metadata
license: apache-2.0
base_model: mnoukhov/pythia410m-sft-tldr
tags:
- trl
- reward-trainer
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: pythia410m-rm-tldr
results: []
pythia410m-rm-tldr
This model is a fine-tuned version of mnoukhov/pythia410m-sft-tldr on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4313
- Accuracy: 0.7928
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-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 1
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.5145 | 0.2006 | 291 | 0.4726 | 0.7659 |
| 0.4356 | 0.4011 | 582 | 0.4641 | 0.7730 |
| 0.3823 | 0.6017 | 873 | 0.4456 | 0.7860 |
| 0.3616 | 0.8022 | 1164 | 0.4313 | 0.7928 |
Framework versions
- Transformers 4.40.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1