Upload config.json
Browse files- config.json +35 -0
config.json
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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": "4.42.4",
|
33 |
+
"use_cache": true,
|
34 |
+
"vocab_size": 256000
|
35 |
+
}
|