Update config.json
Browse files- config.json +10 -8
config.json
CHANGED
|
@@ -8,27 +8,29 @@
|
|
| 8 |
"attn_logit_softcapping": 50.0,
|
| 9 |
"bos_token_id": 2,
|
| 10 |
"cache_implementation": "hybrid",
|
| 11 |
-
"eos_token_id":
|
|
|
|
|
|
|
|
|
|
| 12 |
"final_logit_softcapping": 30.0,
|
| 13 |
"head_dim": 256,
|
| 14 |
"hidden_act": "gelu_pytorch_tanh",
|
| 15 |
"hidden_activation": "gelu_pytorch_tanh",
|
| 16 |
-
"hidden_size":
|
| 17 |
"initializer_range": 0.02,
|
| 18 |
-
"intermediate_size":
|
| 19 |
"max_position_embeddings": 8192,
|
| 20 |
"model_type": "gemma2",
|
| 21 |
-
"num_attention_heads":
|
| 22 |
-
"num_hidden_layers":
|
| 23 |
-
"num_key_value_heads":
|
| 24 |
"pad_token_id": 0,
|
| 25 |
"query_pre_attn_scalar": 256,
|
| 26 |
"rms_norm_eps": 1e-06,
|
| 27 |
"rope_theta": 10000.0,
|
| 28 |
"sliding_window": 4096,
|
| 29 |
-
"sliding_window_size": 4096,
|
| 30 |
"torch_dtype": "bfloat16",
|
| 31 |
-
"transformers_version": "4.
|
| 32 |
"use_cache": true,
|
| 33 |
"vocab_size": 256000
|
| 34 |
}
|
|
|
|
| 8 |
"attn_logit_softcapping": 50.0,
|
| 9 |
"bos_token_id": 2,
|
| 10 |
"cache_implementation": "hybrid",
|
| 11 |
+
"eos_token_id": [
|
| 12 |
+
1,
|
| 13 |
+
107
|
| 14 |
+
],
|
| 15 |
"final_logit_softcapping": 30.0,
|
| 16 |
"head_dim": 256,
|
| 17 |
"hidden_act": "gelu_pytorch_tanh",
|
| 18 |
"hidden_activation": "gelu_pytorch_tanh",
|
| 19 |
+
"hidden_size": 2304,
|
| 20 |
"initializer_range": 0.02,
|
| 21 |
+
"intermediate_size": 9216,
|
| 22 |
"max_position_embeddings": 8192,
|
| 23 |
"model_type": "gemma2",
|
| 24 |
+
"num_attention_heads": 8,
|
| 25 |
+
"num_hidden_layers": 26,
|
| 26 |
+
"num_key_value_heads": 4,
|
| 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": "4.42.4",
|
| 34 |
"use_cache": true,
|
| 35 |
"vocab_size": 256000
|
| 36 |
}
|