metadata
license: other
tags:
- generated_from_trainer
datasets:
- AlekseyKorshuk/amazon-reviews-input-output
metrics:
- accuracy
model-index:
- name: amazon-reviews-input-output-13b
results:
- task:
name: Causal Language Modeling
type: text-generation
dataset:
name: AlekseyKorshuk/amazon-reviews-input-output
type: AlekseyKorshuk/amazon-reviews-input-output
metrics:
- name: Accuracy
type: accuracy
value: 0.040426829268292684
amazon-reviews-input-output-13b
This model is a fine-tuned version of facebook/opt-13b on the AlekseyKorshuk/amazon-reviews-input-output dataset. It achieves the following results on the evaluation set:
- Loss: 2.7168
- Accuracy: 0.0404
- Samples: 100
- Perplexity: 15.1318
- Table: <wandb.data_types.Table object at 0x7f8c9a3fb6d0>
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: 5e-05
- train_batch_size: 1
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1.0
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
3.2134 | 0.96 | 15 | 2.7168 | 0.0404 |
Framework versions
- Transformers 4.25.0.dev0
- Pytorch 1.12.1+cu113
- Datasets 2.3.2
- Tokenizers 0.12.1