Update config.json
Browse files- config.json +11 -6
config.json
CHANGED
@@ -1,10 +1,15 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "damerajee/
|
3 |
"architectures": [
|
4 |
-
"
|
5 |
],
|
6 |
"attention_bias": false,
|
7 |
"attention_dropout": 0.0,
|
|
|
|
|
|
|
|
|
|
|
8 |
"bos_token_id": 1,
|
9 |
"eos_token_id": 2,
|
10 |
"hidden_act": "silu",
|
@@ -12,12 +17,12 @@
|
|
12 |
"initializer_range": 0.02,
|
13 |
"intermediate_size": 11008,
|
14 |
"max_position_embeddings": 4096,
|
15 |
-
"model_type": "
|
16 |
"num_attention_heads": 32,
|
17 |
"num_experts_per_tok": 2,
|
18 |
"num_hidden_layers": 32,
|
19 |
"num_key_value_heads": 32,
|
20 |
-
"num_local_experts":
|
21 |
"output_router_logits": false,
|
22 |
"pretraining_tp": 1,
|
23 |
"rms_norm_eps": 1e-05,
|
@@ -27,7 +32,7 @@
|
|
27 |
"sliding_window": null,
|
28 |
"tie_word_embeddings": false,
|
29 |
"torch_dtype": "float16",
|
30 |
-
"transformers_version": "4.
|
31 |
"use_cache": true,
|
32 |
"vocab_size": 48064
|
33 |
-
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "damerajee/Hixtral-test",
|
3 |
"architectures": [
|
4 |
+
"HixtralForCausalLM"
|
5 |
],
|
6 |
"attention_bias": false,
|
7 |
"attention_dropout": 0.0,
|
8 |
+
"auto_map": {
|
9 |
+
"AutoConfig": "damerajee/Hixtral-test--configuration_Hixtral.HixtralConfig",
|
10 |
+
"AutoModelForCausalLM": "damerajee/Hixtral-test--modeling_Hixtral.HixtralForCausalLM",
|
11 |
+
|
12 |
+
},
|
13 |
"bos_token_id": 1,
|
14 |
"eos_token_id": 2,
|
15 |
"hidden_act": "silu",
|
|
|
17 |
"initializer_range": 0.02,
|
18 |
"intermediate_size": 11008,
|
19 |
"max_position_embeddings": 4096,
|
20 |
+
"model_type": "Llamoe",
|
21 |
"num_attention_heads": 32,
|
22 |
"num_experts_per_tok": 2,
|
23 |
"num_hidden_layers": 32,
|
24 |
"num_key_value_heads": 32,
|
25 |
+
"num_local_experts": 4,
|
26 |
"output_router_logits": false,
|
27 |
"pretraining_tp": 1,
|
28 |
"rms_norm_eps": 1e-05,
|
|
|
32 |
"sliding_window": null,
|
33 |
"tie_word_embeddings": false,
|
34 |
"torch_dtype": "float16",
|
35 |
+
"transformers_version": "4.38.2",
|
36 |
"use_cache": true,
|
37 |
"vocab_size": 48064
|
38 |
+
}
|