tortoise.cpp
tortoise.cpp copied to clipboard
balisujohn
A ggml (C++) re-implementation of tortoise-tts
This task is blocked by the gpt-2 forward pass test being added since this could introduce regressions. https://github.com/balisujohn/tortoise.cpp/issues/5 The task is as follows: measure the runtime of the autoregressive model...
CLVP is a non-essential component of tortoise-tts for filtering for latents that will will yield good generation quality if given to the diffusion model. This task is to export CLVP...
AMD GPUs
Hi, I saw this project currently only supports Cuda. I was wondering if it might be possible to use [HIPIFY](https://github.com/ROCm/HIPIFY) to make it work on AMD GPUs. Do you know...
If people are interested in contributing to tortoise.cpp, a great first task would be getting the tokenizer to always match the tokenization tortoise-tts uses. The tokenizer I'm using in tortoise.cpp...
The "#include numeric" (would add less and greater than but markdown removes them) is necessary for including std::partial_sum() (at line 5219, or, in this fork, 5220). This will fix an...
This is an attempt to rebase on the latest commit of ggml master branch. My primary goal behind it is to add Vulkan/OpenCL support as I only have AMD GPUs....
Everybody realises that's a picture of a turtle? right? :)
Hello, I found a fine-tuned model to handle French : https://huggingface.co/Snowad/French-Tortoise , but the files are in Torch format. Is there any way to convert it to ggml's format ?
Optimize GPT2 inference: Remove redundant `autoregressive_latent_graph` and enable streaming output
Thank you for this excellent implementation. I'd like to suggest an optimization that could significantly speed up inference and enable streaming output. Currently, there are two GPT2 graphs: 1. autoregressive:...
Metadata
Owner
Metadata
A ggml (C++) re-implementation of tortoise-tts