Remove EMA model from Diffusion Policy (#134)

这个提交包含在:
Alexander Soare
2024-05-05 11:26:12 +01:00
提交者 GitHub
父节点 d747195c57
当前提交 f3bba0270d
共有 11 个文件被更改,包括 21 次插入 和 117 次删除

查看文件

@ -1,5 +1,9 @@
# @package _global_
# Defaults for training for the PushT dataset as per https://github.com/real-stanford/diffusion_policy.
# Note: We do not track EMA model weights as we discovered it does not improve the results. See
# https://github.com/huggingface/lerobot/pull/134 for more details.
seed: 100000
dataset_repo_id: lerobot/pusht
@ -91,12 +95,3 @@ policy:
# Inference
num_inference_steps: 100
# ---
# TODO(alexander-soare): Remove these from the policy config.
use_ema: true
ema_update_after_step: 0
ema_min_alpha: 0.0
ema_max_alpha: 0.9999
ema_inv_gamma: 1.0
ema_power: 0.75