Piccolo 27B VL
Piccolo 27B VL is a vision-language model checkpoint with the Qwen3_5ForConditionalGeneration architecture and a vision configuration.
This release contains Safetensors weights: 12 main shards and model-frozen-mtp.safetensors. All 13 weight files are referenced by model.safetensors.index.json. The original model, generation, tokenizer and processor configurations and chat template are included.
See release-manifest.json for file sizes and SHA-256 checksums. This release does not include benchmark results or a validated inference script.
- Downloads last month
- 31
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support