user10383 commited on
Commit
99200e1
·
verified ·
1 Parent(s): cc42726

Upload final Khasi LM files (excluding checkpoints)

Browse files
khasi_sp_transformer.model CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c25ca7ce45754e1c86c651c9bac10919ecd1d69f9dc74a216d9baac61cb86094
3
  size 160907
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4fc0f5537455a36a01efa4386b97432863661bd3c63dad8ea689da732075707
3
  size 160907
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3b74d9ba31c40647e8e8f365960033f36836d57d81f1145811f7d99a72ac5d41
3
- size 33254160
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:48135694d7c532ab85eca020fcfb34fc94510e24fe4fecb6f2f6d55210108499
3
+ size 33647376
tokenizer.model CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c25ca7ce45754e1c86c651c9bac10919ecd1d69f9dc74a216d9baac61cb86094
3
  size 160907
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4fc0f5537455a36a01efa4386b97432863661bd3c63dad8ea689da732075707
3
  size 160907
tokenizer_config.json CHANGED
@@ -1,13 +1,8 @@
1
  {
2
  "vocab_size": 10000,
3
- "actual_vocab_size": 10000,
4
  "model_type": "bpe",
5
  "pad_token_id": 0,
6
  "unk_token_id": 1,
7
  "bos_token_id": 2,
8
- "eos_token_id": 3,
9
- "pad_token": "<pad>",
10
- "unk_token": "<unk>",
11
- "bos_token": "<s>",
12
- "eos_token": "</s>"
13
  }
 
1
  {
2
  "vocab_size": 10000,
 
3
  "model_type": "bpe",
4
  "pad_token_id": 0,
5
  "unk_token_id": 1,
6
  "bos_token_id": 2,
7
+ "eos_token_id": 3
 
 
 
 
8
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f77eabe6259eaf009141558b1de07115969f1e8519da2378b89371869cf11aad
3
  size 5368
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4966bc672152f28cd9a5d7961231e5cea0fc95976ccccbdb206b91f23c7738aa
3
  size 5368