update readme

README.md

@@ -0,0 +1,144 @@
+# DynaTraj — Training and Data Collection (SB3 + dm_control)
+
+This repo provides two entry-point scripts:
+- `sb3_train.py`: Train an RL policy with Stable-Baselines3 on dm_control tasks and periodically save policy checkpoints
+- `sb3_collect.py`: Load specific checkpoints and collect fixed-length trajectories into `.npz` datasets with minimal metadata
+
+---
+
+## Environment & Dependencies
+
+- Python 3.10+
+- Required:
+  - dm_control + MuJoCo: `pip install dm-control mujoco`
+  - Reinforcement learning: `pip install stable-baselines3`
+  - Utilities: `pip install numpy tqdm torch`
+  - Gym API: prefers `gymnasium`, falls back to `gym` (handled automatically)
+- Optional (only if you need on-screen rendering during collection):
+  - OpenCV UI: `pip install opencv-python`
+  - or Matplotlib: `pip install matplotlib`
+
+Note: On headless Linux/remote environments, rendering might require EGL/OSMesa configuration. If you do not render, you do not need any display backend.
+
+---
+
+## Output Layout (Conventions)
+
+- Training checkpoints: `weights/<domain>/<task>/ckpt-<k>.pt`
+  - `k` starts from 1; default save interval is every 10,000 timesteps
+  - Examples: `ckpt-1.pt` ≈ 10k steps, `ckpt-2.pt` ≈ 20k steps
+- Full SB3 model: `weights/sb3_<algo>_<domain>-<task>_seed<seed>_<timestamp>.zip`
+- Collected datasets: `dataset/*.npz` with a companion `*_metadata.pkl`
+
+Default absolute paths in this repo:
+- Checkpoint root: `/home/lau/sim/DynaTraj/weights`
+- Dataset output: `/home/lau/sim/DynaTraj/dataset`
+
+---
+
+## Training
+
+Script: `sb3_train.py`
+
+Supported algorithms (argument is case-sensitive and must be lowercase here):
+- `sac`, `ppo`, `td3`
+
+Common args:
+- `--domain`: dm_control domain, e.g., `cheetah`, `quadruped` (default: `cheetah`)
+- `--task`: task name, e.g., `run`, `walk` (default: `run`)
+- `--algo`: `sac|ppo|td3` (default: `sac`)
+- `--total_timesteps`: total training steps (default: `500000`)
+- `--n_envs`: number of parallel envs (default: `1`; uses sub-process vectorization if >1)
+- `--seed`: random seed (default: `0`)
+- `--device`: `cpu|cuda|auto` (default: `auto`)
+- `--out_dir`: where to save models & checkpoints (default: `/home/lau/sim/DynaTraj/weights`)
+
+Example:
+```bash
+python /home/lau/sim/DynaTraj/train_sb3_dmcontrol.py \
+  --domain cheetah --task run\
+  --algo sac \
+  --total_timesteps 500000 \
+  --out_dir /home/lau/sim/DynaTraj/weights
+```
+
+Checkpointing:
+- A policy checkpoint is saved every 10,000 global timesteps as `ckpt-<k>.pt` (1-based counter).
+- A full SB3 model `.zip` is also saved at the end for plain SB3 loading if needed.
+
+---
+
+## Data Collection
+
+Script: `sb3_collect.py`
+
+Purpose: Run inference with specific training checkpoints and write fixed-length trajectories to `.npz` files plus metadata.
+
+Common args:
+- `--domain` / `--task`: must match training (defaults: `cheetah` / `run`)
+- `--ckpt_root`: root directory of checkpoints (default: `/home/lau/sim/DynaTraj/weights`)
+- `--ckpt_indices`: comma-separated checkpoint indices, e.g., `1,5,10` (note: starts from 1)
+- `--trajectories_per_ckpt`: how many trajectories per checkpoint (default: `5120`)
+- `--steps_per_trajectory`: steps per trajectory (default: `24`)
+- `--out_dir`: dataset output directory (default: `/home/lau/sim/DynaTraj/dataset`)
+- `--device`: inference device (default: `cpu`)
+- `--render`: optional flag to render frames (requires OpenCV or Matplotlib)
+- `--algo`: usually unnecessary. The script reads the real algo name from the checkpoint payload. Only used as a fallback (must be UPPERCASE: `SAC|PPO|TD3`).
+
+Example:
+```bash
+
+python /home/lau/sim/DynaTraj/sb3_collect.py \
+--domain cheetah --task run \
+--algo SAC \
+--ckpt_root /home/lau/sim/DynaTraj/weights \
+--ckpt_indices 1,10,20,30,40,50 \
+--trajectories_per_ckpt 5120 \
+--steps_per_trajectory 24 \
+--out_dir /home/lau/sim/DynaTraj/dataset \
+--device cpu
+```
+
+With rendering (optional):
+```bash
+python /home/lau/sim/DynaTraj/sb3_collect.py \
+--domain cheetah --task run \
+--algo SAC \
+--ckpt_root /home/lau/sim/DynaTraj/weights \
+--ckpt_indices 50 \
+--trajectories_per_ckpt 5120 \
+--steps_per_trajectory 24 \
+--out_dir /home/lau/sim/DynaTraj/dataset \
+--device cpu \
+  --render
+```
+
+Notes:
+- The script searches checkpoints under `ckpt_root/<domain>/<task>/` as `ckpt-<k>.pt`. If you see an error for `ckpt-0.pt`, switch to 1-based indices.
+- If your environment has no display backend, simply omit `--render`.
+
+---
+
+## Output Format
+
+For each `k`:
+- Dataset file: `<out_dir>/sb3_<domain>_<task>_ckpt<kNNN>_<timestamp>.npz`
+- Metadata file: `<out_dir>/sb3_<domain>_<task>_ckpt<kNNN>_<timestamp>_metadata.pkl`
+
+Metadata keys include:
+- `domain`, `task`, `algo`, `seed`
+- `ckpt_index` (the `k` you collected)
+- `trajectories_per_ckpt`, `steps_per_trajectory`
+- `total_trajectories`, `total_steps`
+- `render` (whether rendering was enabled)
+
+The exact arrays inside `.npz` are defined by the internal `TrajectoryBuffer` implementation. Typically they include time-aligned observation/state tensors, actions, rewards, and done flags.
+
+---
+
+## Tips
+
+- Prefer absolute paths (all examples above use absolute paths).
+- Training uses `DummyVecEnv`/`SubprocVecEnv`, automatically flattens dm_control observations and clips actions to env bounds.
+- Collection reconstructs an SB3 policy and loads the `policy_state_dict` from each checkpoint; it does not require the full `.zip` model.
+- Algo-arg case: training expects lowercase (`sac|ppo|td3`), collection fallback uses UPPERCASE (`SAC|PPO|TD3`).

README_dmcontrol_collect.md

@@ -1,87 +0,0 @@
-## dm_control data collection (dmcontrol_collect.py)
-
-### Overview
-This script collects trajectories from DeepMind Control (dm_control) environments using uniformly sampled torque actions in [-1, 1]. Data are saved with `TrajectoryBuffer` as compressed `.npz` along with a metadata `.pkl`.
-
-Collected state per step contains (in order):
-- joint angles (radians)
-- joint angular velocities (rad/s)
-- root position (x, y, z)
-- root linear velocity (vx, vy, vz)
-- root rotation quaternion (qx, qy, qz, qw)
-- root angular velocity (wx, wy, wz)
-- last applied torque (action vector)
-
-### Requirements
-- Python 3.9+
-- dm_control and MuJoCo installed:
-  ```bash
-  pip install dm-control mujoco
-  ```
-
-### Hyperparameters (CLI)
-
-| Name | Type / Default | Description |
-|------|-----------------|-------------|
-| `--domain` | str, default `quadruped` | dm_control domain name, e.g. `quadruped`, `cheetah`. |
-| `--task` | str, default `walk` | dm_control task name, e.g. `walk`, `run`. |
-| `--seed` | int, default `0` | PRNG seed used for env and action sampling. |
-| `--trajectories_per_file` | int, default `512` | Number of trajectories to collect and save in one output file. |
-| `--steps_per_trajectory` | int, default `48` | Number of steps per trajectory segment saved to the dataset. |
-| `--out_dir` | str, default `/home/lau/sim/DynaTraj/dataset` | Directory to store output `.npz` and metadata `.pkl`. |
-| `--render` | flag (bool), default `False` | If set, render frames during collection (tries OpenCV, then matplotlib). |
-
-Notes:
-- Actions are sampled i.i.d. uniformly from [-1, 1] each step and treated as torques.
-- If the model uses a free base, the root quaternion is output as `(x, y, z, w)`.
-
-### Output format
-- Dataset file: `dmcontrol_{domain}_{task}_seed{seed}_{timestamp}.npz`
-- Metadata file: `dmcontrol_{domain}_{task}_seed{seed}_{timestamp}_metadata.pkl`
-
-`npz` keys (all stored by `TrajectoryBuffer`):
-- `obs`: shape `[N, B, T, D_obs]`
-- `ext_obs`: shape `[N, B, T, D_obs]` (same content as `obs` in this script)
-- `action`: shape `[N, B, T, D_act]`
-- `reward`: shape `[N, B, T]`
-- `done`: shape `[N, B, T]`
-
-Where:
-- `N` = number of trajectory segments (equals `trajectories_per_file` for `B=1`)
-- `B` = batch size (this script uses `B=1`)
-- `T` = `steps_per_trajectory`
-- `D_obs` = state dimension described above
-- `D_act` = action dimension from the environment action spec
-
-The metadata `.pkl` contains: domain, task, seed, counts, action bounds, timestamp, and `render` flag.
-
-### Examples
-- Quadruped walk (default):
-  ```bash
-  python /home/lau/sim/DynaTraj/dmcontrol_collect.py
-  ```
-- Cheetah run (planar cheetah):
-  ```bash
-  python /home/lau/sim/DynaTraj/dmcontrol_collect.py --domain cheetah --task run --seed 1 --trajectories_per_file 512 --steps_per_trajectory 48 --out_dir /home/lau/sim/DynaTraj/dataset
-  ```
-- With rendering (requires OpenCV or matplotlib):
-  ```bash
-  python /home/lau/sim/DynaTraj/dmcontrol_collect.py --domain quadruped --task walk --render
-  ```
-
-### Tips
-- Rendering slows down collection; disable `--render` when collecting large datasets.
-- If a task terminates early, the script resets automatically and continues until it reaches the requested number of trajectories.
-- Ensure MuJoCo is set up properly in your environment if dm_control fails to import. 
-
-
-python /home/lau/sim/DynaTraj/sb3_collect.py \
-  --domain cheetah --task run \
-  --algo SAC \
-  --ckpt_root /home/lau/sim/DynaTraj/weights \
-  --ckpt_indices 1,10,20,30,40,50 \
-  --trajectories_per_ckpt 5120 \
-  --steps_per_trajectory 24 \
-  --out_dir /home/lau/sim/DynaTraj/dataset \
-  --device cpu \
-  --render