Hugging Face's logo

esrt
/
unsloth-gemma3-27b-ineq-4qpt

Any-to-Any
Transformers
English
gemma3
text-generation-inference
unsloth
Model card Files
xet
Community
Egemen Sert commited on
Commit
0c6fff0
·
verified ·
1 Parent(s): 52f6204

Merge unsloth text config with gemma 3's original config

Browse files
Files changed (1) hide show
  1. config.json +59 -30
config.json CHANGED
@@ -1,37 +1,66 @@
1
  {
2
  "architectures": [
3
- "Gemma3ForCausalLM"
4
  ],
5
- "attention_bias": false,
6
- "attention_dropout": 0.0,
7
- "attn_logit_softcapping": null,
8
- "bos_token_id": 2,
9
- "cache_implementation": "hybrid",
10
- "eos_token_id": 1,
11
- "final_logit_softcapping": null,
12
- "head_dim": 128,
13
- "hidden_activation": "gelu_pytorch_tanh",
14
- "hidden_size": 5376,
15
  "initializer_range": 0.02,
16
- "intermediate_size": 21504,
17
- "max_position_embeddings": 131072,
18
- "model_type": "gemma3_text",
19
- "num_attention_heads": 32,
20
- "num_hidden_layers": 62,
21
- "num_key_value_heads": 16,
22
- "pad_token_id": 0,
23
- "query_pre_attn_scalar": 168,
24
- "rms_norm_eps": 1e-06,
25
- "rope_local_base_freq": 10000.0,
26
- "rope_scaling": {
27
- "factor": 8.0,
28
- "rope_type": "linear"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
29
  },
30
- "rope_theta": 1000000.0,
31
- "sliding_window": 1024,
32
- "sliding_window_pattern": 6,
33
  "torch_dtype": "bfloat16",
34
- "transformers_version": "4.50.0",
35
- "use_cache": true,
36
- "vocab_size": 262208
 
 
 
 
 
 
 
 
 
 
 
 
 
37
  }
 
1
  {
2
  "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
  ],
5
+ "boi_token_index": 255999,
6
+ "eoi_token_index": 256000,
7
+ "eos_token_id": [
8
+ 1,
9
+ 106
10
+ ],
11
+ "image_token_index": 262144,
 
 
 
12
  "initializer_range": 0.02,
13
+ "mm_tokens_per_image": 256,
14
+ "model_type": "gemma3",
15
+ "text_config": {
16
+ "attention_bias": false,
17
+ "attention_dropout": 0.0,
18
+ "attn_logit_softcapping": null,
19
+ "bos_token_id": 2,
20
+ "cache_implementation": "hybrid",
21
+ "eos_token_id": 1,
22
+ "final_logit_softcapping": null,
23
+ "head_dim": 128,
24
+ "hidden_activation": "gelu_pytorch_tanh",
25
+ "hidden_size": 5376,
26
+ "initializer_range": 0.02,
27
+ "intermediate_size": 21504,
28
+ "max_position_embeddings": 131072,
29
+ "model_type": "gemma3_text",
30
+ "num_attention_heads": 32,
31
+ "num_hidden_layers": 62,
32
+ "num_key_value_heads": 16,
33
+ "pad_token_id": 0,
34
+ "query_pre_attn_scalar": 168,
35
+ "rms_norm_eps": 1e-06,
36
+ "rope_local_base_freq": 10000.0,
37
+ "rope_scaling": {
38
+ "factor": 8.0,
39
+ "rope_type": "linear"
40
+ },
41
+ "rope_theta": 1000000.0,
42
+ "sliding_window": 1024,
43
+ "sliding_window_pattern": 6,
44
+ "torch_dtype": "bfloat16",
45
+ "transformers_version": "4.50.0",
46
+ "use_cache": true,
47
+ "vocab_size": 262208
48
  },
 
 
 
49
  "torch_dtype": "bfloat16",
50
+ "transformers_version": "4.50.0.dev0",
51
+ "vision_config": {
52
+ "attention_dropout": 0.0,
53
+ "hidden_act": "gelu_pytorch_tanh",
54
+ "hidden_size": 1152,
55
+ "image_size": 896,
56
+ "intermediate_size": 4304,
57
+ "layer_norm_eps": 1e-06,
58
+ "model_type": "siglip_vision_model",
59
+ "num_attention_heads": 16,
60
+ "num_channels": 3,
61
+ "num_hidden_layers": 27,
62
+ "patch_size": 14,
63
+ "torch_dtype": "float32",
64
+ "vision_use_head": false
65
+ }
66
  }