vinvcn
vinvcn
Please refer to the description of this feature request to use the redis async. Also refer to this issue for explanation: https://github.com/zilliztech/GPTCache/issues/415
seems like the precision issue? try set the dtype by your GPU type? ```python FlaxWhisperPipline("openai/whisper-large-v2", dtype=jnp.bfloat16, batch_size=16) ```
> we use a10 to inference, also looks slow. Could you provide a brief measurement on that? etc. token/word per second
> You can use other fast inference libraries like FasterTransformers. We will also soon release a high throughput batching backend. Thanks for your work to improve this. Would you please...
It seems the VSCode extension is using a combination of ChatGPT like models and the CodeGeeX models. It seems the CodeGeeX was trained on code data only, not good enough...
I had to set up Jax by following the install locally harder option. That is, install cuda drivers, Cudnn, then run above command to install jax[cuda11_local]. The caveat is you...
chatpdf使用的微调过的模型？很想知道他们怎么做的？