diff --git a/.nojekyll b/.nojekyll index bcd59e80..989d0bf5 100644 --- a/.nojekyll +++ b/.nojekyll @@ -1 +1 @@ -4d062164 \ No newline at end of file +1c3e6b6f \ No newline at end of file diff --git a/blog.html b/blog.html index 2b5dfc49..42fb8421 100644 --- a/blog.html +++ b/blog.html @@ -247,7 +247,7 @@
Personally I have been following ML/AI for the past five years atleast. Last year really was hyped up. Initially there was this releases in DALL-E and it’s open source alternative Craiyon(then named DALLE mini) being viral. During this time so many trending things like Stable Diffusion, Whisper etc came.
One month back, I started seeing my cousins, parents etc. discussing about ChatGPT and how awesome it is. Some of the application apps like Roamaround.ai build on top of ChatGPT was being used. The final nail was last Sunday, when my grandfather called me to read about ChatGPT article in news.
I have seen people get hyped up about advances in ML. Yet I have never heard about the new technology being discussed by any of my family member like this. Hamel Hussein put this really well in his tweet about what is different in this wave of ML advances.
- +The GPT-4 model was released by OpenAI today IST. It’s been a hectic week already TBH, with so many releases already.
- +The initial reaction to this model has been awesome, and lot of people have been reacting about this.
- +You can read more about this release in below resources:
Check out the summary of what was highlights of GPT-4 and it’s developer livestream in this tweet thread by FSDL.
- +Which is heavier, two pounds of brick or one pound of feathers?
GPT-4 seems to give the correct answer at time of writing.
- +Bing-GPT seems to give wrong answer.
- +Chat-GPT seems to now give correct answer, it was previously giving wrong answer.
- +Charles seems to have come with an explaination to this problem, at the time I was writing this article.
“GPT-4” is really a series of models, and the one live in chat now is likely the 0314 checkpoint available via the API. BingChat can’t’ve launched with that checkpoint and likely has branched off in the intervening weeks.”
- +I am an AI Engineer & Team Lead working in Sentient.io, a fast-paced startup based in Singapore. I have multiple years of experience in Python and Machine learning experience. I have contributed to various open source organizations like Keras, DVC, HuggingFace, fast.ai, Swathanthra Malayalam Computing, CloudCV etc.
I like to speak and interact with developer community. I am an active member of fast.ai community and engage in lot of other communities based in Kerala. I have given talk as expert in competence of Machine Learning and core Python in Pycon India, which is the largest gathering of Pythonistas in India for the Python programming language.
-I believe in the old Malayalam proverb, “അണ്ണാൻ കുന്നിനും തന്നാൽ ആവത്”. In English it translates to this meaning - “A small rabbit does what it can do it’s best of abilities, similarly we should also strive to do everything to best of our abilities.”.
+I believe in the old Malayalam proverb, “അണ്ണാങ്കുഞ്ഞും തന്നാലാവത്”. In English it translates to this meaning - “Even a squirrel does what it can do it’s best of abilities, similarly we should also strive to do everything to best of our abilities.”.
Smash that donate button like a true superstar, your contribution to help my work by buying me a coffee would be absolutely pixel-perfect amazing.
I am quite easy to reach out via email or feel free to connect or DM me via linkedIn or twitter.
diff --git a/posts/2022/2022-05-01-audiocnndemo.html b/posts/2022/2022-05-01-audiocnndemo.html index ae8b17d4..4a0d982e 100644 --- a/posts/2022/2022-05-01-audiocnndemo.html +++ b/posts/2022/2022-05-01-audiocnndemo.html @@ -1006,7 +1006,7 @@We trained a ML model which can identify with 54.4% accuracy to classify in a music file which genre it is. It’s not so bad for a baseline model. Dien Hoa Truong has shared some techniques which he learned during Kaggle competition with music genres.
- +Thanks for reading :pray:
diff --git a/posts/2022/2022-07-26-wandbbug.html b/posts/2022/2022-07-26-wandbbug.html index 6b73fad1..7f29fa1a 100644 --- a/posts/2022/2022-07-26-wandbbug.html +++ b/posts/2022/2022-07-26-wandbbug.html @@ -277,9 +277,9 @@Hence I learned this valuable lesson, which fixed my bug in inferencing code for huggingface spaces which I was creating.
- +Zach Mueller also confirmed this is the case.
- + diff --git a/posts/2023/fossunited/kochi.html b/posts/2023/fossunited/kochi.html index ac319d74..d62ff5e0 100644 --- a/posts/2023/fossunited/kochi.html +++ b/posts/2023/fossunited/kochi.html @@ -233,7 +233,7 @@I have seen people get hyped up about advances in ML. Yet I have never heard about the new technology being discussed by any of my family member like this. Hamel Hussein put this really well in his tweet about what is different in this wave of ML advances.
- +The GPT-4 model was released by OpenAI today IST. It’s been a hectic week already TBH, with so many releases already.
- +The initial reaction to this model has been awesome, and lot of people have been reacting about this.
- +You can read more about this release in below resources:
Check out the summary of what was highlights of GPT-4 and it’s developer livestream in this tweet thread by FSDL.
- +Which is heavier, two pounds of brick or one pound of feathers?
GPT-4 seems to give the correct answer at time of writing.
- +Bing-GPT seems to give wrong answer.
- +Chat-GPT seems to now give correct answer, it was previously giving wrong answer.
- +Charles seems to have come with an explaination to this problem, at the time I was writing this article.
“GPT-4” is really a series of models, and the one live in chat now is likely the 0314 checkpoint available via the API. BingChat can’t’ve launched with that checkpoint and likely has branched off in the intervening weeks.”
- +