MediSimplifier — Merged OpenBioLLM-8B

Merged LoRA adapter (r=32, all_attn, rsLoRA) into aaditya/Llama3-OpenBioLLM-8B for medical text simplification.

Built with Meta Llama 3. Released under the Llama 3 Community License.

Base model: aaditya/Llama3-OpenBioLLM-8B (Llama-3-based) Fine-tuning: LoRA r=32, target modules Q/K/V/O, 3 epochs on GuyDor007/medisimplifier-dataset Evaluation: ROUGE-L 0.6638 on 1,001-sample test split

For details, see github.com/deepset01-sys/medisimplifier-nebius.

Downloads last month: 71

Safetensors

Model size

8B params

Tensor type

F16

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for chambul/MediSimplifier-OpenBioLLM-merged

Base model

meta-llama/Meta-Llama-3-8B

Finetuned

aaditya/Llama3-OpenBioLLM-8B

Adapter

(51)

this model