mesolitica
/

llama-600m-hf-32768-fpf

Text Generation

Inference Endpoints

text-generation-inference

Model card Files Files and versions Community

Edit model card

Full Parameter Finetuning 600M 32768 context length Llama2 on Malaysian text

600M derived from first 2 layers 7B model.

README at https://github.com/mesolitica/malaya/tree/5.1/session/llama2#600m-32768-context-length-flash-attention-2

WandB, https://wandb.ai/mesolitica/fpf-Llama-2-600m-32k-hf

Downloads last month: 5

Safetensors

Model size

667M params

Tensor type

F32

·

FP16

·

Collection including mesolitica/llama-600m-hf-32768-fpf

Malaysian Smaller Llama

Non-Official smaller Llama FPF on Malaysian context. • 14 items • Updated 26 days ago