1db30786-aad8-4587-a235-519326aa26bb

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: 0.8320

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: 0.000214
• train_batch_size: 4
• eval_batch_size: 4
• seed: 140
• 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	0.0005	1	4.2553
2.9014	0.0239	50	1.3918
4.0067	0.0478	100	2.3113
3.2874	0.0717	150	1.4772
2.4335	0.0956	200	1.0888
2.7371	0.1194	250	1.1248
2.2539	0.1433	300	0.9772
2.3061	0.1672	350	0.9379
2.0217	0.1911	400	0.8595
2.2005	0.2150	450	0.8373
2.1046	0.2389	500	0.8320

Framework versions

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