Відео

Hi, Thanks for watching. I have run this project on my laptop which is i5 with 16 gb ram. So minimum system requirement i cant say because i haven't check. But to run it smoothly on gpu which have 16 or 24 gb will be enough i am saying according to whisperx model.

We only train the last few layers (classification head, projection layers) or task-specific layers are trained or fine-tuned. In Our case as i have explained in video target modules are q_proj, v_proj (query and value projection). You asked what should you do, I am not able to understand. You have to explain your problem statement then i can assist you.

I have explained in the video. At 06:03 we are using load_data package to load the data from huggingface. In the load_data function we just need to pass the dataset name.

Sure. What exactly the problem you are facing?

@nextGenAIGuy490 hi thanks for replying can you create a new video specifying how to specifically install this system I was having problem with loading the model it is giving none type error. Also there are some errors related to whisperx

@nextGenAIGuy490 it would be really helpful if you help me

@@RohitTalele Do one thing raise a issue request on github. Or send the error screenshot on email. Its easy to do installation i have mention on github as well. And also i have given the whisperx github link from there you can install. But dont worry lets connect through email or github.

​@@nextGenAIGuy490 Hi I have raised issues on GitHub additionally it would be great if you can share your email so that I could send you a specific error screenshot

@@tamilselvan3525 I haven't trained completely because of GPU limitation. So i won't be able to answer. I just wanted to share that its possible to train and how to train. But training time is dependent on dataset, hardware(GPU configuration) and no. of epochs you are training for.

@@nextGenAIGuy490 Okay, thanks.

@@soulaimanebahi741 No, We can't.

Absolutely wrong ! If you dont know say "dont know" . Dont mislead him , fine tuning over video is possible

@@babusd relax bro. Do one thing rather than saying show the proof. During training on llama 3.2 vision model they have used image and text pair. Read the model architecture. And if you know show me where they have written we can Fine-Tune vision model on videos.

It should work for other 2 data as well. You can debug the code find out where exactly the error is. Still if you are not able to figure it out. We can connect throught mail. Or you can watch the tutorial again i have explained each and every code.

Sure, So you have a api key. Now just create a file name .env than paste the api key after writing this variable in .env file. GROQ_API_KEY = "paste_your_api_key". and you are ready to use the project. I will add the .env file on github repository you can clone from there also.

Sure. I will try to include in my future videos.

@nextGenAIGuy490 thank you so much.cant wait more

Glad it helped!

As i have replied on mail. Type in the terminal streamlit run app.py. Also included steps to run this project in readme.md file in github.

Thankyou 😊

By customizing the collate_fn, we are able to control how the data is prepared. we are using it for batch processing, padding bringing data into format to train the model. Its fine to use it.

I highly doubt that this could work. Different models have different chat templates and processing.

Thanks. You can use AutoModelForVision2Seq to load your model. You need to pass your model path and use huggingface access token.

Glad it helped

Thanks

Hi, The project link is given in the description. I also explained in the video how to run this project.

Yeah in cpu its quite slow. But if you use better GPU. You will notice the processing speed will decrease drastically. T4 is good but nvidia gpu with 24 gb ram do the great job.

What is minimum system requirements to run this project on local machine. Like CPU, GPU, RAM, etc?