Dongfu's workspace
Runs
156
State
Notes
User
Tags
Created
Runtime
Sweep
_name_or_path
accelerator_config.even_batches
accelerator_config.split_batches
accelerator_config.use_seedable_sampler
adafactor
adam_beta1
adam_beta2
adam_epsilon
add_cross_attention
architectures
auto_find_batch_size
bf16
bf16_full_eval
chunk_size_feed_forward
dataloader_drop_last
dataloader_num_workers
dataloader_persistent_workers
dataloader_pin_memory
ddp_timeout
debug
disable_tqdm
diversity_penalty
do_eval
do_predict
do_sample
do_train
early_stopping
encoder_no_repeat_ngram_size
eval_delay
eval_steps
evaluation_strategy
fp16
fp16_backend
fp16_full_eval
fp16_opt_level
fsdp
fsdp_config.min_num_params
fsdp_config.xla
fsdp_config.xla_fsdp_grad_ckpt
fsdp_config.xla_fsdp_v2
fsdp_min_num_params
full_determinism
gradient_accumulation_steps
gradient_checkpointing
Finished
-
dongfu
19h 4m 45s
-
MFuyu/llava_siglip_llama3_8b_pretrain_8192
true
false
true
false
0.9
0.999
1.0000e-8
false
["LlavaForConditionalGeneration"]
false
true
false
0
false
64
false
true
1800
[]
false
0
false
false
false
true
false
0
0
500
no
false
auto
false
O1
[]
0
false
false
false
0
false
8
true
Crashed
-
dongfu
1d 7h 22m 27s
-
MFuyu/llava_siglip_llama3_8b_pretrain_8192
true
false
true
false
0.9
0.999
1.0000e-8
false
["LlavaForConditionalGeneration"]
false
true
false
0
false
64
false
true
1800
[]
false
0
false
false
false
true
false
0
0
500
no
false
auto
false
O1
[]
0
false
false
false
0
false
8
true
1-2
of 2