-
Hi, I tried to replicate your speed experiment, I tested the deit_tiny, batch size=1, RTX3090 environment, after a few days of autotune, compared to tensorrt FP16, speed is still slower.
Here are t…
-
Current list of tasks:
- [x] threads > 1 do not work
- [x] batches > 1 do not work
- [x] check object detection task on any model to test TVM integration
- [x] detect TVM version via CK package …
-
### Summary
We propose to expand the capabilities of the Pruvendo Formal Verification Automated Toolkit to include FunC and Tact languages. Our team has successfully developed and implemented a pow…
-
Hi,
I am looking for some debugging options.
The VPC device I have seems to power down if it is not connected to a computer (if I just power it over USB then the lights all come on, few seconds …
-
Refer to TVM tutorial [Bring Your Own Codegen To TVM](https://docs.tvm.ai/dev/relay_bring_your_own_codegen.html), which details how to create a self-defined c source module codegen.
However, ONNX i…
-
`Unknown opcode: b010`. Any assistance in resolving this issue or guidance on how to add support for new opcodes would be greatly appreciated. Thank you!
-
When applying optimization passes in TVM, there is a discrepancy in the results between directly applying opt_a(opt_b(mod)) and using a sequential optimization approach, where seq_ab = tvm.ir.transfor…
-
## Description
lib_tvm.so file too short
http://jenkins.mxnet-ci.amazon-ml.com/blue/organizations/jenkins/mxnet-validation%2Funix-cpu/detail/PR-17494/3/pipeline/
Unrelated PR #17494
### Err…
-
I tried to compile the TVM CUDA kernel on my own computer with Ubuntu16.04. I have docker and Docker gpu runtime installed and they work well for my other projects.
Following the instructions, I tr…
-
mlc-llm/cpp/serve/threaded_engine.cc:283: Check failed: (output_res.IsOk()) is false: Insufficient GPU memory error: The available single GPU memory is 4762.535 MB, which is less than the sum of model…