---
license: apache-2.0
base_model: "Qwen/Qwen3-0.6B"
base_model_relation: finetune
pipeline_tag: text-generation
library_name: gguf
tags:
- quantized
- lora
- litert
- gguf
---

# test-tra

translation model

## Model provenance

This model was modified from [Qwen/Qwen3-0.6B](https://huggingface.co/Qwen/Qwen3-0.6B/tree/c1899de289a04d12100db370d81485cdf75e47ca) at revision `c1899de289a04d12100db370d81485cdf75e47ca`.
These files are exports of LoRA fine-tuned, merged models. They share the verified base revision above; different formats may come from different fine-tuning checkpoints.

## Downloads and runtime guidance

### gguf — Q4_K_M

[model-Q4_K_M.gguf](model-Q4_K_M.gguf)

SHA-256: `a348f00a63243e0afc9a91425c07a2ce712f542686e2bbccb33af5306d9186ab`. Size: 396704448 bytes.

Use with a compatible llama.cpp/GGUF runtime. Q4_K_M is a mixed 4-bit quantization. The upstream chat template is preserved; thinking is not forcibly disabled. Disable thinking in the runtime or chat-template options for this model, which was fine-tuned with no-thinking prompts.

### litertlm — dynamic_wi4_afp32

[model.litertlm](model.litertlm)

SHA-256: `c7eca1622b0a3467d4fa94c8c8e6f8b003cf254f8f7be8ab79abe92fd3b30a8d`. Size: 315485328 bytes.

Use with LiteRT-LM. This export uses dynamic_wi4_afp32 weights and runtime metadata with enableThinking=false.

## Limitations

Fine-tuning and quantization can change model behavior. No general quality, safety, or device compatibility claims are made. Evaluate on your own tasks and target runtime before use. Training examples, private datasets, and logs are not included.

## License and notices

The base license is preserved in LICENSE. See NOTICE for upstream attribution and changes. LiteRT metadata licensing is preserved in licenses/LICENSE-LiteRT-LM.