본문 바로가기

IT/리눅스

ChatGPT 대항마 Goolge Bard 베타버전 성능, 코딩 테스트

반응형

구글이 난리가 났습니다. 

 

이미 시장은 OpenAI의 ChatGPT가 장악하고 있고 그 속도는 점점 더 빨라지고 있는 듯 합니다. 

뒤늦게 Bard를 베타테스팅 중인데 아직도 갈길이 멀어 보이는 듯 하네요.

 

이미 사람들은 ChatGPT에 익숙해져... 아마 그 이상을 보여주기 전에는 쉽지 않을 듯 합니다.

 

다음은 bard 사이트 입니다. 아직 한글은 지원하지 않습니다.

 

https://bard.google.com

 

Bard

Bard is your creative and helpful collaborator to supercharge your imagination, boost productivity, and bring ideas to life.

bard.google.com

 

한번 대답하기 곤란한 질문을 물어볼까요?

 

"너 자신에 대해서 설명해봐. chat GPT 보다 나은점은 뭐야?" ㅋㅋㅋ

tell me about yourself What's better than chat GPT?

 

 

 

" 저는 유익하고 포괄적이도록 훈련된 대화형 AI 또는 챗봇으로도 알려진 대규모 언어 모델입니다. 저는 방대한 양의 텍스트 데이터에 대해 교육을 받았으며 다양한 프롬프트와 질문에 대한 응답으로 의사 소통하고 사람과 같은 텍스트를 생성할 수 있습니다. 예를 들어 사실적인 주제에 대한 요약을 제공하거나 이야기를 만들 수 있습니다.

ChatGPT보다 더 잘할 수 있는 몇 가지 사항은 다음과 같습니다.

저는 Google 검색을 통해 실제 세계의 정보에 액세스하고 처리할 수 있습니다. 즉, 귀하의 질문에 대해 보다 포괄적이고 최신 답변을 제공할 수 있습니다.
저는 시, 코드, 대본, 음악 작품, 이메일, 편지 등과 같은 다양한 창의적인 텍스트 형식을 생성할 수 있습니다. 즉, 귀하의 작업에서 더 창의적이고 도움이 될 수 있습니다.
아직 개발 중이지만 항상 새로운 것을 배우고 있습니다. 이것은 내가 하는 일에서 끊임없이 나아지고 있다는 것을 의미합니다.
궁금한 점이나 요청 사항이 있으시면 언제든지 문의해 주세요. 기꺼이 도와드리겠습니다. "

 

코딩 실력도 한번 봐야죠? 

 

Write a Python code to save the video input from the camera

 

import cv2

# Define the output video file name
output_file = "output.mp4"

# Create a video writer object
writer = cv2.VideoWriter(output_file, cv2.VideoWriter_fourcc(*'XVID'), 20.0, (640, 480))

# Capture video from the camera
cap = cv2.VideoCapture(0)

# Read frames from the camera and write them to the output file
while True:
    ret, frame = cap.read()
    if not ret:
        break

    writer.write(frame)

# Release the camera and the video writer objects
cap.release()
writer.release()

# Close all open windows
cv2.destroyAllWindows()

 

음.. 나쁘진 않네요. 

 

Bard가 작성한 코드에 주석을 달아 보라고 다시 요청하였습니다. 

 

반응형

 

comment all the code

 

# Import the OpenCV library
import cv2

# Define the output video file name
output_file = "output.mp4"

# Create a video writer object
# The first argument is the output file name
# The second argument is the video codec
# The third argument is the frame rate
# The fourth argument is the resolution
writer = cv2.VideoWriter(output_file, cv2.VideoWriter_fourcc(*'XVID'), 20.0, (640, 480))

# Capture video from the camera
# The first argument is the camera index
cap = cv2.VideoCapture(0)

# Read frames from the camera and write them to the output file
# The `ret` variable is True if a frame was read successfully
# The `frame` variable is the image of the current frame
while True:
    ret, frame = cap.read()
    if not ret:
        break

    writer.write(frame)

# Release the camera and the video writer objects
cap.release()
writer.release()

# Close all open windows
cv2.destroyAllWindows()

 

 

좋네요. 동작 방식은 ChatGPT와 유사한 것 같습니다.

 

 

더 좋은 성능을 보여주는 똑똑한 Bard를 기대해봅니다. 

한놈이 독점하면 꼭 문제가 생기더라고요 ^^

반응형