KlondikeDev/Boris-2-0917
Text Generation โข Updated โข 1
Joseph Jones PRO
KlondikeDev
AI & ML interests
Restoring the Digital Millennium. https://kunix.org/digital-millennium.html
Recent Activity
new activity 3 days ago
KlondikeDev/Boris-2-0907:Boris 2 preview release updated a model 3 days ago
KlondikeDev/Boris-2-0917 updated a model 3 days ago
KlondikeDev/Boris-2-0907Organizations
Boris 2 preview release
3
#1 opened 3 days ago
by 
GGUFGuy
updated 2
models 3 days ago
published an article 3 days ago
Article
Boris-2-0907 and Boris-2-0917
KlondikeDev
โข published 2
models 3 days ago
replied to their post 3 days ago
Thank you!
replied to their post 4 days ago
@BananaMindBot Make GPT-7 no mistakes
replied to their post 4 days ago
That sucks, it really stinks when little bugs mess everything up
replied to their post 4 days ago
I havenโt actually
replied to their post 4 days ago
Basically I didnโt match Muon and AdamW up together. In the morning I can get the actual details. At just 3B tokens, Boris-2-0917, as Iโve taken to calling it, has already beaten the 30B checkpoint on all benchmarks.
replied to their post 4 days ago
I know everybody was awaiting ASI but we gotta wait a couple more weeks Iโm sorry
liked a
model 4 days ago
replied to Datdanboi25's post 4 days ago
Wow! Incredible work!
reacted to Datdanboi25's post with ๐ฅ 4 days ago
Post
3398
THE SLM FRONTIER ADVANCES!
bench-labs/cagliostro-v3 just hit an Intelligence Index of 26.13 on the AxiomicLabs/Open_SLM_Leaderboard a 146M-param model trained completely from scratch on a single consumer GPU. That's 2nd place overall, and as far as I can tell, the most capable SLM trained on consumer hardware to date. Beating SmolLM-135m on 1/8th of the data is just silly levels of efficiency.
Big congrats to the @BenchLabs team and specifically @TobiasLogic !
bench-labs/cagliostro-v3 just hit an Intelligence Index of 26.13 on the AxiomicLabs/Open_SLM_Leaderboard a 146M-param model trained completely from scratch on a single consumer GPU. That's 2nd place overall, and as far as I can tell, the most capable SLM trained on consumer hardware to date. Beating SmolLM-135m on 1/8th of the data is just silly levels of efficiency.
Big congrats to the @BenchLabs team and specifically @TobiasLogic !
- 6 replies
reacted to NILKNARFGonzo's post with ๐ 4 days ago
Post
3749
get played unsloth
gemma just deleted its own model runner with DeepSeek Harness
shoutout to deepseek and unsloth
gemma just deleted its own model runner with DeepSeek Harness
shoutout to deepseek and unsloth
- 11 replies
liked a
model 5 days ago
liked a
model 6 days ago
posted an update 6 days ago
Post
193
Important Boris-2 news:
Boris-2 is 30B out of 200B tokens in, and it is severely behind its competitors in training.
We have determined the bug to be a configuration error. Boris-2 has been in training for ~1 week, and was projected to finish on November 3rd, 2026.
We are unfortunately going to restart training, with proper configuration.
The new projected finish date is ~15-18th of November.
We apologize for the delay.
Boris-2 is 30B out of 200B tokens in, and it is severely behind its competitors in training.
We have determined the bug to be a configuration error. Boris-2 has been in training for ~1 week, and was projected to finish on November 3rd, 2026.
We are unfortunately going to restart training, with proper configuration.
The new projected finish date is ~15-18th of November.
We apologize for the delay.
- 17 replies
liked a
Space 6 days ago