WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
    --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
    --seed 0 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 4 --lmc-layer-indices 0 \
    --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8
WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
    --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
    --seed 20 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 4 --lmc-layer-indices 0 \
    --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8
WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
    --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
    --seed 40 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 4 --lmc-layer-indices 0 \
    --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8
WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
    --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
    --seed 0 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 8 --lmc-layer-indices 0\
    --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8
WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
    --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
    --seed 20 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 8 --lmc-layer-indices 0 \
    --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8
WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
    --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
    --seed 40 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 8 --lmc-layer-indices 0 \
    --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8

WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
    --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
    --seed 0 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 16 --lmc-layer-indices 0\
    --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8

WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
    --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
    --seed 20 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 16 --lmc-layer-indices 0 \
    --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8

WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
    --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
    --seed 40 --tgt_len 512 --mem_len 512 --eval_tgt_len 128  --n_head 16 --lmc-layer-indices 0 \
    --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8



# WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
#     --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
#     --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8
# WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
#     --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
#     --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8
# WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
#     --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
#     --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8
# WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
#     --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
#     --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8
# WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
#     --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
#     --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8
# WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
#     --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
#     --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8

# WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
#     --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
#     --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8

# WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
#     --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
#     --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8

# WANDB_MODE=online CUDA_VISIBLE_DEVICES=1  python src/lgmodeling/finetune.py \
#     --model-path /mnt/data/vinhbk/weights/text8/lr0.00025-learnable-step60000-warm0-size24-layer12-embd512-heads8-shared1-routed0-topk0 \
#     --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 /mnt/data/vinhbk/weights/text8/gpt2-finetune --data-path /mnt/data/vinhbk/datasets/text8