374fb82d-aea5-4405-9f9e-52bd8974120d

This model is a fine-tuned version of EleutherAI/pythia-410m-deduped on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.9464

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: 0.000208
• train_batch_size: 4
• eval_batch_size: 4
• seed: 80
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 50
• training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0002	1	6.0454
12.3516	0.0082	50	5.0920
10.314	0.0164	100	4.7079
10.6881	0.0246	150	4.6861
10.3681	0.0328	200	4.3705
10.0248	0.0410	250	4.2340
9.7719	0.0492	300	4.1229
10.0224	0.0573	350	4.0000
9.8306	0.0655	400	3.9712
9.3675	0.0737	450	3.9477
9.4526	0.0819	500	3.9464

Framework versions

• PEFT 0.13.2
• Transformers 4.46.0
• Pytorch 2.5.0+cu124
• Datasets 3.0.1
• Tokenizers 0.20.1