armanakbari4/sdar_4b_opencodeinstruct-dynamic-10epochSansa__mix_BackwardAveraged_0129_0640

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

sdar4bopencodeinstruct-dynamic-10epochSansa_mixBackwardAveraged['opencodeinstructeob']blockNone01290640

This model is a fine-tuned version of training/model/SDAR-4B-Chat-mix on the opencodeinstruct_eob dataset.

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
• —trainbatchsize: 32
• —evalbatchsize: 8
• —seed: 42
• —distributed_type: multi-GPU
• —num_devices: 8
• —gradientaccumulationsteps: 8
• —totaltrainbatch_size: 2048
• —totalevalbatch_size: 64
• —optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
• —lrschedulertype: cosine
• —lrschedulerwarmup_ratio: 0.03
• —num_epochs: 4

Training results

Framework versions

• —Transformers 4.52.4
• —Pytorch 2.8.0+cu128
• —Datasets 3.6.0
• —Tokenizers 0.21.1