8a78138f683dec32f0396cb42fbb5095

This model is a fine-tuned version of facebook/opt-1.3b on the contemmcm/trec dataset. It achieves the following results on the evaluation set:

• Loss: 0.2781
• Data Size: 1.0
• Epoch Runtime: 37.8057
• Accuracy: 0.95
• F1 Macro: 0.9457

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	2.2521	0	1.7135	0.1313	0.0612
No log	1	170	1.9474	0.0078	2.1339	0.3521	0.2714
No log	2	340	0.7812	0.0156	4.4108	0.7479	0.6032
No log	3	510	0.3671	0.0312	7.4641	0.8917	0.8683
No log	4	680	1.1084	0.0625	10.1221	0.6167	0.5575
0.0379	5	850	0.2284	0.125	12.6574	0.9229	0.8891
0.0379	6	1020	0.3588	0.25	17.4619	0.8854	0.8853
0.2995	7	1190	0.3479	0.5	23.9141	0.9083	0.8926
0.2126	8.0	1360	0.2798	1.0	39.7057	0.9146	0.9276
0.1825	9.0	1530	0.2781	1.0	37.8057	0.95	0.9457

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1