Gianloko/apex-coder-1.5b-GGUF

ApexCoder-1.5B · GGUF Q4_K_M

Last updated: 2026-03-20 — Cycle 2

Quantized GGUF (Q4_K_M, 986 MB) for local inference with Ollama or llama.cpp. This is the recommended format for running ApexCoder locally on any hardware.

📊 Evaluation — Cycle 2

Metric	Value
LLM-as-judge (avg)	12.6/15
Perplexity	1.14
Δ vs previous cycle	+12.6
File size	986 MB
Quantization	Q4_K_M (4-bit, good quality/size balance)

By reasoning type

Type	Status	Score	Progress

Cycle history

Cycle	Date	Score	PPL	Δ	vs Published
1	2026-03-20	12.9/15	1.17	+12.9	12.9
2	2026-03-20	12.6/15	1.14	+12.6	13.2

🦙 Ollama — recommended

# Pull and run
ollama pull hf.co/Gianloko/apex-coder-1.5b-GGUF:Q4_K_M
ollama run hf.co/Gianloko/apex-coder-1.5b-GGUF:Q4_K_M

# Or use in a Modelfile for a custom system prompt
cat > Modelfile << 'EOF'
FROM hf.co/Gianloko/apex-coder-1.5b-GGUF:Q4_K_M
SYSTEM "You are ApexCoder, a world-class Salesforce platform expert specializing in Apex, LWC, SOQL, and all Salesforce coded artifacts. You write clean, production-ready, governor-limit-aware code following Salesforce best practices."
PARAMETER temperature 0.1
PARAMETER num_predict 1024
EOF
ollama create apex-coder -f Modelfile
ollama run apex-coder "Write a bulkified Apex trigger on Opportunity..."

🔧 llama.cpp

# Download the GGUF file
wget https://huggingface.co/Gianloko/apex-coder-1.5b-GGUF/resolve/main/apex-coder-1.5b-Q4_K_M.gguf

# Run with llama.cpp
./llama-cli -m apex-coder-1.5b-Q4_K_M.gguf     --system-prompt "You are ApexCoder, a Salesforce expert."     -p "Write a bulkified Apex trigger..."     -n 512 --temp 0.1

Related repos

• 🔗 Merged model — full 16-bit model for transformers
• 🔌 LoRA adapter — lightweight adapter (~150 MB)
• 📦 Training dataset

License

Apache 2.0