Fully finetune on Q, V (enc, dec, cross) matrices
Lee
Speeeed
·
AI & ML interests
None yet
Recent Activity
upvoted a paper about 1 month ago
Integrating Inductive Biases in Transformers via Distillation for Financial Time Series Forecasting upvoted an article 4 months ago
How to Use Multiple GPUs in Hugging Face Transformers: Device Map vs Tensor Parallelism liked a model 5 months ago
MediaTek-Research/Breeze-Guard-26Organizations
models 61
Speeeed/Gemma-2-2B-hella_attn_random10_portion5_lr5e-5
Text Generation • 3B • Updated • 8
Speeeed/Gemma-2-2B-hella_attn_random10_portion5_lr3e-5
Text Generation • 3B • Updated • 8
Speeeed/Gemma-2-2B-hella_attn_random10_portion5_lr1e-4
Text Generation • 3B • Updated • 9
Speeeed/flan-t5-large-piqa_lora-16
0.8B • Updated • 4
Speeeed/flan-t5-large-boolq_lora-16
0.8B • Updated • 4
Speeeed/flan-t5-large-hella_lora-16
0.8B • Updated • 4
Speeeed/flan-t5-large-agnews_lora-16
0.8B • Updated • 4
Speeeed/flan-t5-large-imdb_lora-16
0.8B • Updated • 4
Speeeed/flan-t5-large-finance_lora-16
0.8B • Updated • 7
Speeeed/Llama-3B-Instruct_dolly_full_qkvo_ft
Updated
datasets 0
None public yet