Post
591
Mariusz Kurman
mkurman
AI & ML interests
AI Tech Lead | MD
Recent Activity
liked
a dataset
1 day ago
zou-lab/MedCaseReasoning
liked
a model
5 days ago
Menlo/Jan-nano
liked
a dataset
7 days ago
lingshu-medical-mllm/ReasonMed
Organizations
Posts
17
Post
885
Just released NVAMP Loss!
āļø modification of the cross-entropy loss function designed specifically for training LLMs.
āļø twist on the standard cross-entropy loss by emphasizing the importance of outlier prediction errors and dynamically normalizing token-level variance.
āļø more stable and efficient training, leading to models that generalize better.
Check it out, give it a spin, and let me know what you think!
Licensed under the Apache 2.0 license and ready to use. Happy training! š„š¤
https://github.com/mkurman/nvamp-loss
āļø modification of the cross-entropy loss function designed specifically for training LLMs.
āļø twist on the standard cross-entropy loss by emphasizing the importance of outlier prediction errors and dynamically normalizing token-level variance.
āļø more stable and efficient training, leading to models that generalize better.
Check it out, give it a spin, and let me know what you think!
Licensed under the Apache 2.0 license and ready to use. Happy training! š„š¤
https://github.com/mkurman/nvamp-loss
models
6
mkurman/llama-3.2-MEDIT-3B-o1-GRPO-LLM-Eval
Text Generation
ā¢
Updated
ā¢
11
mkurman/Llama-3.2-MedIT-3B-R1
Updated
ā¢
85
ā¢
1
mkurman/Llama-3.2-MedIT-SUN-2.5B-BT-GRPO
Updated
ā¢
62
ā¢
3
mkurman/Qwen2.5-14B-DeepSeek-R1-1M
Text Generation
ā¢
Updated
ā¢
2.06k
ā¢
52
mkurman/phi4-MedIT-10B-o1
Text Generation
ā¢
Updated
ā¢
92
ā¢
4
mkurman/llama-3.2-MEDIT-3B-o1
Text Generation
ā¢
Updated
ā¢
57
ā¢
12