Code Monkey home page Code Monkey logo

Comments (1)

btaba avatar btaba commented on June 10, 2024

Hi @esraaelelimy , indeed AutoResetWrapper will cache the first_state, but the first_state is sampled with a different rng for each environment. As the number of parallel environments goes up, the diversity of sampled first_states increases. You're right to point out that this is done for performance reasons. To do a reset, you'll have to call reset with a different rng. See the example here:

brax/brax/training/agents/ppo/train.py

Lines 418 to 431 in f9a4d73

for _ in range(max(num_resets_per_eval, 1)):
# optimization
epoch_key, local_key = jax.random.split(local_key)
epoch_keys = jax.random.split(epoch_key, local_devices_to_use)
(training_state, env_state, training_metrics) = (
training_epoch_with_timing(training_state, env_state, epoch_keys)
)
current_step = int(_unpmap(training_state.env_steps))
key_envs = jax.vmap(
lambda x, s: jax.random.split(x[0], s),
in_axes=(0, None))(key_envs, key_envs.shape[1])
# TODO: move extra reset logic to the AutoResetWrapper.
env_state = reset_fn(key_envs) if num_resets_per_eval > 0 else env_state

But we have not done in-depth analysis on some of these hyperparameters (i.e. num_resets_per_eval).

from brax.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.