inference · ollama
Q4_K_M
- Minimum accelerator memory
- 8GB
- Minimum system memory
- 16GB
- Minimum storage
- 10GB
Qwen3 8B
- Availability
- Available
- License
- open-weights
- Parameters
- 8B
- Context
- 131.1K
- Last verified
Verified artifacts
- qwen3:8bQ4_K_M · 5.2 GBContext41KArtifact updatedollama · Ollama Qwen3 library
Runtime support
- ollamaVersion not recordedSupported OSmacos · windows · linuxProfile statussupportedOllama Qwen3 library
- llama-cppVersion not recordedSupported OSmacos · linuxProfile statusexperimentalllama.cpp repository
Execution profile
lora-finetune · llama-cpp
Q8 adapter training
- Minimum accelerator memory
- 16GB
- Minimum system memory
- 32GB
- Minimum storage
- 80GB
full-finetune · vllm
BF16
- Minimum accelerator memory
- 48GB
- Minimum system memory
- 128GB
- Minimum storage
- 250GB
inference · llama-cpp
unconfirmed
- Minimum accelerator memory
- 16GB
- Minimum system memory
- 32GB
- Minimum storage
- 20GB