Akhauriyash's workspace
Runs
9
State
Notes
User
Tags
Created
Runtime
Sweep
checkpoint.continue_training_from_init
checkpoint.dump.every
checkpoint.dump.keep
checkpoint.eval.every
checkpoint.eval.keep
checkpoint.path
data.add_bos
data.add_eos
data.batch_size
data.load_async
data.n_views
data.prefetch_size
data.root_dir
data.seed
data.seq_len
data.sources.dclm_baseline_1.0_shuffled
data.tokenizer.name
data.tokenizer.path
distributed.compile
distributed.compile_cache_size_limit
distributed.detect_anomaly
distributed.dp_replicate
distributed.dp_shard
distributed.float8_filter
distributed.fsdp_type
distributed.matmul_allow_tf32
distributed.model_dtype
distributed.selective_activation_checkpointing
distributed.spawn_method
distributed.tp_size
dump_dir
env.ENABLE_INTRA_NODE_COMM
env.MKL_NUM_THREADS
env.MKL_SERVICE_FORCE_INTEL
env.NCCL_DEBUG
env.NCCL_IB_TIMEOUT
env.OMP_NUM_THREADS
env.TORCH_NCCL_ASYNC_ERROR_HANDLING
env.TORCH_NCCL_AVOID_RECORD_STREAMS
eval.generator.dtype
eval.generator.max_tokens
eval.harness.tasks
gc_collect_freq
grad_acc_steps
Finished
-
akhauriyash
8h 57m 40s
-
false
10000
3
5000
-1
/scratch/ya255/lingua/long_runs/isoxmer_long/checkpoints
true
true
80
true
2
512
/scratch/ya255/lingua/setup/data/
42
1024
100
sp
setup/l2tokenizer/tokenizer.model
true
8
false
1
1
layers\.[0-9]+\.
full_shard
false
bf16
false
forkserver
1
/scratch/ya255/lingua/long_runs/isoxmer_long
1
1
GNU
INFO
22
1
1
1
bf16
1024
["wikitext"]
1000
1
Finished
-
akhauriyash
10h 31m 12s
-
false
10000
3
5000
-1
/scratch/ya255/lingua/long_runs/minGRU/checkpoints
true
true
80
true
2
512
/scratch/ya255/lingua/setup/data/
42
1024
100
sp
setup/l2tokenizer/tokenizer.model
true
8
false
1
1
layers\.[0-9]+\.
full_shard
false
bf16
false
forkserver
1
/scratch/ya255/lingua/long_runs/minGRU
1
1
GNU
INFO
22
1
1
1
bf16
1024
["wikitext"]
1000
1
Finished
-
akhauriyash
12h 7m 19s
-
false
10000
3
5000
-1
/scratch/ya255/lingua/long_runs/hawk/checkpoints
true
true
80
true
2
512
/scratch/ya255/lingua/setup/data/
42
1024
100
sp
setup/l2tokenizer/tokenizer.model
true
8
false
1
1
layers\.[0-9]+\.
full_shard
false
bf16
false
forkserver
1
/scratch/ya255/lingua/long_runs/hawk
1
1
GNU
INFO
22
1
1
1
bf16
1024
["wikitext"]
1000
1
Finished
-
akhauriyash
19h 35m 23s
-
false
10000
3
5000
-1
/scratch/ya255/lingua/long_runs/mamba_longrun/checkpoints
true
true
80
true
2
512
/scratch/ya255/lingua/setup/data/
42
1024
100
sp
setup/l2tokenizer/tokenizer.model
true
8
false
1
1
layers\.[0-9]+\.
full_shard
false
bf16
false
forkserver
1
/scratch/ya255/lingua/long_runs/mamba_longrun
1
1
GNU
INFO
22
1
1
1
bf16
1024
["wikitext"]
1000
1
Finished
-
akhauriyash
19h 31m 6s
-
false
10000
3
5000
-1
/scratch/ya255/lingua/long_runs/AttentiveSSMNoProjCyc4L32_Long/checkpoints
true
true
80
true
2
512
/scratch/ya255/lingua/setup/data/
42
1024
100
sp
setup/l2tokenizer/tokenizer.model
true
8
false
1
1
layers\.[0-9]+\.
full_shard
false
bf16
false
forkserver
1
/scratch/ya255/lingua/long_runs/AttentiveSSMNoProjCyc4L32_Long
1
1
GNU
INFO
22
1
1
1
bf16
1024
["wikitext"]
1000
1
Finished
-
akhauriyash
1d 1h 5m 43s
-
false
10000
3
5000
-1
/scratch/ya255/lingua/long_runs/mamba_long_v2/checkpoints
true
true
80
true
2
512
/scratch/ya255/lingua/setup/data/
42
1024
100
sp
setup/l2tokenizer/tokenizer.model
true
8
false
1
1
layers\.[0-9]+\.
full_shard
false
bf16
false
forkserver
1
/scratch/ya255/lingua/long_runs/mamba_long_v2
1
1
GNU
INFO
22
1
1
1
bf16
1024
["wikitext"]
1000
1
Finished
-
akhauriyash
11h 57m 4s
-
false
10000
3
5000
-1
/scratch/ya255/lingua/long_runs/mamba_long/checkpoints
true
true
80
true
2
512
/scratch/ya255/lingua/setup/data/
42
1024
100
sp
setup/l2tokenizer/tokenizer.model
true
8
false
1
1
layers\.[0-9]+\.
full_shard
false
bf16
false
forkserver
1
/scratch/ya255/lingua/long_runs/mamba_long
1
1
GNU
INFO
22
1
1
1
bf16
1024
["wikitext"]
1000
1
Finished
-
akhauriyash
10h 29s
-
false
10000
3
5000
-1
/scratch/ya255/lingua/long_runs/xmer_long/checkpoints
true
true
80
true
2
512
/scratch/ya255/lingua/setup/data/
42
1024
100
sp
setup/l2tokenizer/tokenizer.model
true
8
false
1
1
layers\.[0-9]+\.
full_shard
false
bf16
false
forkserver
1
/scratch/ya255/lingua/long_runs/xmer_long
1
1
GNU
INFO
22
1
1
1
bf16
1024
["wikitext"]
1000
1
Finished
-
akhauriyash
19h 25m 52s
-
false
10000
3
5000
-1
/scratch/ya255/lingua/long_runs/AttentiveSSMNoProjCyc8L16_Long/checkpoints
true
true
80
true
2
512
/scratch/ya255/lingua/setup/data/
42
1024
100
sp
setup/l2tokenizer/tokenizer.model
true
8
false
1
1
layers\.[0-9]+\.
full_shard
false
bf16
false
forkserver
1
/scratch/ya255/lingua/long_runs/AttentiveSSMNoProjCyc8L16_Long
1
1
GNU
INFO
22
1
1
1
bf16
1024
["wikitext"]
1000
1
1-9
of 9