README.md

metadata

language:
  - en
license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - mtop
model-index:
  - name: t5-small-pointer-mtop
    results: []

t5-small-pointer-mtop

This model is a fine-tuned version of google/mt5-small on the mtop dataset. It achieves the following results on the evaluation set:

• Loss: 0.1064
• Exact Match: 0.7570

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: 0.001
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 32
• total_train_batch_size: 512
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 3000

Training results

Training Loss	Epoch	Step	Validation Loss	Exact Match
2.0944	6.65	200	0.6548	0.0027
0.5307	13.33	400	0.2456	0.2779
0.2388	19.98	600	0.1486	0.4559
0.1459	26.65	800	0.1190	0.5043
0.1011	33.33	1000	0.1117	0.5230
0.0774	39.98	1200	0.1084	0.5374
0.0598	46.65	1400	0.1064	0.5405
0.0478	53.33	1600	0.1147	0.5454
0.0397	59.98	1800	0.1139	0.5472
0.0337	66.65	2000	0.1179	0.5481
0.0286	73.33	2200	0.1243	0.5499
0.0251	79.98	2400	0.1259	0.5481
0.0218	86.65	2600	0.1276	0.5503
0.0197	93.33	2800	0.1309	0.5503
0.0184	99.98	3000	0.1317	0.5503

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.0+cu117
• Datasets 2.7.1
• Tokenizers 0.13.2