Edit model card

Gemma-2-2B_task-1_120-samples_config-1_full

This model is a fine-tuned version of google/gemma-2-2b-it on the GaetanMichelet/chat-60_ft_task-1 and the GaetanMichelet/chat-120_ft_task-1 datasets. It achieves the following results on the evaluation set:

  • Loss: 1.0136

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: 1
  • eval_batch_size: 1
  • seed: 42
  • distributed_type: multi-GPU
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 8
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss
2.4912 1.0 11 2.4122
1.8242 2.0 22 1.8003
1.3712 3.0 33 1.2810
1.0733 4.0 44 1.1023
0.9287 5.0 55 1.0314
0.8686 6.0 66 1.0136
0.6842 7.0 77 1.0365
0.5911 8.0 88 1.0808
0.4827 9.0 99 1.2147
0.3254 10.0 110 1.4074
0.2263 11.0 121 1.6078
0.1939 12.0 132 1.8584
0.1191 13.0 143 1.9537

Framework versions

  • PEFT 0.12.0
  • Transformers 4.44.0
  • Pytorch 2.1.2+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1
Downloads last month
4
Inference API
Unable to determine this model’s pipeline type. Check the docs .

Model tree for GaetanMichelet/Gemma-2-2B_task-1_120-samples_config-1_full

Base model

google/gemma-2-2b
Adapter
(151)
this model

Collection including GaetanMichelet/Gemma-2-2B_task-1_120-samples_config-1_full