-
Dear quanto folks,
I implemented static quantization as suggested in your coding example. When printing the datatypes of the parameters, I found that after quantization all the weights remained in …
-
```
soft Windows [Version 10.0.20348.2762]
(c) Microsoft Corporation. All rights reserved.
C:\Users\Administrator>cd Desktop
C:\Users\Administrator\Desktop>cd max
C:\Users\Administrator\Des…
-
- Add `flox-activations attach --timeout` call in attach-inplace.bash
- echo flox-activations attach --remove in attach-inplace.bash
- Remove logic in the CLI that disables services for in-place mode
-
### Proposal
In the README or a tutorial, add an example of how to collect activations from a single layer of a model.
### Motivation
run_with_cache() uses a ton of memory. Often, for contra…
-
## 🐞Describing the bug
I'm using 8.0b2 to quant my mlpkg, a resnet and transformer model. It crashed after printed 'calibration success'.
Debug:
in **_model_debugger.py--step function:** outputs…
-
### What happened + What you expected to happen
not sure if it is intended. But we cannot pass non-dag-value to be an input of other dag.
### Versions / Dependencies
master
### Reproduction scrip…
-
We may make backwards incompatible changes to how activations work. When we do this, `activate` and any `services` command should error with something like:
```
❌ ERROR: environment XXX has already …
-
I'm trying to intervene the modules in .generate call with some conditions by self-defined prober.
However, the conditions are calculated by the activations before exiting the tracing context. There…
-
when using 'make', this will not build with clang on my system. (Ubuntu 24.04 LTS ) I get an error of :
```
INFO: optimised for -msse3 -mfpmath=sse -mfxsr -DUSE_SSE -march=x86-64-v3
INFO: buil…
-
### 1. System information
- OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Ubuntu 22.04
- TensorFlow installation (pip package or built from source): pip package
- TensorFlow library (v…