--- license: other license_name: other license_link: LICENSE ---

Model Mixed by [Solo Merge Method](https://medium.com/@puffanddmx82/enhancing-language-models-with-dynamic-attention-version-2-84ef8adc3646) Keep in mind that the accuracy of your desired questions may vary for this merge. Regardless of whether the idea of new merge method is good or bad, I believe that the actual result of what i thought is of great significance. Once again, there is no right answer for the famous LLM. The correct answer is what you choose based on your evidence with so many real human random test. It is good to rely on the evaluation result score, but in LLM, the most important thing is what you actually feel after taking your real fact random test. The gap is bigger than I thought... If you keep going with the wrong first button, you could end up in a black hole from which you can never escape... By the time you realize it, it’s already too late... When looking at an LLM, don't trust others, trust yourself by real fact check. ### Models Merged The following models were included in the merge: * [MLP-KTLim/llama-3-Korean-Bllossom-8B](https://huggingface.co/MLP-KTLim/llama-3-Korean-Bllossom-8B) * [NousResearch/Meta-Llama-3-8B-Instruct](https://huggingface.co/NousResearch/Meta-Llama-3-8B-Instruct) * [maum-ai/Llama-3-MAAL-8B-Instruct-v0.1](https://huggingface.co/maum-ai/Llama-3-MAAL-8B-Instruct-v0.1) ## Ollama Create ``` (.venv) jaylee@lees-MacBook-Pro-2 youtube % ./ollama create solo -f ./Modelfile_Q5_K_M transferring model data creating model layer creating template layer creating system layer creating parameters layer creating config layer using already created layer sha256:1acd536b4123837aee2f43ffde8a697f842be5ab4d789ab6787a7887291c4bb3 using already created layer sha256:8ab4849b038cf0abc5b1c9b8ee1443dca6b93a045c2272180d985126eb40bf6f using already created layer sha256:ae2974c64ea5d6f488eeb1b10717a270f48fb3452432589db6f5e60472ae96ac using already created layer sha256:74ef6315972b317734fe01e7e1ad5b49fce1fa8ed3978cb66501ecb8c3a2e984 writing layer sha256:88698c3b47bc90bf85949d927c7555efe424e666ef9bd94550bcbde9c4f94489 writing manifest success ``` ## Ollama Modelfile change based on your preference ``` FROM solo-llama-3-maal-mlp-koen-8b-Q5_K_M.gguf TEMPLATE """{{ if .System }}<|start_header_id|>system<|end_header_id|> {{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|> {{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|> {{ .Response }}<|eot_id|>""" SYSTEM """ 친절한 챗봇으로서 상대방의 요청에 최대한 자세하고 친절하게 답하자. 모든 대답은 한국어(Korean)으로 대답해줘. """ PARAMETER num_keep 24 PARAMETER temperature 0.7 PARAMETER num_predict 3000 PARAMETER stop "<|start_header_id|>" PARAMETER stop "<|end_header_id|>" PARAMETER stop "<|eot_id|>" ```