Hugging Face's logo

pytorch
/
gemma-3-12b-it-FP8

Any-to-Any
Transformers
PyTorch
English
gemma3
torchao
text-generation-inference
Model card Files
xet
Community
jerryzh168 commited on
Commit
2383fb8
·
verified ·
1 Parent(s): 3b4f142

Upload Gemma3ForConditionalGeneration

Browse files
Files changed (6) hide show
  1. config.json +58 -7
  2. generation_config.json +1 -1
  3. pytorch_model-00001-of-00003.bin +2 -2
  4. pytorch_model-00002-of-00003.bin +2 -2
  5. pytorch_model-00003-of-00003.bin +2 -2
  6. pytorch_model.bin.index.json +1 -0
config.json CHANGED
@@ -3,6 +3,7 @@
3
  "Gemma3ForConditionalGeneration"
4
  ],
5
  "boi_token_index": 255999,
 
6
  "eoi_token_index": 256000,
7
  "eos_token_id": [
8
  1,
@@ -13,6 +14,7 @@
13
  "mm_tokens_per_image": 256,
14
  "model_type": "gemma3",
15
  "quantization_config": {
 
16
  "modules_to_not_convert": null,
17
  "quant_method": "torchao",
18
  "quant_type": {
@@ -59,19 +61,71 @@
59
  "_version": 2
60
  }
61
  },
62
- "quant_type_kwargs": {}
 
63
  },
64
  "text_config": {
 
65
  "attention_bias": false,
66
  "attention_dropout": 0.0,
67
  "attn_logit_softcapping": null,
68
- "cache_implementation": "hybrid",
69
  "final_logit_softcapping": null,
70
  "head_dim": 256,
71
  "hidden_activation": "gelu_pytorch_tanh",
72
  "hidden_size": 3840,
73
  "initializer_range": 0.02,
74
  "intermediate_size": 15360,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
75
  "max_position_embeddings": 131072,
76
  "model_type": "gemma3_text",
77
  "num_attention_heads": 16,
@@ -86,15 +140,13 @@
86
  },
87
  "rope_theta": 1000000.0,
88
  "sliding_window": 1024,
89
- "sliding_window_pattern": 6,
90
- "torch_dtype": "bfloat16",
91
  "use_cache": true,
92
  "vocab_size": 262208
93
  },
94
- "torch_dtype": "bfloat16",
95
- "transformers_version": "4.52.0.dev0",
96
  "vision_config": {
97
  "attention_dropout": 0.0,
 
98
  "hidden_act": "gelu_pytorch_tanh",
99
  "hidden_size": 1152,
100
  "image_size": 896,
@@ -105,7 +157,6 @@
105
  "num_channels": 3,
106
  "num_hidden_layers": 27,
107
  "patch_size": 14,
108
- "torch_dtype": "bfloat16",
109
  "vision_use_head": false
110
  }
111
  }
 
3
  "Gemma3ForConditionalGeneration"
4
  ],
5
  "boi_token_index": 255999,
6
+ "dtype": "bfloat16",
7
  "eoi_token_index": 256000,
8
  "eos_token_id": [
9
  1,
 
14
  "mm_tokens_per_image": 256,
15
  "model_type": "gemma3",
16
  "quantization_config": {
17
+ "include_input_output_embeddings": false,
18
  "modules_to_not_convert": null,
19
  "quant_method": "torchao",
20
  "quant_type": {
 
61
  "_version": 2
62
  }
63
  },
64
+ "quant_type_kwargs": {},
65
+ "untie_embedding_weights": false
66
  },
67
  "text_config": {
68
+ "_sliding_window_pattern": 6,
69
  "attention_bias": false,
70
  "attention_dropout": 0.0,
71
  "attn_logit_softcapping": null,
72
+ "dtype": "bfloat16",
73
  "final_logit_softcapping": null,
74
  "head_dim": 256,
75
  "hidden_activation": "gelu_pytorch_tanh",
76
  "hidden_size": 3840,
77
  "initializer_range": 0.02,
78
  "intermediate_size": 15360,
79
+ "layer_types": [
80
+ "sliding_attention",
81
+ "sliding_attention",
82
+ "sliding_attention",
83
+ "sliding_attention",
84
+ "sliding_attention",
85
+ "full_attention",
86
+ "sliding_attention",
87
+ "sliding_attention",
88
+ "sliding_attention",
89
+ "sliding_attention",
90
+ "sliding_attention",
91
+ "full_attention",
92
+ "sliding_attention",
93
+ "sliding_attention",
94
+ "sliding_attention",
95
+ "sliding_attention",
96
+ "sliding_attention",
97
+ "full_attention",
98
+ "sliding_attention",
99
+ "sliding_attention",
100
+ "sliding_attention",
101
+ "sliding_attention",
102
+ "sliding_attention",
103
+ "full_attention",
104
+ "sliding_attention",
105
+ "sliding_attention",
106
+ "sliding_attention",
107
+ "sliding_attention",
108
+ "sliding_attention",
109
+ "full_attention",
110
+ "sliding_attention",
111
+ "sliding_attention",
112
+ "sliding_attention",
113
+ "sliding_attention",
114
+ "sliding_attention",
115
+ "full_attention",
116
+ "sliding_attention",
117
+ "sliding_attention",
118
+ "sliding_attention",
119
+ "sliding_attention",
120
+ "sliding_attention",
121
+ "full_attention",
122
+ "sliding_attention",
123
+ "sliding_attention",
124
+ "sliding_attention",
125
+ "sliding_attention",
126
+ "sliding_attention",
127
+ "full_attention"
128
+ ],
129
  "max_position_embeddings": 131072,
130
  "model_type": "gemma3_text",
131
  "num_attention_heads": 16,
 
140
  },
141
  "rope_theta": 1000000.0,
142
  "sliding_window": 1024,
 
 
143
  "use_cache": true,
144
  "vocab_size": 262208
145
  },
146
+ "transformers_version": "4.56.1",
 
147
  "vision_config": {
148
  "attention_dropout": 0.0,
149
+ "dtype": "bfloat16",
150
  "hidden_act": "gelu_pytorch_tanh",
151
  "hidden_size": 1152,
152
  "image_size": 896,
 
157
  "num_channels": 3,
158
  "num_hidden_layers": 27,
159
  "patch_size": 14,
 
160
  "vision_use_head": false
161
  }
162
  }
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": "4.52.0.dev0"
13
  }
 
9
  "pad_token_id": 0,
10
  "top_k": 64,
11
  "top_p": 0.95,
12
+ "transformers_version": "4.56.1"
13
  }
pytorch_model-00001-of-00003.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a84bb3ff994e0b70ec42e80540cb1fa9b824c68763fc04efce32b3f3f409eabe
3
- size 4961790497
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eca13cfc42db532fc77e6db21640759af07164d3a8e2c56fe8f4cd1b77c1d5ad
3
+ size 4961789025
pytorch_model-00002-of-00003.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c94026b1087f116fdabb48f8d2fae2b188fe296df56f93063332d7006cfe8635
3
- size 4994952519
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0de26e754e3076c28ba87b14f9c91e307ebe5d74d0e042ac671d33e57216f469
3
+ size 4994951559
pytorch_model-00003-of-00003.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3c0dc6b8dbb09051233e850bae140cd2212e0843522f1e1229da72be69e1d0d5
3
- size 3259115760
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eaa4e0d72293432be91d04382345ad50ab11a477dff35233e85b0b90f0cebf98
3
+ size 3259115184
pytorch_model.bin.index.json CHANGED
@@ -1,5 +1,6 @@
1
  {
2
  "metadata": {
 
3
  "total_size": 13215222432
4
  },
5
  "weight_map": {
 
1
  {
2
  "metadata": {
3
+ "total_parameters": 12187325040,
4
  "total_size": 13215222432
5
  },
6
  "weight_map": {