Stack Overflow
Stack Overflow

Questions tagged [llm]

A general tag for large language model (LLM)-related subjects. Please ALWAYS use the more specific tags if available (GPT variants, PaLM , LLaMa, BLOOM, Claude etc..)

A general tag for large language model (LLM)-related subjects. Please ALWAYS use the more specific tags if available (GPT variants, PaLM , LLaMa, BLOOM, Claude etc..)

A large language model is characterized by its large size. Their AI accellerator networks are able to process huge amounts of text data, usually scraped from the internet.

200 questions
0
votes
1 answer

Fit the chat response into a list in GPT API

I'm trying to get the emotion in a text using chatgpt API def infer_feeling(text): prompt = f"What feeling is filled in the following text?\nText: {text}\nFeeling:" response = openai.ChatCompletion.create( model=model, …
mhmtemnacr
  • 185
  • 3
  • 18
0
votes
1 answer

ImportError: cannot import name 'CustomLLM' from 'llama_index.llms'

I'm having difficulties to work with llama_index. I want to load a custom LLM to use it. Fortunately, they have the exact example for my need on their documentation, unfortunately, it does not work! They have these imports in their example: from…
Peyman
  • 3,097
  • 5
  • 33
  • 56
0
votes
0 answers

openai.error.InvalidRequestError: 'text-classification' is not one of ['fine-tune'] - 'purpose'

I hope you are all doing good. I'm having an issue with this code and I don't understand what I'm doing wrong... I'm a newby in all of this so that might be a dumb question. I get this openai.error.InvalidRequestError: 'text-classification' is not…
skwuij
  • 1
0
votes
0 answers

Using Langchain with SQLDatabaseToolkit, create_pandas_dataframe_agent and PythonREPL for data analysis

I would like to use langchain with SQLDatabaseToolkit, create_pandas_dataframe_agent and PythonREPL for data analysis. Do you have a working approach for me? My approach isn't working # Ensure the SQLite database file exists and is…
Karsten
  • 1
  • 1
0
votes
0 answers

Is QLoRA viable to train Falcon-7(40)B-instruct on a new language?

Hei, there. I've been researching the latest technologies, and I understand quite well(I hope I do) that (Q)LoRA is a viable option for finetuning LLMs on QA for specific tasks. I am still wondering if I could use the same technique to finetune an…
Marcus
  • 128
  • 2
  • 12
0
votes
0 answers

How to improve the output of fine tuned Open Llama 7b model for text generation?

I am trying to fine tune a openllama model with huggingface's peft and lora. I fine tuned the model on a specific dataset. However, the output from the model.generate() is very poor for the given input. When I give a whole sentence form the dataset…
Md Tahmid Hasan Fuad
  • 15
  • 5
0
votes
0 answers

Fine-tuning an open-source LLM for a new language?

What are the most suitable open source LLMs and frameworks for fine-tuning? I intend to use this model in a quite specific domain, perhaps a physics mentor for a school. How long might it take (with 3070 Ti 11Gb) to achieve acceptable accuracy for…
cptatoo
  • 11
0
votes
0 answers

LangChain & OpenAI, Comparing Documents

I've done some embeddings which are labeled resumes(CV's), stored them into a pinecone with a cosine similarity search. Is there a way to pass a .pdf file and create a comparison between inputed CV and labeled CV's? i've trying using QAretrival but…
Vandalism
  • 23
  • 4
0
votes
0 answers

LLM(BLOOM) infer with CPU instead GPUs and OOM happens

import torch from transformers import AutoTokenizer, AutoModelForCausalLM import os os.environ["PYTORCH_CUDA_ALLOC_CONF"] = "max_split_size_mb:32" os.environ['CUDA_VISIBLE_DEVICES'] = '0,1,2,3,4,5,6,7' device = torch.device('cuda') tokenizer =…
Yg Meng
  • 53
  • 1
  • 1
  • 8
0
votes
0 answers

Grouping Keywords based on semantic value of the word

I am trying to find a way to group multiple keywords based on their meaning or possible relation to each other. These grouping are based on the semantics of the word and not word similarity. I don't know how to do this as all the solutions that find…
Kaki Master Of Time
  • 1,428
  • 1
  • 21
  • 39
0
votes
0 answers

How to use LangChain to store, load, and query Loaders on LlamaHub?

Can loaders on LlamaHub only use LlamaIndex? How can LangChain be used to store, load, and query it? For example, Mongo Loader. I saw on LlamaHub that it seems that all the examples use LlamaIndex. So what can LlamaHub provide for LangChain? If…
chenkun
  • 45
  • 6
0
votes
1 answer

How to use LangChain to load the index generated by LlamaIndex and perform a query?

I used LlamaIndex to generate an index for a section of text, which is stored in the myindex folder. How should I use LangChain to load it and query it? At present, I can only use LlamaIndex for querying, but this will lack the functionality of…
chenkun
  • 45
  • 6
0
votes
0 answers

LLM Content Generation in Non-English Languages

I am trying to use GPT3 to generate content in non-English languages, including some low-resource languages with an inherently small amount of training data. I can think of two approaches to this challenge. The first one is to use some translation…
perplexedcoder
  • 1
0
votes
0 answers

Llamaindex query working in windows but failing in linux

This code works fine in windows but the same fails in Linux. I get RetryError[] error in linux. I have no idea why this is the case. from llama_index import SimpleDirectoryReader,…
CaptainAmerica
  • 31
  • 4
0
votes
0 answers

Is it valid to evaluate a flan-t5 model on sequences longer than it's max_length of 2048 tokens (assuming I have enough memory)?

I am evaluating the different flan-t5 models with few-shot chain of thought prompts which can go over the 2048 maximum token length. I am under the impression that because T5 uses relative position encoding, that it would be valid (make sense) to do…
Juan Diego Bermeo
  • 1
1 2 3
…
13 14