Questions tagged [gpt]
The gpt tag has no usage guidance.
93
questions
0
votes
0
answers
24
views
data verctorizing of a free text field to get better responses of the prompt
wonder if best practices are existing to approach the solution "what kind of vectorizing of a input filed" is the best for my use-case. Model, size and others can have a significant impact ...
- 1
0
votes
1
answer
36
views
Reinforcement learning
I am working on a sentiment analysis project. I used BERT model for training but lack of data it gives huge overfitting. So after i moved LLM approach to do that. Using LLM finally i got good results....
0
votes
1
answer
62
views
What is the "Extract" token and how is the final Linear layer applied in GPT?
In the manuscript of GPT, the authors have given the following image:
Questions:
What is the final "Extract" (token?)? Is it the "END" token?
How is the final linear layer ...
- 821
0
votes
0
answers
27
views
Is openAI text generation models an extension of embedding models?
we can creating embeddings using below code
...
- 167
1
vote
1
answer
63
views
A question about contextual embeddings in the decoder only transformer architecture (gpt)
I am reading up on the decoder only architecture
Relevant excerpts:
We can use any model that maps token sequences into contextual embeddings (e.g., LSTMs, Transformers):
$$\phi : V^L \to R^{d \times ...
- 821
1
vote
1
answer
150
views
what is the difference between window size and context length of language model?
is window size and context length of language model one and the same thing?
******** following text is added as question with ONLY above text was not allowed *****
I am trying to understand how GPT ...
- 167
0
votes
0
answers
66
views
How does supervised fine-tuning work in InstructGPT?
See Figure 2 from the InstructGPT paper:
I want to know how Step 1 works. Here is one possible algorithm.
Pass the prompt through the model, and compute the negative log of the probability of the ...
- 135
0
votes
1
answer
204
views
Why do GPT models use a transpose of the embedding matrix to convert outputs to logits?
According to section 3.1 of the original GPT paper, GPT right-multiplies the final output vectors (after applying a Transformer decoder model) by the transpose of the embedding matrix, before applying ...
- 135
4
votes
2
answers
4k
views
Why does everyone use BERT in research instead of LLAMA or GPT or PaLM, etc?
It could be that I'm misunderstanding the problems space and the iterations of LLAMA, GPT, and PaLM are all based on BERT like many language models are, but every time I see a new paper in improving ...
- 143
1
vote
1
answer
265
views
How to measure accuracy of GPT model
I am working on a model to build questions automatically from some text
My model will analyse provided article and ask authors questions that can help improving their articles
How can we measure the ...
- 539
0
votes
0
answers
119
views
Disadvantages of using just Vector Stores
I've just started with langchain, and one thing baffles me.
For starters, I'm using an in-memory vector store created from a local csv file
...
- 101
0
votes
0
answers
35
views
What is the input and output of GPT model for fine-tuning?
From my understanding, for the pretraining of GPT model, we need to do next token prediction task.
In this case,
Input -> The GPT models are general-purpose language models that can perform ... (...
- 1
2
votes
1
answer
73
views
How was the token library constructed for ChatGPT / other GPT systems?
I have found literally hundreds of articles on Google with titles like 'What are tokens and how to use them,' but haven't been able to find any information at all on how the token libraries themselves ...
- 153
0
votes
0
answers
24
views
Are there any studies on the actual loss of quality in a quantized GPT model? (eg GPTQ)
I've been trying to think of how different a GPT would perform if we swapped from floats to 8-bit. Given that we take say the top 10 words, how much fidelity is actually lost? How many tokens are ...
- 111
2
votes
2
answers
272
views
Training GPT Model with Swagger Documents, Need Help with Model Fit
We are currently developing an application that performs actions based on user input using 'gpt-3.5-turbo-0613'.
For example, if a user enters the prompt "create a user with the name Potter and ...
- 23
1
vote
1
answer
65
views
Why is the target output in GPT (Andrej Karpathy's course) contains copies of tokens from the input?
The question is based on Andrej Karpathy lecture on training a toy GPT model on Tiny Shakespeare dataset (youtube link). In this model, tokens are single characters with a dictionary of around 60 ...
- 215
0
votes
0
answers
501
views
GPT4All with Langchain generating gibberish in RHEL 8
I was trying to create a pipeline using Langchain and GPT4All (gpt4all-converted.bin). The pipeline ran fine when we tried on a windows system. But now when I am trying to run the same code on a RHEL ...
0
votes
0
answers
22
views
Huggingface not saving model checkpoint
I am trying to train T5 model. This is how my training arguments look like:
...
- 103
0
votes
0
answers
684
views
Creating LLM chatbot using llama-index + langchain
As the title suggests: I'm trying to build a chatbot which his goal should be sort of like "chatgpt".
The chatbot will be installed on Slack workspace and I'm struggling with which scope I ...
- 199
0
votes
0
answers
71
views
Question about GPT-2 causal masking for downstream tuning
I'm currently finetuning a pre-trained GPT-2 model for image captioning. Using a ViT for image encoding, the output embeddings are then fed into the GPT-2 model (from HuggingFace) as ...
0
votes
1
answer
220
views
Passing target text to gpt2 and T5 for fine tuning to learn text generation task
I have text with each line in following format:
<text-1> some text-1 <text-2> some text-2 <text-3> some text-3
I want fine tune model to learn ...
- 103
0
votes
0
answers
192
views
Fine tune Flan T5 to build a SlackBot
I want to create a Slack chatbot that will be able to "somehow" mock a "ChatGPT like" behavior.
I want the bot to be able:
answer questions
understand from context
summarize info
...
- 199
2
votes
1
answer
436
views
dolly v2 - how does it internally learn to follow instructions
this is more a curiosity query than anything else. The git repo for dolly gives us an easy way to swap, the training dataset , to train custom models, as long as we follow the format. I however have ...
- 328
0
votes
1
answer
19
views
Where can I see a summary of tools and techniques for most updated transformer developments?
Since the invention of chatGPT, there are many tools and techniques and variants invented ever since. I want to keep track of these developments and tools but I find no avail. May I know which ...
- 411
1
vote
1
answer
4k
views
Chromadb in my local system
I am trying to install chromadb on my local system but getting following error even after installing Microsoft C++ Build Tools:
Is it because "Chromadb" can only be used in google Collab and ...
3
votes
1
answer
218
views
Use a "ChatGPT like" engine really locally without exposing sensitive data
I was recently playing around with llama_index and llama Hub and found it very easy to be used on my own data set.
While this is ...
- 199
11
votes
1
answer
10k
views
How does an LLM "parameter" relate to a "weight" in a neural network?
I keep reading about how the latest and greatest LLMs have billions of parameters. As someone who is more familiar with standard neural nets but is trying to better understand LLMs, I'm curious if a ...
- 113
1
vote
0
answers
16
views
In ChatGPT, The difference of using reward to guide policy vs using the dataset of reward to train policy?
In ChatGPT, What are the differences of using reward to guide policy vs using the dataset of reward to train policy?
- 691
5
votes
5
answers
900
views
Is using GPT-4 to label data advisable?
If I have a lot of text data that needs to be labeled (e.g. sentiment analysis), and given the high accuracy of GPT-4, could I use it to label data? Or would that introduce bias or some other issues?
- 63
1
vote
1
answer
65
views
What should the numerical values of the <startofsentence> and <endofsentence> token vectors be?
I'm trying to build GPT2 from scratch. I understand how to convert each word in a sentence to its respective token index and each token is then converted to its respective word embedding vector. I ...
0
votes
1
answer
325
views
What are the benefits of changing the temperature parameter in the chatgpt api?
(Cross-posted from: https://ai.stackexchange.com/q/39813/57463 after a week of no satisfactory answer)
TLDR:
What are the benefits of increasing the temperature in the chatGPT API ?
Context (can be ...
- 101
2
votes
1
answer
1k
views
Can I use LLM to explain codebase?
I am a Data Engineer, and I am currently assigned a task to refactor an outdated code and rectify any bugs present. However, I am unable to comprehend the code written in the existing codebase. ...
- 158
3
votes
1
answer
792
views
How do GPT models go from token probabilities to textual outputs?
Suppose GPT-2 or GPT-3 is trying to generate the next token, and it has a probability distribution (after applying softmax to some output logits) for the different possible next tokens. How does it ...
- 135
0
votes
0
answers
50
views
fine tuning open ai model with historical data
i'm trying to understand more about training models and unsure how to approach this problem.
I have a bunch of historical financial data that I would like open ai to use as additional context when ...
- 101
0
votes
1
answer
96
views
Where can I find gpt-35-turbo in the Microsoft Pricing calculator?
Where can I find gpt-35-turbo in the Microsoft Pricing calculator? I don't see gpt-35-turbo in the model list:
https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/ ...
- 5,650
4
votes
2
answers
16k
views
ChatGPT: How to use long texts in prompt?
I like the website chatpdf.com a lot. You can upload a PDF file and then discuss the textual content of the file with the file "itself". It uses ChatGPT.
I would like to program something ...
- 63
0
votes
1
answer
232
views
GPT-2 architecture question
I am currently working on a NLP model that compares two comments and determines which one would be more popular. I have already came up with an architecture - it will be based on GPT-2. But now I am ...
0
votes
0
answers
69
views
How to use LLMs to explain comman lines
I am interested to use LLMs to explain command lines.
For POC, I used GPT-3 and got some really good results:
...
- 101
0
votes
1
answer
667
views
What are the advantages of autoregressive over seq2seq?
Why are recent dialog agents, such as ChatGPT, BlenderBot3, and Sparrow, based on the decoder architecture instead of the encoder-decoder architecture?
I know the difference between the attention of ...
- 3
0
votes
0
answers
32
views
How to extract a relevant paragraph/content from a multiple documents as per the question asked by the user?
I have a multiple similar text file in which part of the answer resides as per my questions and I want to only extract that paragraph from the most similar documents. I have used FAISS for searching ...
0
votes
1
answer
531
views
Words limit using GPT-3 API and fine tune model
In the documentation for GPT-3 API, it says One limitation to keep in mind is that, for most models, a single API request can only process up to 2,048 tokens (roughly 1,500 words) between your prompt ...
- 103
2
votes
1
answer
207
views
Does GPT-3 remember data from prompts used to fine tune it?
I am trying to fine tune a model using OpenAI's fine tuning API. I am passing bodies of text (for example, news paper articles) as prompts and the data I want from it as completions.
Let us consider ...
3
votes
1
answer
2k
views
Does OpenAI and ChatGPT use Scikit Learn?
Scikit Learn is Python's go-to open source package for running common AI and machine learning algorithms. Does OpenAI and its product ChatGPT use or rely on Scikit Learn on its back-end at all? If not,...
- 131
0
votes
0
answers
41
views
Vision Language Models Similar to DeepMind's Flamingo Available for Commercial Purposes?
Does anyone know of any Vision Language Models similar to DeepMind's Flamingo model that will take in a video clip a user provides the model along with a user-defined question in a prompt about what ...
- 221
1
vote
1
answer
123
views
Some answers given by ChatGPT are just beyond ridiculous, what could be the reasons?
Some answers given by ChatGPT are just beyond ridiculous, especially in Chinese (I am Chinese so I ask ChatGPT in both Chinese and English). Here is an example,
The last answer
...
- 113
3
votes
1
answer
154
views
Which Publicly Accessible Large Language Models are Very Similar to OpenAI's ChatGPT?
What other large language models exist or will soon exist that are VERY similar to OpenAI's ChatGPT in the sense of being fine-tuned or otherwise specifically created for conversational tasks ...
- 221
1
vote
1
answer
4k
views
Specifics about ChatGPT's Architecture
Does anyone know of reliable sources that have written about the architecture of OpenAI's ChatGPT - specifically regarding the following?:
Number of hidden layers
Number of attention heads
...
- 221
8
votes
6
answers
14k
views
ChatGPT's Architecture - Decoder Only? Or Encoder-Decoder?
Does ChatGPT use an encoder-decoder architecture, or a decoder-only architecture? I have been coming across Medium and TowardsDataScience articles suggesting that it has an encoder-decoder ...
- 221
2
votes
1
answer
268
views
Does opting out of having my content used for improvement mean there are no other forms of data retention of my content by OpenAI?
Regarding the use of OpenAI API, Terms of Use at OpenAI mentions:
You can opt out of having Content used for improvement by contacting [email protected] with your organization ID. Please note that ...
- 5,650
0
votes
1
answer
241
views
Is there any differences between using text-davinci-003 with the Azure API vs. with the OpenAI API? [closed]
I wonder whether there are any differences between using text-davinci-003 with the Azure API vs. with the OpenAI API.
- 5,650