JW17/Untied-49K-tokenized-20B-v0.1

<!-- 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. -->

Untied-49K-tokenized-20B-v0.1

This model is a fine-tuned version of allenai/OLMo-7B-0724-hf on the gartland/finewebedu-49K-tokenized-20B 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: 0.0006
• —trainbatchsize: 64
• —evalbatchsize: 8
• —seed: 42
• —distributed_type: multi-GPU
• —num_devices: 2
• —gradientaccumulationsteps: 2
• —totaltrainbatch_size: 256
• —totalevalbatch_size: 16
• —optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
• —lrschedulertype: cosine
• —lrschedulerwarmup_steps: 700
• —num_epochs: 1

Training results

Framework versions

• —Transformers 4.49.0
• —Pytorch 2.4.1+cu124
• —Datasets 3.6.0
• —Tokenizers 0.21.2