pszemraj/griffin-v0.01-c3t-8layer-simplewiki
- griffin/recurrent_gemma arch
- claude3 tokenizer (as an HF gpt2 tokenizer)
Model description
This model is a fine-tuned version of ./griffin-1024-c3t-8layer on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 4.1928
- Accuracy: 0.4084
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: 0.0003
- train_batch_size: 4
- eval_batch_size: 4
- seed: 80085
- gradient_accumulation_steps: 32
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.99) and epsilon=1e-07
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_ratio: 0.05
- num_epochs: 2.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 13.2525 | 0.2548 | 100 | 11.9768 | 0.0131 |
| 8.8873 | 0.5095 | 200 | 8.0127 | 0.0357 |
| 7.2457 | 0.7643 | 300 | 6.4508 | 0.0512 |
| 6.3152 | 1.0190 | 400 | 5.6163 | 0.0460 |
| 5.5586 | 1.2738 | 500 | 4.7645 | 0.3650 |
| 5.2936 | 1.5285 | 600 | 4.3919 | 0.3934 |
| 4.8839 | 1.7833 | 700 | 4.1928 | 0.4084 |
Framework versions
- Transformers 4.40.1
- Pytorch 2.2.0+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1
- Downloads last month
- 50
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.