CodeLlama-7b-Instruct-hf_Fi__CMP_TR_size_304_epochs_10_2024-06-22_21-11-23_3558625

This model is a fine-tuned version of codellama/CodeLlama-7b-Instruct-hf on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.9442
  • Accuracy: 0.464
  • Chrf: 0.282
  • Bleu: 0.212
  • Sacrebleu: 0.2
  • Rouge1: 0.473
  • Rouge2: 0.304
  • Rougel: 0.447
  • Rougelsum: 0.467
  • Meteor: 0.474

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: 0.001
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 3407
  • distributed_type: multi-GPU
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-06
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 304
  • training_steps: 3040

Training results

Training Loss Epoch Step Validation Loss Accuracy Chrf Bleu Sacrebleu Rouge1 Rouge2 Rougel Rougelsum Meteor
0.7293 1.0 304 2.8400 0.471 0.109 0.089 0.1 0.318 0.168 0.304 0.297 0.274
0.043 2.0 608 3.2408 0.498 0.051 0.019 0.0 0.162 0.063 0.136 0.142 0.216
0.0514 3.0 912 2.8322 0.478 0.156 0.059 0.1 0.3 0.145 0.284 0.289 0.289
0.0145 4.0 1216 2.5898 0.478 0.101 0.064 0.1 0.263 0.167 0.258 0.258 0.32
0.8203 5.0 1520 2.7395 0.478 0.16 0.049 0.0 0.306 0.114 0.284 0.298 0.27
0.0546 6.0 1824 2.8379 0.458 0.052 0.022 0.0 0.068 0.0 0.056 0.057 0.21
0.0352 7.0 2128 2.6987 0.481 0.165 0.133 0.1 0.356 0.246 0.352 0.355 0.33
0.042 8.0 2432 2.0781 0.481 0.264 0.169 0.2 0.421 0.261 0.403 0.421 0.431
0.0124 9.0 2736 1.9029 0.464 0.293 0.222 0.2 0.466 0.304 0.445 0.465 0.473
0.0382 10.0 3040 1.9442 0.464 0.282 0.212 0.2 0.473 0.304 0.447 0.467 0.474

Framework versions

  • PEFT 0.7.1
  • Transformers 4.37.0
  • Pytorch 2.2.1+cu121
  • Datasets 2.20.0
  • Tokenizers 0.15.2
Downloads last month
2
Inference API
Unable to determine this model’s pipeline type. Check the docs .

Model tree for vdavidr/CodeLlama-7b-Instruct-hf_Fi__CMP_TR_size_304_epochs_10_2024-06-22_21-11-23_3558625

Adapter
(75)
this model