Jamba-1.5-Mini-Full-FT

This model is a fine-tuned version of ai21labs/AI21-Jamba-1.5-Mini on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 0.7744

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: 1e-06
• train_batch_size: 2
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• num_devices: 64
• total_train_batch_size: 128
• total_eval_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.01
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
0.778	1.0	159	0.8063
0.7417	2.0	318	0.7769
0.7136	3.0	477	0.7744

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.1.0
• Datasets 2.18.0
• Tokenizers 0.19.1