69 1 year ago

This model was fine-tuned on a Turkish corpus consisting of 21,399 Wikipedia articles and approximately 5,000 multiple-choice questions sourced from the alibayram/turkish_mmlu dataset.

ollama run ahmets/gemma3-4b-finetuned

Details

1 year ago

8f9e94f38210 · 2.5GB ·

gemma3
·
3.88B
·
Q4_K_M
{ "stop": [ "<end_of_turn>" ], "temperature": 1, "top_k": 64, "top_p": 0
{{- range $i, $_ := .Messages }} {{- $last := eq (len (slice $.Messages $i)) 1 }} {{- if or (eq .Rol

Readme

No readme