|
|
43cd3c41b9
|
Fix gitignore and add LFS
|
2026-04-20 20:44:57 +01:00 |
|
|
|
a83cc9804a
|
Remove pycache and add to gitignore
|
2026-04-13 01:48:36 +01:00 |
|
|
|
dc91b522f0
|
Added models, outputs, data and tensorflow runs
|
2026-04-13 01:47:24 +01:00 |
|
|
|
1cca27e0b8
|
added further documentation across all files
|
2026-04-05 14:19:57 +01:00 |
|
|
|
7fa67af6c0
|
added timestamps to infer.py and tested it
|
2026-04-01 02:33:47 +01:00 |
|
|
|
82e6277cc1
|
Inference working with cli text, csv text, just need to add timings and I am done
|
2026-04-01 01:28:14 +01:00 |
|
|
|
1e8ea39287
|
inference shouldn't need much to complete
|
2026-03-29 00:17:05 +00:00 |
|
|
|
72c27aca13
|
Merge branch 'main' of github.com:charlie-rasberry/6013
Pushed diagram via web
|
2026-03-28 22:34:44 +00:00 |
|
|
|
0af8bff4a8
|
Added comments and made a start on infer.py
|
2026-03-28 22:31:10 +00:00 |
|
Charlie Rasberry
|
6145ca0333
|
Rename architecture_gh.png to architecture.png
|
2026-03-26 18:32:09 +00:00 |
|
Charlie Rasberry
|
ef557e663a
|
Add files via upload
|
2026-03-26 18:31:43 +00:00 |
|
|
|
753723694b
|
Added some comments and readability
|
2026-03-24 18:11:31 +00:00 |
|
|
|
afe61eaaa2
|
Merge branch 'main' of github.com:charlie-rasberry/6013
Merging the delete with reclass_analysis.ipynb and infer.py
|
2026-03-07 18:52:38 +00:00 |
|
|
|
5206e62d95
|
Analysis started, almost complete - compiled some excel sheets from the csv output with notes. Started infer.py, nothing major implemented yet
|
2026-03-07 18:51:15 +00:00 |
|
Charlie Rasberry
|
a8aaa077c7
|
Delete src/__pycache__ directory
|
2026-02-26 20:39:59 +00:00 |
|
|
|
cabf8aa9b5
|
Fixed evaluation indentation and other bugs
|
2026-02-26 20:39:19 +00:00 |
|
|
|
99896c0873
|
Added evaluation pipeline
|
2026-02-26 20:15:19 +00:00 |
|
|
|
96a0c45e84
|
Added implementation for single task roberta, using args for everything made it simple
|
2026-02-26 18:21:13 +00:00 |
|
|
|
01e2142276
|
Fixed a few issues with performance data collection and debugging output, mtl training is ready, moving on to single-task training to compare in write-up
|
2026-02-26 17:40:37 +00:00 |
|
|
|
df6aec7165
|
ignore runs
|
2026-02-23 16:29:44 +00:00 |
|
|
|
9467ea2519
|
ignore tensorboard logs and model checkpoints
|
2026-02-23 16:28:43 +00:00 |
|
|
|
4f0c54fe28
|
Added training loop for the MTL architecture on the original distribution
|
2026-02-23 16:26:48 +00:00 |
|
|
|
7bd68108d0
|
Implemented initial training structure, adding further logic soon including loss, stopping, optimisation and loop
|
2026-02-23 12:54:23 +00:00 |
|
|
|
76d9b8509b
|
Model almost complete, need to work on loss functions soon
|
2026-02-20 19:17:22 +00:00 |
|
|
|
cccd91a680
|
Small bit of progress towards model.py, now building forward()
|
2026-02-20 18:18:17 +00:00 |
|
|
|
61df4e3e26
|
Implemented dataset.py which tokenises and returns tensors, ready to load the model now
|
2026-02-19 22:10:25 +00:00 |
|
|
|
19c0d4bce3
|
Started dataset.py, added the ReviewDataset class and implemented the __init__, __len__ and __getitem__ methods. The __getitem__ method currently just returns the review text, but will be updated to return the tokenized review as a tensor
|
2026-02-19 18:45:55 +00:00 |
|
|
|
19bcf2aa18
|
Started dataset.py, added the ReviewDataset class and implemented the __init__, __len__ and __getitem__ methods. The __getitem__ method currently just returns the review text, but will be updated to return the tokenized review as a tensor
|
2026-02-19 18:41:37 +00:00 |
|
|
|
c5e91b79b2
|
Decided on max_length by finding out how many and which reviews would be truncated (it will be 256 tokens)
|
2026-02-19 01:28:10 +00:00 |
|
|
|
0be7da2dde
|
Finally processed the data fully and tested. Moving on to dataset.py and model.py
|
2026-02-19 00:44:36 +00:00 |
|
|
|
608588f023
|
Preprocessed tagged datasets, fixed CSV formatting issues, and added integrity checks. Also saved mappings for later inference use.
|
2026-02-18 22:36:58 +00:00 |
|
|
|
94a9fa1f17
|
gitignore change
|
2026-02-16 16:51:32 +00:00 |
|
Charlie Rasberry
|
8dbc5e7fc1
|
Remove duplicate repository structure heading
Removed redundant repository structure section.
|
2026-02-16 12:42:16 +00:00 |
|
Charlie Rasberry
|
c006b2fcff
|
Fix formatting in README for repository structure
|
2026-02-16 12:41:53 +00:00 |
|
charlie-rasberry
|
b88504725d
|
cleaned notebooks, finished datalabelling
|
2026-02-16 12:36:29 +00:00 |
|
charlie-rasberry
|
8d3dee6d30
|
House Cleaning
|
2026-01-28 16:41:27 +00:00 |
|
|
|
6cf36faf64
|
Line ending issue with my setup
|
2025-12-19 07:19:02 +00:00 |
|
|
|
487be5cd27
|
Everything is good to go for annotations.
|
2025-12-19 07:14:13 +00:00 |
|
|
|
5b9fbfc75e
|
data processing pipeline now finished just need to annotate reviews
|
2025-11-22 09:41:12 +00:00 |
|
charlie-rasberry
|
45ec02fa46
|
Moving on to multitag.py, sampling complete I think
|
2025-11-12 06:21:16 +00:00 |
|
charlie-rasberry
|
2cbdd55243
|
Fixed get_stratified_sample() and replace broken x() with actual working logic, added sample_with_keywords().
|
2025-11-12 02:05:20 +00:00 |
|
|
|
a178284ffc
|
Added multitag.py (65% complete), preprocess.py (complete), sampler.py (80% complete)
|
2025-11-09 01:45:09 +00:00 |
|
|
|
4d6e2511e6
|
Added multitag, includes preprocess.py, sampler.py and multitag.py(the main gui for labelling/annotation)
|
2025-11-06 17:40:29 +00:00 |
|
|
|
c0d4c13824
|
Ignore large CSV data files
|
2025-11-06 17:39:26 +00:00 |
|
|
|
cf6f2d8371
|
uber review
|
2025-10-08 00:19:49 +01:00 |
|
|
|
3c51a51331
|
initial commit
|
2025-10-07 23:47:48 +01:00 |
|
