Phi-3.5-MultiCap-tool-embedding-past

This model is a fine-tuned version of microsoft/Phi-3.5-mini-instruct on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 0.7571

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: 0.0001
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
1.037	0.1524	50	1.1134
1.0204	0.3048	100	1.0044
0.855	0.4571	150	0.9503
0.8786	0.6095	200	0.9030
0.8936	0.7619	250	0.8664
0.8912	0.9143	300	0.8372
0.7467	1.0667	350	0.8161
0.8009	1.2190	400	0.7984
0.7408	1.3714	450	0.7836
0.6925	1.5238	500	0.7728
0.6949	1.6762	550	0.7653
0.7537	1.8286	600	0.7597
0.7729	1.9810	650	0.7571

Framework versions

• PEFT 0.12.0
• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1