All Questions
Tagged with large-language-model machine-learning
61
questions
0
votes
0
answers
10
views
CUDA Out of Memory Error Despite Having Multiple GPUs
I'm encountering a CUDA out-of-memory error while trying to run a PyTorch model, even though my system has multiple NVIDIA GPUs.
# Load the tokenizer and model
tokenizer = AutoTokenizer....
-2
votes
0
answers
20
views
training help hybrid based model that integrates contextual and numerical features for a classification problem [closed]
I want a critical production RISK analysis problem. So, based on a record I want to risk rank each record from 0 to 5. The training set is fairly imbalanced.
> "0.0 964
> 1.0 393
&...
0
votes
1
answer
127
views
Convert safetensors model format(LLaVA model) into gguf format
I want to do LLaVA inference in ollama, so I need to convert it in gguf file format.
My model has the file format safetensors.(trained with lora)
It seems that ollama supports only llama, but not ...
-1
votes
0
answers
29
views
Implementing Few-Shot Learning without Prompts for Llama2
I am working with the Llama2 model. I have successfully started and fine-tuned the model, and I have also used Few-Shot Prompting with and without LangChain. However, now I am looking for a method ...
2
votes
0
answers
33
views
DSPy can't retrieve passage with text embeddings in ChromaDB
I am working on a RAG application using DSPy and ChromaDB for pdf files.
At first I fetched the text from the pdf and add it to the Chromadb as chunks. Also added the embeddings of the chunks. And ...
0
votes
1
answer
39
views
How to get multimodal embeddings from CLIP model?
I'm hoping to use CLIP to get a single embedding for rows of multimodal (image and text) data.
Say I have the following model:
from PIL import Image
import torch
from transformers import CLIPProcessor,...
1
vote
0
answers
82
views
Fine tune llama3 with message replies like dataset (slack)
I want to fine tune llama3 on a dataset in which the data structure is a list of messages considering the below rules:
there are channels.
in each channel there are messages from all sort of users.
...
0
votes
0
answers
32
views
Pretrained Model Weights Not Updating During DPO Training
I'm trying to apply DPO to a pre-trained model. However, during the training process, the scores given by the pre-trained model and the fine-tuned model are identical, and the loss remains the same ...
-2
votes
1
answer
41
views
Understanding the results of Transformers Learn In Context with Gradient Descent
I'm trying to implement this paper:
https://arxiv.org/pdf/2212.07677
(Here's their code):
https://github.com/google-research/self-organising-systems/tree/master/transformers_learn_icl_by_gd
I'm ...
0
votes
1
answer
101
views
Search for documents with similar texts
I have a document with three attributes: tags, location, and text.
Currently, I am indexing all of them using LangChain/pgvector/embeddings.
I have satisfactory results, but I want to know if there is ...
0
votes
0
answers
21
views
What are the key quality metrics for large language model releases?
I am a first year PhD student working on improving the release practices of Machine Learning Models, especially pre-trained large language models. I want to understand the above concept for a ...
0
votes
1
answer
587
views
langchain RetrievalQA error: ValueError: Missing some input keys: {'query'}
In a RAG project, I am using langchain. When I run the QA chain with query input, this error keep showing up:
----> result = qa_chain({'query': question})
ValueError: Missing some input keys: {'...
1
vote
1
answer
77
views
How to tune LLM to give full length and detailed answers [closed]
I am building an application in which you can select an open source model from a list of models and ask it general questions. I am using searxng to search the web for context. While all of this is ...
0
votes
0
answers
66
views
why gradient scaling used in mix-precision training could lead to bigger Learning Rate in float32 model?
Background:
Gradient Scaling original is used in mix-precision training like (part of model weights are float16 and part of them are float 32), which its purpose is to reduce underflow of small ...
0
votes
0
answers
174
views
How to save QLoRA fine tuned Llama-3-8B model on disk and use it without the need to download base-model again
I want to fine tune llama3-8B model and save the complete Fine-tuned model in local disk.
And, use the Fine-tuned model in local disk for prediction. I don't want to download the based model again ...