Comments (4)
@muhammadfahid51 - have you took a look at the examples, i.e. https://github.com/kpe/bert-for-tf2/blob/master/examples/tpu_movie_reviews.ipynb
I'm not sure whether doing something fancy on top of BERT when fine-tuning brings much (compared to just classifying on the [CLS] activations). I believe it is rather important to try not over-fitting and find a good learning rate schedule.
from bert-for-tf2.
@muhammadfahid51 - what do you mean by "model is training" - does the loss drop over time? What hyper parameters do you use? What is your kernel size in the CNN and how many CNN layers do you use? And how large is your dataset?
I don't see why your model above should not work if using proper hyper-parameters.
from bert-for-tf2.
@kpe Hey,
Training loss does not drop over time and neither the validation loss. About parameters I have changed them alot. Like tried different number of layers, filters, regularization parameters etc but to no avail.
But I think the problem was something else. My data is in Urdu Language and it has got 10k samples of 3 classes with 50(Positive), 30(Negative) and 20(Neutral) percent distribution. I tried the same model on imdb data and it worked well.
But one tweak that remains is,
How does bert handles unknown words. Does it assign all of them to '[UNK]' and then learns them while training or is there some other technique ?
from bert-for-tf2.
@muhammadfahid51 - I believe the BERT models were pre-trained on English only. So until there is a multilingual pre-trained model you might have to pre-train from scratch yourself. To check how the tokenization works with your language, you could tokenize and convert the tokens to ids, and then back.
Currently bert-for-tf2 does not have an example/tools for pre-training from scratch. When pre-training from scratch it might be easier using the workflow from the original google-research/bert. Once you have the pre-trained weights, you could fine-tune with bert-for-tf2.
from bert-for-tf2.
Related Issues (20)
- Custom tokenizer layer HOT 5
- ResourceExhaustedError: OOM when allocating tensor with shape[501153,768] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc [Op:Mul]
- mixed precision HOT 3
- example (gpu_movie_reviews) has some mistake
- Failed to get weights from pretrained google model HOT 2
- Can not load pretrained bert weights when loading chinese_L-12_H-768_A-12/bert_model.ckpt HOT 3
- Paddings must be non-negative
- albert classification error(Failed copying input tensor from GPU in order to run Identity: GPU sync failed [Op:Identity])
- ValueError: Found unexpected keys that do not correspond to any Model output
- More comments for the code
- Can't train BERT with loaded weights on QA Task HOT 3
- Setting unexpected parameter 'name' in Params instance 'Params' HOT 2
- how to using this in functional model
- may be there is some problem work with tf hub
- AttributeError: module 'bert' has no attribute 'Layer'
- type error HOT 5
- Activation after bert-layer differs
- Count of weight not found[196]
- OSS License compatibility question
- tensorflow.python.keras.layer.input_spec should be replaced with tensorflow.keras.layers.InputSpec HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from bert-for-tf2.