Update config.json
Browse files- config.json +48 -47
config.json
CHANGED
@@ -1,47 +1,48 @@
|
|
1 |
-
{
|
2 |
-
"_name_or_path": "./results/Doge-60M",
|
3 |
-
"architectures": [
|
4 |
-
"DogeForCausalLM"
|
5 |
-
],
|
6 |
-
"attention_dropout": 0.0,
|
7 |
-
"auto_map": {
|
8 |
-
"AutoConfig": "configuration_doge.DogeConfig",
|
9 |
-
"AutoModelForCausalLM": "modeling_doge.DogeForCausalLM"
|
10 |
-
},
|
11 |
-
"bos_token_id": 0,
|
12 |
-
"dynamic_mask_ratio": 0.0,
|
13 |
-
"eos_token_id": 1,
|
14 |
-
"expert_retrieval_size": 256,
|
15 |
-
"hidden_act": "silu",
|
16 |
-
"hidden_bias": false,
|
17 |
-
"hidden_dropout": 0.0,
|
18 |
-
"hidden_size": 512,
|
19 |
-
"initializer_range": 0.02,
|
20 |
-
"intermediate_size": 1024,
|
21 |
-
"is_moe": false,
|
22 |
-
"max_position_embeddings": 2048,
|
23 |
-
"model_type": "doge",
|
24 |
-
"num_attention_heads": 4,
|
25 |
-
"num_cdmmoe_experts": 2048,
|
26 |
-
"num_cdmmoe_experts_per_head": 8,
|
27 |
-
"num_cdmmoe_heads": 4,
|
28 |
-
"num_cdmoe_experts": 16348,
|
29 |
-
"num_cdmoe_experts_per_head": 8,
|
30 |
-
"num_cdmoe_heads": 4,
|
31 |
-
"num_channels": 3,
|
32 |
-
"num_hidden_layers": 16,
|
33 |
-
"num_key_value_heads": 2,
|
34 |
-
"pad_token_id": 2,
|
35 |
-
"patch_size": 16,
|
36 |
-
"rms_norm_eps": 1e-06,
|
37 |
-
"rope_scaling": {
|
38 |
-
"factor": 4.0,
|
39 |
-
"original_max_position_embeddings": 2048,
|
40 |
-
"rope_type": "dynamic"
|
41 |
-
},
|
42 |
-
"rope_theta": 10000.0,
|
43 |
-
"
|
44 |
-
"
|
45 |
-
"
|
46 |
-
"
|
47 |
-
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "./results/Doge-60M",
|
3 |
+
"architectures": [
|
4 |
+
"DogeForCausalLM"
|
5 |
+
],
|
6 |
+
"attention_dropout": 0.0,
|
7 |
+
"auto_map": {
|
8 |
+
"AutoConfig": "configuration_doge.DogeConfig",
|
9 |
+
"AutoModelForCausalLM": "modeling_doge.DogeForCausalLM"
|
10 |
+
},
|
11 |
+
"bos_token_id": 0,
|
12 |
+
"dynamic_mask_ratio": 0.0,
|
13 |
+
"eos_token_id": 1,
|
14 |
+
"expert_retrieval_size": 256,
|
15 |
+
"hidden_act": "silu",
|
16 |
+
"hidden_bias": false,
|
17 |
+
"hidden_dropout": 0.0,
|
18 |
+
"hidden_size": 512,
|
19 |
+
"initializer_range": 0.02,
|
20 |
+
"intermediate_size": 1024,
|
21 |
+
"is_moe": false,
|
22 |
+
"max_position_embeddings": 2048,
|
23 |
+
"model_type": "doge",
|
24 |
+
"num_attention_heads": 4,
|
25 |
+
"num_cdmmoe_experts": 2048,
|
26 |
+
"num_cdmmoe_experts_per_head": 8,
|
27 |
+
"num_cdmmoe_heads": 4,
|
28 |
+
"num_cdmoe_experts": 16348,
|
29 |
+
"num_cdmoe_experts_per_head": 8,
|
30 |
+
"num_cdmoe_heads": 4,
|
31 |
+
"num_channels": 3,
|
32 |
+
"num_hidden_layers": 16,
|
33 |
+
"num_key_value_heads": 2,
|
34 |
+
"pad_token_id": 2,
|
35 |
+
"patch_size": 16,
|
36 |
+
"rms_norm_eps": 1e-06,
|
37 |
+
"rope_scaling": {
|
38 |
+
"factor": 4.0,
|
39 |
+
"original_max_position_embeddings": 2048,
|
40 |
+
"rope_type": "dynamic"
|
41 |
+
},
|
42 |
+
"rope_theta": 10000.0,
|
43 |
+
"tie_word_embeddings": true,
|
44 |
+
"torch_dtype": "float32",
|
45 |
+
"transformers_version": "4.48.1",
|
46 |
+
"use_cache": true,
|
47 |
+
"vocab_size": 32768
|
48 |
+
}
|