pretrain core
Browse files- config-0.json +1 -1
- scripts/pretrain_core_model_0.yaml +2 -2
config-0.json
CHANGED
@@ -8,7 +8,7 @@
|
|
8 |
"eos_token_id": 1,
|
9 |
"head_dim": 256,
|
10 |
"hidden_act": "silu",
|
11 |
-
"hidden_size":
|
12 |
"initializer_range": 0.02,
|
13 |
"intermediate_size": 2048,
|
14 |
"max_position_embeddings": 131072,
|
|
|
8 |
"eos_token_id": 1,
|
9 |
"head_dim": 256,
|
10 |
"hidden_act": "silu",
|
11 |
+
"hidden_size": 768,
|
12 |
"initializer_range": 0.02,
|
13 |
"intermediate_size": 2048,
|
14 |
"max_position_embeddings": 131072,
|
scripts/pretrain_core_model_0.yaml
CHANGED
@@ -11,14 +11,14 @@ model_config:
|
|
11 |
padded_vocab_size: 131072
|
12 |
n_layer: 32
|
13 |
n_head: 16
|
14 |
-
n_embd:
|
15 |
n_query_groups: 4
|
16 |
rotary_percentage: 1.0
|
17 |
parallel_residual: False
|
18 |
bias: False
|
19 |
norm_class_name: "RMSNorm"
|
20 |
mlp_class_name: "LLaMAMLP"
|
21 |
-
intermediate_size: 2048 # n_embd *
|
22 |
norm_eps: 1e-5
|
23 |
rope_base: 4300 # https://arxiv.org/pdf/2405.14591
|
24 |
head_size: 256 # n_embd / n_head
|
|
|
11 |
padded_vocab_size: 131072
|
12 |
n_layer: 32
|
13 |
n_head: 16
|
14 |
+
n_embd: 768
|
15 |
n_query_groups: 4
|
16 |
rotary_percentage: 1.0
|
17 |
parallel_residual: False
|
18 |
bias: False
|
19 |
norm_class_name: "RMSNorm"
|
20 |
mlp_class_name: "LLaMAMLP"
|
21 |
+
intermediate_size: 2048 # n_embd * 2.666
|
22 |
norm_eps: 1e-5
|
23 |
rope_base: 4300 # https://arxiv.org/pdf/2405.14591
|
24 |
head_size: 256 # n_embd / n_head
|