137b1282271906927a1dd7e3f9db5202

This model is a fine-tuned version of albert/albert-xlarge-v1 on the contemmcm/amazon_reviews_2013 [cell-phone] dataset. It achieves the following results on the evaluation set:

• Loss: 1.4818
• Data Size: 1.0
• Epoch Runtime: 382.3281
• Accuracy: 0.3853
• F1 Macro: 0.1112

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
No log	0	0	1.6351	0	24.5794	0.2449	0.1624
No log	1	1973	1.5319	0.0078	28.0666	0.3807	0.1188
0.0346	2	3946	1.5539	0.0156	30.2307	0.3853	0.1112
1.5447	3	5919	1.4901	0.0312	35.7225	0.3853	0.1112
1.5078	4	7892	1.4898	0.0625	46.9017	0.3853	0.1112
1.4806	5	9865	1.4833	0.125	69.5398	0.3853	0.1112
1.4873	6	11838	1.4875	0.25	114.1994	0.3853	0.1112
1.4944	7	13811	1.4823	0.5	204.3173	0.3853	0.1112
1.5087	8.0	15784	1.4868	1.0	382.7569	0.3853	0.1112
1.4879	9.0	17757	1.4815	1.0	381.8557	0.3853	0.1112
1.487	10.0	19730	1.4818	1.0	382.9116	0.3853	0.1112
1.4861	11.0	21703	1.4828	1.0	384.5495	0.3853	0.1112
1.475	12.0	23676	1.4807	1.0	382.7869	0.3853	0.1112
1.4787	13.0	25649	1.4813	1.0	383.2169	0.3853	0.1112
1.4624	14.0	27622	1.4805	1.0	383.3159	0.3851	0.1114
1.503	15.0	29595	1.4830	1.0	383.3170	0.3853	0.1112
1.4984	16.0	31568	1.4832	1.0	383.7510	0.3853	0.1112
1.4914	17.0	33541	1.4817	1.0	383.8177	0.3853	0.1112
1.4826	18.0	35514	1.4818	1.0	382.3281	0.3853	0.1112

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1