Qwen
Qwen 3 4B
4B parameters
License
Apache 2.0
Tune type
LORA
Typical cost
~$10
Cold start
Instant
Best for
- edge
- Apache 2.0 small
License terms
Qwen 3 4B is released under the Apache 2.0 license. Yachay has vetted it for commercial use (allowed).
No naming or attribution constraints on derivatives.
How Yachay tunes it
We default to LORA on Google Cloud Vertex AI Tuning. Yachay auto-selects the right GPU for this model — typically A10G based on the 8 GB FP16 footprint. Hyperparameters are customisable at submit (epochs, batch size, learning rate).
Released April 28, 2025. Weights are pre-mirrored in our GCS catalog — your job starts within seconds.
Related models
Same family first, then nearest size across families.
Ready to fine-tune Qwen 3 4B?
Sign in with your Condor umbrella account — from ~$0.15 / M training tokens on production-size jobs. No upfront bond; card on file, charged at completion.