Daizee commited on
Commit
ca35455
·
verified ·
1 Parent(s): 572617d

Update config.json

Browse files
Files changed (1) hide show
  1. config.json +57 -50
config.json CHANGED
@@ -1,55 +1,62 @@
1
  {
2
- "architectures": [
3
- "Gemma3ForConditionalGeneration"
4
- ],
5
- "boi_token_index": 255999,
6
- "eoi_token_index": 256000,
7
- "eos_token_id": 1,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8
  "image_token_index": 262144,
 
 
 
 
 
 
 
 
 
 
 
 
9
  "initializer_range": 0.02,
10
- "mm_tokens_per_image": 256,
11
- "model_type": "gemma3",
12
- "quantization": {
13
- "group_size": 64,
14
- "bits": 4,
15
- "mode": "affine"
16
- },
17
- "quantization_config": {
18
- "group_size": 64,
19
- "bits": 4,
20
- "mode": "affine"
21
- },
22
- "text_config": {
23
- "attention_bias": false,
24
- "attention_dropout": 0.0,
25
- "attn_logit_softcapping": null,
26
- "cache_implementation": "hybrid",
27
- "final_logit_softcapping": null,
28
- "head_dim": 256,
29
- "hidden_activation": "gelu_pytorch_tanh",
30
- "hidden_size": 2560,
31
- "initializer_range": 0.02,
32
- "intermediate_size": 10240,
33
- "max_position_embeddings": 131072,
34
- "model_type": "gemma3_text",
35
- "num_attention_heads": 8,
36
- "num_hidden_layers": 34,
37
- "num_key_value_heads": 4,
38
- "query_pre_attn_scalar": 256,
39
- "rms_norm_eps": 1e-06,
40
- "rope_local_base_freq": 10000.0,
41
- "rope_scaling": {
42
- "factor": 8.0,
43
- "rope_type": "linear"
44
- },
45
- "rope_theta": 1000000.0,
46
- "sliding_window": 1024,
47
- "sliding_window_pattern": 6,
48
- "torch_dtype": "bfloat16",
49
- "use_cache": true,
50
- "vocab_size": 262208
51
  },
 
 
 
52
  "torch_dtype": "bfloat16",
53
- "transformers_version": "4.50.0.dev0",
54
- "use_cache": false
55
- }
 
 
 
 
 
1
  {
2
+ "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
+ ],
5
+ "boi_token_index": 255999,
6
+ "eoi_token_index": 256000,
7
+ "eos_token_id": 1,
8
+ "image_token_index": 262144,
9
+ "initializer_range": 0.02,
10
+ "mm_tokens_per_image": 256,
11
+ "model_type": "gemma3",
12
+ "quantization": {
13
+ "group_size": 64,
14
+ "bits": 4,
15
+ "mode": "affine"
16
+ },
17
+ "quantization_config": {
18
+ "group_size": 64,
19
+ "bits": 4,
20
+ "mode": "affine"
21
+ },
22
+ "vision_config": {
23
+ "model_type": "gemma3_vision",
24
+ "mm_tokens_per_image": 256,
25
  "image_token_index": 262144,
26
+ "boi_token_index": 255999,
27
+ "eoi_token_index": 256000
28
+ },
29
+ "text_config": {
30
+ "attention_bias": false,
31
+ "attention_dropout": 0.0,
32
+ "attn_logit_softcapping": null,
33
+ "cache_implementation": "hybrid",
34
+ "final_logit_softcapping": null,
35
+ "head_dim": 256,
36
+ "hidden_activation": "gelu_pytorch_tanh",
37
+ "hidden_size": 2560,
38
  "initializer_range": 0.02,
39
+ "intermediate_size": 10240,
40
+ "max_position_embeddings": 131072,
41
+ "model_type": "gemma3_text",
42
+ "num_attention_heads": 8,
43
+ "num_hidden_layers": 34,
44
+ "num_key_value_heads": 4,
45
+ "query_pre_attn_scalar": 256,
46
+ "rms_norm_eps": 1e-06,
47
+ "rope_local_base_freq": 10000.0,
48
+ "rope_scaling": {
49
+ "factor": 8.0,
50
+ "rope_type": "linear"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
51
  },
52
+ "rope_theta": 1000000.0,
53
+ "sliding_window": 1024,
54
+ "sliding_window_pattern": 6,
55
  "torch_dtype": "bfloat16",
56
+ "use_cache": true,
57
+ "vocab_size": 262208
58
+ },
59
+ "torch_dtype": "bfloat16",
60
+ "transformers_version": "4.50.0.dev0",
61
+ "use_cache": false
62
+ }