Hugging Face's logo

isbondarev
/
adv_gemma_test

Safetensors
gemma2
llama-factory
Model card Files
xet
Community
isbondarev commited on
Commit
4ae5ce7
·
verified ·
1 Parent(s): 161b84e

Upload Gemma2ForCausalLM

Browse files
Files changed (5) hide show
  1. config.json +24 -53
  2. generation_config.json +2 -4
  3. model-00001-of-00002.safetensors +1 -1
  4. model-00002-of-00002.safetensors +1 -1
  5. model.safetensors.index.json +0 -0
config.json CHANGED
@@ -1,64 +1,35 @@
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
- "hidden_size": 3840,
12
- "image_token_index": 262144,
 
 
 
13
  "initializer_range": 0.02,
14
- "mm_tokens_per_image": 256,
15
- "model_type": "gemma3",
16
- "text_config": {
17
- "attention_bias": false,
18
- "attention_dropout": 0.0,
19
- "attn_logit_softcapping": null,
20
- "cache_implementation": "hybrid",
21
- "final_logit_softcapping": null,
22
- "head_dim": 256,
23
- "hidden_activation": "gelu_pytorch_tanh",
24
- "hidden_size": 3840,
25
- "initializer_range": 0.02,
26
- "intermediate_size": 15360,
27
- "max_position_embeddings": 131072,
28
- "model_type": "gemma3_text",
29
- "num_attention_heads": 16,
30
- "num_hidden_layers": 48,
31
- "num_key_value_heads": 8,
32
- "query_pre_attn_scalar": 256,
33
- "rms_norm_eps": 1e-06,
34
- "rope_local_base_freq": 10000.0,
35
- "rope_scaling": {
36
- "factor": 8.0,
37
- "rope_type": "linear"
38
- },
39
- "rope_theta": 1000000.0,
40
- "sliding_window": 1024,
41
- "sliding_window_pattern": 6,
42
- "torch_dtype": "bfloat16",
43
- "use_cache": true,
44
- "vocab_size": 262208
45
- },
46
  "torch_dtype": "bfloat16",
47
  "transformers_version": "4.51.3",
48
  "use_cache": true,
49
- "vision_config": {
50
- "attention_dropout": 0.0,
51
- "hidden_act": "gelu_pytorch_tanh",
52
- "hidden_size": 1152,
53
- "image_size": 896,
54
- "intermediate_size": 4304,
55
- "layer_norm_eps": 1e-06,
56
- "model_type": "siglip_vision_model",
57
- "num_attention_heads": 16,
58
- "num_channels": 3,
59
- "num_hidden_layers": 27,
60
- "patch_size": 14,
61
- "torch_dtype": "bfloat16",
62
- "vision_use_head": false
63
- }
64
  }
 
1
  {
2
  "architectures": [
3
+ "Gemma2ForCausalLM"
4
  ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "attn_logit_softcapping": 50.0,
8
+ "bos_token_id": 2,
9
+ "cache_implementation": "hybrid",
10
  "eos_token_id": [
11
  1,
12
+ 107
13
  ],
14
+ "final_logit_softcapping": 30.0,
15
+ "head_dim": 256,
16
+ "hidden_act": "gelu_pytorch_tanh",
17
+ "hidden_activation": "gelu_pytorch_tanh",
18
+ "hidden_size": 2304,
19
  "initializer_range": 0.02,
20
+ "intermediate_size": 9216,
21
+ "max_position_embeddings": 8192,
22
+ "model_type": "gemma2",
23
+ "num_attention_heads": 8,
24
+ "num_hidden_layers": 26,
25
+ "num_key_value_heads": 4,
26
+ "pad_token_id": 0,
27
+ "query_pre_attn_scalar": 256,
28
+ "rms_norm_eps": 1e-06,
29
+ "rope_theta": 10000.0,
30
+ "sliding_window": 4096,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
31
  "torch_dtype": "bfloat16",
32
  "transformers_version": "4.51.3",
33
  "use_cache": true,
34
+ "vocab_size": 256000
 
 
 
 
 
 
 
 
 
 
 
 
 
 
35
  }
generation_config.json CHANGED
@@ -1,13 +1,11 @@
1
  {
 
2
  "bos_token_id": 2,
3
  "cache_implementation": "hybrid",
4
- "do_sample": true,
5
  "eos_token_id": [
6
  1,
7
- 106
8
  ],
9
  "pad_token_id": 0,
10
- "top_k": 64,
11
- "top_p": 0.95,
12
  "transformers_version": "4.51.3"
13
  }
 
1
  {
2
+ "_from_model_config": true,
3
  "bos_token_id": 2,
4
  "cache_implementation": "hybrid",
 
5
  "eos_token_id": [
6
  1,
7
+ 107
8
  ],
9
  "pad_token_id": 0,
 
 
10
  "transformers_version": "4.51.3"
11
  }
model-00001-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:836516dbd5c0aa17816f38277acb2b2e771913e11050cb16d788b6684f18c9bd
3
  size 4988025760
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:adeac12e6f8015ec81792abd7fd4aa5222c0b8b58c0e7e9d9b4d30a3e83f134d
3
  size 4988025760
model-00002-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b929df37a9ee538af363034370ddbd35fa8e3c0fe97cf36cfa05e79b741b2e37
3
  size 240691728
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:525367062e201669b2a2b65ae9103f01014d205aa2a9ff8fa09a2c1bcccd697d
3
  size 240691728
model.safetensors.index.json CHANGED
The diff for this file is too large to render. See raw diff