Hugging Face's logo

kontextdev
/
agent-gemma

Image-Text-to-Text
Transformers
Safetensors
English
gemma3n
function-calling
tool-use
on-device
mobile
gemma
litertlm
conversational
Model card Files
xet
 Community
1
macmacmacmac commited on
Commit
4174976
·
verified ·
1 Parent(s): da7f35e

Upload folder using huggingface_hub

Browse files
Files changed (12) hide show
  1. .gitattributes +1 -0
  2. chat_template.jinja +13 -0
  3. config.json +220 -0
  4. generation_config.json +13 -0
  5. model-00001-of-00005.safetensors +3 -0
  6. model-00002-of-00005.safetensors +3 -0
  7. model-00003-of-00005.safetensors +3 -0
  8. model-00004-of-00005.safetensors +3 -0
  9. model-00005-of-00005.safetensors +3 -0
  10. model.safetensors.index.json +0 -0
  11. tokenizer.json +3 -0
  12. tokenizer_config.json +30 -0
.gitattributes CHANGED
@@ -34,3 +34,4 @@ saved_model/**/* 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
  gemma-3n-E2B-it-agent-fixed.litertlm 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
  gemma-3n-E2B-it-agent-fixed.litertlm filter=lfs diff=lfs merge=lfs -text
37
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
chat_template.jinja ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- for message in messages -%}{%- if message.role == 'developer' or message.role == 'system' -%}<start_of_turn>developer
2
+ {{ message.content }}{%- if tools is defined and tools|length > 0 %}
3
+
4
+ Available tools:{%- for tool in tools %}
5
+ <start_function_declaration>{%- if tool.function is defined %}{{ tool.function | tojson }}{%- else %}{{ tool | tojson }}{%- endif %}<end_function_declaration>{%- endfor %}{%- endif %}<end_of_turn>
6
+ {%- elif message.role == 'user' -%}<start_of_turn>user
7
+ {{ message.content }}<end_of_turn>
8
+ {%- elif message.role == 'model' or message.role == 'assistant' -%}<start_of_turn>model
9
+ {%- if message.tool_calls is defined and message.tool_calls -%}{%- for tc in message.tool_calls -%}<start_function_call>call:{{ tc.function.name }}{{ '{' }}{%- for k, v in tc.function.arguments.items() -%}{{ k }}:<escape>{{ v }}<escape>{%- if not loop.last %},{% endif -%}{%- endfor -%}{{ '}' }}<end_function_call>{%- endfor -%}{%- else -%}{{ message.content }}{%- endif -%}<end_of_turn>
10
+ {%- elif message.role == 'tool' -%}<start_of_turn>tool
11
+ {{ message.content }}<end_of_turn>
12
+ {%- endif -%}{%- endfor -%}{%- if add_generation_prompt -%}<start_of_turn>model
13
+ {%- endif -%}
config.json ADDED
@@ -0,0 +1,220 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3nForConditionalGeneration"
4
+ ],
5
+ "audio_config": {
6
+ "conf_attention_chunk_size": 12,
7
+ "conf_attention_context_left": 13,
8
+ "conf_attention_context_right": 0,
9
+ "conf_attention_logit_cap": 50.0,
10
+ "conf_conv_kernel_size": 5,
11
+ "conf_num_attention_heads": 8,
12
+ "conf_num_hidden_layers": 12,
13
+ "conf_reduction_factor": 4,
14
+ "conf_residual_weight": 0.5,
15
+ "dtype": "bfloat16",
16
+ "gradient_clipping": 10000000000.0,
17
+ "hidden_size": 1536,
18
+ "input_feat_size": 128,
19
+ "model_type": "gemma3n_audio",
20
+ "rms_norm_eps": 1e-06,
21
+ "sscp_conv_channel_size": [
22
+ 128,
23
+ 32
24
+ ],
25
+ "sscp_conv_group_norm_eps": 0.001,
26
+ "sscp_conv_kernel_size": [
27
+ [
28
+ 3,
29
+ 3
30
+ ],
31
+ [
32
+ 3,
33
+ 3
34
+ ]
35
+ ],
36
+ "sscp_conv_stride_size": [
37
+ [
38
+ 2,
39
+ 2
40
+ ],
41
+ [
42
+ 2,
43
+ 2
44
+ ]
45
+ ],
46
+ "vocab_offset": 262272,
47
+ "vocab_size": 128
48
+ },
49
+ "audio_soft_tokens_per_image": 188,
50
+ "audio_token_id": 262273,
51
+ "boa_token_id": 256000,
52
+ "boi_token_id": 255999,
53
+ "dtype": "bfloat16",
54
+ "eoa_token_id": 262272,
55
+ "eoi_token_id": 262144,
56
+ "eos_token_id": [
57
+ 1,
58
+ 106
59
+ ],
60
+ "image_token_id": 262145,
61
+ "initializer_range": 0.02,
62
+ "model_type": "gemma3n",
63
+ "text_config": {
64
+ "activation_sparsity_pattern": [
65
+ 0.95,
66
+ 0.95,
67
+ 0.95,
68
+ 0.95,
69
+ 0.95,
70
+ 0.95,
71
+ 0.95,
72
+ 0.95,
73
+ 0.95,
74
+ 0.95,
75
+ 0.0,
76
+ 0.0,
77
+ 0.0,
78
+ 0.0,
79
+ 0.0,
80
+ 0.0,
81
+ 0.0,
82
+ 0.0,
83
+ 0.0,
84
+ 0.0,
85
+ 0.0,
86
+ 0.0,
87
+ 0.0,
88
+ 0.0,
89
+ 0.0,
90
+ 0.0,
91
+ 0.0,
92
+ 0.0,
93
+ 0.0,
94
+ 0.0
95
+ ],
96
+ "altup_active_idx": 0,
97
+ "altup_coef_clip": 120.0,
98
+ "altup_correct_scale": true,
99
+ "altup_num_inputs": 4,
100
+ "attention_bias": false,
101
+ "attention_dropout": 0.0,
102
+ "bos_token_id": 2,
103
+ "dtype": "bfloat16",
104
+ "eos_token_id": 1,
105
+ "final_logit_softcapping": 30.0,
106
+ "head_dim": 256,
107
+ "hidden_activation": "gelu_pytorch_tanh",
108
+ "hidden_size": 2048,
109
+ "hidden_size_per_layer_input": 256,
110
+ "initializer_range": 0.02,
111
+ "intermediate_size": [
112
+ 8192,
113
+ 8192,
114
+ 8192,
115
+ 8192,
116
+ 8192,
117
+ 8192,
118
+ 8192,
119
+ 8192,
120
+ 8192,
121
+ 8192,
122
+ 8192,
123
+ 8192,
124
+ 8192,
125
+ 8192,
126
+ 8192,
127
+ 8192,
128
+ 8192,
129
+ 8192,
130
+ 8192,
131
+ 8192,
132
+ 8192,
133
+ 8192,
134
+ 8192,
135
+ 8192,
136
+ 8192,
137
+ 8192,
138
+ 8192,
139
+ 8192,
140
+ 8192,
141
+ 8192
142
+ ],
143
+ "laurel_rank": 64,
144
+ "layer_types": [
145
+ "sliding_attention",
146
+ "sliding_attention",
147
+ "sliding_attention",
148
+ "sliding_attention",
149
+ "full_attention",
150
+ "sliding_attention",
151
+ "sliding_attention",
152
+ "sliding_attention",
153
+ "sliding_attention",
154
+ "full_attention",
155
+ "sliding_attention",
156
+ "sliding_attention",
157
+ "sliding_attention",
158
+ "sliding_attention",
159
+ "full_attention",
160
+ "sliding_attention",
161
+ "sliding_attention",
162
+ "sliding_attention",
163
+ "sliding_attention",
164
+ "full_attention",
165
+ "sliding_attention",
166
+ "sliding_attention",
167
+ "sliding_attention",
168
+ "sliding_attention",
169
+ "full_attention",
170
+ "sliding_attention",
171
+ "sliding_attention",
172
+ "sliding_attention",
173
+ "sliding_attention",
174
+ "full_attention"
175
+ ],
176
+ "max_position_embeddings": 32768,
177
+ "model_type": "gemma3n_text",
178
+ "num_attention_heads": 8,
179
+ "num_hidden_layers": 30,
180
+ "num_key_value_heads": 2,
181
+ "num_kv_shared_layers": 10,
182
+ "pad_token_id": 0,
183
+ "rms_norm_eps": 1e-06,
184
+ "rope_parameters": {
185
+ "full_attention": {
186
+ "rope_theta": 1000000.0,
187
+ "rope_type": "default"
188
+ },
189
+ "sliding_attention": {
190
+ "rope_theta": 10000.0,
191
+ "rope_type": "default"
192
+ }
193
+ },
194
+ "sliding_window": 512,
195
+ "tie_word_embeddings": true,
196
+ "use_cache": true,
197
+ "vocab_size": 262400,
198
+ "vocab_size_per_layer_input": 262144
199
+ },
200
+ "tie_word_embeddings": true,
201
+ "transformers_version": "5.3.0",
202
+ "vision_config": {
203
+ "architecture": "mobilenetv5_300m_enc",
204
+ "do_pooling": false,
205
+ "dtype": "bfloat16",
206
+ "hidden_size": 2048,
207
+ "initializer_range": 0.02,
208
+ "label_names": [
209
+ "LABEL_0",
210
+ "LABEL_1"
211
+ ],
212
+ "model_args": null,
213
+ "model_type": "gemma3n_vision",
214
+ "num_classes": 2,
215
+ "rms_norm_eps": 1e-06,
216
+ "vocab_offset": 262144,
217
+ "vocab_size": 128
218
+ },
219
+ "vision_soft_tokens_per_image": 256
220
+ }
generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "cache_implementation": "hybrid",
4
+ "do_sample": true,
5
+ "eos_token_id": [
6
+ 1,
7
+ 106
8
+ ],
9
+ "pad_token_id": 0,
10
+ "top_k": 64,
11
+ "top_p": 0.95,
12
+ "transformers_version": "5.3.0"
13
+ }
model-00001-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:335695effdb54cc2a857dca67ccc6ed92ddb7f903d9630eb511d7a24db455fa8
3
+ size 1428679136
model-00002-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:991af92a17279622f17fdcb380fdca68a44fe079b7389ed38d1f2298e45a0469
3
+ size 4026532008
model-00003-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53f31eae71dc484d3f67b21071d9a47fbb46d0f5d18d4b66bf6054a4328aa196
3
+ size 1979071768
model-00004-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:74d72e37e18608f956fa174be7fc063697729c66eccf7b7a50990bb5eef1fc69
3
+ size 1988955896
model-00005-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f3f3a75fab0c7c87d2c36b4530b694c00f1dea583ffbc754ef579b1820b93c39
3
+ size 1455846048
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6aa40b46d55d1fa5f1c0b218de3a695942c620858ba34aa94e2aa705efc6f071
3
+ size 33442552
tokenizer_config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "audio_token": "<audio_soft_token>",
3
+ "backend": "tokenizers",
4
+ "boa_token": "<start_of_audio>",
5
+ "boi_token": "<start_of_image>",
6
+ "bos_token": "<bos>",
7
+ "clean_up_tokenization_spaces": false,
8
+ "eoa_token": "<end_of_audio>",
9
+ "eoi_token": "<end_of_image>",
10
+ "eos_token": "<eos>",
11
+ "image_token": "<image_soft_token>",
12
+ "is_local": true,
13
+ "mask_token": "<mask>",
14
+ "model_max_length": 1000000000000000019884624838656,
15
+ "model_specific_special_tokens": {
16
+ "audio_token": "<audio_soft_token>",
17
+ "boa_token": "<start_of_audio>",
18
+ "boi_token": "<start_of_image>",
19
+ "eoa_token": "<end_of_audio>",
20
+ "eoi_token": "<end_of_image>",
21
+ "image_token": "<image_soft_token>"
22
+ },
23
+ "pad_token": "<pad>",
24
+ "processor_class": "Gemma3nProcessor",
25
+ "sp_model_kwargs": null,
26
+ "spaces_between_special_tokens": false,
27
+ "tokenizer_class": "GemmaTokenizer",
28
+ "unk_token": "<unk>",
29
+ "use_default_system_prompt": false
30
+ }