rshwndsz commited on
Commit
6aba99c
·
verified ·
1 Parent(s): 0723db1

Update config.json

Browse files
Files changed (1) hide show
  1. config.json +8 -53
config.json CHANGED
@@ -1,31 +1,27 @@
1
  {
2
  "architectures": [
3
- "Gemma3ForSequenceClassification"
4
  ],
 
 
 
 
 
 
5
  "attention_bias": false,
6
  "attention_dropout": 0.0,
7
  "attn_logit_softcapping": null,
8
- "boi_token_index": 255999,
9
  "bos_token_id": 2,
10
  "cache_implementation": "hybrid",
11
- "eoi_token_index": 256000,
12
  "eos_token_id": 1,
13
  "final_logit_softcapping": null,
14
  "head_dim": 256,
15
  "hidden_activation": "gelu_pytorch_tanh",
16
  "hidden_size": 1152,
17
- "id2label": {
18
- "0": "LABEL_0"
19
- },
20
- "image_token_index": 262144,
21
  "initializer_range": 0.02,
22
  "intermediate_size": 6912,
23
- "label2id": {
24
- "LABEL_0": 0
25
- },
26
  "max_position_embeddings": 32768,
27
- "mm_tokens_per_image": 256,
28
- "model_type": "gemma3",
29
  "num_attention_heads": 4,
30
  "num_hidden_layers": 26,
31
  "num_key_value_heads": 1,
@@ -37,49 +33,8 @@
37
  "rope_theta": 1000000,
38
  "sliding_window": 512,
39
  "sliding_window_pattern": 6,
40
- "text_config": {
41
- "attention_bias": false,
42
- "attention_dropout": 0.0,
43
- "attn_logit_softcapping": null,
44
- "cache_implementation": "hybrid",
45
- "final_logit_softcapping": null,
46
- "head_dim": 256,
47
- "hidden_activation": "gelu_pytorch_tanh",
48
- "hidden_size": 2304,
49
- "initializer_range": 0.02,
50
- "intermediate_size": 9216,
51
- "max_position_embeddings": 131072,
52
- "model_type": "gemma3_text",
53
- "num_attention_heads": 8,
54
- "num_hidden_layers": 26,
55
- "num_key_value_heads": 4,
56
- "query_pre_attn_scalar": 256,
57
- "rms_norm_eps": 1e-06,
58
- "rope_local_base_freq": 10000.0,
59
- "rope_scaling": null,
60
- "rope_theta": 1000000.0,
61
- "sliding_window": 4096,
62
- "sliding_window_pattern": 6,
63
- "torch_dtype": "bfloat16",
64
- "use_cache": true,
65
- "vocab_size": 262208
66
- },
67
  "torch_dtype": "bfloat16",
68
  "transformers_version": "4.51.3",
69
  "use_cache": false,
70
- "vision_config": {
71
- "attention_dropout": 0.0,
72
- "hidden_act": "gelu_pytorch_tanh",
73
- "hidden_size": 768,
74
- "image_size": 224,
75
- "intermediate_size": 3072,
76
- "layer_norm_eps": 1e-06,
77
- "model_type": "siglip_vision_model",
78
- "num_attention_heads": 12,
79
- "num_channels": 3,
80
- "num_hidden_layers": 12,
81
- "patch_size": 16,
82
- "torch_dtype": "bfloat16"
83
- },
84
  "vocab_size": 262148
85
  }
 
1
  {
2
  "architectures": [
3
+ "Gemma3TextForSequenceClassification"
4
  ],
5
+ "id2label": {
6
+ "0": "LABEL_0"
7
+ },
8
+ "label2id": {
9
+ "LABEL_0": 0
10
+ },
11
  "attention_bias": false,
12
  "attention_dropout": 0.0,
13
  "attn_logit_softcapping": null,
 
14
  "bos_token_id": 2,
15
  "cache_implementation": "hybrid",
 
16
  "eos_token_id": 1,
17
  "final_logit_softcapping": null,
18
  "head_dim": 256,
19
  "hidden_activation": "gelu_pytorch_tanh",
20
  "hidden_size": 1152,
 
 
 
 
21
  "initializer_range": 0.02,
22
  "intermediate_size": 6912,
 
 
 
23
  "max_position_embeddings": 32768,
24
+ "model_type": "gemma3_text",
 
25
  "num_attention_heads": 4,
26
  "num_hidden_layers": 26,
27
  "num_key_value_heads": 1,
 
33
  "rope_theta": 1000000,
34
  "sliding_window": 512,
35
  "sliding_window_pattern": 6,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
36
  "torch_dtype": "bfloat16",
37
  "transformers_version": "4.51.3",
38
  "use_cache": false,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
39
  "vocab_size": 262148
40
  }