985d422df19c3eb6e07b2981b3cc1bd1

This model is a fine-tuned version of albert/albert-xxlarge-v1 on the google/boolq dataset. It achieves the following results on the evaluation set:

• Loss: 0.6653
• Data Size: 1.0
• Epoch Runtime: 60.4685
• Accuracy: 0.6213
• F1 Macro: 0.3832
• Rouge1: 0.6213
• Rouge2: 0.0
• Rougel: 0.6207
• Rougelsum: 0.6210

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
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro	Rouge1	Rouge2	Rougel	Rougelsum
No log	0	0	0.6789	0	6.2974	0.6066	0.4192	0.6069	0.0	0.6060	0.6063
No log	1	294	0.7663	0.0078	6.8289	0.4694	0.4647	0.4694	0.0	0.4692	0.4692
No log	2	588	0.6770	0.0156	7.3395	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
No log	3	882	0.6652	0.0312	8.3195	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
0.0276	4	1176	0.6644	0.0625	10.0768	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
0.055	5	1470	0.6634	0.125	13.2490	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
0.0953	6	1764	0.6673	0.25	20.0815	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
0.6378	7	2058	0.5634	0.5	33.6197	0.7310	0.7251	0.7310	0.0	0.7310	0.7310
0.3789	8.0	2352	0.3837	1.0	60.8351	0.8297	0.8111	0.8294	0.0	0.8294	0.8300
0.2198	9.0	2646	0.4425	1.0	60.6112	0.8238	0.8035	0.8241	0.0	0.8235	0.8238
0.1165	10.0	2940	0.4895	1.0	61.0117	0.8339	0.8225	0.8343	0.0	0.8343	0.8339
0.0881	11.0	3234	0.8399	1.0	60.5086	0.8315	0.8187	0.8318	0.0	0.8309	0.8315
0.6703	12.0	3528	0.6653	1.0	60.4685	0.6213	0.3832	0.6213	0.0	0.6207	0.6210

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.1