Skip to content

Distil* scores update #1490

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 18 commits into
base: dev
Choose a base branch
from
+1,447 −125
Open

Distil* scores update #1490

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
18 commits
Select commit Hold shift + click to select a range
01fde43
DistilBERT links and description added
ayeffkay Jul 13, 2021
b003a53
Distil configs for paraphraser
ayeffkay Jul 13, 2021
e840a0b
Distil scores and configs added
ayeffkay Jul 13, 2021
544508d
Distil configs for paraphraser
ayeffkay Jul 13, 2021
05ae908
Distil configs for rusentiment
ayeffkay Jul 13, 2021
74eeedd
Distil configs for ner
ayeffkay Jul 13, 2021
5ae20e1
Distil squad configs added
ayeffkay Jul 13, 2021
b3dc15c
feat: paraphraser_convers_distilrubert_2L_torch removed torch
IgnatovFedor Jul 30, 2021
6107d77
Merge branch 'dev' of https://github.com/deepmipt/DeepPavlov into dev
IgnatovFedor Jul 30, 2021
54830d1
feat: added test paraphraser_convers_distilrubert_2L
IgnatovFedor Jul 30, 2021
d79aca0
paraphraser_convers_distilrubert_6L renamed
IgnatovFedor Jul 30, 2021
7c120a5
fix html build
IgnatovFedor Jul 30, 2021
d8f53e5
removed torch from names
IgnatovFedor Jul 30, 2021
f390f1b
refactor: datasets for distil
IgnatovFedor Jul 30, 2021
f99ce48
Distilrubert-tiny configs update
ayeffkay Sep 24, 2021
3da37c4
Distilrubert_base: configs updated
ayeffkay Nov 3, 2021
d7618ee
Scores updated
ayeffkay Nov 3, 2021
f111896
Hyperparams changed for distil- ner and squad
ayeffkay Nov 4, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
93 changes: 93 additions & 0 deletions deeppavlov/configs/classifiers/paraphraser_convers_distilrubert_2L.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,93 @@
{
"dataset_reader": {
"class_name": "paraphraser_reader",
"data_path": "{DOWNLOADS_PATH}/paraphraser_data",
"do_lower_case": false
},
"dataset_iterator": {
"class_name": "siamese_iterator",
"seed": 243,
"len_valid": 500
},
"chainer": {
"in": ["text_a", "text_b"],
"in_y": ["y"],
"pipe": [
{
"class_name": "torch_transformers_preprocessor",
"vocab_file": "{TRANSFORMER}",
"do_lower_case": false,
"max_seq_length": 64,
"in": ["text_a", "text_b"],
"out": ["bert_features"]
},
{
"class_name": "torch_transformers_classifier",
"n_classes": 2,
"return_probas": false,
"pretrained_bert": "{TRANSFORMER}",
"save_path": "{MODEL_PATH}/model",
"load_path": "{MODEL_PATH}/model",
"attention_probs_keep_prob": 0.67,
"hidden_keep_prob": 0.0,
"optimizer": "AdamW",
"optimizer_parameters": {
"lr": 9e-05
},
"learning_rate_drop_patience": 3,
"learning_rate_drop_div": 1.5,
"in": [
"bert_features"
],
"in_y": [
"y"
],
"out": [
"predictions"
]
}
],
"out": ["predictions"]
},
"train": {
"epochs": 100,
"batch_size": 64,
"metrics": [
"f1",
"accuracy"
],
"validation_patience": 7,
"val_every_n_batches": 50,
"log_every_n_batches": 50,
"evaluation_targets": [
"train",
"valid",
"test"
],
"tensorboard_log_dir": "{MODEL_PATH}/",
"class_name": "torch_trainer"
},
"metadata": {
"variables": {
"ROOT_PATH": "~/.deeppavlov",
"DOWNLOADS_PATH": "{ROOT_PATH}/downloads",
"TRANSFORMER": "DeepPavlov/distilrubert-tiny-cased-conversational",
"MODELS_PATH": "{ROOT_PATH}/models",
"MODEL_PATH": "{MODELS_PATH}/paraphraser_convers_distilrubert_2L"
},
"download": [
{
"url": "http://files.deeppavlov.ai/deeppavlov_data/classifiers/paraphraser_convers_distilrubert_2L.tar.gz",
"subdir": "{MODELS_PATH}"
},
{
"url": "http://files.deeppavlov.ai/datasets/paraphraser.zip",
"subdir": "{DOWNLOADS_PATH}/paraphraser_data"
},
{
"url": "http://files.deeppavlov.ai/datasets/paraphraser_gold.zip",
"subdir": "{DOWNLOADS_PATH}/paraphraser_data"
}
]
}
}
93 changes: 93 additions & 0 deletions deeppavlov/configs/classifiers/paraphraser_convers_distilrubert_6L.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,93 @@
{
"dataset_reader": {
"class_name": "paraphraser_reader",
"data_path": "{DOWNLOADS_PATH}/paraphraser_data",
"do_lower_case": false
},
"dataset_iterator": {
"class_name": "siamese_iterator",
"seed": 243,
"len_valid": 500
},
"chainer": {
"in": ["text_a", "text_b"],
"in_y": ["y"],
"pipe": [
{
"class_name": "torch_transformers_preprocessor",
"vocab_file": "{TRANSFORMER}",
"do_lower_case": false,
"max_seq_length": 64,
"in": ["text_a", "text_b"],
"out": ["bert_features"]
},
{
"class_name": "torch_transformers_classifier",
"n_classes": 2,
"return_probas": false,
"pretrained_bert": "{TRANSFORMER}",
"save_path": "{MODEL_PATH}/model",
"load_path": "{MODEL_PATH}/model",
"attention_probs_keep_prob": 0.89,
"hidden_keep_prob": 0.44,
"optimizer": "AdamW",
"optimizer_parameters": {
"lr": 5.46e-05
},
"learning_rate_drop_patience": 3,
"learning_rate_drop_div": 1.5,
"in": [
"bert_features"
],
"in_y": [
"y"
],
"out": [
"predictions"
]
}
],
"out": ["predictions"]
},
"train": {
"epochs": 100,
"batch_size": 64,
"metrics": [
"f1",
"accuracy"
],
"validation_patience": 7,
"val_every_n_batches": 50,
"log_every_n_batches": 50,
"evaluation_targets": [
"train",
"valid",
"test"
],
"tensorboard_log_dir": "{MODEL_PATH}/",
"class_name": "torch_trainer"
},
"metadata": {
"variables": {
"ROOT_PATH": "~/.deeppavlov",
"DOWNLOADS_PATH": "{ROOT_PATH}/downloads",
"TRANSFORMER": "DeepPavlov/distilrubert-base-cased-conversational",
"MODELS_PATH": "{ROOT_PATH}/models",
"MODEL_PATH": "{MODELS_PATH}/paraphraser_convers_distilrubert_6L"
},
"download": [
{
"url": "http://files.deeppavlov.ai/deeppavlov_data/classifiers/paraphraser_convers_distilrubert_6L.tar.gz",
"subdir": "{MODELS_PATH}"
},
{
"url": "http://files.deeppavlov.ai/datasets/paraphraser.zip",
"subdir": "{DOWNLOADS_PATH}/paraphraser_data"
},
{
"url": "http://files.deeppavlov.ai/datasets/paraphraser_gold.zip",
"subdir": "{DOWNLOADS_PATH}/paraphraser_data"
}
]
}
}
145 changes: 145 additions & 0 deletions deeppavlov/configs/classifiers/rusentiment_convers_distilrubert_2L.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,145 @@
{
"dataset_reader": {
"class_name": "basic_classification_reader",
"x": "text",
"y": "label",
"data_path": "{DOWNLOADS_PATH}/rusentiment/",
"train": "rusentiment_random_posts.csv",
"test": "rusentiment_test.csv"
},
"dataset_iterator": {
"class_name": "basic_classification_iterator",
"seed": 42,
"split_seed": 23,
"field_to_split": "train",
"split_fields": [
"train",
"valid"
],
"split_proportions": [
0.9,
0.1
]
},
"chainer": {
"in": [
"x"
],
"in_y": [
"y"
],
"pipe": [
{
"class_name": "torch_transformers_preprocessor",
"vocab_file": "{TRANSFORMER}",
"do_lower_case": true,
"max_seq_length": 64,
"in": [
"x"
],
"out": [
"bert_features"
]
},
{
"id": "classes_vocab",
"class_name": "simple_vocab",
"fit_on": [
"y"
],
"save_path": "{MODEL_PATH}/classes.dict",
"load_path": "{MODEL_PATH}/classes.dict",
"in": "y",
"out": "y_ids"
},
{
"in": "y_ids",
"out": "y_onehot",
"class_name": "one_hotter",
"depth": "#classes_vocab.len",
"single_vector": true
},
{
"class_name": "torch_transformers_classifier",
"n_classes": "#classes_vocab.len",
"return_probas": true,
"pretrained_bert": "{TRANSFORMER}",
"save_path": "{MODEL_PATH}/model",
"load_path": "{MODEL_PATH}/model",
"attention_probs_keep_prob": 0.33,
"hidden_keep_prob": 0.67,
"optimizer": "AdamW",
"optimizer_parameters": {
"lr": 3.67e-05
},
"learning_rate_drop_patience": 5,
"learning_rate_drop_div": 1.5,
"in": [
"bert_features"
],
"in_y": [
"y_ids"
],
"out": [
"y_pred_probas"
]
},
{
"in": "y_pred_probas",
"out": "y_pred_ids",
"class_name": "proba2labels",
"max_proba": true
},
{
"in": "y_pred_ids",
"out": "y_pred_labels",
"ref": "classes_vocab"
}
],
"out": [
"y_pred_labels"
]
},
"train": {
"epochs": 100,
"batch_size": 64,
"metrics": [
"f1_weighted",
"f1_macro",
"accuracy",
{
"name": "roc_auc",
"inputs": [
"y_onehot",
"y_pred_probas"
]
}
],
"validation_patience": 5,
"val_every_n_epochs": 1,
"log_every_n_epochs": 1,
"show_examples": false,
"evaluation_targets": [
"train",
"valid",
"test"
],
"tensorboard_log_dir": "{MODEL_PATH}/",
"class_name": "torch_trainer"
},
"metadata": {
"variables": {
"ROOT_PATH": "~/.deeppavlov",
"DOWNLOADS_PATH": "{ROOT_PATH}/downloads",
"TRANSFORMER": "DeepPavlov/distilrubert-tiny-cased-conversational",
"MODELS_PATH": "{ROOT_PATH}/models",
"MODEL_PATH": "{MODELS_PATH}/classifiers/rusentiment_convers_distilrubert_2L"
},
"download": [
{
"url": "http://files.deeppavlov.ai/deeppavlov_data/classifiers/rusentiment_convers_distilrubert_2L.tar.gz",
"subdir": "{MODELS_PATH}/classifiers/"
}
]
}
}
Loading