metadata
library_name: peft
license: other
base_model: deepseek-ai/deepseek-coder-6.7b-base
tags:
- generated_from_trainer
model-index:
- name: lemexp-task1-v3-template_small_nodefs-deepseek-coder-6.7b-base
results: []
lemexp-task1-v3-template_small_nodefs-deepseek-coder-6.7b-base
This model is a fine-tuned version of deepseek-ai/deepseek-coder-6.7b-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1382
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: 0.0004
- train_batch_size: 4
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- total_train_batch_size: 16
- total_eval_batch_size: 8
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 12
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.2943 | 0.2001 | 720 | 0.2140 |
| 0.1986 | 0.4001 | 1440 | 0.1763 |
| 0.1686 | 0.6002 | 2160 | 0.1616 |
| 0.1577 | 0.8002 | 2880 | 0.1504 |
| 0.1477 | 1.0003 | 3600 | 0.1454 |
| 0.1318 | 1.2003 | 4320 | 0.1396 |
| 0.1269 | 1.4004 | 5040 | 0.1409 |
| 0.1263 | 1.6004 | 5760 | 0.1312 |
| 0.1258 | 1.8005 | 6480 | 0.1292 |
| 0.1201 | 2.0006 | 7200 | 0.1226 |
| 0.108 | 2.2006 | 7920 | 0.1280 |
| 0.1055 | 2.4007 | 8640 | 0.1271 |
| 0.1033 | 2.6007 | 9360 | 0.1222 |
| 0.1042 | 2.8008 | 10080 | 0.1169 |
| 0.1032 | 3.0008 | 10800 | 0.1154 |
| 0.0874 | 3.2009 | 11520 | 0.1194 |
| 0.0902 | 3.4009 | 12240 | 0.1132 |
| 0.0891 | 3.6010 | 12960 | 0.1170 |
| 0.0878 | 3.8011 | 13680 | 0.1136 |
| 0.088 | 4.0011 | 14400 | 0.1111 |
| 0.0754 | 4.2012 | 15120 | 0.1146 |
| 0.076 | 4.4012 | 15840 | 0.1115 |
| 0.0762 | 4.6013 | 16560 | 0.1125 |
| 0.0771 | 4.8013 | 17280 | 0.1065 |
| 0.0749 | 5.0014 | 18000 | 0.1085 |
| 0.0642 | 5.2014 | 18720 | 0.1124 |
| 0.0649 | 5.4015 | 19440 | 0.1092 |
| 0.0656 | 5.6016 | 20160 | 0.1073 |
| 0.0673 | 5.8016 | 20880 | 0.1055 |
| 0.065 | 6.0017 | 21600 | 0.1046 |
| 0.0565 | 6.2017 | 22320 | 0.1111 |
| 0.0566 | 6.4018 | 23040 | 0.1106 |
| 0.0567 | 6.6018 | 23760 | 0.1132 |
| 0.057 | 6.8019 | 24480 | 0.1066 |
| 0.0564 | 7.0019 | 25200 | 0.1074 |
| 0.049 | 7.2020 | 25920 | 0.1171 |
| 0.0478 | 7.4021 | 26640 | 0.1096 |
| 0.0481 | 7.6021 | 27360 | 0.1151 |
| 0.0483 | 7.8022 | 28080 | 0.1063 |
| 0.0487 | 8.0022 | 28800 | 0.1124 |
| 0.0411 | 8.2023 | 29520 | 0.1167 |
| 0.0398 | 8.4023 | 30240 | 0.1151 |
| 0.0401 | 8.6024 | 30960 | 0.1173 |
| 0.0411 | 8.8024 | 31680 | 0.1139 |
| 0.0399 | 9.0025 | 32400 | 0.1134 |
| 0.0333 | 9.2026 | 33120 | 0.1223 |
| 0.034 | 9.4026 | 33840 | 0.1184 |
| 0.0345 | 9.6027 | 34560 | 0.1249 |
| 0.0352 | 9.8027 | 35280 | 0.1201 |
| 0.0338 | 10.0028 | 36000 | 0.1183 |
| 0.0288 | 10.2028 | 36720 | 0.1291 |
| 0.0283 | 10.4029 | 37440 | 0.1312 |
| 0.0288 | 10.6029 | 38160 | 0.1287 |
| 0.0289 | 10.8030 | 38880 | 0.1264 |
| 0.0299 | 11.0031 | 39600 | 0.1309 |
| 0.0259 | 11.2031 | 40320 | 0.1409 |
| 0.0254 | 11.4032 | 41040 | 0.1382 |
| 0.0252 | 11.6032 | 41760 | 0.1400 |
| 0.0254 | 11.8033 | 42480 | 0.1382 |
Framework versions
- PEFT 0.14.0
- Transformers 4.47.0
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.1