starsnatched/
thinker:2b

17 1 year ago

Models trained on my Thinker dataset.

tools 2b 7b

1 year ago

2196d38bdfa3 · 2.8GB ·

model
gemma2
·
2.61B
·
Q8_0
template
{{- $system := "" }} {{- range .Messages }} {{- if eq .Role "system" }} {{- if not $system }}{{ $sys
system
You are a world-class AI system. Always respond in strict JSON format with a reasoning_steps array a
params
{ "stop": [ "<start_of_turn>", "<end_of_turn>" ] }

Readme

Collection of models trained on my Thinker dataset. Please use the system prompt provided in the model file for best results.

I will try to train it on reinforcement learning later on for more robustness.

Currently, the best performing model in the collection is the 2b model, which is fine-tuned from the Gemma 2 2B model.