metadata
library_name: transformers
license: other
base_model: llava-hf/llava-v1.6-mistral-7b-hf
tags:
- llama-factory
- full
- generated_from_trainer
model-index:
- name: Compcap_cosi_0_90
results: []
Compcap_cosi_0_90
This model is a fine-tuned version of llava-hf/llava-v1.6-mistral-7b-hf on the Compcap_cosi_0_90 dataset. It achieves the following results on the evaluation set:
- Loss: 0.7912
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-06
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 4
- total_train_batch_size: 256
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 10
- num_epochs: 3.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.9297 | 0.1505 | 50 | 0.9399 |
| 0.8675 | 0.3010 | 100 | 0.8835 |
| 0.8457 | 0.4515 | 150 | 0.8571 |
| 0.8411 | 0.6020 | 200 | 0.8422 |
| 0.8133 | 0.7524 | 250 | 0.8285 |
| 0.8114 | 0.9029 | 300 | 0.8189 |
| 0.7379 | 1.0534 | 350 | 0.8137 |
| 0.7416 | 1.2039 | 400 | 0.8082 |
| 0.7623 | 1.3544 | 450 | 0.8033 |
| 0.7158 | 1.5049 | 500 | 0.7987 |
| 0.7322 | 1.6554 | 550 | 0.7955 |
| 0.7148 | 1.8059 | 600 | 0.7920 |
| 0.71 | 1.9564 | 650 | 0.7894 |
| 0.6651 | 2.1068 | 700 | 0.7947 |
| 0.6684 | 2.2573 | 750 | 0.7936 |
| 0.6826 | 2.4078 | 800 | 0.7928 |
| 0.6711 | 2.5583 | 850 | 0.7920 |
| 0.6757 | 2.7088 | 900 | 0.7915 |
| 0.6791 | 2.8593 | 950 | 0.7913 |
Framework versions
- Transformers 4.45.2
- Pytorch 2.4.0+cu121
- Datasets 2.21.0
- Tokenizers 0.20.3