BLOOM_AAID_structured_train_final

This model is a fine-tuned version of bigscience/bloom-7b1 on the AAID_structured dataset. It achieves the following results on the evaluation set:

• Loss: 0.8177

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: 0.0003
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss
1.7338	0.0109	10	0.9125
0.6341	0.0219	20	0.8568
0.5526	0.0328	30	0.8991
0.5651	0.0438	40	0.8944
0.5392	0.0547	50	0.8796
0.5038	0.0656	60	0.8612
0.4904	0.0766	70	0.8335
0.476	0.0875	80	0.8787
0.4819	0.0984	90	0.8442
0.4376	0.1094	100	0.8534
0.4443	0.1203	110	0.8331
0.44	0.1313	120	0.8530
0.4362	0.1422	130	0.8594
0.4273	0.1531	140	0.8177
0.438	0.1641	150	0.8450
0.4234	0.1750	160	0.8484
0.4254	0.1859	170	0.8263
0.4074	0.1969	180	0.8609
0.4209	0.2078	190	0.8371
0.4182	0.2188	200	0.8382
0.3883	0.2297	210	0.8301
0.3993	0.2406	220	0.8259
0.4082	0.2516	230	0.8462
0.3757	0.2625	240	0.8523

Framework versions

• PEFT 0.11.1
• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1