Edit model card

Gemma-2-2B_task-2_180-samples_config-2_full

This model is a fine-tuned version of google/gemma-2-2b-it on the GaetanMichelet/chat-60_ft_task-2, the GaetanMichelet/chat-120_ft_task-2 and the GaetanMichelet/chat-180_ft_task-2 datasets. It achieves the following results on the evaluation set:

  • Loss: 0.8710

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: 16
  • total_train_batch_size: 16
  • 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
1.3182 0.9412 8 1.3144
1.2253 2.0 17 1.1693
1.1196 2.9412 25 1.0760
0.9289 4.0 34 0.9662
0.8623 4.9412 42 0.9212
0.8786 6.0 51 0.8973
0.861 6.9412 59 0.8845
0.8112 8.0 68 0.8751
0.7491 8.9412 76 0.8714
0.7439 10.0 85 0.8710
0.7136 10.9412 93 0.8754
0.6546 12.0 102 0.8866
0.6539 12.9412 110 0.9093
0.632 14.0 119 0.9347
0.543 14.9412 127 0.9885
0.4869 16.0 136 1.0262
0.4926 16.9412 144 1.0435

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-2_180-samples_config-2_full

Base model

google/gemma-2-2b
Adapter
(157)
this model

Collection including GaetanMichelet/Gemma-2-2B_task-2_180-samples_config-2_full