Uploaded March 2026 | Updated September 2026, 2 weeks ago
Bojan Jakimovski, an ML engineer, took Arcee AI's open-source Trinity Mini model and turned it into a biomedical specialist — extracting drug-protein relationships from scientific papers. No massive team. No million-dollar budget. Just open weights, a clever training technique called RLVR, and a weekend of GPU time.
⭐️⭐️⭐️ More content on Substack at airealist.ai ⭐️⭐️⭐️
In this video, I break down exactly how it works: the Mixture of Experts architecture behind Trinity Mini, why Reinforcement Learning with Verifiable Rewards (RLVR) beats traditional fine-tuning for domain specialization, how the GRPO algorithm (the same one behind DeepSeek R1) trains a model to reason step by step, and how LoRA makes it possible to specialize a 26B-parameter model for under $50.
Whether you're an ML engineer, a researcher, or just curious about where open-source AI is headed, this is a practical, no-hype walkthrough of a pattern you can replicate in your own domain.
Bojan Jakimovski's blog → shekswess.github.io
Bojan's LinkedIn → linkedin.com/in/bojan-jakimovski
*** MODELS
Trinity-Mini-DrugProt-Think (LoRA adapter) → huggingface.co/lokahq/Trinity-Mini-DrugProt-Think
Arcee Trinity Mini (base model) → huggingface.co/arcee-ai/Trinity-Mini
Arcee Trinity Mini Base (pre-SFT) → huggingface.co/arcee-ai/Trinity-Mini-Base
Trinity Mini on OpenRouter (free tier) → openrouter.ai/arcee-ai/trinity-mini:free
Trinity Mini on OpenRouter (paid API) → openrouter.ai/arcee-ai/trinity-mini
*** CODE & CONFIGS
Full training repo (configs, metrics, deployment) → github.com/LokaHQ/Trinity-Mini-DrugProt-Think
12 experiment TOML configs → github.com/LokaHQ/Trinity-Mini-DrugProt-Think/tree/main/experiments/configs/rl
Training metrics CSVs → github.com/LokaHQ/Trinity-Mini-DrugProt-Think/tree/main/data
Deploying on Amazon SageMaker (Loka blog) → medium.com/loka-engineering/deploying-trinity-mini-drugprot-think-on-amazon-sagemaker-ai-9e1c1c430ce9
***DATASETS
DrugProt on Hugging Face (bigbio) → huggingface.co/datasets/bigbio/drugprot
DrugProt Parquet (OpenMed) → huggingface.co/datasets/OpenMed/drugprot-parquet
*** TOOLS & LIBRARIES
Hugging Face Transformers → github.com/huggingface/transformers
PEFT (LoRA & adapters) → github.com/huggingface/peft
TRL (GRPOTrainer) → github.com/huggingface/trl
Prime Intellect (hosted GRPO training) → primeintellect.ai
Prime Intellect Verifiers (RL environments) → github.com/PrimeIntellect-ai/verifiers
Bojan Jakimovski, an ML engineer, took Arcee AI's open-source Trinity Mini model and turned it into a biomedical specialist — extracting drug-protein relationships from scientific papers. No massive team. No million-dollar budget. Just open weights, a clever training technique called RLVR, and a weekend of GPU time.
⭐️⭐️⭐️ More content on Substack at airealist.ai ⭐️⭐️⭐️
In this video, I break down exactly how it works: the Mixture of Experts architecture behind Trinity Mini, why Reinforcement Learning with Verifiable Rewards (RLVR) beats traditional fine-tuning for domain specialization, how the GRPO algorithm (the same one behind DeepSeek R1) trains a model to reason step by step, and how LoRA makes it possible to specialize a 26B-parameter model for under $50.
Whether you're an ML engineer, a researcher, or just curious about where open-source AI is headed, this is a practical, no-hype walkthrough of a pattern you can replicate in your own domain.
Bojan Jakimovski's blog → shekswess.github.io
Bojan's LinkedIn → linkedin.com/in/bojan-jakimovski
*** MODELS
Trinity-Mini-DrugProt-Think (LoRA adapter) → huggingface.co/lokahq/Trinity-Mini-DrugProt-Think
Arcee Trinity Mini (base model) → huggingface.co/arcee-ai/Trinity-Mini
Arcee Trinity Mini Base (pre-SFT) → huggingface.co/arcee-ai/Trinity-Mini-Base
Trinity Mini on OpenRouter (free tier) → openrouter.ai/arcee-ai/trinity-mini:free
Trinity Mini on OpenRouter (paid API) → openrouter.ai/arcee-ai/trinity-mini
*** CODE & CONFIGS
Full training repo (configs, metrics, deployment) → github.com/LokaHQ/Trinity-Mini-DrugProt-Think
12 experiment TOML configs → github.com/LokaHQ/Trinity-Mini-DrugProt-Think/tree/main/experiments/configs/rl
Training metrics CSVs → github.com/LokaHQ/Trinity-Mini-DrugProt-Think/tree/main/data
Deploying on Amazon SageMaker (Loka blog) → medium.com/loka-engineering/deploying-trinity-mini-drugprot-think-on-amazon-sagemaker-ai-9e1c1c430ce9
***DATASETS
DrugProt on Hugging Face (bigbio) → huggingface.co/datasets/bigbio/drugprot
DrugProt Parquet (OpenMed) → huggingface.co/datasets/OpenMed/drugprot-parquet
*** TOOLS & LIBRARIES
Hugging Face Transformers → github.com/huggingface/transformers
PEFT (LoRA & adapters) → github.com/huggingface/peft
TRL (GRPOTrainer) → github.com/huggingface/trl
Prime Intellect (hosted GRPO training) → primeintellect.ai
Prime Intellect Verifiers (RL environments) → github.com/PrimeIntellect-ai/verifiers
