Add reproducible training Dockerfile

Dockerfile

@@ -0,0 +1,21 @@
+FROM pytorch/pytorch:2.5.1-cuda12.4-cudnn9-runtime
+
+WORKDIR /workspace/nomen-ai
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    git git-lfs build-essential && \
+    rm -rf /var/lib/apt/lists/*
+
+COPY requirements.txt /workspace/nomen-ai/requirements.txt
+RUN pip install --upgrade pip && pip install -r requirements.txt
+
+COPY . /workspace/nomen-ai
+RUN pip install -e .
+
+ENV TRACKIO_PROJECT=nomen-ai
+ENV TRACKIO_SPACE_ID=krystv/ml-intern-nomen001
+ENV BASE_MODEL=Qwen/Qwen2.5-1.5B-Instruct
+ENV SFT_DATASET=krystv/nomen-ai-sft
+ENV DPO_DATASET=krystv/nomen-ai-dpo
+
+CMD ["bash", "scripts/train_all_colab.sh"]