Skip to content

Commit

Permalink
Merge pull request #17 from Eclectic-Sheep/feature/refactoring
Browse files Browse the repository at this point in the history 
feat: update homepage
  • Loading branch information
@belerico
belerico authored Sep 26, 2023
2 parents 3de3c15 + f3b40c0 commit b46837f
Show file tree
Hide file tree
Showing 3 changed files with 3,143 additions and 4,419 deletions.
119 changes: 66 additions & 53 deletions _includes/introduction.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -40,38 +40,45 @@ <h2 class="custom">Train and deploy Reinforcement Learning agents faster with Sh
<code data-index="0" class="active">
<span class="token-wrapper">
<span class="token bright">
sheeprl ppo
<span style="white-space: nowrap;">--env_id=CartPole-v1</span>
python sheeprl exp=ppo
<span style="white-space: nowrap;">algo=ppo algo.name=ppo env=gym env.id=CartPole-v1</span>
</span>
</span>
</code>

<code data-index="1">
<span class="token-wrapper">
<span class="token bright">lightning run model sheeprl.py</span>
<span class="token">ppo</span>
<span class="token">
python sheeprl.py exp=ppo
<span style="white-space: nowrap;">algo=ppo algo.name=ppo env=gym env.id=CartPole-v1</span>
</span>
<span class="token bright">
<span style="white-space: nowrap;">fabric.devices=2</span>
</span>
</span>
</code>

<code data-index="2">
<span class="token-wrapper">
<span class="token">lightning run model</span>
<span class="token bright">
<span style="white-space: nowrap;">--devices=2</span>
<span style="white-space: nowrap;">--accelerator=gpu</span>
<span class="token">
python sheeprl.py exp=ppo
<span style="white-space: nowrap;">algo=ppo algo.name=ppo env=gym env.id=CartPole-v1</span>
<span style="white-space: nowrap;">fabric.devices=2</span>
</span>
<span class="token">sheeprl.py ppo</span>
<span class="token bright">
<span style="white-space: nowrap;">fabric.accelerator=gpu</span>
</span>
</span>
</code>

<code data-index="3">
<span class="token-wrapper">
<span class="token">lightning run model sheeprl.py</span>
<span class="token">python sheeprl.py</span>
<span class="token bright">
sac
<span style="white-space: nowrap;">--env_id=LunarLanderContinuous-v2</span>
<span style="white-space: nowrap;">--total_steps=2000000</span>
<span style="white-space: nowrap;">--capture_video</span>
exp=sac algo=sac algo.name=sac env=gym
<span style="white-space: nowrap;">env.id=LunarLanderContinuous-v2</span>
<span style="white-space: nowrap;">total_steps=2000000</span>
<span style="white-space: nowrap;">env.capture_video=True</span>
</span>
</span>
</code>
Expand Down Expand Up @@ -126,67 +133,73 @@ <h2 class="custom" style="margin-top: 20px;">Visualize your trained agent behavi
<span>
<code data-index="4" class="active">
<span class="token-wrapper">
<span class="token">lightning run model
<span style="white-space: nowrap;">--devices="2,3"</span>
<span style="white-space: nowrap;">--accelerator=gpu</span>
sheeprl.py ppo_atari
</span>
<span class="token">python sheeprl.py exp=ppo</span>
<span class="token bright">
<span style="white-space: nowrap;">--env_id=PongNoFrameskip-v4</span>
<span style="white-space: nowrap;">fabric.devices=[2,3]</span>
</span>
<span class="token">
<span style="white-space: nowrap;">--anneal_lr</span>
<span style="white-space: nowrap;">--anneal_ent_coef</span>
<span style="white-space: nowrap;">--anneal_clip_coef</span>
<span style="white-space: nowrap;">fabric.accelerator=gpu</span>
</span>
<span class="token bright">
<span style="white-space: nowrap;">cnn_keys.encoder=[rgb]</span>
<span style="white-space: nowrap;">env=atari</span>
<span style="white-space: nowrap;">env.id=PongNoFrameskip-v4</span>
<span style="white-space: nowrap;">algo.optimizer.lr=2.5e-4</span>
</span>
<span class="token">
<span style="white-space: nowrap;">--ent_coef=0.01</span>
<span style="white-space: nowrap;">--clip_coef=0.1</span>
<span style="white-space: nowrap;">--rollout_steps=128</span>
<span style="white-space: nowrap;">--lr=2.5e-4</span>
<span style="white-space: nowrap;">--update_epochs=3</span>
<span style="white-space: nowrap;">algo.anneal_lr=True</span>
<span style="white-space: nowrap;">algo.anneal_ent_coef=True</span>
<span style="white-space: nowrap;">algo.anneal_clip_coef=True</span>
</span>
<span class="token">
<span style="white-space: nowrap;">--num_envs=8</span>
<span style="white-space: nowrap;">--per_rank_batch_size=128</span>
<span style="white-space: nowrap;">--total_steps=40000000</span>
<span style="white-space: nowrap;">algo.ent_coef=0.01</span>
<span style="white-space: nowrap;">algo.clip_coef=0.1</span>
<span style="white-space: nowrap;">algo.rollout_steps=128</span>
<span style="white-space: nowrap;">algo.update_epochs=3</span>
</span>
<span class="token bright">
<span style="white-space: nowrap;">--capture_video</span>
<span style="white-space: nowrap;">env.num_envs=8</span>
<span style="white-space: nowrap;">per_rank_batch_size=128</span>
<span style="white-space: nowrap;">total_steps=40000000</span>
</span>
<span class="token">
<span style="white-space: nowrap;">env.capture_video=True</span>
</span>
</span>
</code>

<code data-index="5">
<span class="token-wrapper">
<span class="token">
lightning run model
<span style="white-space: nowrap;">--devices="2,3"</span>
<span style="white-space: nowrap;">--accelerator=gpu</span>
sheeprl.py ppo_atari
python sheeprl.py exp=ppo
<span style="white-space: nowrap;">fabric.devices=[2,3]</span>
<span style="white-space: nowrap;">fabric.accelerator=gpu</span>

</span>
<span class="token bright">
<span style="white-space: nowrap;">--env_id=BreakoutNoFrameskip-v4</span>
<span style="white-space: nowrap;">env=atari</span>
<span style="white-space: nowrap;">env.id=BreakoutNoFrameskip-v4</span>
<span style="white-space: nowrap;">cnn_keys.encoder=[rgb]</span>
</span>
<span class="token">
<span style="white-space: nowrap;">--anneal_lr</span>
<span style="white-space: nowrap;">--anneal_ent_coef</span>
<span style="white-space: nowrap;">--anneal_clip_coef</span>
<span style="white-space: nowrap;">algo.anneal_lr=True</span>
<span style="white-space: nowrap;">algo.anneal_ent_coef=True</span>
<span style="white-space: nowrap;">algo.anneal_clip_coef=True</span>
</span>
<span class="token">
<span style="white-space: nowrap;">--ent_coef=0.01</span>
<span style="white-space: nowrap;">--clip_coef=0.1</span>
<span style="white-space: nowrap;">--rollout_steps=128</span>
<span style="white-space: nowrap;">--lr=2.5e-4</span>
<span style="white-space: nowrap;">--update_epochs=3</span>
</span>
<span class="token">
<span style="white-space: nowrap;">--num_envs=8</span>
<span style="white-space: nowrap;">--per_rank_batch_size=128</span>
<span style="white-space: nowrap;">--total_steps=40000000</span>
<span style="white-space: nowrap;">algo.ent_coef=0.01</span>
<span style="white-space: nowrap;">algo.clip_coef=0.1</span>
<span style="white-space: nowrap;">algo.rollout_steps=128</span>
<span style="white-space: nowrap;">algo.optimizer.lr=2.5e-4</span>
<span style="white-space: nowrap;">algo.update_epochs=3</span>
</span>
<span class="token bright">
<span style="white-space: nowrap;">--capture_video</span>
<span style="white-space: nowrap;">env.num_envs=8</span>
<span style="white-space: nowrap;">per_rank_batch_size=128</span>
<span style="white-space: nowrap;">total_steps=40000000</span>
</span>
<span class="token">
<span style="white-space: nowrap;">env.capture_video=True</span>
</span>
</span>
</code>
Expand Down
Loading

0 comments on commit b46837f

Please sign in to comment.