File size: 2,525 Bytes
ce5618e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
export http_proxy=http://192.168.16.76:18000 https_proxy=http://192.168.16.76:18000

source /home/susc/.ssh/sichang/openpi/.venv/bin/activate
GIT_LFS_SKIP_SMUDGE=1 uv sync
GIT_LFS_SKIP_SMUDGE=1 uv pip install -e .

uv pip install tensorflow tensorflow_datasets
uv run examples/libero/convert_libero_data_to_lerobot.py --data_dir /home/susc/.ssh/sichang/openpi/modified_libero_rlds --push_to_hub

# pi0_fast_libero_low_mem_finetune
# base model: pi0_fast_base
uv run scripts/compute_norm_stats.py --config-name pi0_fast_libero_low_mem_finetune
XLA_PYTHON_CLIENT_MEM_FRACTION=0.9 uv run scripts/train.py pi0_fast_libero_low_mem_finetune --exp-name=pi0_fast_LoRA_finetune --overwrite
uv run scripts/serve_policy.py policy:checkpoint --policy.config=pi0_fast_libero_low_mem_finetune --policy.dir=checkpoints/pi0_fast_libero_low_mem_finetune/pi0_fast_LoRA_finetune/10000

# pi0_fast_libero
# base model: pi0_fast_base
uv run scripts/compute_norm_stats.py --config-name pi0_fast_libero
XLA_PYTHON_CLIENT_MEM_FRACTION=0.9 uv run scripts/train.py pi0_fast_libero --exp-name=pi0_fast_full_finetune --overwrite
uv run scripts/serve_policy.py policy:checkpoint --policy.config=pi0_fast_libero --policy.dir=checkpoints/pi0_fast_libero/pi0_fast_full_finetune/50000


# pi0_libero_low_mem_finetune
# base model: pi0_base
uv run scripts/compute_norm_stats.py --config-name pi0_libero_low_mem_finetune
XLA_PYTHON_CLIENT_MEM_FRACTION=0.9 uv run scripts/train.py pi0_libero_low_mem_finetune --exp-name=pi0_LoRA_finetune --overwrite
uv run scripts/serve_policy.py policy:checkpoint --policy.config=pi0_libero_low_mem_finetune --policy.dir=checkpoints/pi0_libero_low_mem_finetune/pi0_LoRA_finetune/20000


# pi0_fast_libero_low_mem_finetune_oft
uv run scripts/compute_norm_stats.py --config-name pi0_fast_libero_low_mem_finetune
XLA_PYTHON_CLIENT_MEM_FRACTION=0.9 uv run scripts/train_oft.py pi0_fast_libero_low_mem_finetune --exp-name=pi0_fast_oft_finetune --overwrite



rm -rf /home/susc/.cache/openpi/openpi-assets/checkpoints/pi0_fast_libero
rm -rf /home/susc/.cache/openpi/openpi-assets/checkpoints/pi0_base
rm -rf /home/susc/.cache/openpi/openpi-assets/checkpoints/pi0_base/params/


#Evaluate
#terminal 1

# Run the simulation
python examples/libero/main.py

# To run with glx for Mujoco instead (use this if you have egl errors):
MUJOCO_GL=glx python examples/libero/main.py

#terminal 2 (run this one terminal first)
# Run the server
source /home/susc/.ssh/sichang/openpi/.venv/bin/activate
uv run scripts/serve_policy.py --env LIBERO