The MTEB(por) benchmark for Brazilian Portuguese (leaderboard + task datasets) and the Colibri embedding model that tops it among small models.
Tardelli Stekel
tardellirs
AI & ML interests
None yet
Recent Activity
updated a model about 11 hours ago
tardellirs/qwen3.5-2b-caa-mix updated a model about 13 hours ago
tardellirs/qwen3.5-0.8b-caa-mix published a model about 14 hours ago
tardellirs/qwen3.5-2b-caa-mixOrganizations
models 19
tardellirs/qwen3.5-2b-caa-mix
Text Generation • 2B • Updated
tardellirs/qwen3.5-0.8b-caa-mix
Text Generation • 0.8B • Updated
tardellirs/qwen3-0.6b-caa-mix
Text Generation • 0.6B • Updated • 34
tardellirs/gemma-3-270m-caa-mix
Text Generation • 0.3B • Updated • 23
tardellirs/gemma-3-1b-caa-mix
Text Generation • 1.0B • Updated • 40
tardellirs/gemma-3-1b-caa-ptbr
Text Generation • 1.0B • Updated • 69
tardellirs/Maestro-Retriever-Llama-3.1-8B
Feature Extraction • Updated • 19 • 1
tardellirs/colibri-embed-ptbr
Sentence Similarity • 0.2B • Updated • 3.8k • 103
tardellirs/embeddinggemma-pt-br-128k
Sentence Similarity • 0.2B • Updated • 81 • 1
tardellirs/embeddinggemma-pt-br-16k
Sentence Similarity • 0.1B • Updated • 80 • 1