File size: 1,207 Bytes
c881908 |
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 |
---
language: en
license: apache-2.0
model-index:
- name: CartPole-v1__functional_dqn__0__1671474891
---
DQN model applied to the this discrete environments CartPole-v1
## Model Description
The model was trained from the CleanRl library using the DQN algorithm
## Intended Use & Limitation
The model is intended to be used for the following environments CartPole-v1
and understand the implication of Quantization on this type of model from a pretrained state## Training Procdure
### Training Hyperparameters
```
The folloing hyperparameters were used during training:
- exp_name: functional_dqn
- seed: 0
- torch_deterministic: True
- cuda: False
- track: True
- wandb_project_name: cleanRL
- wandb_entity: compress_rl
- capture_video: False
- env_id: CartPole-v1
- total_timesteps: 500000
- learning_rate: 0.00025
- buffer_size: 10000
- gamma: 0.99
- target_network_frequency: 500
- batch_size: 128
- start_e: 1
- end_e: 0.05
- exploration_fraction: 0.5
- learning_starts: 10000
- train_frequency: 10
- optimizer: Adam
- wandb_project: cleanrl
```
### Framework and version
```
Pytorch 1.12.1+cu102
gym 0.23.1
Weights and Biases 0.13.3
Hugging Face Hub 0.11.1
|