 Alex Cheema
|
6b1c8635fc
ensure exo dir on start, retry with exp backoff on file downloads
|
7 mesiacov pred |
|
Alex Cheema
|
e6b4f2993c
fix prompt output spacing in tui
|
7 mesiacov pred |
|
Alex Cheema
|
3675804f4d
throttle repo progress events and only send them out if something changed
|
7 mesiacov pred |
|
Alex Cheema
|
96f1aecb05
only in_progress if any given file is in_progress
|
7 mesiacov pred |
|
Alex Cheema
|
23a5030604
even if part of a file is downloaded it may not be in_progress
|
7 mesiacov pred |
|
Alex Cheema
|
31b56e862f
make a singleton thread pool executor for tinygrad since we always want it to run on the same thread
|
7 mesiacov pred |
|
Alex Cheema
|
9f6c688d62
update tinygrad
|
7 mesiacov pred |
|
Alex Cheema
|
4887be5103
parallelise model loading
|
7 mesiacov pred |
|
Alex Cheema
|
141de0d011
increase chatgpt api response timeout to 900 seconds
|
7 mesiacov pred |
 Alex Cheema
|
837ed5d980
Merge pull request #648 from exo-explore/modelasyncload
|
7 mesiacov pred |
|
Alex Cheema
|
9c1bea97e8
fix embed_tokens for last layer in qwen models
|
7 mesiacov pred |
|
Alex Cheema
|
af171f06fa
propagate prompts to other nodes so they can display them, cleaner prompt/output output
|
7 mesiacov pred |
|
Alex Cheema
|
edfa53a4c2
Merge pull request #646 from exo-explore/modelasyncload
|
7 mesiacov pred |
|
Alex Cheema
|
4a5b80a958
make sure mlx stuff is on separate thread non blocking
|
7 mesiacov pred |
|
Alex Cheema
|
92d1bc01de
Merge pull request #645 from exo-explore/modelasyncload
|
7 mesiacov pred |
|
Alex Cheema
|
6662d5668c
load mlx model shard on mlx thread so it doesnt block
|
7 mesiacov pred |
|
Alex Cheema
|
a0d673fa3a
Merge pull request #640 from exo-explore/simpledownload
|
7 mesiacov pred |
|
Alex Cheema
|
7c649085a1
fix eta/speed for resuming an existing download, using the session downloaded bytes
|
7 mesiacov pred |
|
Alex Cheema
|
90e0e2761f
ignore not_started progress updates
|
7 mesiacov pred |
|
Alex Cheema
|
265586f7b4
set timeout on get too
|
7 mesiacov pred |
|
Alex Cheema
|
4748bb7dc7
increase file download timeout to 30min
|
7 mesiacov pred |
|
Alex Cheema
|
ae770db4f3
increase download chunks to 1MB
|
7 mesiacov pred |
|
Alex Cheema
|
82f75d0ccf
increase hf download http timeout 15 mins for large downloads
|
7 mesiacov pred |
|
Alex Cheema
|
295f41c5cc
increase bench job timeout to give enough time to download
|
7 mesiacov pred |
|
Alex Cheema
|
19a27c5bfd
HF_HOME -> EXO_HOME
|
7 mesiacov pred |
|
Alex Cheema
|
d7ca9b7732
show each node id in the tinychat topology viz
|
7 mesiacov pred |
|
Alex Cheema
|
b349e48b0d
fix visual bug where frontend would show the full hf repo size, but in some cases that includes redundant files so we should use the model index in those cases too
|
7 mesiacov pred |
|
Alex Cheema
|
21586063f6
use llama-3.2-1b in tinygrad test
|
7 mesiacov pred |
|
Alex Cheema
|
277d63d860
special case when a model doesnt have a model index file, then use wildcard for allow_patterns
|
7 mesiacov pred |
|
Alex Cheema
|
74379ef671
log download logs with DEBUG>=6 very verbose
|
7 mesiacov pred |
