atsuki-yamaguchi
commited on
Commit
•
b94e6e7
1
Parent(s):
2a3007e
Upload config.json with huggingface_hub
Browse files- config.json +35 -35
config.json
CHANGED
@@ -1,36 +1,36 @@
|
|
1 |
{
|
2 |
-
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "google/gemma-2-9b",
|
3 |
+
"architectures": [
|
4 |
+
"Gemma2ForMultiCausalLM"
|
5 |
+
],
|
6 |
+
"attention_bias": false,
|
7 |
+
"attention_dropout": 0.0,
|
8 |
+
"attn_logit_softcapping": 50.0,
|
9 |
+
"bos_token_id": 2,
|
10 |
+
"cache_implementation": "hybrid",
|
11 |
+
"copy_lm_head": true,
|
12 |
+
"eos_token_id": 1,
|
13 |
+
"final_logit_softcapping": 30.0,
|
14 |
+
"head_dim": 256,
|
15 |
+
"hidden_act": "gelu_pytorch_tanh",
|
16 |
+
"hidden_activation": "gelu_pytorch_tanh",
|
17 |
+
"hidden_size": 3584,
|
18 |
+
"initializer_range": 0.02,
|
19 |
+
"intermediate_size": 14336,
|
20 |
+
"max_position_embeddings": 8192,
|
21 |
+
"model_type": "gemma2",
|
22 |
+
"num_attention_heads": 16,
|
23 |
+
"num_hidden_layers": 42,
|
24 |
+
"num_key_value_heads": 8,
|
25 |
+
"num_lm_heads": 1,
|
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 |
+
"sliding_window_size": 4096,
|
32 |
+
"torch_dtype": "float16",
|
33 |
+
"transformers_version": "4.44.0.dev0",
|
34 |
+
"use_cache": true,
|
35 |
+
"vocab_size": 256104
|
36 |
+
}
|