lora_evo_ta_all_layers_15

This model is a fine-tuned version of togethercomputer/evo-1-8k-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.5456

Model description

Trained on single ID token 5K dataset filtered to 8k sequences

lora_alpha = 128

lora_dropout = 0.1

lora_r = 128

epochs = 3

learning rate = 3e-4

warmup_steps=200

gradient_accumulation_steps = 1

train_batch = 2

eval_batch = 2

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: 0.0003
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• lr_scheduler_warmup_steps: 200
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
2.694	0.625	2500	2.5997
2.5703	1.25	5000	2.5697
2.5341	1.875	7500	2.5471
2.4969	2.5	10000	2.5456

Framework versions

• PEFT 0.11.1
• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1