OLMo icon indicating copy to clipboard operation
OLMo copied to clipboard

I'm interested in OLMo-twin, but I found no more information except its name.

Open HuXinjing opened this issue 1 year ago • 2 comments

❓ The question

So, where do i find more information about OLMo-twin, such as its backbone.

HuXinjing avatar Mar 04 '24 11:03 HuXinjing

@HuXinjing Both the OLMo-7B and OLMo-7B-Twin-2T have the same architecture, and have been trained from scratch on the same data. You can find the model here: https://huggingface.co/allenai/OLMo-7B-Twin-2T

AkshitaB avatar Mar 04 '24 22:03 AkshitaB

That's to say "twin" here means it is the sibling of OLMo-7b? I misunderstood it. 

---Original--- From: "Akshita @.> Date: Tue, Mar 5, 2024 06:30 AM To: @.>; Cc: @.@.>; Subject: Re: [allenai/OLMo] I'm interested in OLMo-twin, but I found no moreinformation except its name. (Issue #479)

@HuXinjing Both the OLMo-7B and OLMo-7B-Twin-2T have the same architecture, and have been trained from scratch on the same data. You can find the model here: https://huggingface.co/allenai/OLMo-7B-Twin-2T

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>

HuXinjing avatar Mar 05 '24 02:03 HuXinjing

@HuXinjing - correct, same models, except for the hardware they're trained on: Twin is trained on LUMI (AMD) while the non-twin is on Mosaic (NVIDIA). Please reopen this if you have followup questions.

dumitrac avatar May 08 '24 00:05 dumitrac