metadata
base_model: >-
RefalMachine/llama3_extended_darulm_20_05_24_part1-2_64000_bpe_mean_init_03_07_24
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: llama3_extended_darulm_20_05_24_part1-2_64000_bpe_part1_lr2e4_bs256
results: []
llama3_extended_darulm_20_05_24_part1-2_64000_bpe_part1_lr2e4_bs256
This model is a fine-tuned version of RefalMachine/llama3_extended_darulm_20_05_24_part1-2_64000_bpe_mean_init_03_07_24 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.1597
- Accuracy: 0.5489
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: 4
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 32
- total_train_batch_size: 128
- total_eval_batch_size: 128
- optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-05
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 1.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.4497 | 0.09 | 2000 | 2.2890 | 0.5302 |
| 2.4021 | 0.18 | 4000 | 2.2496 | 0.5356 |
| 2.3883 | 0.28 | 6000 | 2.2251 | 0.5390 |
| 2.3684 | 0.37 | 8000 | 2.2056 | 0.5416 |
| 2.3547 | 0.46 | 10000 | 2.1889 | 0.5438 |
| 2.3271 | 0.55 | 12000 | 2.1759 | 0.5459 |
| 2.3153 | 0.64 | 14000 | 2.1664 | 0.5476 |
| 2.3175 | 0.73 | 16000 | 2.1618 | 0.5485 |
| 2.3013 | 0.83 | 18000 | 2.1599 | 0.5488 |
| 2.3075 | 0.92 | 20000 | 2.1597 | 0.5489 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.3.0a0+6ddf5cf85e.nv24.04
- Datasets 2.18.0
- Tokenizers 0.15.2