google/flan_t5_base_ledgar/README.md

metadata

license: apache-2.0
base_model: google/flan-t5-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: flan_t5_base_ledgar
    results: []

flan_t5_base_ledgar

This model is a fine-tuned version of google/flan-t5-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5092
• Accuracy: 0.8685
• F1 Macro: 0.7955
• F1 Micro: 0.8685

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: 0.0005
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• distributed_type: multi-GPU
• num_devices: 2
• total_train_batch_size: 64
• total_eval_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	F1 Micro
1.2824	0.11	100	1.0243	0.7464	0.5723	0.7464
0.8961	0.21	200	0.8572	0.7743	0.6359	0.7743
0.8233	0.32	300	0.7788	0.7968	0.6756	0.7968
0.7722	0.43	400	0.7432	0.8014	0.6689	0.8014
0.7739	0.53	500	0.6933	0.8135	0.7020	0.8135
0.7435	0.64	600	0.6871	0.8137	0.7028	0.8137
0.6877	0.75	700	0.6751	0.8169	0.7012	0.8169
0.6968	0.85	800	0.6471	0.8225	0.7249	0.8225
0.6218	0.96	900	0.6366	0.8219	0.7264	0.8219
0.5101	1.07	1000	0.6130	0.8378	0.7388	0.8378
0.5411	1.17	1100	0.6088	0.8375	0.7392	0.8375
0.5918	1.28	1200	0.5864	0.8449	0.7577	0.8449
0.4785	1.39	1300	0.5917	0.8391	0.7460	0.8391
0.4372	1.49	1400	0.5790	0.8409	0.7563	0.8409
0.5022	1.6	1500	0.5868	0.8437	0.7524	0.8437
0.5391	1.71	1600	0.5613	0.8447	0.7520	0.8447
0.4971	1.81	1700	0.5466	0.8545	0.7702	0.8545
0.4686	1.92	1800	0.5260	0.8566	0.7774	0.8566
0.349	2.03	1900	0.5416	0.8565	0.7721	0.8565
0.3216	2.13	2000	0.5441	0.8565	0.7763	0.8565
0.3531	2.24	2100	0.5444	0.8553	0.7753	0.8553
0.3276	2.35	2200	0.5380	0.8604	0.7796	0.8604
0.3372	2.45	2300	0.5231	0.8634	0.7831	0.8634
0.3227	2.56	2400	0.5210	0.8651	0.7872	0.8651
0.2987	2.67	2500	0.5188	0.8665	0.7910	0.8665
0.3354	2.77	2600	0.5150	0.8666	0.7931	0.8666
0.3103	2.88	2700	0.5103	0.8681	0.7942	0.8681
0.3248	2.99	2800	0.5092	0.8685	0.7955	0.8685

Framework versions

• Transformers 4.39.0.dev0
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2