eyad-silx commited on
Commit
cfcce5e
·
verified ·
1 Parent(s): a1330e0

Upload folder using huggingface_hub

Browse files
config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "gelu",
3
+ "architectures": [
4
+ "LNNModel"
5
+ ],
6
+ "dt": 1.0,
7
+ "expert_dim": 2048,
8
+ "hidden_size": 8192,
9
+ "initializer_range": 0.02,
10
+ "lambda_res": 0.0,
11
+ "model_type": "lnn",
12
+ "moe_load_balance_loss_weight": 0.01,
13
+ "num_experts": 128,
14
+ "num_experts_per_tok": 2,
15
+ "num_hidden_layers": 96,
16
+ "pmb_num_blocks": 1024,
17
+ "pmb_slots_per_block": 4096,
18
+ "pmb_top_k": 1,
19
+ "torch_dtype": "bfloat16",
20
+ "transformers_version": "4.35.2",
21
+ "use_moe": true,
22
+ "use_pmb": true,
23
+ "vocab_size": 128815
24
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|begin▁of▁sentence|>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|end▁of▁sentence|>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<|end▁of▁sentence|>",
18
+ "lstrip": false,
19
+ "normalized": true,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ }
23
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
The diff for this file is too large to render. See raw diff