Tukan-1.1B-Chat-reasoning-sft-COLA

This model is a fine-tuned version of TinyLlama/TinyLlama-1.1B-Chat-v1.0 on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 1.0480

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: 6
• eval_batch_size: 3
• seed: 42
• distributed_type: multi-GPU
• gradient_accumulation_steps: 20
• total_train_batch_size: 120
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
1.0669	0.24	10	1.0635
1.0701	0.47	20	1.0547
1.0753	0.71	30	1.0493
1.0579	0.94	40	1.0480

Framework versions

• Transformers 4.36.2
• Pytorch 2.2.0a0+gitd925d94
• Datasets 2.14.6
• Tokenizers 0.15.0