# Segment Anything Model (SAM) inference **URL:** \ **Category:** Amazon SageMaker\ **Created:** [August 2, 2023, 9:40am UTC](https://discuss.huggingface.co/t/segment-anything-model-sam-inference/49114 "2023-08-02T09:40:47Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![francisurias](https://avatars.discourse-cdn.com/v4/letter/f/b3f665/32.png) [@francisurias](https://discuss.huggingface.co/u/francisurias)\ **Post date:** [August 2, 2023, 9:40am UTC](https://discuss.huggingface.co/t/segment-anything-model-sam-inference/49114/1 "2023-08-02T09:40:47Z")
Hi, I’m working on a project that uses SAM to do image segmentation. I’m trying to deploy the “sam-vit-large” model (found here: [facebook/sam-vit-large · Hugging Face](https://huggingface.co/facebook/sam-vit-large)) on AWS SageMaker using the code given in the deployment section. I have managed to deploy the endpoint on SageMaker, but I’m not sure what the payload/input I need to give the endpoint is; would it be in json format? If so, what would be the properties? Thanks! ---
**Author:** ![harryneal](https://avatars.discourse-cdn.com/v4/letter/h/f9ae1b/32.png) [@harryneal](https://discuss.huggingface.co/u/harryneal)\ **Post date:** [October 16, 2023, 2:43pm UTC](https://discuss.huggingface.co/t/segment-anything-model-sam-inference/49114/2 "2023-10-16T14:43:35Z")
Did you find an answer to this? I would love to know what you found out if so. Thanks! ---
**Author:** ![radames](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/radames/32/28246_2.png) [@radames](https://discuss.huggingface.co/u/radames)\ **Post date:** [October 19, 2023, 4:29am UTC](https://discuss.huggingface.co/t/segment-anything-model-sam-inference/49114/3 "2023-10-19T04:29:50Z")
hi it looks like the HuggingFace sagemaker needs to update the transformers images, currently the latest version support is `4.28.1` and SAM is only integrated on transformers `4.29.0` cc @philschmid for an update. Also my guess is SAM my need a custom inference handler, please look [https://github.com/huggingface/notebooks/blob/main/sagemaker/17\_custom\_inference\_script/sagemaker-notebook.ipynb](https://github.com/huggingface/notebooks/blob/main/sagemaker/17_custom_inference_script/sagemaker-notebook.ipynb) ---
**Author:** ![payamahv](https://avatars.discourse-cdn.com/v4/letter/p/d6d6ee/32.png) [@payamahv](https://discuss.huggingface.co/u/payamahv)\ **Post date:** [October 26, 2023, 6:41pm UTC](https://discuss.huggingface.co/t/segment-anything-model-sam-inference/49114/5 "2023-10-26T18:41:57Z")
I am trying to pass an model to this model like this in sagemaker with open(“/root/aws-sam/02.JPG”, “rb”) as f: data = f.read() #raw\_image = Image.open(data).convert(“RGB”) response = runtime.invoke\_endpoint(EndpointName=ENDPOINT\_NAME, ContentType=‘image/jpeg’, #image/jpeg image/x-image’ Body=data) I am getting this error 2023-10-26T09:42:33.354-07:00 Copy 2023-10-26T16:42:33,265 [INFO] W-9000-facebook\_\_sam-vit-large-stdout com.amazonaws.ml.mms.wlm.WorkerLifeCycle - raise KeyError(key) 2023-10-26T16:42:33,265 [INFO] W-9000-facebook\_\_sam-vit-large-stdout com.amazonaws.ml.mms.wlm.WorkerLifeCycle - raise KeyError(key) AllTraffic/i-0dad3e9d16762a67f 2023-10-26T09:42:33.354-07:00 Copy 2023-10-26T16:42:33,265 [INFO] W-9000-facebook\_\_sam-vit-large-stdout com.amazonaws.ml.mms.wlm.WorkerLifeCycle - KeyError: ‘sam’ 2023-10-26T16:42:33,265 [INFO] W-9000-facebook\_\_sam-vit-large-stdout com.amazonaws.ml.mms.wlm.WorkerLifeCycle - KeyError: ‘sam’ AllTraffic/i-0dad3e9d16762a67f 2023-10-26T09:42:37.363-07:00 Copy 2023-10-26T16:42:33,265 [INFO] W-9000-facebook\_\_sam-vit-large-stdout com.amazonaws.ml.mms.wlm.WorkerLifeCycle - Model facebook\_\_sam-vit-large loaded io\_fd=966296fffe2e23f7-00000015-00000000-a9415483c6ddb70e-6cd401cc 2023-10-26T16:42:33,265 [INFO] W-9000-facebook\_\_sam-vit-large-stdout com.amazonaws.ml.mms.wlm.WorkerLifeCycle - Model facebook\_\_sam-vit-large loaded io\_fd=966296fffe2e23f7-00000015-00000000-a9415483c6ddb70e-6cd401cc