Upload config.json with huggingface_hub
Browse files- config.json +55 -0
config.json
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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 |
+
"quantization": {
|
16 |
+
"group_size": 64,
|
17 |
+
"bits": 3
|
18 |
+
},
|
19 |
+
"quantization_config": {
|
20 |
+
"group_size": 64,
|
21 |
+
"bits": 3
|
22 |
+
},
|
23 |
+
"text_config": {
|
24 |
+
"attention_bias": false,
|
25 |
+
"attention_dropout": 0.0,
|
26 |
+
"attn_logit_softcapping": null,
|
27 |
+
"cache_implementation": "hybrid",
|
28 |
+
"final_logit_softcapping": null,
|
29 |
+
"head_dim": 128,
|
30 |
+
"hidden_activation": "gelu_pytorch_tanh",
|
31 |
+
"hidden_size": 5376,
|
32 |
+
"initializer_range": 0.02,
|
33 |
+
"intermediate_size": 21504,
|
34 |
+
"max_position_embeddings": 131072,
|
35 |
+
"model_type": "gemma3_text",
|
36 |
+
"num_attention_heads": 32,
|
37 |
+
"num_hidden_layers": 62,
|
38 |
+
"num_key_value_heads": 16,
|
39 |
+
"query_pre_attn_scalar": 168,
|
40 |
+
"rms_norm_eps": 1e-06,
|
41 |
+
"rope_local_base_freq": 10000.0,
|
42 |
+
"rope_scaling": {
|
43 |
+
"factor": 8.0,
|
44 |
+
"rope_type": "linear"
|
45 |
+
},
|
46 |
+
"rope_theta": 1000000.0,
|
47 |
+
"sliding_window": 1024,
|
48 |
+
"sliding_window_pattern": 6,
|
49 |
+
"torch_dtype": "bfloat16",
|
50 |
+
"use_cache": true,
|
51 |
+
"vocab_size": 262208
|
52 |
+
},
|
53 |
+
"torch_dtype": "bfloat16",
|
54 |
+
"transformers_version": "4.50.0.dev0"
|
55 |
+
}
|