bin/speculative-news.sh

#!/bin/bash

input="$(< /dev/stdin)"
template="$(< $HOME/Work/ai-strategy/template-summary.txt)"

llama-speculative \
    --model $HOME/.ai/models/llama/idm/Nous-Hermes-13B-GGUF/nous-hermes-llama2-13b.gguf.q6_K.bin \
    --model-draft $HOME/.ai/models/llama/TheBloke/orca_mini_v3_7B-GGUF/orca_mini_v3_7b.Q4_K_M.gguf \
    --grammar-file ../../grammars/json_arr.gbnf \
    --escape \
    --n-gpu-layers 1 \
    --threads 4 \
    --n-predict 512 \
    --ctx-size 4096 \
    --draft 16 \
    --temp -1 \
    --prompt "${input}\n\n${template}\n\n"