Mamy Ratsimbazafy
mratsim
AI & ML interests
Slop Hunter, Creative Sampler, Compressor of Weights, Bearer of the Quantizer Torch
Recent Activity
new activity
about 8 hours ago
mratsim/GLM-4.7-EXL3:Optimized Quant between 3bpw and 3.5bpw
new activity
about 8 hours ago
XiaomiMiMo/MiMo-V2-Flash:Can this be loaded in transformers? FP8Linear: size mismatch for weight_scale_inv
liked
a dataset
about 9 hours ago
ammarnasr/the-stack-rust-clean
Organizations
None yet
Text Adventure, Roleplay, Dungeon Master and Creative Writing models based on GLM-4.5-Air finetunes
Text Adventure, Roleplay, Dungeon Master and Creative Writing models based on Mistral-Small-3.x finetunes
Coding/Programming/Agentic focused models
Text Adventure, Roleplay, Dungeon Master and Creative Writing models based on Mistral-Large finetunes
Text Adventure, Roleplay, Dungeon Master and Creative Writing models based on Llama3.3 finetunes
-
mratsim/Anubis-70B-v1.1-NVFP4
Text Generation • 41B • Updated • 5 • 1 -
mratsim/Anubis-70B-v1.1-NVFP4A16
Text Generation • 41B • Updated • 5 -
mratsim/Dungeonmaster-V2.2-Expanded-LLaMa-70B-NVFP4
Text Generation • 41B • Updated • 39 -
mratsim/Dungeonmaster-V2.2-Expanded-LLaMa-70B-NVFP4A16
Text Generation • 41B • Updated • 6
General purpose models in the 32B to 70B parameters range
Text Adventure, Roleplay, Dungeon Master and Creative Writing models based on Mistral-Large finetunes
Text Adventure, Roleplay, Dungeon Master and Creative Writing models based on GLM-4.5-Air finetunes
Text Adventure, Roleplay, Dungeon Master and Creative Writing models based on Llama3.3 finetunes
-
mratsim/Anubis-70B-v1.1-NVFP4
Text Generation • 41B • Updated • 5 • 1 -
mratsim/Anubis-70B-v1.1-NVFP4A16
Text Generation • 41B • Updated • 5 -
mratsim/Dungeonmaster-V2.2-Expanded-LLaMa-70B-NVFP4
Text Generation • 41B • Updated • 39 -
mratsim/Dungeonmaster-V2.2-Expanded-LLaMa-70B-NVFP4A16
Text Generation • 41B • Updated • 6
Text Adventure, Roleplay, Dungeon Master and Creative Writing models based on Mistral-Small-3.x finetunes
General purpose models in the 32B to 70B parameters range
Coding/Programming/Agentic focused models