Hugging Face's logo

unsloth
/
gemma-3n-E4B-it-unsloth-bnb-4bit

Image-Text-to-Text
Transformers
Safetensors
English
gemma3n
gemma3
unsloth
gemma
google
conversational
4-bit precision
bitsandbytes
Model card Files
xet
Community
1
danielhanchen commited on
Commit
3035ff2
·
verified ·
1 Parent(s): a580ff6

Upload folder using huggingface_hub

Browse files
Files changed (7) hide show
  1. config.json +43 -5
  2. generation_config.json +1 -1
  3. model-00001-of-00003.safetensors +2 -2
  4. model-00002-of-00003.safetensors +2 -2
  5. model-00003-of-00003.safetensors +2 -2
  6. model.safetensors.index.json +0 -0
  7. preprocessor_config.json +1 -0
config.json CHANGED
@@ -23,6 +23,7 @@
23
  32
24
  ],
25
  "sscp_conv_eps": 0.001,
 
26
  "sscp_conv_kernel_size": [
27
  [
28
  3,
@@ -80,12 +81,12 @@
80
  "modality_router",
81
  "embedding_projection",
82
  "input_proj_linear",
83
- "laurel",
84
- "per_layer_input_gate",
85
  "per_layer_projection",
86
  "audio_tower",
87
  "embed_vision",
88
- "embed_audio"
 
 
89
  ],
90
  "llm_int8_threshold": 6.0,
91
  "load_in_4bit": true,
@@ -143,7 +144,43 @@
143
  "hidden_size": 2048,
144
  "hidden_size_per_layer_input": 256,
145
  "initializer_range": 0.02,
146
- "intermediate_size": 16384,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
147
  "laurel_rank": 64,
148
  "layer_types": [
149
  "sliding_attention",
@@ -200,7 +237,7 @@
200
  "vocab_size_per_layer_input": 262144
201
  },
202
  "torch_dtype": "bfloat16",
203
- "transformers_version": "4.53.0.dev0",
204
  "unsloth_fixed": true,
205
  "vision_config": {
206
  "architecture": "mobilenetv5_300m_enc",
@@ -211,6 +248,7 @@
211
  "LABEL_0",
212
  "LABEL_1"
213
  ],
 
214
  "model_type": "gemma3n_vision",
215
  "num_classes": 2,
216
  "rms_norm_eps": 1e-06,
 
23
  32
24
  ],
25
  "sscp_conv_eps": 0.001,
26
+ "sscp_conv_group_norm_eps": 0.001,
27
  "sscp_conv_kernel_size": [
28
  [
29
  3,
 
81
  "modality_router",
82
  "embedding_projection",
83
  "input_proj_linear",
 
 
84
  "per_layer_projection",
85
  "audio_tower",
86
  "embed_vision",
87
+ "embed_audio",
88
+ "altup_projections",
89
+ "altup_unembed_projections"
90
  ],
91
  "llm_int8_threshold": 6.0,
92
  "load_in_4bit": true,
 
144
  "hidden_size": 2048,
145
  "hidden_size_per_layer_input": 256,
146
  "initializer_range": 0.02,
147
+ "intermediate_size": [
148
+ 16384,
149
+ 16384,
150
+ 16384,
151
+ 16384,
152
+ 16384,
153
+ 16384,
154
+ 16384,
155
+ 16384,
156
+ 16384,
157
+ 16384,
158
+ 16384,
159
+ 16384,
160
+ 16384,
161
+ 16384,
162
+ 16384,
163
+ 16384,
164
+ 16384,
165
+ 16384,
166
+ 16384,
167
+ 16384,
168
+ 16384,
169
+ 16384,
170
+ 16384,
171
+ 16384,
172
+ 16384,
173
+ 16384,
174
+ 16384,
175
+ 16384,
176
+ 16384,
177
+ 16384,
178
+ 16384,
179
+ 16384,
180
+ 16384,
181
+ 16384,
182
+ 16384
183
+ ],
184
  "laurel_rank": 64,
185
  "layer_types": [
186
  "sliding_attention",
 
237
  "vocab_size_per_layer_input": 262144
238
  },
239
  "torch_dtype": "bfloat16",
240
+ "transformers_version": "4.53.0",
241
  "unsloth_fixed": true,
242
  "vision_config": {
243
  "architecture": "mobilenetv5_300m_enc",
 
248
  "LABEL_0",
249
  "LABEL_1"
250
  ],
251
+ "model_args": null,
252
  "model_type": "gemma3n_vision",
253
  "num_classes": 2,
254
  "rms_norm_eps": 1e-06,
generation_config.json CHANGED
@@ -9,5 +9,5 @@
9
  "pad_token_id": 0,
10
  "top_k": 64,
11
  "top_p": 0.95,
12
- "transformers_version": "4.53.0.dev0"
13
  }
 
9
  "pad_token_id": 0,
10
  "top_k": 64,
11
  "top_p": 0.95,
12
+ "transformers_version": "4.53.0"
13
  }
model-00001-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c9f7231d6c9a5e21ba03180dbfc5e0c2fea6e75509af536d72a87c8d20adce2b
3
- size 3764063294
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:267f89a70730a171920867d4045e936c1912d12eea66681dca8331f53d2511f4
3
+ size 3723417366
model-00002-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4cad8a612d448068c5108ba8a0d569a0847d6f63284c2d14ca49c4a1ad681ea5
3
- size 4997088662
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:409d2f4311d680891dccf82c45cafef6e445e9cf9a2933fd978a25225da4e237
3
+ size 4987233092
model-00003-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4dac0124d84e6d36998a5910a745ef1ea8a91c06c3b35ab465355feb01c6c075
3
- size 1101342016
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0ed9c89b1adfd85a3815f1085e8cad5e351cc299407710a1296ed9e6097d38f3
3
+ size 1148535480
model.safetensors.index.json CHANGED
The diff for this file is too large to render. See raw diff
 
preprocessor_config.json CHANGED
@@ -3,6 +3,7 @@
3
  "data_format": "channels_first",
4
  "default_to_square": false,
5
  "device": null,
 
6
  "dither": 0.0,
7
  "do_center_crop": null,
8
  "do_convert_rgb": null,
 
3
  "data_format": "channels_first",
4
  "default_to_square": false,
5
  "device": null,
6
+ "disable_grouping": null,
7
  "dither": 0.0,
8
  "do_center_crop": null,
9
  "do_convert_rgb": null,