Norod78 commited on
Commit
d6e6407
1 Parent(s): 574871e

Hebrew gpt_neo small model ckpt-80500 commit from $USER

Browse files
config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPTNeoForCausalLM"
5
+ ],
6
+ "attention_dropout": 0,
7
+ "attention_layers": [
8
+ "global",
9
+ "global",
10
+ "global",
11
+ "global",
12
+ "global",
13
+ "global",
14
+ "global",
15
+ "global",
16
+ "global",
17
+ "global",
18
+ "global",
19
+ "global"
20
+ ],
21
+ "attention_types": [
22
+ [
23
+ [
24
+ "global"
25
+ ],
26
+ 12
27
+ ]
28
+ ],
29
+ "bos_token_id": 50256,
30
+ "embed_dropout": 0,
31
+ "eos_token_id": 50256,
32
+ "gradient_checkpointing": false,
33
+ "hidden_size": 768,
34
+ "initializer_range": 0.02,
35
+ "intermediate_size": null,
36
+ "layer_norm_epsilon": 1e-05,
37
+ "max_position_embeddings": 2048,
38
+ "model_type": "gpt_neo",
39
+ "num_heads": 12,
40
+ "num_layers": 12,
41
+ "resid_dropout": 0,
42
+ "summary_activation": null,
43
+ "summary_first_dropout": 0.1,
44
+ "summary_proj_to_labels": true,
45
+ "summary_type": "cls_index",
46
+ "summary_use_proj": true,
47
+ "transformers_version": "4.5.1",
48
+ "use_cache": true,
49
+ "vocab_size": 50257,
50
+ "window_size": 256
51
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e609d2fc34d1d25d1445c926ca7941e15d644cd1339de501d891112eddf97f2e
3
+ size 551187055
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "<|startoftext|>", "eos_token": "<|endoftext|>", "unk_token": "<|unknown|>"}
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"do_lower_case": false, "max_len": 1024, "bos_token": "<|startoftext|>", "eos_token": "<|endoftext|>", "unk_token": "<|endoftext|>", "special_tokens_map_file": "special_tokens_map.json", "full_tokenizer_file": null}
vocab.json ADDED
The diff for this file is too large to render. See raw diff