Update config.json
Browse files- config.json +7 -6
config.json
CHANGED
@@ -16,21 +16,22 @@
|
|
16 |
"head_dim": 256,
|
17 |
"hidden_act": "gelu_pytorch_tanh",
|
18 |
"hidden_activation": "gelu_pytorch_tanh",
|
19 |
-
"hidden_size":
|
20 |
"initializer_range": 0.02,
|
21 |
-
"intermediate_size":
|
22 |
"max_position_embeddings": 8192,
|
23 |
"model_type": "gemma2",
|
24 |
-
"num_attention_heads":
|
25 |
-
"num_hidden_layers":
|
26 |
-
"num_key_value_heads":
|
27 |
"pad_token_id": 0,
|
28 |
"query_pre_attn_scalar": 256,
|
29 |
"rms_norm_eps": 1e-06,
|
30 |
"rope_theta": 10000.0,
|
31 |
"sliding_window": 4096,
|
|
|
32 |
"torch_dtype": "bfloat16",
|
33 |
-
"transformers_version": "4.
|
34 |
"use_cache": true,
|
35 |
"vocab_size": 256000
|
36 |
}
|
|
|
16 |
"head_dim": 256,
|
17 |
"hidden_act": "gelu_pytorch_tanh",
|
18 |
"hidden_activation": "gelu_pytorch_tanh",
|
19 |
+
"hidden_size": 3584,
|
20 |
"initializer_range": 0.02,
|
21 |
+
"intermediate_size": 14336,
|
22 |
"max_position_embeddings": 8192,
|
23 |
"model_type": "gemma2",
|
24 |
+
"num_attention_heads": 16,
|
25 |
+
"num_hidden_layers": 42,
|
26 |
+
"num_key_value_heads": 8,
|
27 |
"pad_token_id": 0,
|
28 |
"query_pre_attn_scalar": 256,
|
29 |
"rms_norm_eps": 1e-06,
|
30 |
"rope_theta": 10000.0,
|
31 |
"sliding_window": 4096,
|
32 |
+
"sliding_window_size": 4096,
|
33 |
"torch_dtype": "bfloat16",
|
34 |
+
"transformers_version": "4.44.2",
|
35 |
"use_cache": true,
|
36 |
"vocab_size": 256000
|
37 |
}
|