Commit History

Unloading models + current evals
05b3b2d

Yeyito_gpu commited on

merge changes
22c2b9c

Yeyito commited on

Current evals and queue
c2639d2

Yeyito commited on

reverted again
1eac270

Yeyito commited on

nvm text doesn't work like that
e41a7dd

Yeyito commited on

winogrande warnings
6e3353a

Yeyito commited on

fixed issue with bagel model
5866f1e

Yeyito commited on

new evals
818c1fc

Yeyito commited on

new evals
b9b5c7f

Yeyito commited on

Current queue
233fcbe

Yeyito commited on

current evals
2f2a7f2

Yeyito commited on

current evals
5538403

Yeyito commited on

Error with Yi fix
f2b6c4c

Yeyito commited on

current eval and queue
5aa9ac8

Yeyito commited on

current queue
fc731aa

Yeyito commited on

current eval
91ac6f1

Yeyito commited on

Added How does this work? To the about
e87b0e5

Yeyito commited on

Fixing lack of ref_model
3c1fdd7

Yeyito commited on

reverted tokenizer fix, will be solved manually upon testing, current evals
5213c6c

Yeyito commited on

current eval
39963b1

Yeyito commited on

forgot to remove an imort
f915caf

Yeyito commited on

Load tokenizer from parent model & app.py fixes
ca453e8

Yeyito commited on

Mistral cache
98d650b

Yeyito_gpu commited on

ref model and tuple error fix
5e2e1fb

Yeyito commited on

Current queue
d74f5ae

Yeyito commited on

Reference model full implementation
9259578

Yeyito commited on

Reference model selection box
3163ba6

Yeyito commited on

mistralai default ref, queue hotfix
9bf390d

Yeyito commited on

small python fix
f27ee6f

Yeyito commited on

gsm8k fix, queue, ref_model column
4e4454b

Yeyito commited on

current evals num_z=50
ca8085c

Yeyito commited on

re0commented evaluations
d0fa4b0

Yeyito commited on

new evals num_z=50
6ebbf31

Yeyitogpu commited on

fixed tuple oversight
b4d145e

Yeyitogpu commited on

current evals with num_z=50
56589b2

Yeyito commited on

current eval
00a99c6

Yeyitogpu commited on

We use the same code in here to test the models outside of this space
20ed3ea

Yeyito commited on

Current evaluations (num_z = 50)
0c9031b

Yeyito commited on

num_z=50 saves and current evaluations
95f8f54

Yeyito_Cloud commited on

No more restarting space, fixed format
807834f

Yeyito commited on

num_z = 50 for all
28487fe

Yeyito commited on

It was and not or πŸ™‚
f6d7d3d

Yeyito commited on

More aggressive OOM prevention, num_z reduced to 50 for gsm8k and mmlu
98282cd

Yeyito commited on

Trying to prevent OOM on mmlu and gsm8k by halving seq len
a0bf640

Yeyito commited on

Meant to do last commit on run.py not app.py
b56563d

Yeyito commited on

Trying to stop OOMs on MMLU and GSM8K by halving seq len
c13858c

Yeyito commited on

Update .gitignore
4996c9a

Yeyito commited on

Avoiding re-loading already loaded models. Stated unload functionality as not-implemented.
b28ad14

Yeyito commited on

Upload 16 files
2a135fe

Yeyito commited on

Caching model1 responses for 2x speedup
5b3849a

Yeyito commited on