Djuunaa's picture

Djuunaa

djuna

·

AI & ML interests

None yet

Recent Activity

liked a model about 8 hours ago

OpenResearcher/OpenResearcher-30B-A3B

liked a model about 8 hours ago

Klingspor/StarPO-4B

reacted to danielhanchen's post with 🔥 8 days ago

We collaborated with Hugging Face to enable you to train MoE models 12× faster with 35% less VRAM via our new Triton kernels (no accuracy loss). 🤗 Train gpt-oss locally on 12.8GB VRAM with our free notebooks: https://unsloth.ai/docs/new/faster-moe

View all activity

Organizations

djuna 's Spaces 6

mergekit-gui

Merge and upload models using a YAML config

Gradient Array Slider

Gradient Array Slider (for mergekit)

Replete Qwen2 1.5B Instruct

Llama Pro Dry Run

Calculate duplicated layers for model expansion

Chat Template Tester

Experiment with Chat Template

Generate Mergkit For Upscale LLM

Generate YAML config for model slicing

djuna (Djuunaa)

Djuunaa's picture

Djuunaa

djuna

·

AI & ML interests

None yet

Recent Activity

liked a model about 8 hours ago

OpenResearcher/OpenResearcher-30B-A3B

liked a model about 8 hours ago

Klingspor/StarPO-4B

reacted to danielhanchen's post with 🔥 8 days ago

We collaborated with Hugging Face to enable you to train MoE models 12× faster with 35% less VRAM via our new Triton kernels (no accuracy loss). 🤗 Train gpt-oss locally on 12.8GB VRAM with our free notebooks: https://unsloth.ai/docs/new/faster-moe

View all activity

Organizations

djuna 's Spaces 6

mergekit-gui

Merge and upload models using a YAML config

Gradient Array Slider

Gradient Array Slider (for mergekit)

Replete Qwen2 1.5B Instruct

Llama Pro Dry Run

Calculate duplicated layers for model expansion

Chat Template Tester

Experiment with Chat Template

Generate Mergkit For Upscale LLM

Generate YAML config for model slicing