dataset: timdettmers/openassistant-guanaco
model: PY007/TinyLlama-1.1B-intermediate-step-480k-1T
Trained from the 1T token Tinyllama.
pipe format is: pipe(f"<s>[INST] {prompt} [/INST]")
dataset: timdettmers/openassistant-guanaco
model: PY007/TinyLlama-1.1B-intermediate-step-480k-1T
Trained from the 1T token Tinyllama.
pipe format is: pipe(f"<s>[INST] {prompt} [/INST]")