<!-- 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. -->

gpt-neox-20b-8k-longtuning2

This model was finetuned on books.

Model description

Base model is GPT-Neox-20B as indicated in the name.

Intended uses & limitations

Research and development

Training and evaluation data

N/A

Training procedure

Running run_clm on the book dataset

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 4
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 8
• total_train_batch_size: 256
• total_eval_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 100000

Framework versions

• Transformers 4.27.4
• Pytorch 2.0.0
• Datasets 2.11.0
• Tokenizers 0.11.0