Akarshan Biswas
@qnixsynapse.bsky.social
Stalking on ML updates.
Engineering inference @ Menlo
Engineering inference @ Menlo
"That's intense" 😂
January 28, 2025 at 3:25 PM
"That's intense" 😂
I haven't seen this level of roast in my life LOL. 😂
January 27, 2025 at 3:29 PM
I haven't seen this level of roast in my life LOL. 😂
The media in India is hilarious man! And no, here Gemma isn't making things up.
January 23, 2025 at 1:42 PM
The media in India is hilarious man! And no, here Gemma isn't making things up.
On the other news, ChatGPT is down.
January 23, 2025 at 12:38 PM
On the other news, ChatGPT is down.
The official finetune of Meta's LLaMA 3.2 3B thinks, strawberry problem is simple and then confidently tells the wrong answer.... LMAO!
January 23, 2025 at 12:32 PM
The official finetune of Meta's LLaMA 3.2 3B thinks, strawberry problem is simple and then confidently tells the wrong answer.... LMAO!
Understanding English like this would have been a bit difficult unless you know Mandarin!
"can't help we discussing" -> "帮不上我们讨论" -> "doesn’t help our discussion"
Mandarin is such a language which is easy on the grammar and very hard on the writing.
Saw this comment on Github.
"can't help we discussing" -> "帮不上我们讨论" -> "doesn’t help our discussion"
Mandarin is such a language which is easy on the grammar and very hard on the writing.
Saw this comment on Github.
January 15, 2025 at 7:12 AM
Understanding English like this would have been a bit difficult unless you know Mandarin!
"can't help we discussing" -> "帮不上我们讨论" -> "doesn’t help our discussion"
Mandarin is such a language which is easy on the grammar and very hard on the writing.
Saw this comment on Github.
"can't help we discussing" -> "帮不上我们讨论" -> "doesn’t help our discussion"
Mandarin is such a language which is easy on the grammar and very hard on the writing.
Saw this comment on Github.
YouTube started showing "video summaries". A must needed feature which can tackle those clickbaity video!! I do that usually with my local LLMs but this is very welcome. Should be on every video.
January 8, 2025 at 4:14 PM
YouTube started showing "video summaries". A must needed feature which can tackle those clickbaity video!! I do that usually with my local LLMs but this is very welcome. Should be on every video.
Thanks to openwebui, I have "Gemma live" on my GPU poor desktop.
Mt. Everest is meters feet tall indeed!
Mt. Everest is meters feet tall indeed!
January 8, 2025 at 2:38 PM
Thanks to openwebui, I have "Gemma live" on my GPU poor desktop.
Mt. Everest is meters feet tall indeed!
Mt. Everest is meters feet tall indeed!
January 3, 2025 at 12:32 PM
Their MLA is also interesting IMO.
December 27, 2024 at 2:21 PM
Their MLA is also interesting IMO.
Interesting app. I have been blocked by 5 people and I have never interacted with any one of them.
All within the past 48 hours.
All within the past 48 hours.
November 28, 2024 at 1:44 PM
Interesting app. I have been blocked by 5 people and I have never interacted with any one of them.
All within the past 48 hours.
All within the past 48 hours.
Guess the model please. 🤣
November 28, 2024 at 1:38 PM
Guess the model please. 🤣
Interesting! Source: huggingface.co/spaces/PR-Pu...
November 26, 2024 at 5:24 PM
Interesting! Source: huggingface.co/spaces/PR-Pu...
I wonder if big companies keep signs like this. 😀
#MachineLearning
#MachineLearning
November 26, 2024 at 1:18 PM
I wonder if big companies keep signs like this. 😀
#MachineLearning
#MachineLearning
Grok came up with "Xcretion" as a new name for "tweet" after Twitter's rebrand. I was surprised because it really defines what X is now currently.
Turns out, it was already "tweeted" by someone else a year ago!
Can LLMs come up with original punchlines? That's the question we need to ask.
Turns out, it was already "tweeted" by someone else a year ago!
Can LLMs come up with original punchlines? That's the question we need to ask.
November 25, 2024 at 12:01 PM
Grok came up with "Xcretion" as a new name for "tweet" after Twitter's rebrand. I was surprised because it really defines what X is now currently.
Turns out, it was already "tweeted" by someone else a year ago!
Can LLMs come up with original punchlines? That's the question we need to ask.
Turns out, it was already "tweeted" by someone else a year ago!
Can LLMs come up with original punchlines? That's the question we need to ask.
