hf-transformers-bot
commited on
Commit
•
a6d5daf
1
Parent(s):
3d59243
Upload tiny models for TapasForMaskedLM
Browse files- config.json +56 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +10 -0
- tokenizer_config.json +28 -0
- vocab.txt +0 -0
config.json
ADDED
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"agg_temperature": 1.0,
|
3 |
+
"aggregation_labels": null,
|
4 |
+
"aggregation_loss_importance": 0.8,
|
5 |
+
"aggregation_loss_weight": 1.0,
|
6 |
+
"aggregation_temperature": 1.0,
|
7 |
+
"allow_empty_column_selection": false,
|
8 |
+
"answer_loss_cutoff": 100,
|
9 |
+
"answer_loss_importance": 0.001,
|
10 |
+
"architectures": [
|
11 |
+
"TapasForMaskedLM"
|
12 |
+
],
|
13 |
+
"attention_probs_dropout_prob": 0.1,
|
14 |
+
"average_approximation_function": "ratio",
|
15 |
+
"average_logits_per_cell": true,
|
16 |
+
"cell_selection_preference": 0.5,
|
17 |
+
"disable_per_token_loss": false,
|
18 |
+
"hidden_act": "gelu",
|
19 |
+
"hidden_dropout_prob": 0.1,
|
20 |
+
"hidden_size": 32,
|
21 |
+
"huber_loss_delta": 25.0,
|
22 |
+
"init_cell_selection_weights_to_zero": true,
|
23 |
+
"initializer_range": 0.02,
|
24 |
+
"intermediate_size": 37,
|
25 |
+
"layer_norm_eps": 1e-12,
|
26 |
+
"max_num_columns": 32,
|
27 |
+
"max_num_rows": 64,
|
28 |
+
"max_position_embeddings": 512,
|
29 |
+
"model_type": "tapas",
|
30 |
+
"no_aggregation_label_index": null,
|
31 |
+
"num_aggregation_labels": 4,
|
32 |
+
"num_attention_heads": 4,
|
33 |
+
"num_hidden_layers": 5,
|
34 |
+
"pad_token_id": 0,
|
35 |
+
"positive_label_weight": 10.0,
|
36 |
+
"positive_weight": 10.0,
|
37 |
+
"reset_position_index_per_cell": true,
|
38 |
+
"select_one_column": true,
|
39 |
+
"torch_dtype": "float32",
|
40 |
+
"transformers_version": "4.28.0.dev0",
|
41 |
+
"type_vocab_sizes": [
|
42 |
+
3,
|
43 |
+
256,
|
44 |
+
256,
|
45 |
+
2,
|
46 |
+
256,
|
47 |
+
256,
|
48 |
+
10
|
49 |
+
],
|
50 |
+
"use_answer_as_supervision": true,
|
51 |
+
"use_gumbel_for_agg": false,
|
52 |
+
"use_gumbel_for_aggregation": false,
|
53 |
+
"use_gumbel_for_cells": false,
|
54 |
+
"use_normalized_answer_loss": false,
|
55 |
+
"vocab_size": 30522
|
56 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1937a7edb4feb18b8adaeb72eefa79880e064bfc92a61248df8b6f8a79441eeb
|
3 |
+
size 4398865
|
special_tokens_map.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
"[EMPTY]"
|
4 |
+
],
|
5 |
+
"cls_token": "[CLS]",
|
6 |
+
"mask_token": "[MASK]",
|
7 |
+
"pad_token": "[PAD]",
|
8 |
+
"sep_token": "[SEP]",
|
9 |
+
"unk_token": "[UNK]"
|
10 |
+
}
|
tokenizer_config.json
ADDED
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
"[EMPTY]"
|
4 |
+
],
|
5 |
+
"cell_trim_length": -1,
|
6 |
+
"clean_up_tokenization_spaces": true,
|
7 |
+
"cls_token": "[CLS]",
|
8 |
+
"do_basic_tokenize": true,
|
9 |
+
"do_lower_case": true,
|
10 |
+
"drop_rows_to_fit": false,
|
11 |
+
"empty_token": "[EMPTY]",
|
12 |
+
"mask_token": "[MASK]",
|
13 |
+
"max_column_id": null,
|
14 |
+
"max_question_length": null,
|
15 |
+
"max_row_id": null,
|
16 |
+
"min_question_length": null,
|
17 |
+
"model_max_length": 512,
|
18 |
+
"never_split": null,
|
19 |
+
"pad_token": "[PAD]",
|
20 |
+
"sep_token": "[SEP]",
|
21 |
+
"special_tokens_map_file": "/home/runner/.cache/huggingface/hub/models--google--tapas-base-finetuned-sqa/snapshots/81916d20eef75766aeae71b9487fd615017b0413/special_tokens_map.json",
|
22 |
+
"strip_accents": null,
|
23 |
+
"strip_column_names": false,
|
24 |
+
"tokenize_chinese_chars": true,
|
25 |
+
"tokenizer_class": "TapasTokenizer",
|
26 |
+
"unk_token": "[UNK]",
|
27 |
+
"update_answer_coordinates": false
|
28 |
+
}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|