Commit History

:recycle: take refering url from the gradio environment
a18eddc
Running

wxgeorge commited on

:sparkles: updating model list.
ac7e364

wxgeorge commited on

:beatle: correct twitter handle.
f1ecb17

wxgeorge commited on

:wrench: drop reflection. add Nemotron. make default model.
f02037a

wxgeorge commited on

:see_no_evil: hide unused logos.
75d7eaa

wxgeorge commited on

:beetle: adding missing logo.
f903097

wxgeorge commited on

:pencil: correct twitter handle!
01edef7

wxgeorge commited on

:lock: don't accept inference requests for models not on the list
a9b1f7f

wxgeorge commited on

:wrench: revert to a different model each day.
674f62d

wxgeorge commited on

:see_no_evil: hide unused functions to avoid cluttering api pane.
554cf75

wxgeorge commited on

:lipstick: make logo bigger and more prominent, conclude with some calls to action.
fcd14c4

wxgeorge commited on

:recycle: refactor larger model whitelisting.
3fa9161

wxgeorge commited on

:sparkles: bring back l3.1-8b models.
c7ff178

wxgeorge commited on

:wrench: put README content in the right place for easier recreation.
3793467

wxgeorge commited on

:sparkles: update model list.
6c352bd

wxgeorge commited on

:truck: drop "working" from model cache script name.
ceeab78

wxgeorge commited on

annotate qwen 2
306918a

wxgeorge commited on

:sparkles: include Qwen2.5-72B
0a89ae4

wxgeorge commited on

:goal_net: fail to start if API key is missing.
7f61b1d

wxgeorge commited on

:pencil2: updating README
bdde565

wxgeorge commited on

:wrench: updating model list
5e8c4b7

wxgeorge commited on

:wrench: apply reflection system prompt only to Reflection 70B
bd9ae66

wxgeorge commited on

Fix html output
1ce20f1
verified

m8than commited on

:fire: revert manual chat templating for reflection now that it's working in featherless backend.
6f983da

wxgeorge commited on

Update app.py
2433428
verified

m8than commited on

Update app.py
7719c51
verified

m8than commited on

Update app.py
8c568be
verified

m8than commited on

Changed the concurrency limit.
b4b16a1
verified

m8than commited on

:pencil2: lead copy tuning
8abfddb

wxgeorge commited on

:sparkles: add Reflection-Llama to the annotations.
43df791

wxgeorge commited on

:heavy_plus_sign: I really only want transformers but just adding it seems to break HF?
ae83cd8

wxgeorge commited on

:poop: cheesy "de"chatformatization of response.
4c36b18

wxgeorge commited on

:sparkles: support mattshumer's Reflection
30bad6e

wxgeorge commited on

:sparkels: add button to facilitate returning to model card.
68492c3

wxgeorge commited on

:lipstick: keep chat interface filling the screen
34e11d5

wxgeorge commited on

:chart_with_upwards_trend: associate app attribution with inference request.
77ee232

wxgeorge commited on

:sparkles: update model cache constructor to include all models
988b5a0

wxgeorge commited on

:rocket: update model list
0810fbd

wxgeorge commited on

:wrench: update README to annotate only smaller models.
f3dd871

wxgeorge commited on

:wrench: update model listing to avoid unintentionally listing larger models.
2018dd8

wxgeorge commited on

:sparkles: make initial model choice change day over day.
a1c24d9

wxgeorge commited on

:sparkles: model list update.
7302e17

wxgeorge commited on

:sparkles: updating model list.
cc133a4

wxgeorge commited on

:wrench: include unhealthy models in model cache as we expect this state to be transient.
8f28494

wxgeorge commited on

:heavy_plus_sign: revving model list.
ae4c273

wxgeorge commited on

:heavy_plus_sign: update list of supported models.
2c5c7ab

wxgeorge commited on

:wrench: ensure model-cache is in sorted order for easier git inspection.
ccf3a62

wxgeorge commited on

:truck: move model-cache.json back to root ... needed for space.
77d9d41

wxgeorge commited on

:sparkles: add annotations for missing llama2-solar-10b7 models
2d50157

wxgeorge commited on

:sparkles: script to auto-generate model annotations.
caba5f3

wxgeorge commited on