distilbert-finetuned-russian

lngrid2020/distilbert-finetuned-russian

distilbert-finetuned-russian is a text classification model from lngrid2020. Use it when you need a label for a piece of text. It is set up for transformers. The card lists the license as apache-2.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

15% of all-time downloads

All-time downloads

Public

Parameters

54.3M

1.3 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors217 MB · 100%

At a glance

Task: Text Classification
Library: transformers
License: apache-2.0
Model type: distilbert
Access: Public
Created: Nov 14, 2023
Updated: Nov 14, 2023
SHA: 62460e4b

Try a prompt

Base models

DmitryPogrebnoy/distilbert-base-russian-cased

Model card

From the Hugging Face model README

distilbert-finetuned-russian

This model is a fine-tuned version of DmitryPogrebnoy/distilbert-base-russian-cased on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.2469
F1: 0.8280
Roc Auc: 0.8853
Accuracy: 0.7576

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.4883	0.14	100	0.3524	0.7132	0.8036	0.6324
0.3065	0.29	200	0.2797	0.7704	0.8391	0.7005
0.2765	0.43	300	0.2655	0.7793	0.8437	0.7078
0.2571	0.57	400	0.2457	0.7913	0.8530	0.7248
0.2518	0.72	500	0.2333	0.7986	0.8593	0.7255
0.2471	0.86	600	0.2439	0.7880	0.8522	0.7225
0.2324	1.0	700	0.2209	0.8049	0.8636	0.7410
0.2223	1.15	800	0.2262	0.8040	0.8615	0.7388
0.207	1.29	900	0.2197	0.8084	0.8679	0.7404
0.213	1.43	1000	0.2147	0.8136	0.8683	0.7488
0.2137	1.58	1100	0.2146	0.8102	0.8646	0.7447
0.2078	1.72	1200	0.2143	0.8187	0.8768	0.7481
0.2055	1.86	1300	0.2092	0.8118	0.8696	0.7422
0.2107	2.01	1400	0.2098	0.8128	0.8713	0.7461
0.1742	2.15	1500	0.2183	0.8220	0.8787	0.7573
0.1624	2.29	1600	0.2094	0.8240	0.8801	0.7594
0.1632	2.44	1700	0.2163	0.8231	0.8821	0.7530
0.1736	2.58	1800	0.2065	0.8230	0.8750	0.7616
0.1615	2.72	1900	0.2135	0.8172	0.8725	0.7522
0.1728	2.87	2000	0.2054	0.8296	0.8872	0.7562
0.1685	3.01	2100	0.2146	0.8243	0.8834	0.7512
0.1323	3.15	2200	0.2216	0.8263	0.8826	0.7589
0.123	3.3	2300	0.2234	0.8202	0.8802	0.7488
0.1316	3.44	2400	0.2199	0.8238	0.8810	0.7558
0.119	3.58	2500	0.2230	0.8285	0.8824	0.7659
0.1292	3.73	2600	0.2258	0.8226	0.8790	0.7573
0.1303	3.87	2700	0.2272	0.8225	0.8830	0.7499
0.1114	4.01	2800	0.2420	0.8207	0.8815	0.7485
0.0931	4.16	2900	0.2459	0.8222	0.8803	0.7556
0.0983	4.3	3000	0.2434	0.8264	0.8849	0.7551
0.0934	4.44	3100	0.2485	0.8261	0.8844	0.7567
0.0939	4.59	3200	0.2532	0.8174	0.8784	0.7458
0.0976	4.73	3300	0.2498	0.8248	0.8823	0.7571
0.098	4.87	3400	0.2469	0.8280	0.8853	0.7576

Framework versions

Transformers 4.35.0
Pytorch 1.12.1+cu116
Datasets 2.4.0
Tokenizers 0.14.1

Task: Text Classification
Library: transformers
Type: distilbert
License: apache-2.0
Created: Nov 14, 2023
Updated: Nov 14, 2023