SiamGPT: Quality-First Fine-Tuning for Stable Thai Text Generation
This addresses the deployment challenge for Thai language users by improving instruction adherence and linguistic stability, though it is incremental as it builds on existing models and fine-tuning methods.
The paper tackled the problem of unstable Thai text generation in open-weights large language models by fine-tuning Qwen3-32B with a Quality-First strategy, resulting in SiamGPT-32B achieving the strongest overall performance on the SEA-HELM benchmark among similar-scale open-weights Thai models.
Open-weights large language models remain difficult to deploy for Thai due to unstable generation under complex instructions, despite strong English performance. To mitigate these limitations, We present SiamGPT-32B, an open-weights model based on Qwen3-32B, fine-tuned with a Quality-First strategy emphasizing curated supervision over data scale. The fine-tuning pipeline combines translated high-complexity English instruction data with a Thai-adapted AutoIF framework for instruction and linguistic constraints. Using supervised fine-tuning only, without continual pretraining or corpus expansion, SiamGPT-32B improves instruction adherence, multi-turn robustness, and linguistic stability. Evaluations on the SEA-HELM benchmark show that SiamGPT-32B achieves the strongest overall performance among similar-scale open-weights Thai models, with consistent gains in instruction following, multi-turn dialogue, and natural language understanding.