Template updates

Rho FR3 Duo

rho-fr3-duo · policy template by Microsoft · Franka FR3 Duo · finetune first · 30 Hz. Each version below is fixed once published, newest first.

seq init rho-fr3-duo gives the stable version, seq init rho-fr3-duo@beta the beta, and seq init rho-fr3-duo@v<N> any version below that is not withdrawn. A deployment made from this template names the version it runs in seq policy versions and on its page in the console; seq policy upgrade moves it to the newer stable one.

Performance

No runs yet.

Versions

  1. v4stablebeta

    As v3; policy.py says, as the README does, how a worker's CPU and memory are billed.

    Needs sequence-ai 0.56.0 or later to deploy it (pip install -U sequence-ai); its client runs on 0.56.0 or later with its video encoder (pip install -U "sequence-ai[observe]").

    • v3's files; only policy.py's comment on CPU and memory changed: the machine its card sets (an L4's 8 cores and 32 GiB), or what it declares where that is more.
    Use v4
    rho-fr3-duo v4shell
    # start a policy from this version, then deploy it
    # my_policy_user: the name you choose
    seq init rho-fr3-duo@v4 my_policy_user
    seq deploy my_policy_user
    
    # or move a deployment made from this template to it
    seq policy upgrade my_policy_user --to rho-fr3-duo@v4
    
    # on the robot's machine: this version's client.py and its README
    pip install -U sequence-ai
    seq init rho-fr3-duo@v4 my_policy_user

    In the console, Deploy deploys this version: the console deploys the stable one.

  2. v3

    As v2; the README says a worker's CPU and memory are billed as the machine its card sets.

    Needs sequence-ai 0.56.0 or later to deploy it (pip install -U sequence-ai); its client runs on 0.56.0 or later with its video encoder (pip install -U "sequence-ai[observe]").

    • v2's files; only the README's billing of CPU and memory changed: the machine its card sets (an L4's 8 cores and 32 GiB), or what it declares where that is more.
    Use v3
    rho-fr3-duo v3shell
    # start a policy from this version, then deploy it
    # my_policy_user: the name you choose
    seq init rho-fr3-duo@v3 my_policy_user
    seq deploy my_policy_user
    
    # or move a deployment made from this template to it
    seq policy upgrade my_policy_user --to rho-fr3-duo@v3
    
    # on the robot's machine: this version's client.py and its README
    pip install -U sequence-ai
    seq init rho-fr3-duo@v3 my_policy_user
  3. v2

    Microsoft's Rho for the Franka FR3 Duo as before; the README names the frame both poses are in, as Rho's report does.

    Needs sequence-ai 0.56.0 or later to deploy it (pip install -U sequence-ai); its client runs on 0.56.0 or later with its video encoder (pip install -U "sequence-ai[observe]").

    • Both poses are in a common world frame centered on the dual-arm mount; which arm comes first is still not stated.
    • The weights are the head of the checkpoint's pull request 1, which Microsoft merged into main on 2026-11-06.
    Use v2
    rho-fr3-duo v2shell
    # start a policy from this version, then deploy it
    # my_policy_user: the name you choose
    seq init rho-fr3-duo@v2 my_policy_user
    seq deploy my_policy_user
    
    # or move a deployment made from this template to it
    seq policy upgrade my_policy_user --to rho-fr3-duo@v2
    
    # on the robot's machine: this version's client.py and its README
    pip install -U sequence-ai
    seq init rho-fr3-duo@v2 my_policy_user
  4. v1

    Microsoft's Rho for the Franka FR3 Duo on an L4: a starting point to finetune on your own task.

    Needs sequence-ai 0.56.0 or later to deploy it (pip install -U sequence-ai); its client runs on 0.56.0 or later with its video encoder (pip install -U "sequence-ai[observe]").

    Use v1
    rho-fr3-duo v1shell
    # start a policy from this version, then deploy it
    # my_policy_user: the name you choose
    seq init rho-fr3-duo@v1 my_policy_user
    seq deploy my_policy_user
    
    # or move a deployment made from this template to it
    seq policy upgrade my_policy_user --to rho-fr3-duo@v1
    
    # on the robot's machine: this version's client.py and its README
    pip install -U sequence-ai
    seq init rho-fr3-duo@v1 my_policy_user