from modal import App, Image, Volume
import modal
import json
model_name = "Learnable"
app = App(f"Train Model {model_name} with Text8")
# Build image with all local dependencies added directly
image = (
    Image.from_registry("nvidia/cuda:12.4.0-devel-ubuntu22.04",add_python="3.10")  # <--- This is the required 'tag'
    .pip_install(["torch", "flax", "pandas", "tqdm","optax", "dataclasses", "argparse","matplotlib", 
                  "scikit-learn","wandb","timm","torchvision","datasets","transformers","timm"]) 
    .run_commands("""pip install --upgrade "jax[cuda]" -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html""")
    .add_local_file("pyproject.toml", "/root/pyproject.toml")
    .add_local_dir("src", "/root/src")
)
# Shared volume for saving outputs or checkpoints
volume1 = Volume.from_name("weights", create_if_missing=True)
volume2 = Volume.from_name("datasets", create_if_missing=True)
volume3 = Volume.from_name("plots", create_if_missing=True)
volume4 = Volume.from_name("results", create_if_missing=True)
@app.function(
    image=image,
    gpu="H100:1",
    timeout=3600 * 24,
    volumes={"/root/weights/":volume1,"/root/datasets/":volume2,"/root/plots/":volume3,"/root/results/":volume4},
)
def run_command():
    import os
    os.system("pip install -e .")
    os.environ["NVIDIA_TF32_OVERRIDE"] = "0"
    os.environ["JAX_DEFAULT_MATMUL_PRECISION"] = "highest" 
    os.environ["CUBLAS_WORKSPACE_CONFIG"] = ":4096:8"
    os.system("""
    WANDB_MODE=online CUDA_VISIBLE_DEVICES=0  python src/lgmodeling/finetune.py \
        --model-path /root/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0/best_60000 \
        --seed 0 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 8  --lmc-layer-indices 0 1 2 3 4 5 6 7 8 9 10 11\
        --learning-rate 0.00025 --batch-size 24 --max_step 60000 --warmup_step 0 --dataset text8 \
        --wandb-project LMC-Attention --wandb-group "GPT2-Text8-FFN" --wandb-entity "vinh-bui0512-hcmut"\
        --model-save-dir /root/weights/text8/gpt2-finetune --data-path /root/datasets/text8
    """)
    os.system("""
    WANDB_MODE=online CUDA_VISIBLE_DEVICES=0  python src/lgmodeling/finetune.py \
        --model-path /root/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0/best_60000 \
        --seed 20 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 8  --lmc-layer-indices 0 1 2 3 4 5 6 7 8 9 10 11\
        --learning-rate 0.00025 --batch-size 24 --max_step 60000 --warmup_step 0 --dataset text8 \
        --wandb-project LMC-Attention --wandb-group "GPT2-Text8-FFN" --wandb-entity "vinh-bui0512-hcmut"\
        --model-save-dir /root/weights/text8/gpt2-finetune --data-path /root/datasets/text8
    """)
    os.system("""
    WANDB_MODE=online CUDA_VISIBLE_DEVICES=0  python src/lgmodeling/finetune.py \
        --model-path /root/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0/best_60000 \
        --seed 40 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 8  --lmc-layer-indices 0 1 2 3 4 5 6 7 8 9 10 11\
        --learning-rate 0.00025 --batch-size 24 --max_step 60000 --warmup_step 0 --dataset text8 \
        --wandb-project LMC-Attention --wandb-group "GPT2-Text8-FFN" --wandb-entity "vinh-bui0512-hcmut"\
        --model-save-dir /root/weights/text8/gpt2-finetune --data-path /root/datasets/text8
    """)
    os.system("""
    WANDB_MODE=online CUDA_VISIBLE_DEVICES=0  python src/lgmodeling/finetune.py \
        --model-path /root/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0/best_60000 \
        --seed 0 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 16  --lmc-layer-indices 0 1 2 3 4 5 6 7 8 9 10 11\
        --learning-rate 0.00025 --batch-size 24 --max_step 60000 --warmup_step 0 --dataset text8 \
        --wandb-project LMC-Attention --wandb-group "GPT2-Text8-FFN" --wandb-entity "vinh-bui0512-hcmut"\
        --model-save-dir /root/weights/text8/gpt2-finetune --data-path /root/datasets/text8
    """)
    os.system("""
    WANDB_MODE=online CUDA_VISIBLE_DEVICES=0  python src/lgmodeling/finetune.py \
        --model-path /root/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0/best_60000 \
        --seed 20 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 16  --lmc-layer-indices 0 1 2 3 4 5 6 7 8 9 10 11\
        --learning-rate 0.00025 --batch-size 24 --max_step 60000 --warmup_step 0 --dataset text8 \
        --wandb-project LMC-Attention --wandb-group "GPT2-Text8-FFN" --wandb-entity "vinh-bui0512-hcmut"\
        --model-save-dir /root/weights/text8/gpt2-finetune --data-path /root/datasets/text8
    """)
    os.system("""
    WANDB_MODE=online CUDA_VISIBLE_DEVICES=0  python src/lgmodeling/finetune.py \
        --model-path /root/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0/best_60000 \
        --seed 40 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 16  --lmc-layer-indices 0 1 2 3 4 5 6 7 8 9 10 11\
        --learning-rate 0.00025 --batch-size 24 --max_step 60000 --warmup_step 0 --dataset text8 \
        --wandb-project LMC-Attention --wandb-group "GPT2-Text8-FFN" --wandb-entity "vinh-bui0512-hcmut"\
        --model-save-dir /root/weights/text8/gpt2-finetune --data-path /root/datasets/text8
    """)
    os.system("""
    WANDB_MODE=online CUDA_VISIBLE_DEVICES=0  python src/lgmodeling/finetune.py \
        --model-path /root/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0/best_60000 \
        --seed 0 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 4  --lmc-layer-indices 0 1 2 3 4 5 6 7 8 9 10 11\
        --learning-rate 0.00025 --batch-size 24 --max_step 60000 --warmup_step 0 --dataset text8 \
        --wandb-project LMC-Attention --wandb-group "GPT2-Text8-FFN" --wandb-entity "vinh-bui0512-hcmut"\
        --model-save-dir /root/weights/text8/gpt2-finetune --data-path /root/datasets/text8
    """)
    os.system("""
    WANDB_MODE=online CUDA_VISIBLE_DEVICES=0  python src/lgmodeling/finetune.py \
        --model-path /root/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0/best_60000 \
        --seed 20 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 4  --lmc-layer-indices 0 1 2 3 4 5 6 7 8 9 10 11\
        --learning-rate 0.00025 --batch-size 24 --max_step 60000 --warmup_step 0 --dataset text8 \
        --wandb-project LMC-Attention --wandb-group "GPT2-Text8-FFN" --wandb-entity "vinh-bui0512-hcmut"\
        --model-save-dir /root/weights/text8/gpt2-finetune --data-path /root/datasets/text8
    """)
    os.system("""
    WANDB_MODE=online CUDA_VISIBLE_DEVICES=0  python src/lgmodeling/finetune.py \
        --model-path /root/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0/best_60000 \
        --seed 40 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 4  --lmc-layer-indices 0 1 2 3 4 5 6 7 8 9 10 11\
        --learning-rate 0.00025 --batch-size 24 --max_step 60000 --warmup_step 0 --dataset text8 \
        --wandb-project LMC-Attention --wandb-group "GPT2-Text8-FFN" --wandb-entity "vinh-bui0512-hcmut"\
        --model-save-dir /root/weights/text8/gpt2-finetune --data-path /root/datasets/text8
    """)

if __name__ == "__main__":
    with app.run():
        run_command.remote()