Downloads · 30 days
0
herta_lora_v0.1
Vexxations/herta_lora_v0.1
herta_lora_v0.1 is a machine learning model from Vexxations. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as apache-2.0.
- Developed by: Vexxations - License: apache-2.0 - Finetuned from model : unsloth/llama-3-8b-Instruct-bnb-4bit
Access
Public
Updated Feb 12, 2025
Repo size
185 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors168 MB · 91%
At a glance
- Library
- transformers
- License
- apache-2.0
- Access
- Public
- Created
- Feb 12, 2025
- Updated
- Feb 12, 2025
- SHA
- 023fbe22
Base models
Model card
From the Hugging Face model README
Uploaded model
- Developed by: Vexxations
- License: apache-2.0
- Finetuned from model : unsloth/llama-3-8b-Instruct-bnb-4bit
LoRA adapters for Hertabot v0.1, a conversational chatbot finetuned from llama-3-8b-instruct for deploying on Discord. This is an incredibly early-stage model that will be replaced as soon as improvements are made - and there are PLENTY
Planned additions:
- Discord action integration: allow for usage of images, emojis, gifs and other features on discord
- Memory: remember users by name, traits inferred from messages
- User recognition: distinguish users in context and address them accordingly
- Enhanced persona believability: ensure Herta has no knowledge of "real life" subjects or distances herself from it (technically Earth is insignificant to her, and she's from the future)
- Improved system prompt: Prompt atm is using Deepseekslop and I'd rather look over it and finetune it myself
- Etc.
This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.
- Library
- transformers
- License
- apache-2.0
- Languages
- en
- Created
- Feb 12, 2025
- Updated
- Feb 12, 2025