Commit History

revert to best working version
86518f0

emar commited on

forget import
981a527

emar commited on

attempting some control
39596d4

emar commited on

revert to working simplicity
64b21dd

emar commited on

testing sliders
b594447

emar commited on

reverting ifaces
eebc9a6

emar commited on

added sliders
eddf0b4

emar commited on

added sliders
5561bd8

emar commited on

ee
67b46b4

emar commited on

return to bge
d829f59

emar commited on

fixed omitted import
b7463ad

emar commited on

added reranker
b69cdc9

emar commited on

trying ada embeddings
58442f5

emar commited on

updated message trying different context
66c3bb5

emar commited on

yaay
bb34f7d

emar commited on

attempt meta llama
9ac9011

emar commited on

edit
da93e12

emar commited on

reduce to simplicity?
6c79873

emar commited on

test 2
4b52a00

emar commited on

some modifications to hopefully fix gradio errors
25a6760

emar commited on

test 2
4b94278

emar commited on

attempting longer context time
fa8e75c

emar commited on

more attempt
d90a012

emar commited on

test
36ad3c7

emar commited on

test
f3b743f

emar commited on

back to shitty cpu for now
b0752bd

emar commited on

test
09070fb

emar commited on

tweak
e0e7586

emar commited on

test zero2
3216f02

emar commited on

test zero 2
5dd7727

emar commited on

backup
d6fc481

emar commited on

changed requirements
247a7ac

emar commited on

trying to use chatgpt version
935059d

emar commited on

trying for zero space
4f71b75

emar commited on

attempt 2
fc62077

emar commited on

hard code flash-attention
04e4814

emar commited on

attempting phi3mini again
7ca1231

emar commited on

updated text
92e8cd5

emar commited on

switching back to the shitty version
f7c2d40

emar commited on

trying to see if llama3 will load in the cheapo tier
b71f192

emar commited on

trying to see if qwen fits.
7b07f28

emar commited on

reverted to tinyllama while I think about finding a better demo piece
ca8e8ef

emar commited on

re-enabled quantz
84d9622

emar commited on

retrying llama3
93dcada

emar commited on

removed quantization since this is a free cpu
99f0ceb

emar commited on

switched to a smaller and more open model
685c250

emar commited on

added log in
b0b3ade

emar commited on

flashattn being mean
134e284

emar commited on

flashattn update
6852d6c

emar commited on

oops torch
aabb3de

emar commited on