DedeProGames PRO

ORCIM IVON IA
...NOOS

ETNOZIROH GNOL ED OCITNEGA OLEDOM OSSON

EKAF SI CITNEGA IVON

ESNECIL 0.2 EHCAPA EHT REDNU DESAELER SI LEDOM EHT .TNUOC RETEMARAP STI OT EVITALER SKSAT GNIREENIGNE ERAWTFOS DLROW-LAER DNA GNIDOC LANIMRET CITNEGA NI HTGNERTS RALUCITRAP GNITCELFER — LAVEWALC NO 8.76 DNA ,DNOMAID AQPG NO 9.68 ,ORP HCNEB-EWS NO 1.94 ,DEIFIREV HCNEB-EWS NO 5.37 ,)EDOC EDUALC( 1.2 HCNEB-LANIMRET NO 8.05 GNIDULCNI — KRAMHCNEB DETROPER YREVE YLRAEN SSORCA B13-4-AMMEG REGRAL EHT NEVE DNA ,B9-5.3NEWQ ,B9-0.1-HTINRO ,B9-5.1-HTINRO GNIMROFREPTUO ,SSALC THGIEW STI EVOBA LLEW SEHCNUP B9-REDOCXO ,EZIS TCAPMOC STI ETIPSED .ERUTCETIHCRA TNENOPMOC DNA CIGOL IU FO GNIDNATSREDNU PEED HTIW ,LEDOM B01-BUS A ROF GNINOSAER DNE-TNORF GNORTS YLBATON STIBIHXE DNA ,SETIRWER LLUF NAHT REHTAR SFFID TIDE LAMINIM NEVIRD-CITSONGAID-PSL DNA SNRETTAP ETIRW-EROFEB-DAER EKIL SROIVAHEB YREVOCER-RORRE SNRAEL ,SESABEDOC ELIF-ITLUM XELPMOC GNILDNAH ROF WODNIW TXETNOC EVITAN NEKOT-441,262 LLUF A HTIW SPIHS TI .STNEMNORIVNE XEDOC DNA ,EDOCNEPO ,EDOC EDUALC SSORCA 3.5-MLG DNA 1.5-ELBAF YB DETARENEG SEIROTCEJART CITNEGA REITNORF MORF DELLITSID ,GNINOSAER CITNEGA DNA ,NOITAMOTUA LANIMRET ,GNIDOC CITNEGA NOZIROH-GNOL ROF B9-5.3NEWQ NO TLIUB LEDOM GNIDOC RETEMARAP-NOILLIB-9 THGIEWTHGIL A ,MLLNOIRO YB DEPOLEVED SI B9-REDOCXO

ORCIM IVON IA
...NOOS

ETNOZIROH GNOL ED OCITNEGA OLEDOM OSSON

Please dont spoil it to the others @GGUFGuy

I type soo fast

It cna takr multiple sensions of trainingz but at least we will have a good model

i can just slower

Time isnt quality

QEDOC ANANAB DNIM
NOOS

ik but you have compute?

Why cant you use Molab's RTX Pro 6000

I tought they are partners
But since you're so worried about it, I'll leave Supralabs, seeing as I was working with BananMind first. 😅

Edit: Dont worry, i will never leave SupraLabs, im working with both organizations to create powerful models
Since @Banaxi-Tech asked me to help him with his agentic coding and reasoning model CodeQ, inspired by GRM, my family of reasoning and agentic coding models, im helping him with this awesome project that i have mastered
BananaMind-CodeQ is an lightweight, multi-size, long-horizon agentic coding and reasoning models finetuned to agentic coding, and to compete with models such as Qwen3.8-27B, Qwen3.5-9B, and Spark-X1.7B