unparallelled commited on
Commit
8ab8623
·
verified ·
1 Parent(s): 30fcbeb

Upload folder using huggingface_hub

Browse files
logs/l3long_run17/janitor.log ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [15:29:53] freed 928 MB, 45G avail
2
+ [15:44:54] freed 928 MB, 54G avail
3
+ [15:49:54] freed 928 MB, 57G avail
4
+ [15:54:54] freed 928 MB, 57G avail
5
+ [15:59:54] freed 928 MB, 57G avail
6
+ [16:04:54] freed 928 MB, 57G avail
7
+ [16:09:55] freed 928 MB, 57G avail
8
+ [16:14:55] freed 928 MB, 57G avail
9
+ [16:19:55] freed 928 MB, 57G avail
10
+ [16:24:55] freed 928 MB, 57G avail
11
+ [16:29:55] freed 928 MB, 57G avail
12
+ [16:34:55] freed 928 MB, 57G avail
13
+ [16:39:56] freed 928 MB, 57G avail
14
+ [16:44:56] freed 928 MB, 41G avail
15
+ [16:54:56] freed 928 MB, 41G avail
16
+ [17:09:56] freed 928 MB, 33G avail
17
+ [17:14:57] freed 928 MB, 36G avail
18
+ [17:19:57] freed 928 MB, 38G avail
19
+ [17:24:57] freed 928 MB, 41G avail
20
+ [17:29:57] freed 928 MB, 41G avail
21
+ [17:34:57] freed 928 MB, 41G avail
22
+ [17:39:57] freed 928 MB, 41G avail
23
+ [17:44:58] freed 928 MB, 41G avail
24
+ [17:49:58] freed 928 MB, 41G avail
25
+ [17:54:58] freed 928 MB, 41G avail
26
+ [17:59:58] freed 928 MB, 41G avail
27
+ [18:04:58] freed 928 MB, 41G avail
28
+ [18:09:59] freed 928 MB, 41G avail
29
+ [18:14:59] freed 928 MB, 41G avail
30
+ [18:19:59] freed 928 MB, 40G avail
logs/l3long_run17/l3long.log ADDED
The diff for this file is too large to render. See raw diff
 
logs/l3long_run17/master.log ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [06:54:49] waiting for the running b3 experiment to release the GPUs
2
+ [07:06:18] waiting for the running b3 experiment to release the GPUs
3
+ [09:43:20] === long b3 arm: START ===
4
+ [09:43:20] parent run7' (same as l3@400, so l3@400 vs this isolates STEP COUNT)
5
+ [09:43:20] steps=4000 stable=4000 -> decay_start=4100 (> 4000, so NO decay: flat 3.0e-4)
6
+ [09:43:20] ckpt every 25 steps, keep 2 (= 34.56 GB), free 83G
7
+ [09:43:20] tokens 2097152000 = 70.0% of b3's 2.93 B
8
+ [09:43:22] trainer exited rc=1
9
+ [09:43:22] run number: unknown
10
+ [09:43:22] LR trace: at step 100
11
+ [09:43:22] at step 2000 (must equal step 100)
12
+ [09:43:22] at the last logged step
13
+ [09:43:22] to continue: ./b3_long_run.sh resume <run>
14
+ [09:46:00] === long b3 arm: START ===
15
+ [09:46:00] parent run7' (same as l3@400, so l3@400 vs this isolates STEP COUNT)
16
+ [09:46:00] steps=4000 stable=3800 -> decay_start=3900; flat 3.0e-4 for steps 100..3899, then 100-step tail
17
+ [09:46:01] ckpt every 25 steps, keep 2 (= 34.56 GB), free 83G
18
+ [09:46:01] tokens 2097152000 = 70.0% of b3's 2.93 B
19
+ [13:17:43] RESUMING run 17 (anchor pinned to 0 so warmup does not restart)
20
+ [13:17:43] steps=4000 stable=3800 -> decay_start=3900; flat 3.0e-4 for steps 100..3899, then 100-step tail
21
+ [13:17:43] ckpt every 25 steps, keep 2 (= 34.56 GB), free 2G
22
+ [13:17:43] tokens 2097152000 = 70.0% of b3's 2.93 B
23
+ [16:49:01] RESUMING run 17 (anchor pinned to 0 so warmup does not restart)
24
+ [16:49:01] steps=4000 stable=3800 -> decay_start=3900; flat 3.0e-4 for steps 100..3899, then 100-step tail
25
+ [16:49:01] ckpt every 25 steps, keep 2 (= 34.56 GB), free 41G
26
+ [16:49:01] tokens 2097152000 = 70.0% of b3's 2.93 B
logs/l3long_run17/runnum ADDED
@@ -0,0 +1 @@
 
 
1
+ 17