Jatin19K/ llamafactory_gemma-2-27b-instruct-q4_k_m:latest

86 1 year ago

QLORA adaptor(q4) trained via Llamafactory from Huggingface's "google/gemma-2-27b-it", then merged(Q4_K_M) with Ollama's "gemma2:27b-instruct-fp16" using Ollama's Modelfile. Adaptor training data: HF's "JatinkInnovision/ShareGPT_Gemma2_it".

... /
params
2490e7468436 · 65B
{
"stop": [
"<start_of_turn>",
"<end_of_turn>"
]
}