8057787115cea7a35a21adff8257c2b8

This model is a fine-tuned version of facebook/opt-2.7b on the contemmcm/trec dataset. It achieves the following results on the evaluation set:

• Loss: 0.4806
• Data Size: 0.25
• Epoch Runtime: 29.5295
• Accuracy: 0.8896
• F1 Macro: 0.8602

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	2.0083	0	2.2461	0.1812	0.1132
No log	1	170	1.0066	0.0078	2.9566	0.625	0.4449
No log	2	340	0.4294	0.0156	5.7032	0.8938	0.7475
No log	3	510	0.9452	0.0312	10.2384	0.6937	0.6915
No log	4	680	1.7359	0.0625	13.9943	0.2958	0.2710
0.0787	5	850	0.6879	0.125	19.7794	0.7604	0.6332
0.0787	6	1020	0.4806	0.25	29.5295	0.8896	0.8602

Framework versions

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