xiaomenshen commited on
Commit
7464d8a
·
verified ·
1 Parent(s): 929aec0

Upload 2 files

Browse files

add kimi k25 eagle3

Files changed (2) hide show
  1. config.json +41 -0
  2. model.safetensors +3 -0
config.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLMEagle3"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 163584,
8
+ "draft_vocab_size": 32000,
9
+ "dtype": "bfloat16",
10
+ "eagle_config": {
11
+ "eagle_aux_hidden_state_layer_ids": [
12
+ 2,
13
+ 30,
14
+ 58
15
+ ],
16
+ "use_aux_hidden_state": true
17
+ },
18
+ "eos_token_id": 163585,
19
+ "head_dim": 128,
20
+ "hidden_act": "silu",
21
+ "hidden_size": 7168,
22
+ "initializer_range": 0.02,
23
+ "intermediate_size": 18432,
24
+ "max_position_embeddings": 262144,
25
+ "max_window_layers": 48,
26
+ "mlp_bias": false,
27
+ "model_type": "llama",
28
+ "num_attention_heads": 64,
29
+ "num_hidden_layers": 1,
30
+ "num_key_value_heads": 64,
31
+ "pretraining_tp": 1,
32
+ "rms_norm_eps": 1e-06,
33
+ "rope_scaling": null,
34
+ "rope_theta": 1000000.0,
35
+ "sliding_window": null,
36
+ "tie_word_embeddings": false,
37
+ "transformers_version": "4.57.1",
38
+ "use_cache": true,
39
+ "use_sliding_window": false,
40
+ "vocab_size": 163840
41
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:128c9f0f306feed7f8447412b4607733d4eb5d9432d35229e22003532aa74dfe
3
+ size 2382319104