.safetensors",
362
+ "vision_tower.encoder.layers.2.self_attn.k_proj.weight": "model.safetensors",
363
+ "vision_tower.encoder.layers.2.self_attn.out_proj.bias": "model.safetensors",
364
+ "vision_tower.encoder.layers.2.self_attn.out_proj.weight": "model.safetensors",
365
+ "vision_tower.encoder.layers.2.self_attn.q_proj.bias": "model.safetensors",
366
+ "vision_tower.encoder.layers.2.self_attn.q_proj.weight": "model.safetensors",
367
+ "vision_tower.encoder.layers.2.self_attn.v_proj.bias": "model.safetensors",
368
+ "vision_tower.encoder.layers.2.self_attn.v_proj.weight": "model.safetensors",
369
+ "vision_tower.encoder.layers.20.layer_norm1.bias": "model.safetensors",
370
+ "vision_tower.encoder.layers.20.layer_norm1.weight": "model.safetensors",
371
+ "vision_tower.encoder.layers.20.layer_norm2.bias": "model.safetensors",
372
+ "vision_tower.encoder.layers.20.layer_norm2.weight": "model.safetensors",
373
+ "vision_tower.encoder.layers.20.mlp.fc1.bias": "model.safetensors",
374
+ "vision_tower.encoder.layers.20.mlp.fc1.weight": "model.safetensors",
375
+ "vision_tower.encoder.layers.20.mlp.fc2.bias": "model.safetensors",
376
+ "vision_tower.encoder.layers.20.mlp.fc2.weight": "model.safetensors",
377
+ "vision_tower.encoder.layers.20.self_attn.k_proj.bias": "model.safetensors",
378
+ "vision_tower.encoder.layers.20.self_attn.k_proj.weight": "model.safetensors",
379
+ "vision_tower.encoder.layers.20.self_attn.out_proj.bias": "model.safetensors",
380
+ "vision_tower.encoder.layers.20.self_attn.out_proj.weight": "model.safetensors",
381
+ "vision_tower.encoder.layers.20.self_attn.q_proj.bias": "model.safetensors",
382
+ "vision_tower.encoder.layers.20.self_attn.q_proj.weight": "model.safetensors",
383
+ "vision_tower.encoder.layers.20.self_attn.v_proj.bias": "model.safetensors",
384
+ "vision_tower.encoder.layers.20.self_attn.v_proj.weight": "model.safetensors",
385
+ "vision_tower.encoder.layers.21.layer_norm1.bias": "model.safetensors",
386
+ "vision_tower.encoder.layers.21.layer_norm1.weight": "model.safetensors",
387
+ "vision_tower.encoder.layers.21.layer_norm2.bias": "model.safetensors",
388
+ "vision_tower.encoder.layers.21.layer_norm2.weight": "model.safetensors",
389
+ "vision_tower.encoder.layers.21.mlp.fc1.bias": "model.safetensors",
390
+ "vision_tower.encoder.layers.21.mlp.fc1.weight": "model.safetensors",
391
+ "vision_tower.encoder.layers.21.mlp.fc2.bias": "model.safetensors",
392
+ "vision_tower.encoder.layers.21.mlp.fc2.weight": "model.safetensors",
393
+ "vision_tower.encoder.layers.21.self_attn.k_proj.bias": "model.safetensors",
394
+ "vision_tower.encoder.layers.21.self_attn.k_proj.weight": "model.safetensors",
395
+ "vision_tower.encoder.layers.21.self_attn.out_proj.bias": "model.safetensors",
396
+ "vision_tower.encoder.layers.21.self_attn.out_proj.weight": "model.safetensors",
397
+ "vision_tower.encoder.layers.21.self_attn.q_proj.bias": "model.safetensors",
398
+ "vision_tower.encoder.layers.21.self_attn.q_proj.weight": "model.safetensors",
399
+ "vision_tower.encoder.layers.21.self_attn.v_proj.bias": "model.safetensors",
400
+ "vision_tower.encoder.layers.21.self_attn.v_proj.weight": "model.safetensors",
401
+ "vision_tower.encoder.layers.22.layer_norm1.bias": "model.safetensors",
402
+ "vision_tower.encoder.layers.22.layer_norm1.weight": "model.safetensors",
403
+ "vision_tower.encoder.layers.22.layer_norm2.bias": "model.safetensors",
404
+ "vision_tower.encoder.layers.22.layer_norm2.weight": "model.safetensors",
405
+ "vision_tower.encoder.layers.22.mlp.fc1.bias": "model.safetensors",
406
+ "vision_tower.encoder.layers.22.mlp.fc1.weight": "model.safetensors",
407
+ "vision_tower.encoder.layers.22.mlp.fc2.bias": "model.safetensors",
408
+ "vision_tower.encoder.layers.22.mlp.fc2.weight": "model.safetensors",
409
+ "vision_tower.encoder.layers.22.self_attn.k_proj.bias": "model.safetensors",
410
+ "vision_tower.encoder.layers.22.self_attn.k_proj.weight": "model.safetensors",
411
+ "vision_tower.encoder.layers.22.self_attn.out_proj.bias": "model.safetensors",
412
+ "vision_tower.encoder.layers.22.self_attn.out_proj.weight": "model.safetensors",
413
+ "vision_tower.encoder.layers.22.self_attn.q_proj.bias": "model.safetensors",
414
+ "vision_tower.encoder.layers.22.self_attn.q_proj.weight": "model.safetensors",
415
+ "vision_tower.encoder.layers.22.self_attn.v_proj.bias": "model.safetensors",
416
+ "vision_tower.encoder.layers.22.self_attn.v_proj.weight": "model.safetensors",
417
+ "vision_tower.encoder.layers.23.layer_norm1.bias": "model.safetensors",
418
+ "vision_tower.encoder.layers.23.layer_norm1.weight": "model.safetensors",
419
+ "vision_tower.encoder.layers.23.layer_norm2.bias": "model.safetensors",
420
+ "vision_tower.encoder.layers.23.layer_norm2.weight": "model.safetensors",
421
+ "vision_tower.encoder.layers.23.mlp.fc1.bias": "model.safetensors",
422
+ "vision_tower.encoder.layers.23.mlp.fc1.weight": "model.safetensors",
423
+ "vision_tower.encoder.layers.23.mlp.fc2.bias": "model.safetensors",
424
+ "vision_tower.encoder.layers.23.mlp.fc2.weight": "model.safetensors",
425
+ "vision_tower.encoder.layers.23.self_attn.k_proj.bias": "model.safetensors",
426
+ "vision_tower.encoder.layers.23.self_attn.k_proj.weight": "model.safetensors",
427
+ "vision_tower.encoder.layers.23.self_attn.out_proj.bias": "model.safetensors",
428
+ "vision_tower.encoder.layers.23.self_attn.out_proj.weight": "model.safetensors",
429
+ "vision_tower.encoder.layers.23.self_attn.q_proj.bias": "model.safetensors",
430
+ "vision_tower.encoder.layers.23.self_attn.q_proj.weight": "model.safetensors",
431
+ "vision_tower.encoder.layers.23.self_attn.v_proj.bias": "model.safetensors",
432
+ "vision_tower.encoder.layers.23.self_attn.v_proj.weight": "model.safetensors",
433
+ "vision_tower.encoder.layers.24.layer_norm1.bias": "model.safetensors",
434
+ "vision_tower.encoder.layers.24.layer_norm1.weight": "model.safetensors",
435
+ "vision_tower.encoder.layers.24.layer_norm2.bias": "model.safetensors",
436
+ "vision_tower.encoder.layers.24.layer_norm2.weight": "model.safetensors",
437
+ "vision_tower.encoder.layers.24.mlp.fc1.bias": "model.safetensors",
438
+ "vision_tower.encoder.layers.24.mlp.fc1.weight": "model.safetensors",
439
+ "vision_tower.encoder.layers.24.mlp.fc2.bias": "model.safetensors",
440
+ "vision_tower.encoder.layers.24.mlp.fc2.weight": "model.safetensors",
441
+ "vision_tower.encoder.layers.24.self_attn.k_proj.bias": "model.safetensors",
442
+ "vision_tower.encoder.layers.24.self_attn.k_proj.weight": "model.safetensors",
443
+ "vision_tower.encoder.layers.24.self_attn.out_proj.bias": "model.safetensors",
444
+ "vision_tower.encoder.layers.24.self_attn.out_proj.weight": "model.safetensors",
445
+ "vision_tower.encoder.layers.24.self_attn.q_proj.bias": "model.safetensors",
446
+ "vision_tower.encoder.layers.24.self_attn.q_proj.weight": "model.safetensors",
447
+ "vision_tower.encoder.layers.24.self_attn.v_proj.bias": "model.safetensors",
448
+ "vision_tower.encoder.layers.24.self_attn.v_proj.weight": "model.safetensors",
449
+ "vision_tower.encoder.layers.25.layer_norm1.bias": "model.safetensors",
450
+ "vision_tower.encoder.layers.25.layer_norm1.weight": "model.safetensors",
451
+ "vision_tower.encoder.layers.25.layer_norm2.bias": "model.safetensors",
452
+ "vision_tower.encoder.layers.25.layer_norm2.weight": "model.safetensors",
453
+ "vision_tower.encoder.layers.25.mlp.fc1.bias": "model.safetensors",
454
+ "vision_tower.encoder.layers.25.mlp.fc1.weight": "model.safetensors",
455
+ "vision_tower.encoder.layers.25.mlp.fc2.bias": "model.safetensors",
456
+ "vision_tower.encoder.layers.25.mlp.fc2.weight": "model.safetensors",
457
+ "vision_tower.encoder.layers.25.self_attn.k_proj.bias": "model.safetensors",
458
+ "vision_tower.encoder.layers.25.self_attn.k_proj.weight": "model.safetensors",
459
+ "vision_tower.encoder.layers.25.self_attn.out_proj.bias": "model.safetensors",
460
+ "vision_tower.encoder.layers.25.self_attn.out_proj.weight": "model.safetensors",
461
+ "vision_tower.encoder.layers.25.self_attn.q_proj.bias": "model.safetensors",
462
+ "vision_tower.encoder.layers.25.self_attn.q_proj.weight": "model.safetensors",
463
+ "vision_tower.encoder.layers.25.self_attn.v_proj.bias": "model.safetensors",
464
+ "vision_tower.encoder.layers.25.self_attn.v_proj.weight": "model.safetensors",
465
+ "vision_tower.encoder.layers.26.layer_norm1.bias": "model.safetensors",
466
+ "vision_tower.encoder.layers.26.layer_norm1.weight": "model.safetensors",
467
+ "vision_tower.encoder.layers.26.layer_norm2.bias": "model.safetensors",
468
+ "vision_tower.encoder.layers.26.layer_norm2.weight": "model.safetensors",
469
+ "vision_tower.encoder.layers.26.mlp.fc1.bias": "model.safetensors",
470
+ "vision_tower.encoder.layers.26.mlp.fc1.weight": "model.safetensors",
471
+ "vision_tower.encoder.layers.26.mlp.fc2.bias": "model.safetensors",
472
+ "vision_tower.encoder.layers.26.mlp.fc2.weight": "model.safetensors",
473
+ "vision_tower.encoder.layers.26.self_attn.k_proj.bias": "model.safetensors",
474
+ "vision_tower.encoder.layers.26.self_attn.k_proj.weight": "model.safetensors",
475
+ "vision_tower.encoder.layers.26.self_attn.out_proj.bias": "model.safetensors",
476
+ "vision_tower.encoder.layers.26.self_attn.out_proj.weight": "model.safetensors",
477
+ "vision_tower.encoder.layers.26.self_attn.q_proj.bias": "model.safetensors",
478
+ "vision_tower.encoder.layers.26.self_attn.q_proj.weight": "model.safetensors",
479
+ "vision_tower.encoder.layers.26.self_attn.v_proj.bias": "model.safetensors",
480
+ "vision_tower.encoder.layers.26.self_attn.v_proj.weight": "model.safetensors",
481
+ "vision_tower.encoder.layers.3.layer_norm1.bias": "model.safetensors",
482
+ "vision_tower.encoder.layers.3.layer_norm1.weight": "model.safetensors",
483
+ "vision_tower.encoder.layers.3.layer_norm2.bias": "model.safetensors",
484
+ "vision_tower.encoder.layers.3.layer_norm2.weight": "model.safetensors",
485
+ "vision_tower.encoder.layers.3.mlp.fc1.bias": "model.safetensors",
486
+ "vision_tower.encoder.layers.3.mlp.fc1.weight": "model.safetensors",
487
+ "vision_tower.encoder.layers.3.mlp.fc2.bias": "model.safetensors",
488
+ "vision_tower.encoder.layers.3.mlp.fc2.weight": "model.safetensors",
489
+ "vision_tower.encoder.layers.3.self_attn.k_proj.bias": "model.safetensors",
490
+ "vision_tower.encoder.layers.3.self_attn.k_proj.weight": "model.safetensors",
491
+ "vision_tower.encoder.layers.3.self_attn.out_proj.bias": "model.safetensors",
492
+ "vision_tower.encoder.layers.3.self_attn.out_proj.weight": "model.safetensors",
493
+ "vision_tower.encoder.layers.3.self_attn.q_proj.bias": "model.safetensors",
494
+ "vision_tower.encoder.layers.3.self_attn.q_proj.weight": "model.safetensors",
495
+ "vision_tower.encoder.layers.3.self_attn.v_proj.bias": "model.safetensors",
496
+ "vision_tower.encoder.layers.3.self_attn.v_proj.weight": "model.safetensors",
497
+ "vision_tower.encoder.layers.4.layer_norm1.bias": "model.safetensors",
498
+ "vision_tower.encoder.layers.4.layer_norm1.weight": "model.safetensors",
499
+ "vision_tower.encoder.layers.4.layer_norm2.bias": "model.safetensors",
500
+ "vision_tower.encoder.layers.4.layer_norm2.weight": "model.safetensors",
501
+ "vision_tower.encoder.layers.4.mlp.fc1.bias": "model.safetensors",
502
+ "vision_tower.encoder.layers.4.mlp.fc1.weight": "model.safetensors",
503
+ "vision_tower.encoder.layers.4.mlp.fc2.bias": "model.safetensors",
504
+ "vision_tower.encoder.layers.4.mlp.fc2.weight": "model.safetensors",
505
+ "vision_tower.encoder.layers.4.self_attn.k_proj.bias": "model.safetensors",
506
+ "vision_tower.encoder.layers.4.self_attn.k_proj.weight": "model.safetensors",
507
+ "vision_tower.encoder.layers.4.self_attn.out_proj.bias": "model.safetensors",
508
+ "vision_tower.encoder.layers.4.self_attn.out_proj.weight": "model.safetensors",
509
+ "vision_tower.encoder.layers.4.self_attn.q_proj.bias": "model.safetensors",
510
+ "vision_tower.encoder.layers.4.self_attn.q_proj.weight": "model.safetensors",
511
+ "vision_tower.encoder.layers.4.self_attn.v_proj.bias": "model.safetensors",
512
+ "vision_tower.encoder.layers.4.self_attn.v_proj.weight": "model.safetensors",
513
+ "vision_tower.encoder.layers.5.layer_norm1.bias": "model.safetensors",
514
+ "vision_tower.encoder.layers.5.layer_norm1.weight": "model.safetensors",
515
+ "vision_tower.encoder.layers.5.layer_norm2.bias": "model.safetensors",
516
+ "vision_tower.encoder.layers.5.layer_norm2.weight": "model.safetensors",
517
+ "vision_tower.encoder.layers.5.mlp.fc1.bias": "model.safetensors",
518
+ "vision_tower.encoder.layers.5.mlp.fc1.weight": "model.safetensors",
519
+ "vision_tower.encoder.layers.5.mlp.fc2.bias": "model.safetensors",
520
+ "vision_tower.encoder.layers.5.mlp.fc2.weight": "model.safetensors",
521
+ "vision_tower.encoder.layers.5.self_attn.k_proj.bias": "model.safetensors",
522
+ "vision_tower.encoder.layers.5.self_attn.k_proj.weight": "model.safetensors",
523
+ "vision_tower.encoder.layers.5.self_attn.out_proj.bias": "model.safetensors",
524
+ "vision_tower.encoder.layers.5.self_attn.out_proj.weight": "model.safetensors",
525
+ "vision_tower.encoder.layers.5.self_attn.q_proj.bias": "model.safetensors",
526
+ "vision_tower.encoder.layers.5.self_attn.q_proj.weight": "model.safetensors",
527
+ "vision_tower.encoder.layers.5.self_attn.v_proj.bias": "model.safetensors",
528
+ "vision_tower.encoder.layers.5.self_attn.v_proj.weight": "model.safetensors",
529
+ "vision_tower.encoder.layers.6.layer_norm1.bias": "model.safetensors",
530
+ "vision_tower.encoder.layers.6.layer_norm1.weight": "model.safetensors",
531
+ "vision_tower.encoder.layers.6.layer_norm2.bias": "model.safetensors",
532
+ "vision_tower.encoder.layers.6.layer_norm2.weight": "model.safetensors",
533
+ "vision_tower.encoder.layers.6.mlp.fc1.bias": "model.safetensors",
534
+ "vision_tower.encoder.layers.6.mlp.fc1.weight": "model.safetensors",
535
+ "vision_tower.encoder.layers.6.mlp.fc2.bias": "model.safetensors",
536
+ "vision_tower.encoder.layers.6.mlp.fc2.weight": "model.safetensors",
537
+ "vision_tower.encoder.layers.6.self_attn.k_proj.bias": "model.safetensors",
538
+ "vision_tower.encoder.layers.6.self_attn.k_proj.weight": "model.safetensors",
539
+ "vision_tower.encoder.layers.6.self_attn.out_proj.bias": "model.safetensors",
540
+ "vision_tower.encoder.layers.6.self_attn.out_proj.weight": "model.safetensors",
541
+ "vision_tower.encoder.layers.6.self_attn.q_proj.bias": "model.safetensors",
542
+ "vision_tower.encoder.layers.6.self_attn.q_proj.weight": "model.safetensors",
543
+ "vision_tower.encoder.layers.6.self_attn.v_proj.bias": "model.safetensors",
544
+ "vision_tower.encoder.layers.6.self_attn.v_proj.weight": "model.safetensors",
545
+ "vision_tower.encoder.layers.7.layer_norm1.bias": "model.safetensors",
546
+ "vision_tower.encoder.layers.7.layer_norm1.weight": "model.safetensors",
547
+ "vision_tower.encoder.layers.7.layer_norm2.bias": "model.safetensors",
548
+ "vision_tower.encoder.layers.7.layer_norm2.weight": "model.safetensors",
549
+ "vision_tower.encoder.layers.7.mlp.fc1.bias": "model.safetensors",
550
+ "vision_tower.encoder.layers.7.mlp.fc1.weight": "model.safetensors",
551
+ "vision_tower.encoder.layers.7.mlp.fc2.bias": "model.safetensors",
552
+ "vision_tower.encoder.layers.7.mlp.fc2.weight": "model.safetensors",
553
+ "vision_tower.encoder.layers.7.self_attn.k_proj.bias": "model.safetensors",
554
+ "vision_tower.encoder.layers.7.self_attn.k_proj.weight": "model.safetensors",
555
+ "vision_tower.encoder.layers.7.self_attn.out_proj.bias": "model.safetensors",
556
+ "vision_tower.encoder.layers.7.self_attn.out_proj.weight": "model.safetensors",
557
+ "vision_tower.encoder.layers.7.self_attn.q_proj.bias": "model.safetensors",
558
+ "vision_tower.encoder.layers.7.self_attn.q_proj.weight": "model.safetensors",
559
+ "vision_tower.encoder.layers.7.self_attn.v_proj.bias": "model.safetensors",
560
+ "vision_tower.encoder.layers.7.self_attn.v_proj.weight": "model.safetensors",
561
+ "vision_tower.encoder.layers.8.layer_norm1.bias": "model.safetensors",
562
+ "vision_tower.encoder.layers.8.layer_norm1.weight": "model.safetensors",
563
+ "vision_tower.encoder.layers.8.layer_norm2.bias": "model.safetensors",
564
+ "vision_tower.encoder.layers.8.layer_norm2.weight": "model.safetensors",
565
+ "vision_tower.encoder.layers.8.mlp.fc1.bias": "model.safetensors",
566
+ "vision_tower.encoder.layers.8.mlp.fc1.weight": "model.safetensors",
567
+ "vision_tower.encoder.layers.8.mlp.fc2.bias": "model.safetensors",
568
+ "vision_tower.encoder.layers.8.mlp.fc2.weight": "model.safetensors",
569
+ "vision_tower.encoder.layers.8.self_attn.k_proj.bias": "model.safetensors",
570
+ "vision_tower.encoder.layers.8.self_attn.k_proj.weight": "model.safetensors",
571
+ "vision_tower.encoder.layers.8.self_attn.out_proj.bias": "model.safetensors",
572
+ "vision_tower.encoder.layers.8.self_attn.out_proj.weight": "model.safetensors",
573
+ "vision_tower.encoder.layers.8.self_attn.q_proj.bias": "model.safetensors",
574
+ "vision_tower.encoder.layers.8.self_attn.q_proj.weight": "model.safetensors",
575
+ "vision_tower.encoder.layers.8.self_attn.v_proj.bias": "model.safetensors",
576
+ "vision_tower.encoder.layers.8.self_attn.v_proj.weight": "model.safetensors",
577
+ "vision_tower.encoder.layers.9.layer_norm1.bias": "model.safetensors",
578
+ "vision_tower.encoder.layers.9.layer_norm1.weight": "model.safetensors",
579
+ "vision_tower.encoder.layers.9.layer_norm2.bias": "model.safetensors",
580
+ "vision_tower.encoder.layers.9.layer_norm2.weight": "model.safetensors",
581
+ "vision_tower.encoder.layers.9.mlp.fc1.bias": "model.safetensors",
582
+ "vision_tower.encoder.layers.9.mlp.fc1.weight": "model.safetensors",
583
+ "vision_tower.encoder.layers.9.mlp.fc2.bias": "model.safetensors",
584
+ "vision_tower.encoder.layers.9.mlp.fc2.weight": "model.safetensors",
585
+ "vision_tower.encoder.layers.9.self_attn.k_proj.bias": "model.safetensors",
586
+ "vision_tower.encoder.layers.9.self_attn.k_proj.weight": "model.safetensors",
587
+ "vision_tower.encoder.layers.9.self_attn.out_proj.bias": "model.safetensors",
588
+ "vision_tower.encoder.layers.9.self_attn.out_proj.weight": "model.safetensors",
589
+ "vision_tower.encoder.layers.9.self_attn.q_proj.bias": "model.safetensors",
590
+ "vision_tower.encoder.layers.9.self_attn.q_proj.weight": "model.safetensors",
591
+ "vision_tower.encoder.layers.9.self_attn.v_proj.bias": "model.safetensors",
592
+ "vision_tower.encoder.layers.9.self_attn.v_proj.weight": "model.safetensors",
593
+ "vision_tower.post_layernorm.bias": "model.safetensors",
594
+ "vision_tower.post_layernorm.weight": "model.safetensors"
595
+ }
596
+ }
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 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": null,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|startoftext|>",
5
+ "clean_up_tokenization_spaces": true,
6
+ "eos_token": "<|im_end|>",
7
+ "extra_special_tokens": [],
8
+ "image_end_token": "<|image_end|>",
9
+ "image_start_token": "<|image_start|>",
10
+ "image_thumbnail": "<|img_thumbnail|>",
11
+ "image_token": "<image>",
12
+ "is_local": true,
13
+ "legacy": false,
14
+ "model_max_length": 1000000000000000019884624838656,
15
+ "model_specific_special_tokens": {
16
+ "image_end_token": "<|image_end|>",
17
+ "image_start_token": "<|image_start|>",
18
+ "image_thumbnail": "<|img_thumbnail|>",
19
+ "image_token": "<image>"
20
+ },
21
+ "pad_token": "<|pad|>",
22
+ "processor_class": "Lfm2VlProcessor",
23
+ "return_token_type_ids": false,
24
+ "sp_model_kwargs": {},
25
+ "spaces_between_special_tokens": false,
26
+ "tokenizer_class": "TokenizersBackend",
27
+ "use_default_system_prompt": false,
28
+ "use_fast": true
29
+ }