Commit History

Changed 'large' model to Phi 3 Mini gguf 128k. Added requirements file for cpu. Put prompts in separate file.
232a079
Running

seanpedrickcase commited on

Attempted Gradio downgrade to 3.50.2 to resolve version conflicts
15476e8

seanpedrickcase commited on

CPU Flan inference is crashing, so trying to revert to previous package versions that worked
85b6613

seanpedrickcase commited on

Removed reference to ctransformers
5cdf399

seanpedrickcase commited on

Upgraded gradio and packages to latest. Switched Ctransformers with Llama.cpp Python
f301d67

seanpedrickcase commited on

Minor dependency updates, increased max allowable layers for GPU
8aa3ebb

Sean-Case commited on

Added temperature slider, more stringent checks for document relevance
6a76923

Sean-Case commited on

Added slider for changing number of passages out
d53332d

Sean-Case commited on

Improved prompting for csv/excel files
ee77123

Sean-Case commited on

put back pip install requirement for gradio 3.42.0 for likeData compatibility
a462256

Sean-Case commited on

Added csv/Excel file support
d4b0a2c

Sean-Case commited on

Put sources behind accordion, minor change to model params. Gitignore update
c2ff47a

Sean-Case commited on

Hid gpu layers behind accordion
7339026

Sean-Case commited on

Build fail when gpu_layers > 0, so will set to 0 at start and modify in app.
0a7a8db

Sean-Case commited on

Made GPU layer bar invisible. Changed default GPU layers when CUDA available
8249fd3

Sean-Case commited on

Upgraded large model to Mistral OpenOrca 7B Q4. More checks for empty questions.
994ad90

Sean-Case commited on

Changed embedding model, added reference to chat model on front page
275393f

Sean-Case commited on

Improved advanced model prompt, added stop generation button. context prompt improvements
d213c15

Sean-Case commited on

Stop generation button. Better model load. Trying one source that's longer.
d5a8385

Sean-Case commited on

Setting gpu_layer slider to false to prevent accidental misuse
9aef340

Sean-Case commited on

Removed bootstrapper.py as it may confuse HF
febdc08

Sean-Case commited on

gpu_layers should now update correctly. Added code for creating distribution.
114048b

Sean-Case commited on

Downgraded Gradio back to 3.42.0 in case that's an issue.
9795699

Sean-Case commited on

Hopefully fixed Orca load cuda error. Loads in Orca on app load. Later gradio version installed.
1365c48

Sean-Case commited on

Temporarily disabled Orca Mini load until Cuda compatibility sorted out
67af224

Sean-Case commited on

Explicitly installs required gradio version upon loading app
e78df68

Sean-Case commited on

Added working like buttons. Added model choice. Modified requirements txt
2e536f9

Sean-Case commited on

added bot icon with white background
84b25ff

Sean-Case commited on

Cleaned up code a bit, added user icons, thumbs up/down
aa0ad5d

Sean-Case commited on

Reduced chat box size further
0b0054b

Sean-Case commited on

Reduced chat box height
302ada4

Sean-Case commited on

Moved around submit and clear chat buttons
30689f9

Sean-Case commited on

Added support for Mistral Orca
e4df9f2

Sean-Case commited on

Attempt to switch to Orca Mini GGUF
d2ddc62

Sean-Case commited on

Tidied up intro text
f6036ad

Sean-Case commited on

Updated intro text
26e86cf

Sean-Case commited on

Disabled chat memory as I think it is not needed for most user queries (max_memory_length = 0)
ed48a70

Sean-Case commited on

Changed theme, example set now disappears upon new file load
71c040a

Sean-Case commited on

Adapted code to keep newly-loaded vectorstores within local user state
102df35

Sean-Case commited on

New file uploads no longer write a global variable so hopefully not shared between sessions
bc459f6

Sean-Case commited on

Added reference to source on app. Minor text changes.
9118536

Sean-Case commited on

Updated web ingest. Added some warnings to intro text
ae4a7ec

Sean-Case commited on

Added to intro, temp disabled web page load, address pdf load issue
41ed1b7

Sean-Case commited on

Remove ctransformers model load (not needed yet). Used faster keyword model
1ae7b34

Sean-Case commited on

Added nltk wordnet download to chatfuncs
9a3229c

Sean-Case commited on

Added keybert, span_marker, gensim to reqs
8fe4ba4

Sean-Case commited on

Added ctransformers to requirements.txt
2409e6a

Sean-Case commited on

Added GPT4all to requirements.txt
8881c9a

Sean-Case commited on

Changed docx requirement to python-docx for compatibility
a2212b7

Sean-Case commited on

Added docx to requirements
949ef5c

Sean-Case commited on