diff --git a/data/envs/metaworld/generate_dataset.py b/data/envs/metaworld/generate_dataset.py
index dc51b4f..ed99a4a 100644
--- a/data/envs/metaworld/generate_dataset.py
+++ b/data/envs/metaworld/generate_dataset.py
@@ -4,7 +4,9 @@ from typing import Dict, Optional
 import gym
 import metaworld
 import numpy as np
+import pandas as pd
 import torch
+from datasets import Dataset
 from huggingface_hub import HfApi, repocard, upload_folder
 from sample_factory.algo.learning.learner import Learner
 from sample_factory.algo.sampling.batched_sampling import preprocess_actions
@@ -12,11 +14,7 @@ from sample_factory.algo.utils.action_distributions import argmax_actions
 from sample_factory.algo.utils.env_info import extract_env_info
 from sample_factory.algo.utils.make_env import make_env_func_batched
 from sample_factory.algo.utils.rl_utils import make_dones, prepare_and_normalize_obs
-from sample_factory.cfg.arguments import (
-    load_from_checkpoint,
-    parse_full_cfg,
-    parse_sf_args,
-)
+from sample_factory.cfg.arguments import load_from_checkpoint, parse_full_cfg, parse_sf_args
 from sample_factory.envs.env_utils import register_env
 from sample_factory.model.actor_critic import create_actor_critic
 from sample_factory.model.model_utils import get_rnn_size
@@ -165,10 +163,8 @@ def create_dataset(cfg: Config):
     # Create dataset
     dataset_size = 100_000
     dataset = {
-        "observations": np.zeros(
-            (dataset_size, *env.observation_space["obs"].shape), dtype=env.observation_space["obs"].dtype
-        ),
-        "actions": np.zeros((dataset_size, *env.action_space.shape), env.action_space.dtype),
+        "observations": np.zeros((dataset_size, *env.observation_space["obs"].shape), dtype=np.float32),
+        "actions": np.zeros((dataset_size, *env.action_space.shape), np.float32),
         "dones": np.zeros((dataset_size,), bool),
         "rewards": np.zeros((dataset_size,), np.float32),
     }
@@ -206,6 +202,13 @@ def create_dataset(cfg: Config):
 
     env.close()
 
+    # Convert dict of numpy array to pandas dataframe
+#    dataset = Dataset.from_dict(dataset)
+#    dataset.create_config_id
+    # Set the card of the dataset
+#    dataset.card = f""""""
+#    dataset.push_to_hub("qgallouedec/prj_gia_dataset_metaworld_assembly_v2_1111_demo")
+
     # Save dataset
     repo_path = f"{cfg.train_dir}/datasets/{cfg.experiment}"
     os.makedirs(repo_path, exist_ok=True)
diff --git a/data/envs/metaworld/generate_dataset_all.sh b/data/envs/metaworld/generate_dataset_all.sh
index 802bf5c..3cc4f97 100755
--- a/data/envs/metaworld/generate_dataset_all.sh
+++ b/data/envs/metaworld/generate_dataset_all.sh
@@ -1,34 +1,6 @@
 #!/bin/bash
 
 ENVS=(
-    assembly
-    basketball
-    bin-picking
-    box-close
-    button-press-topdown
-    button-press-topdown-wall
-    button-press
-    button-press-wall
-    coffee-button
-    coffee-pull
-    coffee-push
-    dial-turn
-    disassemble
-    door-close
-    door-lock
-    door-open
-    door-unlock
-    drawer-close
-    drawer-open
-    faucet-close
-    faucet-open
-    hammer
-    hand-insert
-    handle-press-side
-    handle-press
-    handle-pull-side
-    handle-pull
-    lever-pull
     peg-insert-side
     peg-unplug-side
     pick-out-of-hole
@@ -40,19 +12,8 @@ ENVS=(
     plate-slide
     push-back
     push
-    push-wall
-    reach
-    reach-wall
-    shelf-place
-    soccer
-    stick-pull
-    stick-push
-    sweep-into
-    sweep
-    window-close
-    window-open
 )
 
 for ENV in "${ENVS[@]}"; do
-    python generate_dataset.py --env $ENV-v2 --experiment $ENV-v2 --train_dir=./train_dir 
+    python generate_dataset.py --env $ENV-v2 --experiment $ENV-v2 --train_dir=./train_dir
 done