Hugging Face's logo

Spaces:
Sellopale
/
EdgeTAM
Runtime error

App Files Community
EdgeTAM

Commit History

  1. increase GPU duration

    chongzhou
    88a526c

  2. move cached feature to CUDA

    chongzhou
    45c2c68

  3. doublecheck inference_state["device"]

    chongzhou
    7209747

  4. quick fix in init_state

    chongzhou
    7563789

  5. set device in inference_state manually

    chongzhou
    917decd

  6. make predictor global and remove bf16

    chongzhou
    282a45a

  7. autocast

    chongzhou
    db1d5c6

  8. rollback sam2_base and sam2_video_predictor

    chongzhou
    3ccde9c

  9. move segment_with_points to CPU

    chongzhou
    cf4b18a

  1. carefully move between CPU and GPU

    chongzhou
    238c545

  2. upload examples

    chongzhou
    129e201

  3. trim examples

    chongzhou
    0c484bb

  4. offload_state_to_cpu

    chongzhou
    3e65e00

  1. change to decorator

    chongzhou
    4277d6f

  1. remove demo.queue()

    chongzhou
    7af8abe

  2. build_ext

    chongzhou
    87d758c

  3. remove torch.cuda.is_available

    chongzhou
    3cc85d3

  4. fix inference_state["device"]

    chongzhou
    4d4080b

  5. remove predictor from state

    chongzhou
    defc201

  6. state init

    chongzhou
    93472c9

  1. move inference_state to gr.state

    chongzhou
    dd6a79c

  1. session_hash

    chongzhou
    1f52c1d

  2. get session_id

    chongzhou
    b8b46fe

  3. typo

    chongzhou
    c306abe

  4. make gr.State individual

    chongzhou
    3da2a0c

  5. move inference_states out of gr.State

    chongzhou
    0a7fba1

  6. move model to cpu when not using ZeroGPU

    chongzhou
    113b7b2

  1. Initial commit

    chongzhou
    9bc4638

  2. initial commit

    meta-bot
    Verified signature
    e80adb9