 Alex Cheema
|
ca64456260
fix broken links in README
|
7 months ago |
 Alex Cheema
|
87e08f89f1
Merge pull request #203 from exo-explore/non_blocking
|
8 months ago |
|
Alex Cheema
|
874886abc4
simplify mlx non blocking
|
8 months ago |
|
Alex Cheema
|
e616d4e86b
run realize on the result in tinygrad
|
8 months ago |
|
Alex Cheema
|
9345684b38
closely match prev impl mlx non blocking
|
8 months ago |
|
Alex Cheema
|
d6e661fd69
match previous impl with np.array in mlx
|
8 months ago |
|
Alex Cheema
|
caf9b57a2a
trigger ci
|
8 months ago |
|
Alex Cheema
|
84187113de
add a test for hf get_weight_map
|
8 months ago |
|
Alex Cheema
|
4ec613d4e8
simplify tinygrad non blocking
|
8 months ago |
|
Alex Cheema
|
6342384df4
Merge branch 'main' into non_blocking
|
8 months ago |
|
Alex Cheema
|
a1a0ffac55
add tinychat option for llama-3.1-70b-bf16
|
8 months ago |
|
Alex Cheema
|
de19f0ab42
Merge branch 'main' into non_blocking
|
8 months ago |
|
Alex Cheema
|
2948a83448
add llama-3.1-70b-bf16 model option
|
8 months ago |
|
Alex Cheema
|
11dd952d26
use set for shard specific patterns
|
8 months ago |
|
Alex Cheema
|
ea3322dea4
remove comment
|
8 months ago |
|
Alex Cheema
|
e0fda94d20
use sets for shard specific patterns
|
8 months ago |
|
Alex Cheema
|
b239c8a6d0
Merge branch 'main' into non_blocking
|
8 months ago |
|
Alex Cheema
|
8f65e1e697
fix weight_map resolution. previously we were always defaulting to allow pattern *.safetensors
|
8 months ago |
|
Alex Cheema
|
6881722b72
simplify non-blocking mlx inference
|
8 months ago |
|
Alex Cheema
|
9db16f8dca
use a queue for non-blocking mlx inference
|
8 months ago |
|
Alex Cheema
|
0ca5c26094
run mlx inference engine on a single thread too
|
8 months ago |
|
Alex Cheema
|
58f535d0b0
formatting
|
8 months ago |
|
Alex Cheema
|
2950373d36
experiment with tinygrad on its own thread, so it doesnt block event loop
|
8 months ago |
|
Alex Cheema
|
41f0a22e76
DEBUG>=8 for SendOpaqueStatus logs
|
8 months ago |
|
Alex Cheema
|
01cc6a4c9d
fix Mistral-Large special case when we pass in a path
|
8 months ago |
|
Alex Cheema
|
41dd700ff4
less aggressive logs for opaque status / download progress. too much spam
|
8 months ago |
|
Alex Cheema
|
35aba75be6
Merge pull request #194 from exo-explore/better_networking
|
8 months ago |
|
Alex Cheema
|
4537d614aa
circleci use tee to output logs in realtime as well as capture them
|
8 months ago |
|
Alex Cheema
|
56c1bf9a95
consistent remove _secs / -secs suffix
|
8 months ago |
|
Alex Cheema
|
f342cdcae9
get rid of -secs suffix
|
8 months ago |
