Hugging Face's logo

drwlf
/
medgemma-4b-it-abliterated

Text Generation
Safetensors
English
gemma3
large-language-model
medical
instruction-following
axolotl
lora
abliteration
medgemma
conversational
8-bit precision
bitsandbytes
Model card Files
xet
Community
1
drwlf commited on
Commit
d6a9cb8
·
verified ·
1 Parent(s): 8728121

Add abliterated medgemma-4b-it model

Browse files
Files changed (6) hide show
  1. config.json +90 -29
  2. generation_config.json +3 -4
  3. model-00001-of-00002.safetensors +2 -2
  4. model-00002-of-00002.safetensors +2 -2
  5. model.safetensors.index.json +0 -0
  6. tokenizer.json +2 -2
config.json CHANGED
@@ -1,37 +1,98 @@
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": 256,
13
- "hidden_activation": "gelu_pytorch_tanh",
14
- "hidden_size": 2560,
15
  "initializer_range": 0.02,
16
- "intermediate_size": 10240,
17
- "max_position_embeddings": 131072,
18
- "model_type": "gemma3_text",
19
- "num_attention_heads": 8,
20
- "num_hidden_layers": 34,
21
- "num_key_value_heads": 4,
22
- "pad_token_id": 0,
23
- "query_pre_attn_scalar": 256,
24
- "rms_norm_eps": 1e-06,
25
- "rope_local_base_freq": 10000,
26
- "rope_scaling": {
27
- "factor": 8.0,
28
- "rope_type": "linear"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
29
  },
30
- "rope_theta": 1000000,
31
- "sliding_window": 1024,
32
- "sliding_window_pattern": 6,
33
  "torch_dtype": "bfloat16",
34
  "transformers_version": "4.52.4",
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
+ "cache_implementation": "hybrid",
20
+ "final_logit_softcapping": null,
21
+ "head_dim": 256,
22
+ "hidden_activation": "gelu_pytorch_tanh",
23
+ "hidden_size": 2560,
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 10240,
26
+ "layer_types": [
27
+ "sliding_attention",
28
+ "sliding_attention",
29
+ "sliding_attention",
30
+ "sliding_attention",
31
+ "sliding_attention",
32
+ "full_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "sliding_attention",
36
+ "sliding_attention",
37
+ "sliding_attention",
38
+ "full_attention",
39
+ "sliding_attention",
40
+ "sliding_attention",
41
+ "sliding_attention",
42
+ "sliding_attention",
43
+ "sliding_attention",
44
+ "full_attention",
45
+ "sliding_attention",
46
+ "sliding_attention",
47
+ "sliding_attention",
48
+ "sliding_attention",
49
+ "sliding_attention",
50
+ "full_attention",
51
+ "sliding_attention",
52
+ "sliding_attention",
53
+ "sliding_attention",
54
+ "sliding_attention",
55
+ "sliding_attention",
56
+ "full_attention",
57
+ "sliding_attention",
58
+ "sliding_attention",
59
+ "sliding_attention",
60
+ "sliding_attention"
61
+ ],
62
+ "max_position_embeddings": 131072,
63
+ "model_type": "gemma3_text",
64
+ "num_attention_heads": 8,
65
+ "num_hidden_layers": 34,
66
+ "num_key_value_heads": 4,
67
+ "query_pre_attn_scalar": 256,
68
+ "rms_norm_eps": 1e-06,
69
+ "rope_local_base_freq": 10000,
70
+ "rope_scaling": {
71
+ "factor": 8.0,
72
+ "rope_type": "linear"
73
+ },
74
+ "rope_theta": 1000000,
75
+ "sliding_window": 1024,
76
+ "sliding_window_pattern": 6,
77
+ "torch_dtype": "bfloat16",
78
+ "use_cache": true,
79
+ "vocab_size": 262208
80
  },
 
 
 
81
  "torch_dtype": "bfloat16",
82
  "transformers_version": "4.52.4",
83
+ "vision_config": {
84
+ "attention_dropout": 0.0,
85
+ "hidden_act": "gelu_pytorch_tanh",
86
+ "hidden_size": 1152,
87
+ "image_size": 896,
88
+ "intermediate_size": 4304,
89
+ "layer_norm_eps": 1e-06,
90
+ "model_type": "siglip_vision_model",
91
+ "num_attention_heads": 16,
92
+ "num_channels": 3,
93
+ "num_hidden_layers": 27,
94
+ "patch_size": 14,
95
+ "torch_dtype": "bfloat16",
96
+ "vision_use_head": false
97
+ }
98
  }
generation_config.json CHANGED
@@ -1,11 +1,10 @@
1
  {
2
- "cache_implementation": "hybrid",
3
- "do_sample": true,
4
  "eos_token_id": [
5
  1,
6
  106
7
  ],
8
- "top_k": 64,
9
- "top_p": 0.95,
10
  "transformers_version": "4.52.4"
11
  }
 
1
  {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 2,
4
  "eos_token_id": [
5
  1,
6
  106
7
  ],
8
+ "pad_token_id": 0,
 
9
  "transformers_version": "4.52.4"
10
  }
model-00001-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:642906c828c96364d979cdad174fbac2d8405a883e26b6ab0b7d48e268a9e488
3
- size 4960531344
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f3a3ac88bb1c5ec31c4661c0978ec21f370ab5f7d52c662aef3a138f82218e6
3
+ size 4961251752
model-00002-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:47564cd9ef58fc718b8c9805da0d41012d25719c666ce2db0ee089afbf2ec99b
3
- size 2800046672
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f8c8de9dedd6b93f39a7e124b603cbd59840a195a34bdb58977274dc285d934a
3
+ size 4981531360
model.safetensors.index.json CHANGED
The diff for this file is too large to render. See raw diff
 
tokenizer.json CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d786405177734910d7a3db625c2826640964a0b4e5cdbbd70620ae3313a01bef
3
- size 33384722
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4667f2089529e8e7657cfb6d1c19910ae71ff5f28aa7ab2ff2763330affad795
3
+ size 33384568