Add collaborativeagents module and update gitignore

- Add collaborativeagents subproject with adapters, agents, and evaluation modules
- Update .gitignore to exclude large binary files (.whl, .tar), wandb logs, and results

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

1 files changed, 30 insertions, 0 deletions

diff --git a/collaborativeagents/training/train_sft_lora.sbatch b/collaborativeagents/training/train_sft_lora.sbatch
new file mode 100644
index 0000000..9781dd0
--- /dev/null
+++ b/collaborativeagents/training/train_sft_lora.sbatch
@@ -0,0 +1,30 @@
+#!/bin/bash
+#SBATCH --job-name=sft_lora
+#SBATCH --output=/projects/bfqt/users/yurenh2/ml-projects/personalization-user-model/sft_lora_%j.out
+#SBATCH --error=/projects/bfqt/users/yurenh2/ml-projects/personalization-user-model/sft_lora_%j.err
+#SBATCH --time=12:00:00
+#SBATCH --account=bfqt-delta-gpu
+#SBATCH --partition=gpuA100x4
+#SBATCH --gres=gpu:1
+#SBATCH --cpus-per-task=8
+#SBATCH --mem=64G
+
+# LoRA SFT Training - Minimal Hardware (1 GPU)
+# Expected: ~2-4 hours for 4 epochs on single H200
+
+cd /projects/bfqt/users/yurenh2/ml-projects/personalization-user-model/collaborativeagents/training
+
+source /u/yurenh2/miniforge3/etc/profile.d/conda.sh
+conda activate eval
+
+echo "=== Starting LoRA SFT Training ==="
+echo "GPU: $(nvidia-smi --query-gpu=name --format=csv,noheader | head -1)"
+echo "Config: llama_factory_lora_config.yaml"
+echo "Start time: $(date)"
+
+# Run training with LLaMA-Factory
+llamafactory-cli train llama_factory_lora_config.yaml
+
+echo "=== Training Complete ==="
+echo "End time: $(date)"
+echo "Output: outputs/sft_reflection_lora"