-
Restore GK v12b safetensors
-
Push GK v10 CoT model from backup repo
-
Push GK v12b no-think
-
Push GK v14 r32 no-think v12 merged
-
Push GK v13 r32 checkpoint 7000
-
Push GK v12b no-think
-
push v13 no think
-
Push GK v12b small
-
Push GK v12 no-think
-
Push GK v12 no-think
-
Upload chat_template.jinja with huggingface_hub
-
Upload generation_config.json with huggingface_hub
-
push v10 model cot
-
change the chat template
-
Push GK v8 checkpoint 7800
-
Push GK v6 r32 no-think checkpoint 6600
-
Push GK v7 MMLU-Pro CoT only model
-
Push GK v6 GRPO hard7k checkpoint 2000
-
v6 no think, r32 lr5e5, ckpt6600
-
Push GK v3 clean 30k 30k merged model think on+ cot
-
Upload qwen3_gk_nemscience30k_mmluprocot30k_pool2k_v3_merged
-
Upload qwen3_gk_nemscience30k_mmluprocot30k_pool2k_v3_merged
-
Upload generation_config.json with huggingface_hub
-
Upload folder using huggingface_hub
-
Push GK 53k short-think LoRA r32 merged model
-
Push best GK model with thinking enabled chat template
-
Upload folder using huggingface_hub
-
Upload folder using huggingface_hub
-
Upload folder using huggingface_hub
-
Upload folder using huggingface_hub
-
Upload folder using huggingface_hub
-
Upload folder using huggingface_hub
-
initial commit
