# Quantized models and legalization pass

**URL:** <https://discuss.tvm.apache.org/t/quantized-models-and-legalization-pass/8253>\
**Category:** Development\
**Created:** [October 22, 2020, 10:34pm UTC](https://discuss.tvm.apache.org/t/quantized-models-and-legalization-pass/8253 "2020-10-22T22:34:22Z")\
**Posts on this page:** 1\
**Showing post:** 9

<div class="post-metadata">

**Author:** ![giuseros](https://discuss.tvm.apache.org/user_avatar/discuss.tvm.apache.org/giuseros/32/803_2.png) [@giuseros](https://discuss.tvm.apache.org/u/giuseros)\
**Post date:** [October 29, 2020, 2:53pm UTC](https://discuss.tvm.apache.org/t/quantized-models-and-legalization-pass/8253/9 "2020-10-29T14:53:46Z")

</div>

Hi @FrozenGene, @anijain2305  
I can confirm that this works 🥳! Very good! Now we can implement algorithms like QNNPack and let the tuner try them together! Thanks both guys!

As for the API change, I agree with @FrozenGene that maybe it would be cleaner adding `tinfos` to the `qnn_conv2d_legalize` signature.

I have a related question, that I always meant to ask: in `conv2d_alter_layout`, we don’t execute the function if the current configuration is a Fallback. Do you guys know why? And also, what should the behavior be in `legalize`? I am referring to this code:

```auto
    _, outs = relay.backend.compile_engine.select_implementation(
        relay.op.get("nn.conv2d"), attrs, tinfos, out_type, target
    )
    workload = autotvm.task.get_workload(outs)
    if workload is None:
        # The best implementation is not an AutoTVM template,
        # we then assume it's not necessary to alter this op.
        return None
    cfg = dispatch_ctx.query(target, workload)
    if cfg.is_fallback: # if is fallback, clear query cache and return None
        autotvm.task.clear_fallback_cache(target, workload)
        return None

    topi_tmpl = workload[0]

```

In theory, we could gather `topi_tmpl` directly from the first parameter returned by `relay.backend.compile_engine.select_implementation`. But instead, in `conv2d_alter_layout` we query the `dispatch_ctx` for the current configuration and, if it is a Fallback, we return `None`. To sum up the follow-up questions are:

- Why this behavior is there?
- What should we do in `legalize`? Simply return back a default legalization?

Thanks once more for your help!

---

_[View the full topic](https://discuss.tvm.apache.org/t/quantized-models-and-legalization-pass/8253)._