Questions tagged [finetuning]
The finetuning tag has no usage guidance.
89
questions
1
vote
2
answers
87
views
Fine-tune GPT on sketch data (stroke-3)
These past days I have started a personal project where I would like to build a model that, given an uncompleted sketch, it can finish it.
I was planning on using some pretrained models that are ...
1
vote
1
answer
264
views
How to further fine-tune a transformer NLP model on domain specific dataset, after general fine-tuning
I would like to fine-tune a pre-trained BERT-like model for a semantic similarity analysis task in the fashion of the SNLI/MNLI task (i.e. classify sentence pairs to "entailment" or "...
0
votes
1
answer
53
views
Detection of musical instruments using Yamnet
My goal is to detect musical instruments with AI (machine learning).
I'm currently using the Yamnet model to make inferences, but it has a very wide range of categories, for example, "Growling&...
4
votes
4
answers
11k
views
Understanding alpha parameter tuning in LORA paper
I was reading the LORA paper https://arxiv.org/pdf/2106.09685.pdf a thing I don’t understand is section 4.1, where the updates are updated by alpha, where alpha is a constant in r. It is said that ...
0
votes
1
answer
333
views
Fine-tuning LLM with limited documents and hierarchy
Hello LLM enthusiasts.
I am wondering w.r.t. a neighbouring project if there are state of the art approaches to fine tune a model if:
the realm of documents is limited (still more than just a few),
...
0
votes
1
answer
91
views
Fine-tune zero-shot classification model multi-label
I started a small project where I am trying to fine-tune a zero-shot classification model on a proprietary dataset. I was thinking to use the NLI approach, building contradiction and entailment ...
0
votes
1
answer
12
views
How could I further improve my detection model?
I am currently tasked with the creation of a model which is able to detect burrows of small rodents from high resolution aerial images.
My labeled images are look like this:
My results now are the ...
0
votes
0
answers
10
views
Is FLAN-t5 a good MT model at this stage for baseline comparision?
I'm currently working on a research project which involves collecting some information about baselines of Machine Translation Tasks (preferably WMT '19, WMT '14 datasets), and wanted to get a broad ...
0
votes
0
answers
29
views
Encoding Numerical Data in Text Sequences for Pre-training BERT from scratch
I have a time-series-based app usage logs dataset with columns such as:
App: the app being used (categorical)
...
0
votes
1
answer
2k
views
Incompatible shapes (None, 1) and (None, 5) with Keras VGGFace Finetuning
Categories to learn and predict:
df.race.unique()
array(['0', '1', '3', '2', '4'], dtype=object)
Data:
...
0
votes
0
answers
16
views
Instrument classification on a full audio stem with Yamnet fine-tuned with NSynth dataset
My goal is to feed my model an audio track, or more specifically a stem, and have it predict the instrument of this audio file.
I have done feature extraction on the Yamnet model, trained it with the ...
0
votes
0
answers
36
views
Recommendation System: Two-Tower Model Underperforming Simple Embedding Average Baseline
I'm trying to build a recommendation on a dataset of product purchases. The dataset consists of roughly 4 Amazon products that a particular user has bought (in sequence). I want to use the first 3 ...
0
votes
1
answer
97
views
Can I add a new output class to a decoder and train only the final layer?
I am wondering how to approach a project, where I would like to increase the number of output classes of an already trained network. I have very good reason to believe that the model has already ...
0
votes
1
answer
601
views
How to fine-tune GPT-J with small dataset
Firstly, thank you so much for looking at this post. I could really use some help.
I have followed this guide as closely as possible: https://github.com/kingoflolz/mesh-transformer-jax
I'm trying to ...
0
votes
1
answer
100
views
How to use new dataset on a pretrained neural network model?
I have built a dataset that I would like to pass to a pretrained model in oder to perform some predictions. I am looking for some steps/processes to guide me in this. Should I fine tune?If so what ...
0
votes
0
answers
185
views
NaN grad norm even with a stable loss and gradient
Currently I am working on a custom fine-tune of several code LLMs and while working on the DeepSeekCoder I encountered a strange behaviour.
When training the model earlier or later the loss goes to ...
0
votes
0
answers
21
views
How can I make my Hugging Face fine-tuned model's config.json file reference a specific revision/commit from the original pretrained model?
I uploaded this model: https://huggingface.co/pamessina/CXRFE, which is a fine-tuned version of this model: https://huggingface.co/microsoft/BiomedVLP-CXR-BERT-specialized
Unfortunately, CXR-BERT-...
1
vote
1
answer
143
views
Train on multi-domains, then fine-tune on specific domain
Would it make sense to first train a model on images from multiple domains, and then do "fine-tuning" on one specific domain to improve its performance on it?
For instance, one could train an object ...
2
votes
1
answer
736
views
Why not using linear regression for finetuning the last layer of a neural network?
In transfer learning, often only the last layer of the network is retrained using gradient descent.
However, the last layer of a common neural network performs only a linear transformation, so why do ...
2
votes
2
answers
2k
views
Should you care about truncation and padding in an LLM even if it has a very large tokenizer.max_length so that truncation will never happen?
I want to find out the role of truncation and padding in Huggingface Transformers pretrained models and/or any fine-tuning models on top. Taking a large language model like the German GPT2 shows that ...
2
votes
1
answer
243
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 ...
1
vote
1
answer
822
views
Is it possible to fine-tuning BERT by training it on multiple datasets? (Each dataset having it's own purpose)
BERT can be fine-tuned on a dataset for a specific task. Is it possible to fine-tune it on all these datasets for different tasks and then be utilized for these tasks instead of fine-tuning a BERT ...
0
votes
0
answers
10
views
Validation accuracy dip and recovery when restarting training
i was fine-tuning this large language model with Stochastic Gradient Descent and mid epoch i stopped training, and saved the model weights. Then at a later time, reloaded the weights and restarted the ...
3
votes
1
answer
76
views
Since LoRA parameters are randomly initialized, shouldn't that mean that initially breaks a models output?
I have just tried using LoRA on Llama 3 8B and I found without doing any fine tuning it performed pretty well on my dataset. But then I realized that surely the LoRA parameters are randomly ...
0
votes
0
answers
58
views
Fine tuning a model for question-answering task without context on the dataset
I have a small dataset (contains around 2.5k question answer pairs) which I would like to train a T-5 base model on.
The code examples that I have came across fine-tune with a dataset which has ...
1
vote
2
answers
203
views
How can I fine tune a model to detect digits, used to detect denominations of currency notes
So the task at hand is to detect the denomination of any currency banknote. The dataset I have is about 2k images of each denomination (12 in total). An example banknote (after noise removal, erosion ,...
1
vote
0
answers
26
views
Train a LLM to learn the entropy of the use case
I want to train a LLM (prefered Llama-2-13b) to learn the entropy of german texts - to be specific sports news. I use perplexity as training metric and want to check the training success after the ...
0
votes
0
answers
44
views
How is a causal language model correctly fine-tuned?
I want to fine-tune an SLM like Phi-2 through the huggingface API. I am in doubt how to achieve that, because I see two ways to do that and I am wondering which is the correct way.
The task is just a ...
0
votes
1
answer
41
views
How can I get the list of pretrained large language models?
Is there any place I can get the list of pre-trained large language models in a neat way? Despite the most common ones like gpt, BARD, llama2, which llm do you suggest that can be used for RAG and ...
0
votes
1
answer
24
views
How to choose ideal pretrained model for fine-tuning?
I started to work with LLMs lately and want to know how people choose their pre-trained models in their fine-tuning tasks? What is the criteria to choose the base model and which factors affect?
0
votes
0
answers
81
views
Can I fine tune MedPaLM model
Is it possible to fine-tune MedPaLM and MedPaLM2; Google's llms trained using PaLM specialized for medical domain. Can we fine-tune these models further to get more specialized models?
1
vote
1
answer
36
views
Is it methodologically correct to use the data to be used for finetuning in the pretrain phase of the BERT model?
Let us assume the training of a BERT model.
An initial pre-train is performed with a large data set A.
Subsequently a finetuning is performed with a dataset B which is part of A, but now with labels ...
0
votes
1
answer
56
views
Getting a free and unknown answer to a question against a fine-tuned text generation model trained on many essays and their few questions and answers
Aim
I want to fine-tune a text generation model with essays of changing size and then ask each of these input texts a few questions. I already have a wider range of question-answer pairs at hand for ...
3
votes
1
answer
1k
views
Fine-tuning a pre-trained LLM for question-answering
Objective
My goal is to fine-tune a pre-trained LLM on a dataset about Manchester United's (MU's) 2021/22 season (they had a poor season). I want to be able to prompt the fine-tuned model with ...
0
votes
0
answers
11
views
Help with a MaNet finetuning (binary semantic segmentation task)
Introduction: I am currently working on a computer vision problem, I have satellite images and I have to detect a particular archeological structure (Tell). I have access to the previously made ...
1
vote
2
answers
731
views
How can you get a Huggingface fine-tuning model with the Trainer class from your own text where you can set the arguments for truncation and padding?
I want to find out the role of truncation and padding in Huggingface Transformers pretrained models and any fine-tuning model on top of that. Therefore I played around with these parameters, but I ...
0
votes
1
answer
210
views
Outdated Transformers TextDataset class drops last block when text overlaps. Replace by datasets Dataset class as input of Trainer train_dataset?
Why I try to replace the transformers TextDataset class with datasets Dataset class
I stumbled upon this when I tried to make ...
0
votes
0
answers
27
views
why does my multi-modal model can not learn anything?
I have a multi-modal model. I want to train it using the Pytorch Framework. I have a balanced dataset. I have approximately 150 samples for each client. (I had preprocessed my text data.)
when I train ...
0
votes
0
answers
68
views
RLHF fine-tune llama2 in vertex ai
I have fine tune RLHF with Vertex AI Pipeline. But deployed model not showing in model registry. Why?
code i have used:
...
1
vote
0
answers
79
views
Could someone help with fine-tuning dolphin-2.2.1?
Could someone help with fine-tuning dolphin-2.2.1?
I have a problem with training: my train\loss - 0 and validation\loss - 0.000... after 800-1000 steps and this is overfitting
...
0
votes
0
answers
30
views
How to control and optimize optuna
i'm here with a pretty open question.
I'm using Optuna to fine tune a Catboost Regressor and i've found, trying a set of parameters by hand, that the "best params" it outputs are not the ...
1
vote
0
answers
132
views
Fine-tuning Hugging Face’s Llama Model with Unlabelled Data from PDFs from niche domain
I’m unsure about the next steps. Specifically, I have the following questions:
How can I prepare my unlabelled data for the fine-tuning process?
What’s the best way to fine-tune the Llama model with ...
2
votes
1
answer
3k
views
How to download and fine tune a large language model locally?
I want to fine tune a model locally, not using HuggingFace or any other third party tool. Basically, I want:
Download a trained model (Llama-2, Falcon, whatever is easiest).
Fine-tune it locally with ...
0
votes
0
answers
78
views
Fine-tuning MT5 for making it more like ChatGPT
I am trying to fine-tune a model which works like ChatGPT for Punjabi language, using the mt5-base, however I am not sure if I should go ahead with it since it does not even generate text and when I ...
0
votes
0
answers
30
views
For Finetuning Llama 2, what form of data is required?
I am working on a customer service chatbot project. I have files of different product's manuals on which i need to train the LLM which is Llama 2. As these are manuals so they not in Q/A form. In this ...
0
votes
0
answers
26
views
TFRobertaSequenceClassification for Address Normalization task
I have dataset with two column: one with faulty addresses, and other with correct addresses. I want to train a model such that, I can use it later for correcting all the incoming faulty addresses.
I ...
1
vote
2
answers
180
views
Training model using BERT
I have generated dataset using chat gpt. Dataset has 9000 data recodes. It's 6 class sentiment analysis. classes are 0,1,2,3,4,5
I used 3000 recodes for training, 1200 recods for validation and ...
0
votes
0
answers
31
views
Is it the right approach to select the model when it gives highest accuracy on validation dataset?
I am training the Densenet121 Model on an image dataset. I divided the dataset into 80% for training and 20% for testing.
Then I further divided the training data into 85% for training and 15% for ...
0
votes
0
answers
90
views
Can LLM fine-tuning be used to improve a language?
I'm Danish, and with all the excitement around open LLM models, I'm feeling a little left out.
Take Llama 2, for example - it was trained on a very small Danish dataset. Just enough to learn the words ...
1
vote
0
answers
222
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 ...