Update config.json
Browse files- config.json +13 -13
config.json
CHANGED
|
@@ -12,15 +12,15 @@
|
|
| 12 |
"intermediate_size": 8192,
|
| 13 |
"num_hidden_layers": 28,
|
| 14 |
"num_attention_heads": 24,
|
| 15 |
-
|
| 16 |
"head_dim": 128,
|
| 17 |
"max_position_embeddings": 131072,
|
| 18 |
-
|
| 19 |
"type": "linear",
|
| 20 |
"factor": 32.0
|
| 21 |
},
|
| 22 |
"rope_theta": 500000.0,
|
| 23 |
-
|
| 24 |
"initializer_range": 0.02,
|
| 25 |
"rms_norm_eps": 1e-05,
|
| 26 |
"layer_norm_eps": 1e-05,
|
|
@@ -30,17 +30,17 @@
|
|
| 30 |
"tie_word_embeddings": true,
|
| 31 |
"use_cache": true,
|
| 32 |
"shared_input_output_embedding": true,
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
| 37 |
"ffn_dropout_prob": 0.1,
|
| 38 |
-
|
| 39 |
-
|
| 40 |
"use_rope": true,
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
"output_hidden_states": false,
|
| 45 |
-
|
| 46 |
}
|
|
|
|
| 12 |
"intermediate_size": 8192,
|
| 13 |
"num_hidden_layers": 28,
|
| 14 |
"num_attention_heads": 24,
|
| 15 |
+
"num_key_value_heads": 8,
|
| 16 |
"head_dim": 128,
|
| 17 |
"max_position_embeddings": 131072,
|
| 18 |
+
"rope_scaling": {
|
| 19 |
"type": "linear",
|
| 20 |
"factor": 32.0
|
| 21 |
},
|
| 22 |
"rope_theta": 500000.0,
|
| 23 |
+
"hidden_act": "silu",
|
| 24 |
"initializer_range": 0.02,
|
| 25 |
"rms_norm_eps": 1e-05,
|
| 26 |
"layer_norm_eps": 1e-05,
|
|
|
|
| 30 |
"tie_word_embeddings": true,
|
| 31 |
"use_cache": true,
|
| 32 |
"shared_input_output_embedding": true,
|
| 33 |
+
"attention_dropout": 0.0,
|
| 34 |
+
"hidden_dropout_prob": 0.1,
|
| 35 |
+
"attn_dropout_prob": 0.1,
|
| 36 |
+
"residual_dropout_prob": 0.0,
|
| 37 |
"ffn_dropout_prob": 0.1,
|
| 38 |
+
"output_projection": true,
|
| 39 |
+
"position_embedding_type": "rotary",
|
| 40 |
"use_rope": true,
|
| 41 |
+
"gradient_checkpointing": false,
|
| 42 |
+
"multi_query_attention": false,
|
| 43 |
+
"output_attentions": false,
|
| 44 |
"output_hidden_states": false,
|
| 45 |
+
"layerdrop": 0.0
|
| 46 |
}
|