jinaai/jina-embeddings-v5-text-small
Feature Extraction • 0.6B • Updated • 293k • 189
lxww301
lxww301
AI & ML interests
None yet
Recent Activity
liked a model about 21 hours ago
jinaai/jina-embeddings-v5-text-small liked a model about 2 months ago
ContextualAI/ctxl-rerank-v2-instruct-multilingual-6b liked a model 3 months ago
LilaRest/gemma-4-31B-it-NVFP4-turboOrganizations
liked a
model about 21 hours ago
liked a
model about 2 months ago
liked a
model 3 months ago
vllm compatibility
1
#1 opened 3 months ago
by lxww301
Self-editing context for long-horizon retrieval: practical deployment questions
5
#4 opened 3 months ago
by 
O96a
upvoted a paper 3 months ago
liked 2
models 3 months ago
liked a
dataset 11 months ago
liked a
model 11 months ago
commented a
paper about 1 year ago
reacted to codelion's post with 🔥 about 1 year ago
Post
2429
🚀 Just published: "OpenEvolve: Open-Source Evolutionary Code Optimization with Real-World GPU Kernel Discovery"
We built the first open-source implementation of Google's AlphaEvolve system and used it to automatically discover GPU kernel optimizations that outperform human engineers!
Key results:
- 21.8% average decode speed improvement on Apple Silicon
- 36.7% improvement on long-context transformer attention
- Discovered novel vectorization patterns and 2-pass softmax algorithm
The system evolved a Metal kernel for Qwen3's Grouped Query Attention from a basic 3-pass implementation into something with sophisticated Apple Silicon optimizations that would take experts months to discover manually. The evolved kernel automatically found the optimal vec<T,8> operations for 128-dim attention heads and fused softmax computation with value accumulation.
Really excited about the potential here - imagine evolutionary algorithms automatically discovering optimizations across all our AI infrastructure. What would you want to optimize with this approach?
Full write-up: https://huggingface.co/blog/codelion/openevolve-gpu-kernel-discovery
GitHub: https://github.com/codelion/openevolve
#AI #MachineLearning #GPU #OpenSource #Evolution #CodeOptimization #TransformerOptimization
We built the first open-source implementation of Google's AlphaEvolve system and used it to automatically discover GPU kernel optimizations that outperform human engineers!
Key results:
- 21.8% average decode speed improvement on Apple Silicon
- 36.7% improvement on long-context transformer attention
- Discovered novel vectorization patterns and 2-pass softmax algorithm
The system evolved a Metal kernel for Qwen3's Grouped Query Attention from a basic 3-pass implementation into something with sophisticated Apple Silicon optimizations that would take experts months to discover manually. The evolved kernel automatically found the optimal vec<T,8> operations for 128-dim attention heads and fused softmax computation with value accumulation.
Really excited about the potential here - imagine evolutionary algorithms automatically discovering optimizations across all our AI infrastructure. What would you want to optimize with this approach?
Full write-up: https://huggingface.co/blog/codelion/openevolve-gpu-kernel-discovery
GitHub: https://github.com/codelion/openevolve
#AI #MachineLearning #GPU #OpenSource #Evolution #CodeOptimization #TransformerOptimization
- 1 reply
liked a
model about 1 year ago
liked 6
models over 1 year ago
RekaAI/reka-flash-3
21B • Updated • 214 • 391
google/gemma-3-27b-it
Image-Text-to-Text • 27B • Updated • 869k • • 1.99k
moonshotai/Moonlight-16B-A3B-Instruct
Text Generation • 16B • Updated • 38.9k • 195
QuixiAI/DeepSeek-V3-AWQ
Text Generation • 671B • Updated • 411 • 35
xwen-team/Xwen-7B-Chat
Text Generation • 8B • Updated • 50 • • 32
mistralai/Mistral-Small-24B-Instruct-2501
24B • Updated • 65.4k • 960
liked a
dataset over 1 year ago