Indrajit Bhosale

icon indicating an email [email protected]

icon company Nvidia icon location Santa Clara icon location Systems, Networks, OS, C/C++, Golang

Results 14 comments of Indrajit Bhosale

TensorRT model low throughput

Hi @rs-ixz , we are able to reproduce this. We will update you as soon as we have a RCA/Fix/WAR. CC @tanmayv25

TensorRT model low throughput

Hi @rs-ixz , Yes we are focusing on System Shared Memory, we will provide an estimate after an RCA. Currently we are actively in the process of RCA. Thanks, Indrajit

Update L0_trace for DLIS-6462

> Can you please remove line 567 `RET=0`? This line resets `RET` thus failures before that were not captured. cc @GuanLuo Resolved

Enhance OTEL testing to capture and verify Cancellation Requests and Non-Decoupled model inference.

> Could you please attach a picture of a trace as displayed in jaeger with cancelled request. > > Another question, have you considered cases when request was cancelled, when...

after calling unload_model capi, the memory is not completely released

CC @kthui

Issue with FP16 Output for Half Precision Model

CC @tanmayv25

PermissionError: [Errno 13] Permission denied: '/home/triton-server'

Hello @tapansstardog thanks for reaching out, 1. Can you confirm you are following the https://github.com/triton-inference-server/server/blob/main/deploy/k8s-onprem/README.md tutorial? 2. Can you create the /home/triton-server with appropriate permissions and try again? Thanks

Does python model support venv-pack for Creating Custom Execution Environments

CC @nv-kmcgill53 @mc-nv

Sending two "load" requests to server makes it load twice

Hi @ShuaiShao93 , thanks a lot for reaching out. Can you provide with the following details 1. What type of model/backend? 2. Can you reproduce this behavior with other types...

TensorRT model low throughput

Hello @rs-ixz , We repro-ed your issue on a 8GPU setup with concurrency set to 20 in Perf Analyzer. **Fix: Use --pinned-memory-pool-byte-size at triton startup set the size to a...