Merge branch 'main' of github.com:fabianHueni/UNIBE_MSGAI_HS25_Project

anaylse: visualized distributions of timing metrics and analysed correlation inferencetime vs inputsize

fix: fixed qwen3 check to check first if service.getModelName() is defined

Merge branch 'main' of https://github.com/fabianHueni/UNIBE_MSGAI_HS25_Project

results: added results of 3 models with my lenovo büro laptop

analyse: updated analysis

Merge branch 'main' of github.com:fabianHueni/UNIBE_MSGAI_HS25_Project

analyse: started with data analysis

resuklts add resuklts of gemma-3-270m on lenovo büro device

results: add results of 4 on device models, ran on macbook air M2

results: add results of three on device models

Merge remote-tracking branch 'origin/main'

fix: typo in description

fix: now the csv parsing ignores comma delimiters if they are inside quotes, to ensure all context is parsed correctly

fix: checked first 1500 columns of dataset on marks with the added jupyter notebook

chore: checked first 1000 samples for correct quotations manually

fix: minor typos

feat: reduce experiment setup to 500 items

Merge remote-tracking branch 'origin/main'

feat: remove f1 from csv download

Merge branch 'main' of github.com:fabianHueni/UNIBE_MSGAI_HS25_Project

feat: changed maxTokens for cloud and on-device from 500 to 50

feat: open linkedin links in new tab

Merge branch 'main' of github.com:fabianHueni/UNIBE_MSGAI_HS25_Project

feat: disable thinking for qwen model

feat: add dataset item id to logs and csv result exports

Merge branch 'main' of github.com:fabianHueni/UNIBE_MSGAI_HS25_Project

feat: enhance download of csv after experiment completion

feat: added id to boolq dataset

fix: re-add reload dataset method to scheduler

chore: added comments

feat: added stripping method for thinking tokens, increased maxNewTokens to 500, set enableThinking to false in onDeviceService

fix: made csv parsing more robust, changed prompt

feat: wait until all items of experiment are run to download stats, split downloads in stats and raw data

chore: uncomment not used models in selection

chore: renamed interarrival time method

download flickering (#1)

feat: add granit models and finish model selection

feat: add qwen 3 4B to cloud model selection

feat: add option to specify model quantization in on device model selection

Merge branch 'main' of https://github.com/fabianHueni/UNIBE_MSGAI_HS25_Project

feat: added a button to start experiments with 1000 requests

feat: add queues for on device and cloud processing to measure queue time

feat: track queueing time, inference time, and total latency separately. the timestamp for start inference should be taken somewhere else

fix: changed prompt

Merge branch 'main' of https://github.com/fabianHueni/UNIBE_MSGAI_HS25_Project

Merge branch 'main' of https://github.com/fabianHueni/UNIBE_MSGAI_HS25_Project

fix: changed the prompt to include the passage/context as well and moved the full prompt to scheduler.js file

feat: add exponential interarrival time

chore: removed unnecessary files