ettin-map-1b-singlefold

This model is a fine-tuned version of jhu-clsp/ettin-encoder-1b on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2284
• Accuracy: 0.9137
• Map@3: 0.9559

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: 2e-05
• train_batch_size: 48
• eval_batch_size: 48
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 192
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Map@3
4.65	0.5006	1000	0.3880	0.8524	0.9233
1.2856	1.0010	2000	0.3011	0.8845	0.9407
0.9807	1.5016	3000	0.2610	0.9029	0.9504
0.8878	2.0020	4000	0.2404	0.9085	0.9533
0.7293	2.5026	5000	0.2344	0.9118	0.9549
0.7067	3.0030	6000	0.2290	0.9140	0.9561
0.6482	3.5036	7000	0.2284	0.9137	0.9559

Framework versions

• Transformers 4.56.2
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1