salesGPT_v1

<!-- 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. -->

salesGPT_v1

salesGPT_v1_1

Extension of salesGPT_v1: https://huggingface.co/goendalf666/salesGPT_v1 In this version different LoRa configurations are tested. The backbone of this model is phi1_5: https://huggingface.co/microsoft/phi-1_5

References

The main model is salesGPT_v2: https://huggingface.co/goendalf666/salesGPT_v2 To see the code for the generation, inference and dataset generation see the github repo: https://github.com/tom813/salesGPT_foundation

The datasets that are used to fine tune this model are: https://huggingface.co/datasets/goendalf666/sales-conversations-instruction-base https://huggingface.co/datasets/goendalf666/sales-textbook_for_convincing_and_selling https://huggingface.co/datasets/goendalf666/sales-conversations-2

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0002
• train_batch_size: 2
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• training_steps: 80000

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.0.dev20230829+cu121
• Datasets 2.14.5
• Tokenizers 0.13.3