Update README.md

README.md

@@ -101,7 +101,7 @@ bash merge_model.sh      # Step 3: Merge FSDP checkpoints
 
 Training data: [MBZUAI/medix-rl-data](https://huggingface.co/datasets/MBZUAI/medix-rl-data) (~51K train, ~2.5K test samples)
 
-See [`training/README.md`](training/README.md) for detailed setup, configuration options, and per-model scripts.
+See [`training/README.md`](https://github.com/mbzuai-oryx/MediX-R1/blob/main/training/README.md) for detailed setup, configuration options, and per-model scripts.
 
 ## Evaluation
 
@@ -115,7 +115,7 @@ bash eval.sh             # Run all phases: generate, evaluate, score
 
 Supports self-hosted judge models via vLLM or [OpenRouter](https://openrouter.ai/) as a remote alternative. Results can be submitted to the [MediX Leaderboard](https://medix.cvmbzuai.com/leaderboard).
 
-See [`eval/README.md`](eval/README.md) for task selection, CLI reference, and MMMU-Medical evaluation.
+See [`eval/README.md`](https://github.com/mbzuai-oryx/MediX-R1/blob/main/eval/README.md) for task selection, CLI reference, and MMMU-Medical evaluation.
 
 ---