Comments (6)
You also need to enter these command (including training command) in the inference:
lang=ruby #programming language
lr=5e-5
batch_size=32
beam_size=10
source_length=256
target_length=128
data_dir=../dataset
output_dir=model/$lang
train_file=$data_dir/$lang/train.jsonl
dev_file=$data_dir/$lang/valid.jsonl
epochs=10
pretrained_model=microsoft/codebert-base #Roberta: roberta-base
batch_size=64
dev_file=$data_dir/$lang/valid.jsonl
test_file=$data_dir/$lang/test.jsonl
test_model=$output_dir/checkpoint-best-bleu/pytorch_model.bin #checkpoint for test
from codexglue.
@guody5 thank your for your answer,
but if I run it with all parameters, I will have to do the training?
I want to run just inference, like in the Readme.
Training+Inference works, but just Inference for a trained model fails.
from codexglue.
lang=ruby #programming language
lr=5e-5
batch_size=32
beam_size=10
source_length=256
target_length=128
data_dir=../dataset
output_dir=model/$lang
train_file=$data_dir/$lang/train.jsonl
dev_file=$data_dir/$lang/valid.jsonl
epochs=10
pretrained_model=microsoft/codebert-base #Roberta: roberta-base
batch_size=64
dev_file=$data_dir/$lang/valid.jsonl
test_file=$data_dir/$lang/test.jsonl
test_model=$output_dir/checkpoint-best-bleu/pytorch_model.bin #checkpoint for test
python run.py --do_test --model_type roberta --model_name_or_path microsoft/codebert-base --load_model_path $test_model --dev_filename $dev_file --test_filename $test_file --output_dir $output_dir --max_source_length $source_length --max_target_length $target_length --beam_size $beam_size --eval_batch_size $batch_size
from codexglue.
In your error logs:
model_name_or_path should be "microsoft/codebert-base" but get '/models/pytorch_model.bin'
from codexglue.
According to your log, you should run the following command:
python run.py --do_test --model_type roberta --model_name_or_path microsoft/codebert-base --load_model_path /models/pytorch_model.bin --test_filename /dataset/test.jsonl --output_dir /experiment/output --max_source_length 256 --max_target_length 128 --beam_size 10 --eval_batch_size 32
from codexglue.
@guoday Thank you very much!
Works like a charm. I did not read properly that pretrained model and load_model were seperate things.
from codexglue.
Related Issues (20)
- 403 Forbidden error for Code-To-Text data files HOT 6
- Question related to fine tuning pretrained models for Defect-Detection task
- clone detection reproduction,CodeBert pipeline MAP@R only 76.64
- The CodeBlue evaluation script about code-to-code translation
- About CodeBLEU
- not a mach-o file when run code bleu
- javascript keys for CodeBLEU HOT 1
- [Code Completion - Token level] About eval_acc function
- [codecompletion-token] split function in code/dataset.py
- Missing data in ConCode HOT 2
- this line is wrong HOT 2
- The mlm and mlm_probability arguments in the run.py are not effective.
- Save_total_limit argument not used in run.py
- When resuming from a saved checkpoint, the train_dataloader doesn't resume from the same saved step.
- idx_file.txt is not effectively updated with the current epoch.
- Convert gradient accumulation with Accelerate
- no test file of webquery_predictions
- CloseTesting answer dataset is empty
- Question about text-code evaluation
- No submission response/feedback
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 codexglue.