rust
rust copied to clipboard
Rust language bindings for TensorFlow
Minimal example code: ```rust use tensorflow::{Scope, Session, SessionOptions, SessionRunArgs, Variable}; fn main() { let mut s = Scope::new_root_scope(); let variable = Variable::builder() .const_initial_value([1i64]) .build(&mut s) .unwrap(); let session = Session::new(&SessionOptions::new(),...
Title says all
I'm trying to use `raw_ops::decode_image` to load an image directly from a `u8` slice (as opposed to from file as per the example), but it seems I must first convert...
this is my code ```rust fn tf_ops_range() { // Create a range op that produces [0, 1, 2, 3, 4, 5] let mut scope = tensorflow::Scope::new_root_scope(); let start = tensorflow::ops::constant(0_i32,...
If one restores a model with SavedModelBundle::load() and obtains a graph, there is no way to get Variables from it. One can obtain an Operation with Graph::operation_by_name_required, but not a...
So I recently added `tensorflow` to a rust project that had an external openssl dependency (`reqwests` and `paho-mqtt`) and I immediately started seeing segfaults. The strange thing is that these...
- remove protobuf version restriction - fix CI Closes #414
In Cargo dependency versions are deduplicated across the entire workspace (per each major version). The hardcoded dependency `protobuf` in tensorflow forces all users of tensorflow to use exactly this version...
I saw some models available on [tensorflow models](https://tfhub.dev/tensorflow) but as a beginner, I am not very good at using rust to load these models for image recognition