--- license: llama3.1 language: - ko - en base_model: - NousResearch/Meta-Llama-3.1-8B-Instruct distill_model: - deepseek-ai/DeepSeek-R1-Distill-Llama-8B korean_model: - sh2orc/Llama-3.1-Korean-8B-Instruct pipeline_tag: text-generation --- ## Merged below called "SISaAI" - [Meta Llama](https://huggingface.co/NousResearch/Meta-Llama-3.1-8B-Instruct) - [Korean Llama](https://huggingface.co/sh2orc/Llama-3.1-Korean-8B-Instruct) - [DeepSeek R1 Distill Llama](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) There is no such thing as a flawless system. It's about using it appropriately and reasonably without pushing it to its limits. ## Modelfile ``` FROM ./Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q5_K_M.gguf TEMPLATE """{{ if .Messages }} {{- if or .System .Tools }}<|start_header_id|>system<|end_header_id|> {{- if .System }} {{ .System }} {{- end }} {{- if .Tools }} You are a helpful assistant with tool calling capabilities. When you receive a tool call response, use the output to format an answer to the original user question. {{- end }} {{- end }}<|eot_id|> {{- range $i, $_ := .Messages }} {{- $last := eq (len (slice $.Messages $i)) 1 }} {{- if eq .Role "user" }}<|start_header_id|>user<|end_header_id|> {{- if and $.Tools $last }} Given the following functions, please respond with a JSON for a function call with its proper arguments that best answers the given prompt. Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}. Do not use variables. {{ $.Tools }} {{- end }} {{ .Content }}<|eot_id|>{{ if $last }}<|start_header_id|>assistant<|end_header_id|> {{ end }} {{- else if eq .Role "assistant" }}<|start_header_id|>assistant<|end_header_id|> {{- if .ToolCalls }} {{- range .ToolCalls }}{"name": "{{ .Function.Name }}", "parameters": {{ .Function.Arguments }}}{{ end }} {{- else }} {{ .Content }}{{ if not $last }}<|eot_id|>{{ end }} {{- end }} {{- else if eq .Role "tool" }}<|start_header_id|>ipython<|end_header_id|> {{ .Content }}<|eot_id|>{{ if $last }}<|start_header_id|>assistant<|end_header_id|> {{ end }} {{- end }} {{- end }} {{- else }} {{- 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|> {{ end }}{{ .Response }}{{ if .Response }}<|eot_id|>{{ end }}""" PARAMETER stop <|start_header_id|> PARAMETER stop <|end_header_id|> PARAMETER stop <|eot_id|> ``` ## Test ``` ollama create sisaai -f ./modelfile_sisaai ollama run sisaai:latest ``` ``` jaylee@lees-MacBook-Pro-2 merge % ollama run sisaai:latest >>> 대한민국 수도는 서울입니다. ``` ``` >>> 대한민국 대통령은 현재 대한민국의 대통령은 윤석열입니다. (2023년 5월 4일 기준) ``` ``` >>> 피보나치 수열 파이썬 코드 피보나치 수열은 0과 1로 시작하는 수열이며, 이어서 다음 수는 이전 두 수의 합입니다. 예를 들어: 0, 1, 1, 2, 3, 5, 8, 13, ... 파이썬에서 피보나치 수열을 구현할 수 있습니다. ### 방법 1: 반복문 python def fibonacci(n): if n <= 0: return None elif n == 1: return 0 elif n == 2: return 1 else: a, b = 0, 1 for _ in range(3, n+1): a, b = b, a + b return b # 예제: 10번째 피보나치 수 print(fibonacci(10)) # 55 ### 방법 2: 재귀함수 python def fibonacci(n): if n <= 0: return None elif n == 1: return 0 else: return fibonacci(n-1) + fibonacci(n-2) # 예제: 10번째 피보나치 수 print(fibonacci(10)) # 55 ### 방법 3:-generator python def fibonacci(): a, b = 0, 1 while True: yield a a, b = b, a + b # 예제: 첫 5개의 피보나치 수 f = fibonacci() for _ in range(5): print(next(f)) # 0, 1, 1, 2, 3 ### 방법 4: 이터레이션 python def fibonacci(n): if n <= 0: return None a, b = 0, 1 result = [] for _ in range(n): result.append(a) a, b = b, a + b return result # 예제: 10개의 피보나치 수 print(fibonacci(10)) # [0, 1, 1, 2, 3, 5, 8, 13, 21, 34] ### 방법 5: 매개변수 n의 효율성 python def fibonacci(n): if n <= 0: return None elif n == 1: return 0 elif n == 2: return 1 else: a, b = 0, 1 for i in range(3, n+1): a, b = b, a + b return b # 예제: 10000번째 피보나치 수 print(fibonacci(10000)) ### 메모: 이중 재귀에 대한 성능 상의 문제 재귀적 방법은 큰 n에 대해 비효율적일 수 있습니다. 반복적인 방법이 더 좋습니다. 피보나치 수열을 연산하고 싶으면, [이곳](https://github.com/SeongHyunYoon/fibonacci)에서 Python의 다양한 구현법을 확인할 수 있습니다. ``` ``` >>> third eye blind 가수 소개 Third Eye Blind는 미국의 록 밴드입니다. 1993년에 결성되었으며, 그들은 90년대와 2000년대의 록 음악 장르에서 주목받은 바 있습니다. ### 멤버 - 스티븐 제네 (Steve Jenkins): 보컬, 리더 - 아론 앙드류스 (Aaron Andreoli): 기타, 백보컬 - 켄데사스 윌리엄슨 (Kendall Williamson): 베이스 - 콜린 세이크 (Colin Sears): 드럼 ### 데뷔와 성공 Third Eye Blind는 1997년에 데뷔 음반 "Third Eye Blind"을发布했습니다. 이 음반은 대히트가 되었고, "Semi-Charmed Life", "How's It Going to Be", "Jumper", "Goodbye for Goodness' Sake" 등 다양한 히트곡들을 발표했습니다. ### 두 번째와 세 번째 음반 두 번째 음반 "Out of the Veil"은 2000년에 발매되었고, 세 번째 음반 "The Third Eye Blind Collection"은 2002년에서 2003년에 발매되었습니다. 이 밴드는 또한 여러 개의 EP와 싱글 음반들을 발표했습니다. ### 해체 Third Eye Blind는 2005년에 활동을 중단했지만, 2019년에 복귀를 선언하고 "Warriors"를 발표했습니다. 그들은 계속해서 새로운 곡들을 발표하고 콘서트도 열고 있습니다. ### 음악风格 Third Eye Blind의 음악은 록, 팝, 일렉트로닉, 힙합 등 다양한 장르를 혼합한 것이 특징입니다. 보컬 스티븐 제네의 목소리는 유니크하며, 그들은 세련된 리듬과 멜로디를 가지고 있습니다. ### 인물介绍 - 스티븐 제네 (Steve Jenkins): Third Eye Blind의 리더이자 보컬이며, 노래에 많은 영감을 주는 사람입니다. 그는 또한 음악 프로듀서와 작곡가이기도 합니다. - 아론 앙드류스 (Aaron Andreoli): 기타리스트이자 백보컬로, 밴드에서 중요한 역할을 합니다. 그는 또한 여러 음반에 참여한 적이 있습니다. Third Eye Blind는 그들의 독특하고 매력적인 음악으로 많은 팬들을 이끌고 있습니다. 그들은 90년대와 2000년대의 록 음악사에 중요한 역할을 했습니다. >>> 영어로 Third Eye Blind is an American rock band formed in 1993. They gained popularity in the 1990s and 2000s, particularly with their hit songs like "Semi-Charmed Life" and "Jumper". ### Members - Steve Jenkins: Lead vocals, rhythm guitar - Aaron Andreoli: Lead guitar, backing vocals - Kendall Williamson: Bass guitar - Colin Sears: Drums ### Debut and Success Third Eye Blind released their debut album in 1997, which became a huge hit. They also released several successful singles such as "How's It Going to Be", "Jumper", and "Goodbye for Goodness' Sake". ### Subsequent Albums Their second album, "Out of the Veil", was released in 2000. The third album, "The Third Eye Blind Collection", came out in 2002-2003. ### Hiatus and Reunion Third Eye Blind went on hiatus in 2005 but announced their reunion in 2019. They have since released new music and continued to tour. ### Musical Style Third Eye Blind's music combines elements of rock, pop, electronic, and hip-hop. The band is known for their catchy melodies and Steve Jenkins' distinctive vocals. ### Notable Individuals - Steve Jenkins: Lead singer, songwriter, and producer who brings a unique energy to the band. - Aaron Andreoli: Guitarist and backing vocalist who contributes to the band's sound. Third Eye Blind remains a beloved rock band with a dedicated fan base. Their music has had a significant impact on the 90s and early 2000s rock scene. ``` ``` >>> 뉴욕 여행 일정을 짜줘 뉴욕 여행에 대해 알려드립니다. ### 첫째날 - 아침: 에이플펀드 타워에서 시작합니다. 유리 천장과 주변의 아름다운 경관을欣赏하고, 메트로폴리탄 미술관을 방문해 세계 유명한 예술작품들을 만나보세요. - 점심: 세인트 패트릭스 카페에서 아이스크림과 샌드위치를 즐기고, 하계 파티가 있는 Central Park를 방문하세요. 천연 호수와 나무들이 아름답습니다. - 저녁: 브로드웨이 뮤지컬을 보러 가거나 Times Square로 이동해 뉴욕의 밤모습을 즐ぎ세요. ### 둘째날 - 아침: Statue of Liberty Cruises를 타고 자유의 여신을 방문하고, 9/11 메모리얼 박스를 방문하세요. - 점심: Wall Street에서 스마일링 인코어로 유명한 피자집에 가서 뉴욕 특유의 패스트푸드를 맛보세요. - 저녁: SoHo로 이동해 쇼핑을 하거나, Tribeca로 가면 이탈리아와 스페인의 음식을 즐길 수 있습니다. ### 셋째날 - 아침: 브롱크스에 있는 뉴욕주립대학교로 가서 아트빌딩을 방문하세요. 전 세계적으로 유명한 현대 미술관입니다. - 점심: 하이브리드 공룡 (Hybrid T-Rex)과 사파리 버스 (Safari Bus)를 타고 Central Park를 거닐어 보세요. 야생 동물들처럼 느껴집니다. - 저녁: 한정식을 즐기거나, 플랫IRON Steakhouse에서 스테이크와 와인으로 끝내세요. ### 넷째날 - 아침: 하계의 뉴욕 시티를 걸으며 브루클린 브리지 (Brooklyn Bridge)과 Manhattan Skyline을 보세요. - 점심: Coney Island Beach에 가서 해변과 놀이기구, 식당들을 즐기고, 파라다이스 패크 (Paradise Park)와 세인트 루카스 공원 (St. Luke's Place)을 방문하세요 . - 저녁: 하계의 영화제를 보거나, 뉴욕 시티의 야경을欣赏하고, 해병대 기념관 (The Intrepid Sea, Air & Space Museum)에서 끝내세요. ### 다섯째날 - 아침: 브루클린에 있는 Brooklyn Museum을 방문하세요. 전 세계적으로 유명한 미술관입니다. - 점심: Williamsburg에서 먹는 뉴욕 특유의 스미스 하우스 (Smith House)와 Smorgasborg를 방문하세요. - 저녁: 메이시스 스타디움 (Macy's Stadium)과 Flatbush Ave.로 가면 야외 영화제를 보거나, Manhattan Bridge에서 끝내세요. ### 여섯째날 - 아침: 유니온 스퀘어 공원 (Union Square Park)에서 시작합니다. - 점심: 유니온 스퀘어에서 점심을 즐기고, 5 avenue와 23 street에 가면 아시안 식당들을 만나보세요. - 저녁: 뉴욕 시티의 밤모습을欣赏하고, Broadway를 통해 마지막 뮤지컬을 보세요. 이 여행 일정은 뉴욕의 유명한 장소들을 방문하는 것이며, 각 지역마다不同的 경험과 문화를 즐길 수 있습니다. 여유롭게 여행해 주세요! ``` ## Provided Quants | Link | Type | Size/GB | Notes | |:-----|:-----|--------:|:------| | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q2_K.gguf) | Q2_K | 3.3 | | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q3_K_S.gguf) | Q3_K_S | 3.8 | | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q3_K_M.gguf) | Q3_K_M | 4.1 | lower quality | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q3_K_L.gguf) | Q3_K_L | 4.4 | | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.IQ4_XS.gguf) | IQ4_XS | 4.6 | | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q4_K_S.gguf) | Q4_K_S | 4.8 | fast, recommended | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q4_K_M.gguf) | Q4_K_M | 5.0 | fast, recommended | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q5_K_S.gguf) | Q5_K_S | 5.7 | | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q5_K_M.gguf) | Q5_K_M | 5.8 | | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q6_K.gguf) | Q6_K | 6.7 | very good quality | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.Q8_0.gguf) | Q8_0 | 8.6 | fast, best quality | | [GGUF](https://huggingface.co/mradermacher/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF/resolve/main/Llama-3.1-SISaAI-Ko-merge-8B-Instruct.f16.gguf) | f16 | 16.2 | 16 bpw, overkill | Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better): ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png) ``` @article{Llama-3.1-SISaAI-Ko-merge-8B-Instruct, title={asiansoul/Llama-3.1-SISaAI-Ko-merge-8B-Instruct-GGUF}, author={Asiansoul called "Twp Eye Blind"}, merged={Asiansoul called "Twp Eye Blind"}, year={2025}, url = {https://huggingface.co/asiansoul/Llama-3.1-SISaAI-Ko-merge-8B-Instruct} } ```