metadata
license: gemma
library_name: peft
tags:
- alignment-handbook
- trl
- sft
- generated_from_trainer
base_model: google/gemma-2b
datasets:
- llama-duo/synth_summarize_dataset_dedup
model-index:
- name: gemma2b-summarize-gemini1_5flash-128k
results: []
gemma2b-summarize-gemini1_5flash-128k
This model is a fine-tuned version of google/gemma-2b on the llama-duo/synth_summarize_dataset_dedup dataset. It achieves the following results on the evaluation set:
- Loss: 2.5573
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: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 2
- 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: 0.1
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.0978 | 1.0 | 104 | 2.4831 |
| 0.9985 | 2.0 | 208 | 2.4666 |
| 0.9543 | 3.0 | 312 | 2.4561 |
| 0.92 | 4.0 | 416 | 2.4799 |
| 0.9016 | 5.0 | 520 | 2.4990 |
| 0.8871 | 6.0 | 624 | 2.5250 |
| 0.8635 | 7.0 | 728 | 2.5363 |
| 0.8535 | 8.0 | 832 | 2.5546 |
| 0.845 | 9.0 | 936 | 2.5566 |
| 0.853 | 10.0 | 1040 | 2.5573 |
Framework versions
- PEFT 0.11.1
- Transformers 4.41.2
- Pytorch 2.3.1+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1