bfdc60fb6740862f336f736ef527de54

This model is a fine-tuned version of distilbert/distilbert-base-cased on the contemmcm/trec dataset. It achieves the following results on the evaluation set:

• Loss: 0.2482
• Data Size: 1.0
• Epoch Runtime: 5.6921
• Accuracy: 0.9667
• F1 Macro: 0.9697

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	1.8391	0	0.7147	0.1292	0.0762
No log	1	170	1.7637	0.0078	0.8989	0.2292	0.0645
No log	2	340	1.6896	0.0156	0.8727	0.3146	0.1477
No log	3	510	1.5418	0.0312	1.0700	0.4208	0.2462
No log	4	680	0.9865	0.0625	1.2198	0.7708	0.6587
0.07	5	850	0.3853	0.125	1.5433	0.9062	0.7618
0.07	6	1020	0.1820	0.25	2.1153	0.9521	0.9273
0.2245	7	1190	0.1425	0.5	3.2005	0.9625	0.9477
0.1449	8.0	1360	0.1217	1.0	5.6069	0.9688	0.9635
0.0811	9.0	1530	0.1168	1.0	5.3368	0.9771	0.9722
0.0356	10.0	1700	0.1567	1.0	5.4563	0.9646	0.9586
0.007	11.0	1870	0.2152	1.0	5.3901	0.9688	0.9715
0.0094	12.0	2040	0.2157	1.0	5.4178	0.9646	0.9542
0.0112	13.0	2210	0.2482	1.0	5.6921	0.9667	0.9697

Framework versions

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