candle
candle copied to clipboard
Minimalist ML framework for Rust
Benchmarks did not compile.
I encountered the following problem. There is no problem with using python's transformers: ``` Error: WithBacktrace { inner: Msg("unsupported op_type Mod for op NodeProto { input: [\"/embeddings/patch_embeddings/Gather_1_output_0\", \"/encoder/layers.0/blocks.0/Constant_2_output_0\"], output: [\"/encoder/layers.0/blocks.0/Mod_output_0\"],...
Command: ``` cargo run --example stable-diffusion --features=metal -- --prompt "a cosmonaut on a horse (hd, realistic, high-def)" --img2img [path_to_img] ``` Stacktrace ``` thread 'main' panicked at /Users/worm/.cargo/registry/src/index.crates.io-6f17d22bba15001f/foreign-types-shared-0.3.1/src/lib.rs:72:9: assertion failed: !ptr.is_null()...
Hello, When loading weights from a pretrained PyTorch model using the `VarBuilder::get` method, I've noticed that most of the loaded weights for different layers match correctly, except for a 2...
Using the stable diffusion example running SDXL on CUDA vs Metal. Creating the image on a RTX 4000 ADA using Cuda takes about 1 second per step. Creating the image...
I tried Mixtral on my 128GB M1 Mac. The process state flips between `running` and `stuck`. The cli output from the example is not progressing past this poing: ```sh Running...
Hi Team, Looks like Gemma 2 is not supported by candle yet
Motivation: The current `QTensor::quantize` quantizes the `src` tensor onto the same device as `src`. This behavior is OK for most use cases, but there is a specific condition where this...
Running the new Flux model on Mac M1 Metal running the example ``` cargo run --features metal --example flux -r -- \ --height 1024 --width 1024 \ --prompt "a rusty...