Spaces:
Sleeping
Sleeping
MaroueneA
commited on
Commit
•
8e57f90
1
Parent(s):
24953d0
Initial commit of my Gradio NLP app
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- app.py +123 -0
- models/UBC-NLP/ARBERT/config.json +33 -0
- models/UBC-NLP/ARBERT/model.safetensors +3 -0
- models/UBC-NLP/ARBERT/special_tokens_map.json +7 -0
- models/UBC-NLP/ARBERT/tokenizer.json +0 -0
- models/UBC-NLP/ARBERT/tokenizer_config.json +57 -0
- models/UBC-NLP/ARBERT/vocab.txt +0 -0
- models/UBC-NLP/MARBERT/config.json +33 -0
- models/UBC-NLP/MARBERT/model.safetensors +3 -0
- models/UBC-NLP/MARBERT/special_tokens_map.json +7 -0
- models/UBC-NLP/MARBERT/tokenizer.json +0 -0
- models/UBC-NLP/MARBERT/tokenizer_config.json +57 -0
- models/UBC-NLP/MARBERT/vocab.txt +0 -0
- models/bert-offensive/checkpoint-1000/config.json +27 -0
- models/bert-offensive/checkpoint-1000/model.safetensors +3 -0
- models/bert-offensive/checkpoint-1000/optimizer.pt +3 -0
- models/bert-offensive/checkpoint-1000/rng_state.pth +3 -0
- models/bert-offensive/checkpoint-1000/scheduler.pt +3 -0
- models/bert-offensive/checkpoint-1000/special_tokens_map.json +7 -0
- models/bert-offensive/checkpoint-1000/tokenizer.json +0 -0
- models/bert-offensive/checkpoint-1000/tokenizer_config.json +55 -0
- models/bert-offensive/checkpoint-1000/trainer_state.json +35 -0
- models/bert-offensive/checkpoint-1000/training_args.bin +3 -0
- models/bert-offensive/checkpoint-1000/vocab.txt +0 -0
- models/bert-offensive/checkpoint-1500/config.json +27 -0
- models/bert-offensive/checkpoint-1500/model.safetensors +3 -0
- models/bert-offensive/checkpoint-1500/optimizer.pt +3 -0
- models/bert-offensive/checkpoint-1500/rng_state.pth +3 -0
- models/bert-offensive/checkpoint-1500/scheduler.pt +3 -0
- models/bert-offensive/checkpoint-1500/special_tokens_map.json +7 -0
- models/bert-offensive/checkpoint-1500/tokenizer.json +0 -0
- models/bert-offensive/checkpoint-1500/tokenizer_config.json +55 -0
- models/bert-offensive/checkpoint-1500/trainer_state.json +54 -0
- models/bert-offensive/checkpoint-1500/training_args.bin +3 -0
- models/bert-offensive/checkpoint-1500/vocab.txt +0 -0
- models/bert-offensive/checkpoint-2000/config.json +27 -0
- models/bert-offensive/checkpoint-2000/model.safetensors +3 -0
- models/bert-offensive/checkpoint-2000/optimizer.pt +3 -0
- models/bert-offensive/checkpoint-2000/rng_state.pth +3 -0
- models/bert-offensive/checkpoint-2000/scheduler.pt +3 -0
- models/bert-offensive/checkpoint-2000/special_tokens_map.json +7 -0
- models/bert-offensive/checkpoint-2000/tokenizer.json +0 -0
- models/bert-offensive/checkpoint-2000/tokenizer_config.json +55 -0
- models/bert-offensive/checkpoint-2000/trainer_state.json +61 -0
- models/bert-offensive/checkpoint-2000/training_args.bin +3 -0
- models/bert-offensive/checkpoint-2000/vocab.txt +0 -0
- models/bert-offensive/checkpoint-2500/config.json +27 -0
- models/bert-offensive/checkpoint-2500/model.safetensors +3 -0
- models/bert-offensive/checkpoint-2500/optimizer.pt +3 -0
- models/bert-offensive/checkpoint-2500/rng_state.pth +3 -0
app.py
ADDED
@@ -0,0 +1,123 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
import pandas as pd
|
3 |
+
from transformers import AutoModelForSequenceClassification, AutoTokenizer
|
4 |
+
from datasets import load_dataset
|
5 |
+
from sklearn.metrics import accuracy_score, precision_recall_fscore_support, confusion_matrix
|
6 |
+
import torch
|
7 |
+
from sentence_transformers import SentenceTransformer
|
8 |
+
import umap
|
9 |
+
from sklearn.manifold import TSNE
|
10 |
+
import matplotlib.pyplot as plt
|
11 |
+
import seaborn as sns
|
12 |
+
import numpy as np
|
13 |
+
import tempfile
|
14 |
+
from collections import Counter
|
15 |
+
import os
|
16 |
+
|
17 |
+
# Set the temporary directory for Gradio to use
|
18 |
+
os.environ['GRADIO_TEMP_DIR'] = '/home/marwen/gradio_tmp'
|
19 |
+
|
20 |
+
# Load the models and their tokenizers
|
21 |
+
model_paths = {
|
22 |
+
"roberta-base-offensive": "./models/roberta-base-offensive",
|
23 |
+
"distilbert-base-uncased-offensive": "./models/distilbert-base-uncased-offensive",
|
24 |
+
"bert-offensive": "./models/bert-offensive",
|
25 |
+
"deberta-offensive": "./models/deberta-offensive"
|
26 |
+
}
|
27 |
+
|
28 |
+
models = {name: AutoModelForSequenceClassification.from_pretrained(path) for name, path in model_paths.items()}
|
29 |
+
tokenizers = {name: AutoTokenizer.from_pretrained(path) for name, path in model_paths.items()}
|
30 |
+
|
31 |
+
# Load the dataset
|
32 |
+
dataset = load_dataset("tweet_eval", "offensive")
|
33 |
+
|
34 |
+
# Initialize Sentence Transformer for embedding generation
|
35 |
+
model_embedding = SentenceTransformer('all-MiniLM-L6-v2')
|
36 |
+
|
37 |
+
def encode(texts, tokenizer):
|
38 |
+
return tokenizer(texts, padding="max_length", truncation=True, max_length=128, return_tensors="pt")
|
39 |
+
|
40 |
+
def predict(model, inputs):
|
41 |
+
model.eval()
|
42 |
+
with torch.no_grad():
|
43 |
+
outputs = model(**inputs)
|
44 |
+
preds = outputs.logits.argmax(-1).cpu().numpy()
|
45 |
+
return preds
|
46 |
+
|
47 |
+
def calculate_metrics(labels, preds):
|
48 |
+
accuracy = accuracy_score(labels, preds)
|
49 |
+
precision, recall, f1, _ = precision_recall_fscore_support(labels, preds, average='binary')
|
50 |
+
conf_matrix = confusion_matrix(labels, preds)
|
51 |
+
return accuracy, precision, recall, f1, conf_matrix
|
52 |
+
|
53 |
+
def generate_confusion_matrix(conf_matrix, model_name):
|
54 |
+
plt.figure(figsize=(5, 4))
|
55 |
+
sns.heatmap(conf_matrix, annot=True, fmt="d")
|
56 |
+
plt.title(f'Confusion Matrix: {model_name}')
|
57 |
+
plt.ylabel('Actual')
|
58 |
+
plt.xlabel('Predicted')
|
59 |
+
plt.tight_layout()
|
60 |
+
temp_file = tempfile.NamedTemporaryFile(delete=False, suffix='.png')
|
61 |
+
plt.savefig(temp_file.name)
|
62 |
+
plt.close()
|
63 |
+
return temp_file.name
|
64 |
+
|
65 |
+
def generate_embeddings_and_plot(categories):
|
66 |
+
all_texts = sum(categories.values(), [])
|
67 |
+
embeddings = model_embedding.encode(all_texts)
|
68 |
+
umap_reducer = umap.UMAP(n_neighbors=15, n_components=2, metric='cosine')
|
69 |
+
umap_embeddings = umap_reducer.fit_transform(embeddings)
|
70 |
+
tsne_embeddings = TSNE(n_components=2, perplexity=30).fit_transform(embeddings)
|
71 |
+
|
72 |
+
def plot_embeddings(embeddings, title, file_suffix):
|
73 |
+
plt.figure(figsize=(10, 8))
|
74 |
+
colors = {"correct_both": "green", "incorrect_both": "red", "correct_model1_only": "blue", "correct_model2_only": "orange"}
|
75 |
+
for category, color in colors.items():
|
76 |
+
indices = [i for i, text in enumerate(all_texts) if text in categories[category]]
|
77 |
+
plt.scatter(embeddings[indices, 0], embeddings[indices, 1], label=category, color=color, alpha=0.6)
|
78 |
+
plt.legend()
|
79 |
+
plt.title(title)
|
80 |
+
plt.xlabel('Component 1')
|
81 |
+
plt.ylabel('Component 2')
|
82 |
+
temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=f'_{file_suffix}.png')
|
83 |
+
plt.savefig(temp_file.name)
|
84 |
+
plt.close()
|
85 |
+
return temp_file.name
|
86 |
+
|
87 |
+
umap_plot_path = plot_embeddings(umap_embeddings, "UMAP Projection of Text Categories", "umap")
|
88 |
+
tsne_plot_path = plot_embeddings(tsne_embeddings, "t-SNE Projection of Text Categories", "tsne")
|
89 |
+
return umap_plot_path, tsne_plot_path
|
90 |
+
|
91 |
+
def setup_gradio_interface():
|
92 |
+
with gr.Blocks() as demo:
|
93 |
+
gr.Markdown("## Model Comparison and Text Analysis")
|
94 |
+
with gr.Row():
|
95 |
+
model1_input = gr.Dropdown(list(model_paths.keys()), label="Model 1")
|
96 |
+
model2_input = gr.Dropdown(list(model_paths.keys()), label="Model 2")
|
97 |
+
submit_button = gr.Button("Compare")
|
98 |
+
|
99 |
+
metrics_output = gr.Dataframe()
|
100 |
+
|
101 |
+
with gr.Row():
|
102 |
+
model1_cm_output = gr.Image(label="Confusion Matrix for Model 1")
|
103 |
+
model2_cm_output = gr.Image(label="Confusion Matrix for Model 2")
|
104 |
+
|
105 |
+
with gr.Row():
|
106 |
+
umap_visualization_output = gr.Image(label="UMAP Text Categorization Visualization")
|
107 |
+
tsne_visualization_output = gr.Image(label="t-SNE Text Categorization Visualization")
|
108 |
+
|
109 |
+
def update_interface(model1, model2):
|
110 |
+
metrics_df, conf_matrix1, conf_matrix2 = compare_models(model1, model2)
|
111 |
+
umap_plot_path, tsne_plot_path = generate_embeddings_and_plot(metrics_df)
|
112 |
+
return metrics_df, conf_matrix1, conf_matrix2, umap_plot_path, tsne_plot_path
|
113 |
+
|
114 |
+
submit_button.click(
|
115 |
+
update_interface,
|
116 |
+
inputs=[model1_input, model2_input],
|
117 |
+
outputs=[metrics_output, model1_cm_output, model2_cm_output, umap_visualization_output, tsne_visualization_output]
|
118 |
+
)
|
119 |
+
|
120 |
+
return demo
|
121 |
+
|
122 |
+
demo = setup_gradio_interface()
|
123 |
+
demo.launch(share=True)
|
models/UBC-NLP/ARBERT/config.json
ADDED
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "UBC-NLP/ARBERT",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"directionality": "bidi",
|
9 |
+
"gradient_checkpointing": false,
|
10 |
+
"hidden_act": "gelu",
|
11 |
+
"hidden_dropout_prob": 0.1,
|
12 |
+
"hidden_size": 768,
|
13 |
+
"initializer_range": 0.02,
|
14 |
+
"intermediate_size": 3072,
|
15 |
+
"layer_norm_eps": 1e-12,
|
16 |
+
"max_position_embeddings": 512,
|
17 |
+
"model_type": "bert",
|
18 |
+
"num_attention_heads": 12,
|
19 |
+
"num_hidden_layers": 12,
|
20 |
+
"pad_token_id": 0,
|
21 |
+
"pooler_fc_size": 768,
|
22 |
+
"pooler_num_attention_heads": 12,
|
23 |
+
"pooler_num_fc_layers": 3,
|
24 |
+
"pooler_size_per_head": 128,
|
25 |
+
"pooler_type": "first_token_transform",
|
26 |
+
"position_embedding_type": "absolute",
|
27 |
+
"problem_type": "single_label_classification",
|
28 |
+
"torch_dtype": "float32",
|
29 |
+
"transformers_version": "4.39.2",
|
30 |
+
"type_vocab_size": 2,
|
31 |
+
"use_cache": true,
|
32 |
+
"vocab_size": 100000
|
33 |
+
}
|
models/UBC-NLP/ARBERT/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f5721a08143fe631ff017403f0a526b5d570444ebab9af55869f50815112efd5
|
3 |
+
size 651395072
|
models/UBC-NLP/ARBERT/special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
models/UBC-NLP/ARBERT/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
models/UBC-NLP/ARBERT/tokenizer_config.json
ADDED
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"4": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_basic_tokenize": true,
|
47 |
+
"do_lower_case": true,
|
48 |
+
"mask_token": "[MASK]",
|
49 |
+
"model_max_length": 1000000000000000019884624838656,
|
50 |
+
"never_split": null,
|
51 |
+
"pad_token": "[PAD]",
|
52 |
+
"sep_token": "[SEP]",
|
53 |
+
"strip_accents": null,
|
54 |
+
"tokenize_chinese_chars": true,
|
55 |
+
"tokenizer_class": "BertTokenizer",
|
56 |
+
"unk_token": "[UNK]"
|
57 |
+
}
|
models/UBC-NLP/ARBERT/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
models/UBC-NLP/MARBERT/config.json
ADDED
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "UBC-NLP/MARBERT",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"directionality": "bidi",
|
9 |
+
"gradient_checkpointing": false,
|
10 |
+
"hidden_act": "gelu",
|
11 |
+
"hidden_dropout_prob": 0.1,
|
12 |
+
"hidden_size": 768,
|
13 |
+
"initializer_range": 0.02,
|
14 |
+
"intermediate_size": 3072,
|
15 |
+
"layer_norm_eps": 1e-12,
|
16 |
+
"max_position_embeddings": 512,
|
17 |
+
"model_type": "bert",
|
18 |
+
"num_attention_heads": 12,
|
19 |
+
"num_hidden_layers": 12,
|
20 |
+
"pad_token_id": 0,
|
21 |
+
"pooler_fc_size": 768,
|
22 |
+
"pooler_num_attention_heads": 12,
|
23 |
+
"pooler_num_fc_layers": 3,
|
24 |
+
"pooler_size_per_head": 128,
|
25 |
+
"pooler_type": "first_token_transform",
|
26 |
+
"position_embedding_type": "absolute",
|
27 |
+
"problem_type": "single_label_classification",
|
28 |
+
"torch_dtype": "float32",
|
29 |
+
"transformers_version": "4.39.2",
|
30 |
+
"type_vocab_size": 2,
|
31 |
+
"use_cache": true,
|
32 |
+
"vocab_size": 100000
|
33 |
+
}
|
models/UBC-NLP/MARBERT/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:874e85391a27658550ccfde41216bf0d1bf732472e4461d8ba7876a4dcd1b2c6
|
3 |
+
size 651395072
|
models/UBC-NLP/MARBERT/special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
models/UBC-NLP/MARBERT/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
models/UBC-NLP/MARBERT/tokenizer_config.json
ADDED
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"4": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_basic_tokenize": true,
|
47 |
+
"do_lower_case": true,
|
48 |
+
"mask_token": "[MASK]",
|
49 |
+
"model_max_length": 1000000000000000019884624838656,
|
50 |
+
"never_split": null,
|
51 |
+
"pad_token": "[PAD]",
|
52 |
+
"sep_token": "[SEP]",
|
53 |
+
"strip_accents": null,
|
54 |
+
"tokenize_chinese_chars": true,
|
55 |
+
"tokenizer_class": "BertTokenizer",
|
56 |
+
"unk_token": "[UNK]"
|
57 |
+
}
|
models/UBC-NLP/MARBERT/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
models/bert-offensive/checkpoint-1000/config.json
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "bert-base-uncased",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"gradient_checkpointing": false,
|
9 |
+
"hidden_act": "gelu",
|
10 |
+
"hidden_dropout_prob": 0.1,
|
11 |
+
"hidden_size": 768,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"intermediate_size": 3072,
|
14 |
+
"layer_norm_eps": 1e-12,
|
15 |
+
"max_position_embeddings": 512,
|
16 |
+
"model_type": "bert",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 12,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"position_embedding_type": "absolute",
|
21 |
+
"problem_type": "single_label_classification",
|
22 |
+
"torch_dtype": "float32",
|
23 |
+
"transformers_version": "4.39.3",
|
24 |
+
"type_vocab_size": 2,
|
25 |
+
"use_cache": true,
|
26 |
+
"vocab_size": 30522
|
27 |
+
}
|
models/bert-offensive/checkpoint-1000/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cd658ae002182ad99ffbb9982d5d85515c0c49949eac2d6d4eb86c479d67955b
|
3 |
+
size 437958648
|
models/bert-offensive/checkpoint-1000/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:22c5eb3685348006d842a33f8c23a222816e6efa7e3dbf1212516a44add594c0
|
3 |
+
size 876038394
|
models/bert-offensive/checkpoint-1000/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:89ed3ecad87f77a0ed60753e4ffd969dbef607419cab5d7c3471fc6b16360766
|
3 |
+
size 14244
|
models/bert-offensive/checkpoint-1000/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3cdc454fd78e2054130f997d6bf748bc81368c748607a351ea7de15a408d26e6
|
3 |
+
size 1064
|
models/bert-offensive/checkpoint-1000/special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
models/bert-offensive/checkpoint-1000/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
models/bert-offensive/checkpoint-1000/tokenizer_config.json
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"100": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"101": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"102": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"103": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_lower_case": true,
|
47 |
+
"mask_token": "[MASK]",
|
48 |
+
"model_max_length": 512,
|
49 |
+
"pad_token": "[PAD]",
|
50 |
+
"sep_token": "[SEP]",
|
51 |
+
"strip_accents": null,
|
52 |
+
"tokenize_chinese_chars": true,
|
53 |
+
"tokenizer_class": "BertTokenizer",
|
54 |
+
"unk_token": "[UNK]"
|
55 |
+
}
|
models/bert-offensive/checkpoint-1000/trainer_state.json
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.6711409395973155,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1000,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.34,
|
13 |
+
"grad_norm": 3.309788227081299,
|
14 |
+
"learning_rate": 4.440715883668904e-05,
|
15 |
+
"loss": 0.5291,
|
16 |
+
"step": 500
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.67,
|
20 |
+
"grad_norm": 8.38113784790039,
|
21 |
+
"learning_rate": 3.881431767337807e-05,
|
22 |
+
"loss": 0.4834,
|
23 |
+
"step": 1000
|
24 |
+
}
|
25 |
+
],
|
26 |
+
"logging_steps": 500,
|
27 |
+
"max_steps": 4470,
|
28 |
+
"num_input_tokens_seen": 0,
|
29 |
+
"num_train_epochs": 3,
|
30 |
+
"save_steps": 500,
|
31 |
+
"total_flos": 423493688042880.0,
|
32 |
+
"train_batch_size": 8,
|
33 |
+
"trial_name": null,
|
34 |
+
"trial_params": null
|
35 |
+
}
|
models/bert-offensive/checkpoint-1000/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ff52faf0d3a27635d14f511f07bc80871b57c3a7826358862dda78696619e8f5
|
3 |
+
size 4856
|
models/bert-offensive/checkpoint-1000/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
models/bert-offensive/checkpoint-1500/config.json
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "bert-base-uncased",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"gradient_checkpointing": false,
|
9 |
+
"hidden_act": "gelu",
|
10 |
+
"hidden_dropout_prob": 0.1,
|
11 |
+
"hidden_size": 768,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"intermediate_size": 3072,
|
14 |
+
"layer_norm_eps": 1e-12,
|
15 |
+
"max_position_embeddings": 512,
|
16 |
+
"model_type": "bert",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 12,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"position_embedding_type": "absolute",
|
21 |
+
"problem_type": "single_label_classification",
|
22 |
+
"torch_dtype": "float32",
|
23 |
+
"transformers_version": "4.39.3",
|
24 |
+
"type_vocab_size": 2,
|
25 |
+
"use_cache": true,
|
26 |
+
"vocab_size": 30522
|
27 |
+
}
|
models/bert-offensive/checkpoint-1500/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b13c8461e29b1c84d3992da1258c4e67c9df1faa1b7d2a6d3866c8b4ab31264e
|
3 |
+
size 437958648
|
models/bert-offensive/checkpoint-1500/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5a5c46aedd4909a23c89b0d3396dc7938f781b6289de190c8bb1fa275715998e
|
3 |
+
size 876038394
|
models/bert-offensive/checkpoint-1500/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:17a5e93a639b18c5080b3206380db8bdf3c9bf762e376ba68470d9bd63251d88
|
3 |
+
size 14244
|
models/bert-offensive/checkpoint-1500/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0bd70d973811a00199980200e5162073c3f4bfdfc513c94464ee46ef32af9473
|
3 |
+
size 1064
|
models/bert-offensive/checkpoint-1500/special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
models/bert-offensive/checkpoint-1500/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
models/bert-offensive/checkpoint-1500/tokenizer_config.json
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"100": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"101": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"102": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"103": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_lower_case": true,
|
47 |
+
"mask_token": "[MASK]",
|
48 |
+
"model_max_length": 512,
|
49 |
+
"pad_token": "[PAD]",
|
50 |
+
"sep_token": "[SEP]",
|
51 |
+
"strip_accents": null,
|
52 |
+
"tokenize_chinese_chars": true,
|
53 |
+
"tokenizer_class": "BertTokenizer",
|
54 |
+
"unk_token": "[UNK]"
|
55 |
+
}
|
models/bert-offensive/checkpoint-1500/trainer_state.json
ADDED
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 1.0067114093959733,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1500,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.34,
|
13 |
+
"grad_norm": 3.309788227081299,
|
14 |
+
"learning_rate": 4.440715883668904e-05,
|
15 |
+
"loss": 0.5291,
|
16 |
+
"step": 500
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.67,
|
20 |
+
"grad_norm": 8.38113784790039,
|
21 |
+
"learning_rate": 3.881431767337807e-05,
|
22 |
+
"loss": 0.4834,
|
23 |
+
"step": 1000
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 1.0,
|
27 |
+
"eval_accuracy": 0.8244186046511628,
|
28 |
+
"eval_f1": 0.700990099009901,
|
29 |
+
"eval_loss": 0.40849873423576355,
|
30 |
+
"eval_precision": 0.6679245283018868,
|
31 |
+
"eval_recall": 0.7375,
|
32 |
+
"eval_runtime": 1.415,
|
33 |
+
"eval_samples_per_second": 607.771,
|
34 |
+
"eval_steps_per_second": 76.325,
|
35 |
+
"step": 1490
|
36 |
+
},
|
37 |
+
{
|
38 |
+
"epoch": 1.01,
|
39 |
+
"grad_norm": 4.857753753662109,
|
40 |
+
"learning_rate": 3.3221476510067115e-05,
|
41 |
+
"loss": 0.4743,
|
42 |
+
"step": 1500
|
43 |
+
}
|
44 |
+
],
|
45 |
+
"logging_steps": 500,
|
46 |
+
"max_steps": 4470,
|
47 |
+
"num_input_tokens_seen": 0,
|
48 |
+
"num_train_epochs": 3,
|
49 |
+
"save_steps": 500,
|
50 |
+
"total_flos": 635261087615520.0,
|
51 |
+
"train_batch_size": 8,
|
52 |
+
"trial_name": null,
|
53 |
+
"trial_params": null
|
54 |
+
}
|
models/bert-offensive/checkpoint-1500/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ff52faf0d3a27635d14f511f07bc80871b57c3a7826358862dda78696619e8f5
|
3 |
+
size 4856
|
models/bert-offensive/checkpoint-1500/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
models/bert-offensive/checkpoint-2000/config.json
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "bert-base-uncased",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"gradient_checkpointing": false,
|
9 |
+
"hidden_act": "gelu",
|
10 |
+
"hidden_dropout_prob": 0.1,
|
11 |
+
"hidden_size": 768,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"intermediate_size": 3072,
|
14 |
+
"layer_norm_eps": 1e-12,
|
15 |
+
"max_position_embeddings": 512,
|
16 |
+
"model_type": "bert",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 12,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"position_embedding_type": "absolute",
|
21 |
+
"problem_type": "single_label_classification",
|
22 |
+
"torch_dtype": "float32",
|
23 |
+
"transformers_version": "4.39.3",
|
24 |
+
"type_vocab_size": 2,
|
25 |
+
"use_cache": true,
|
26 |
+
"vocab_size": 30522
|
27 |
+
}
|
models/bert-offensive/checkpoint-2000/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d84caca90c708c6d9b08f796f62f3fe3770d1ebbe5dc4708d787c870652e882a
|
3 |
+
size 437958648
|
models/bert-offensive/checkpoint-2000/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:79f977dfb26bf16852ccff9864f65c99dfca153be14bcd7280af5f7a05a0d808
|
3 |
+
size 876038394
|
models/bert-offensive/checkpoint-2000/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e57db07813490de347f6b29da476f84f897496ba4cc251212c6d017a619dce80
|
3 |
+
size 14244
|
models/bert-offensive/checkpoint-2000/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dd68bf4dff9924c9178e7ecdd91ec247aec55edc72b3ca0231603fd33cef86d3
|
3 |
+
size 1064
|
models/bert-offensive/checkpoint-2000/special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
models/bert-offensive/checkpoint-2000/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
models/bert-offensive/checkpoint-2000/tokenizer_config.json
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"100": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"101": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"102": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"103": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_lower_case": true,
|
47 |
+
"mask_token": "[MASK]",
|
48 |
+
"model_max_length": 512,
|
49 |
+
"pad_token": "[PAD]",
|
50 |
+
"sep_token": "[SEP]",
|
51 |
+
"strip_accents": null,
|
52 |
+
"tokenize_chinese_chars": true,
|
53 |
+
"tokenizer_class": "BertTokenizer",
|
54 |
+
"unk_token": "[UNK]"
|
55 |
+
}
|
models/bert-offensive/checkpoint-2000/trainer_state.json
ADDED
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 1.342281879194631,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 2000,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.34,
|
13 |
+
"grad_norm": 3.309788227081299,
|
14 |
+
"learning_rate": 4.440715883668904e-05,
|
15 |
+
"loss": 0.5291,
|
16 |
+
"step": 500
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.67,
|
20 |
+
"grad_norm": 8.38113784790039,
|
21 |
+
"learning_rate": 3.881431767337807e-05,
|
22 |
+
"loss": 0.4834,
|
23 |
+
"step": 1000
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 1.0,
|
27 |
+
"eval_accuracy": 0.8244186046511628,
|
28 |
+
"eval_f1": 0.700990099009901,
|
29 |
+
"eval_loss": 0.40849873423576355,
|
30 |
+
"eval_precision": 0.6679245283018868,
|
31 |
+
"eval_recall": 0.7375,
|
32 |
+
"eval_runtime": 1.415,
|
33 |
+
"eval_samples_per_second": 607.771,
|
34 |
+
"eval_steps_per_second": 76.325,
|
35 |
+
"step": 1490
|
36 |
+
},
|
37 |
+
{
|
38 |
+
"epoch": 1.01,
|
39 |
+
"grad_norm": 4.857753753662109,
|
40 |
+
"learning_rate": 3.3221476510067115e-05,
|
41 |
+
"loss": 0.4743,
|
42 |
+
"step": 1500
|
43 |
+
},
|
44 |
+
{
|
45 |
+
"epoch": 1.34,
|
46 |
+
"grad_norm": 20.212223052978516,
|
47 |
+
"learning_rate": 2.7628635346756153e-05,
|
48 |
+
"loss": 0.3753,
|
49 |
+
"step": 2000
|
50 |
+
}
|
51 |
+
],
|
52 |
+
"logging_steps": 500,
|
53 |
+
"max_steps": 4470,
|
54 |
+
"num_input_tokens_seen": 0,
|
55 |
+
"num_train_epochs": 3,
|
56 |
+
"save_steps": 500,
|
57 |
+
"total_flos": 846855820558080.0,
|
58 |
+
"train_batch_size": 8,
|
59 |
+
"trial_name": null,
|
60 |
+
"trial_params": null
|
61 |
+
}
|
models/bert-offensive/checkpoint-2000/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ff52faf0d3a27635d14f511f07bc80871b57c3a7826358862dda78696619e8f5
|
3 |
+
size 4856
|
models/bert-offensive/checkpoint-2000/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
models/bert-offensive/checkpoint-2500/config.json
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "bert-base-uncased",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"gradient_checkpointing": false,
|
9 |
+
"hidden_act": "gelu",
|
10 |
+
"hidden_dropout_prob": 0.1,
|
11 |
+
"hidden_size": 768,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"intermediate_size": 3072,
|
14 |
+
"layer_norm_eps": 1e-12,
|
15 |
+
"max_position_embeddings": 512,
|
16 |
+
"model_type": "bert",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 12,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"position_embedding_type": "absolute",
|
21 |
+
"problem_type": "single_label_classification",
|
22 |
+
"torch_dtype": "float32",
|
23 |
+
"transformers_version": "4.39.3",
|
24 |
+
"type_vocab_size": 2,
|
25 |
+
"use_cache": true,
|
26 |
+
"vocab_size": 30522
|
27 |
+
}
|
models/bert-offensive/checkpoint-2500/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f58457eeb8339abe1f0ee26937443a95a8a25aae3c4fe5f569327ab3fd87f48c
|
3 |
+
size 437958648
|
models/bert-offensive/checkpoint-2500/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:77406637943250708e3999e599a8b10ef39a486f27cfa07a01bbff9f9a76e71b
|
3 |
+
size 876038394
|
models/bert-offensive/checkpoint-2500/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5ccff0626d2d12f751c557becd9d0ba8acfabed972b2c6065108c51d3e8d4d12
|
3 |
+
size 14244
|