 Alex Cheema
|
d2ed4c2a16
disable tinygrad infernece engine test waiting Waiting on https://github.com/tinygrad/tinygrad/issues/5549
|
1 éve |
|
Alex Cheema
|
115aab0d6c
cache tinygrad models in github workflow
|
1 éve |
|
Alex Cheema
|
a4cc667754
async model downloading fixes #30
|
1 éve |
|
Alex Cheema
|
e49924e1b9
add chatgpt-api-response-timeout-secs flag, set this to 20 mins in test
|
1 éve |
|
Alex Cheema
|
7dd7ccab91
do one request to load the model then another to check the response
|
1 éve |
|
Alex Cheema
|
144af10686
separate discovery and chatgpt api integration test
|
1 éve |
|
Alex Cheema
|
93df43d07f
redundant sh
|
1 éve |
|
Alex Cheema
|
bf7aa51ba1
rename to discovery integration test as thats all it checks
|
1 éve |
|
Alex Cheema
|
b9a2c0f730
fix tests
|
1 éve |
|
Alex Cheema
|
d9516d2eb4
insstall in workflow
|
1 éve |
|
Alex Cheema
|
8efd656303
set different api ports so they dont conlict
|
1 éve |
|
Alex Cheema
|
8dd17fe00b
integration test with discovery
|
1 éve |
|
Alex Cheema
|
4d962ffcaa
fix hardcoded path in debug_inference_engine
|
1 éve |
|
Alex Cheema
|
30ab126c08
fix test_inference_engine
|
1 éve |
|
Alex Cheema
|
56e5e34ec9
fix invalid escape sequence exo_text
|
1 éve |
|
Alex Cheema
|
62a2407352
github workflow: use python3 consistently
|
1 éve |
|
Alex Cheema
|
ba1916a388
github workflow for tests
|
1 éve |
|
Alex Cheema
|
10a043773f
check for the last file that downloads in case it fails part way through
|
1 éve |
|
Alex Cheema
|
1475c735c9
fix inference_state serialization. related: #40 #44 #45
|
1 éve |
|
Alex Cheema
|
e18549e9e1
rm print
|
1 éve |
|
Alex Cheema
|
0c5a927f9a
spacing in viz
|
1 éve |
|
Alex Cheema
|
9fa0cb1ab1
add gpu poor/rich bar in panel. fixes #33
|
1 éve |
|
Alex Cheema
|
5b8f127bf4
fix opaque broadcast
|
1 éve |
|
Alex Cheema
|
a342e1abd8
add web url and chatgpt api endpoint to panel (fixes #43), fix a rounding error in the partition to shard mapping implementation
|
1 éve |
|
Alex Cheema
|
8939f888eb
remove spammy log
|
1 éve |
|
Alex Cheema
|
d9484906a3
remove the spammy logs
|
1 éve |
|
Alex Cheema
|
dd09c59719
fix issues with chatgpt api where it would generate too long output. avoid nonlocal
|
1 éve |
|
Alex Cheema
|
4b592f9d45
exo topology visualisation that shows the topology of the network, device capabilities and the currently active node using opaque statuses. fixes #36. ready for #33
|
1 éve |
|
Alex Cheema
|
35177690bd
by default find an ephemeral node port fixes #35, more robust topology updates. both fix #15 and #14
|
1 éve |
|
Alex Cheema
|
54c98607ef
more robust grpc discovery with asyncio and proper error handling, add flops to device capabilities. fixes #23 and progress on #33
|
1 éve |
