MathIR parser LoRA v4 (E6)

Continue of AzJoeLesz/mathir-parser-qwen25-7b-lora-v2 for 100 steps (lr 5e-5) on mathir-synth-hu revision v4. v3 is deprecated (1/23 gold, status dropped). Eval with the same system prompt as training/mathir_sft/hf_job_eval_lora_v4.py.

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support