shnguo/distilbert-base-uncased-finetuned-cate-classfication

Model Card for Model ID

Thesaurus category mapping model based on Transformer architecture

Model Details

Model Description

• Developed by: shnguo
• Funded by [optional]: gramtech
• Shared by [optional]: shnguo
• Model type: classfication
• Language(s) (NLP): English
• License: MIT
• Finetuned from model [optional]: distilbert

Training Details

Evaluation

• {'eval_loss': 0.028821110725402832, 'eval_accuracy': 0.9941857972232465, 'eval_f1': 0.9940715947938238, 'eval_runtime': 54.5075, 'eval_samples_per_second': 2139.358, 'eval_steps_per_second': 66.872, 'epoch': 2.0}
• {'train_runtime': 20346.1721, 'train_samples_per_second': 561.671, 'train_steps_per_second': 17.552, 'train_loss': 0.2077474573270054, 'epoch': 2.0}
• wandb: \ 0.462 MB of 0.462 MB uploaded
• wandb:
• wandb: Run summary:
• wandb: eval/accuracy 0.99419
• wandb: eval/f1 0.99407
• wandb: eval/loss 0.02882
• wandb: eval/runtime 54.5075
• wandb: eval/samples_per_second 2139.358
• wandb: eval/steps_per_second 66.872
• wandb: total_flos 3.2915880370021428e+16
• wandb: train/epoch 2.0
• wandb: train/global_step 357122
• wandb: train/grad_norm 0.01881
• wandb: train/learning_rate 0.0
• wandb: train/loss 0.0443
• wandb: train_loss 0.20775
• wandb: train_runtime 20346.1721
• wandb: train_samples_per_second 561.671
• wandb: train_steps_per_second 17.552
• wandb:
• wandb: 🚀 View run feasible-voice-8 at: https://wandb.ai/shn-guo/distilbert-base-uncased-finetuned-cate/runs/lygbrbzx
• wandb: ⭐️ View project at: https://wandb.ai/shn-guo/distilbert-base-uncased-finetuned-cate
• wandb: Synced 6 W&B file(s), 0 media file(s), 2 artifact file(s) and 0 other file(s)
• wandb: Find logs at: ./wandb/run-20240618_133925-lygbrbzx/logs