tamil-llama-7b-instruct-quantized-ASR-output-fine-tuning

This model is a fine-tuned version of abhinand/tamil-llama-7b-instruct-v0.1 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 5.3527

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: 0.0002
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• lr_scheduler_warmup_ratio: 0.03
• training_steps: 1500

Training results

Training Loss	Epoch	Step	Validation Loss
2.4659	2.58	500	4.2383
1.9248	5.17	1000	4.6944
1.4112	7.75	1500	5.3527

Framework versions

• PEFT 0.8.2
• Transformers 4.37.0
• Pytorch 2.1.2
• Datasets 2.16.1
• Tokenizers 0.15.1