fivetech/Harbour

Upload Qwen3.6-35B-A3B-LoRA-Q4_K_M.gguf with huggingface_hub

Update training hardware details: DGX Spark, 7h49min

Add GitHub repo link to README

v2: add chat_template.jinja

v2: add tokenizer_config.json

v2: add tokenizer.json

v2: add LoRA adapter (1.8GB)

v2: add adapter config

v2: Harbour/FWH Coder - Qwen3.5-35B-A3B LoRA (5004 examples, eval_loss=0.479)

Upload test_battery_continue.py with huggingface_hub

Upload test_battery_100.py with huggingface_hub

Upload test_battery.py with huggingface_hub

Upload train_unsloth.py with huggingface_hub

Upload README.md with huggingface_hub

Upload train.py with huggingface_hub

Upload generate_dataset.py with huggingface_hub

Upload dataset_stats.json with huggingface_hub

Upload harbour_dataset_full.jsonl with huggingface_hub

Upload harbour_val.jsonl with huggingface_hub

Upload harbour_train.jsonl with huggingface_hub

initial commit