kobigbird-test45-48294747
Browse files- .gitignore +1 -0
- config.json +35 -0
- pytorch_model.bin +3 -0
- runs/Jan11_01-24-00_6229f9ab93bc/1673400254.27673/events.out.tfevents.1673400254.6229f9ab93bc.427.1 +3 -0
- runs/Jan11_01-24-00_6229f9ab93bc/events.out.tfevents.1673400254.6229f9ab93bc.427.0 +3 -0
- special_tokens_map.json +9 -0
- tokenizer.json +0 -0
- tokenizer_config.json +18 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
.gitignore
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
checkpoint-*/
|
config.json
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "monologg/kobigbird-bert-base",
|
3 |
+
"architectures": [
|
4 |
+
"BigBirdForQuestionAnswering"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"attention_type": "block_sparse",
|
8 |
+
"block_size": 64,
|
9 |
+
"bos_token_id": 5,
|
10 |
+
"classifier_dropout": null,
|
11 |
+
"eos_token_id": 6,
|
12 |
+
"gradient_checkpointing": false,
|
13 |
+
"hidden_act": "gelu_new",
|
14 |
+
"hidden_dropout_prob": 0.1,
|
15 |
+
"hidden_size": 768,
|
16 |
+
"initializer_range": 0.02,
|
17 |
+
"intermediate_size": 3072,
|
18 |
+
"layer_norm_eps": 1e-12,
|
19 |
+
"max_position_embeddings": 4096,
|
20 |
+
"model_type": "big_bird",
|
21 |
+
"num_attention_heads": 12,
|
22 |
+
"num_hidden_layers": 12,
|
23 |
+
"num_random_blocks": 3,
|
24 |
+
"pad_token_id": 0,
|
25 |
+
"position_embedding_type": "absolute",
|
26 |
+
"rescale_embeddings": false,
|
27 |
+
"sep_token_id": 3,
|
28 |
+
"tokenizer_class": "BertTokenizer",
|
29 |
+
"torch_dtype": "float32",
|
30 |
+
"transformers_version": "4.25.1",
|
31 |
+
"type_vocab_size": 2,
|
32 |
+
"use_bias": true,
|
33 |
+
"use_cache": true,
|
34 |
+
"vocab_size": 32500
|
35 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a7024c49717426e14ba823d4b715fb48607f19c7fafd0256a0ec3ff16a9081db
|
3 |
+
size 471657925
|
runs/Jan11_01-24-00_6229f9ab93bc/1673400254.27673/events.out.tfevents.1673400254.6229f9ab93bc.427.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a06c9c5a5c8147cba111aae25b0822dfd5bb8572de577d3bafb53d7a56ba2245
|
3 |
+
size 5539
|
runs/Jan11_01-24-00_6229f9ab93bc/events.out.tfevents.1673400254.6229f9ab93bc.427.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e4fec9b4c349dbdab7d44f6546929209df9bc68ad6411a366cff7a943b4d3245
|
3 |
+
size 5180
|
special_tokens_map.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<s>",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"eos_token": "</s>",
|
5 |
+
"mask_token": "[MASK]",
|
6 |
+
"pad_token": "[PAD]",
|
7 |
+
"sep_token": "[SEP]",
|
8 |
+
"unk_token": "[UNK]"
|
9 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<s>",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"do_basic_tokenize": true,
|
5 |
+
"do_lower_case": false,
|
6 |
+
"eos_token": "</s>",
|
7 |
+
"mask_token": "[MASK]",
|
8 |
+
"model_max_length": 4096,
|
9 |
+
"name_or_path": "monologg/kobigbird-bert-base",
|
10 |
+
"never_split": null,
|
11 |
+
"pad_token": "[PAD]",
|
12 |
+
"sep_token": "[SEP]",
|
13 |
+
"special_tokens_map_file": "/root/.cache/huggingface/hub/models--monologg--kobigbird-bert-base/snapshots/ceacda477e20abef2c929adfa4a07c6f811323be/special_tokens_map.json",
|
14 |
+
"strip_accents": null,
|
15 |
+
"tokenize_chinese_chars": true,
|
16 |
+
"tokenizer_class": "BertTokenizer",
|
17 |
+
"unk_token": "[UNK]"
|
18 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:af4b18ba1518ef6124ab9905d9695a49a24d80ad51a5a1d061397e26ae2b8ce2
|
3 |
+
size 3451
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|