[Apache TVM Discuss] [Questions] Why stop quantize after first ``nn.global_avg_pool2d``

@register_annotate_function("nn.global_avg_pool2d") def global_avg_pool2d_rewrite(ref_call, new_args, ctx): """Rewrite function for global_avg_pool2d for stopping quantize""" if quantize_context().check_to_skip(ref_call): return None expr, x_kind = _g

[Apache TVM Discuss] [Questions] Is Module thread-safe?

Thank you for your reply --- [Visit Topic](https://discuss.tvm.apache.org/t/is-module-thread-safe/2759/14) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/bc47618c7b2

[Apache TVM Discuss] [Questions] [CUDA]Got Error: CUDA ERROR LAUNCH OUT OF RESOURCES

In some sense, I decreased batch size, it usually worked out. Did you find some solution to solve this problem? I mean let module.run() without getting out of cuda memory error? --- [Visit Topic](https://discuss.tvm.apache.org/t/cuda-got-error-cuda-error-launch-out-of-resources/4173/6)

[Apache TVM Discuss] [Questions] Is Module thread-safe?

graph runtime functions are not thread safe since they access per runtime state --- [Visit Topic](https://discuss.tvm.apache.org/t/is-module-thread-safe/2759/13) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https

[Apache TVM Discuss] [Questions] How dose TVM elimitate calls of conv weights layout_transform?

Yes weight layout transformation should be optimized by constant folding. --- [Visit Topic](https://discuss.tvm.apache.org/t/how-dose-tvm-elimitate-calls-of-conv-weights-layout-transform/8208/5) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from

[Apache TVM Discuss] [Questions] Do we have to identify tvm.nd.array context?

for example, if we use `ctx=cuda`, both input datas are run correctly. I saw that default tvm.nd.array context is cpu(0), >input = tvm.nd.array(shape) but, I designate context "cuda" > input = tvm.nd.array(shape, ctx) both inputs are run correctly, but it have small performance diffs. **Ca

[Apache TVM Discuss] [Questions] Is Module thread-safe?

is "GraphRuntime::Run" threadsafe ? --- [Visit Topic](https://discuss.tvm.apache.org/t/is-module-thread-safe/2759/12) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/

[Apache TVM Discuss] [Questions] About Configspace in autotvm

What will be the default tile factor in the loop for any operations --- [Visit Topic](https://discuss.tvm.apache.org/t/about-configspace-in-autotvm/8049/3) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://di

[Apache TVM Discuss] [Questions] Is there a better way to see stmt when we write a tvm pass?

When we write tvm pass, always it's a little hard to see the result directly and clearly, is there a good way to see stmt and faciliate the pass writing and debugging? --- [Visit Topic](https://discuss.tvm.apache.org/t/is-there-a-better-way-to-see-stmt-when-we-write-a-tvm-pass/8216/1) to

[Apache TVM Discuss] [Questions] Is multi-threading supported when TVM target SGX?

I'm using Fortanix SGX to run models in SGX enclave. Multi-threading is supported in SGX, but I cannot make my program run in multi-threading. Changing the value of TVM_NUM_THREADS is useless. --- [Visit Topic](https://discuss.tvm.apache.org/t/is-multi-threading-supported-when-tvm-target-