metadata
library_name: transformers
license: other
base_model: Qwen/Qwen2.5-7B-Instruct
tags:
- llama-factory
- full
- generated_from_trainer
model-index:
- name: agent_trek
results: []
agent_trek
This model is a fine-tuned version of /ceph/home/muhan01/huggingfacemodels/Qwen2.5-7B-Instruct on the agent_trek dataset. It achieves the following results on the evaluation set:
- Loss: 0.2893
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: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- total_eval_batch_size: 8
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 1
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.3906 | 0.1352 | 200 | 0.4129 |
| 0.3823 | 0.2704 | 400 | 0.3794 |
| 0.3507 | 0.4056 | 600 | 0.3506 |
| 0.3065 | 0.5408 | 800 | 0.3299 |
| 0.3042 | 0.6760 | 1000 | 0.3088 |
| 0.2977 | 0.8112 | 1200 | 0.2946 |
| 0.2908 | 0.9464 | 1400 | 0.2895 |
Framework versions
- Transformers 4.57.1
- Pytorch 2.7.1+cu126
- Datasets 4.0.0
- Tokenizers 0.22.2-rc0