Text Generation
Transformers
PyTorch
TensorBoard
Safetensors
bloom
Eval Results
text-generation-inference
Inference Endpoints

Hardware Requirements for Fine Tuning.

#197
by Haziqsayyed - opened

I wanted to know the hardware requirements for
a. just running this model
b. for fine-tuning.
I want to use it to convert sentences to mathematical expressions. Although the model is able to do that already. I have tried using the hosted API here on hugging face. I wanted to know how can I fine-tune it such that whenever I pass in a sentence it gives me back the expression, without having to tell it what to do.

Can someone also help me with the following, please?

  1. Guide to fine-tuning, including information on dataset format, loading the dataset, etc.
  2. Since the model is autoregressive can I give it a prompt like this
    "
  3. sentence: expression
  4. sentence: expression
    Using the above examples as reference convert the following sentences to expressions:
    sentence:
    "
    I want to save the model at this point and then later just pass the sentence I want. is this possible? if yes what will be the compute required?

Sign up or log in to comment