Listar2000's workspace
Runs
4
State
Notes
User
Tags
Created
Runtime
Sweep
actor_rollout_ref.actor._target_
actor_rollout_ref.actor.checkpoint._target_
actor_rollout_ref.actor.checkpoint.async_save
actor_rollout_ref.actor.checkpoint.load_contents
actor_rollout_ref.actor.checkpoint.save_contents
actor_rollout_ref.actor.clip_ratio
actor_rollout_ref.actor.clip_ratio_c
actor_rollout_ref.actor.clip_ratio_high
actor_rollout_ref.actor.clip_ratio_low
actor_rollout_ref.actor.entropy_checkpointing
actor_rollout_ref.actor.entropy_coeff
actor_rollout_ref.actor.entropy_from_logits_with_chunking
actor_rollout_ref.actor.freeze_vision_tower
actor_rollout_ref.actor.fsdp_config._target_
actor_rollout_ref.actor.fsdp_config.entropy_checkpointing
actor_rollout_ref.actor.fsdp_config.entropy_from_logits_with_chunking
actor_rollout_ref.actor.fsdp_config.forward_only
actor_rollout_ref.actor.fsdp_config.forward_prefetch
actor_rollout_ref.actor.fsdp_config.fsdp_size
actor_rollout_ref.actor.fsdp_config.model_dtype
actor_rollout_ref.actor.fsdp_config.offload_policy
actor_rollout_ref.actor.fsdp_config.optimizer_offload
actor_rollout_ref.actor.fsdp_config.param_offload
actor_rollout_ref.actor.fsdp_config.reshard_after_forward
actor_rollout_ref.actor.fsdp_config.strategy
actor_rollout_ref.actor.fsdp_config.ulysses_sequence_parallel_size
actor_rollout_ref.actor.fsdp_config.use_orig_params
actor_rollout_ref.actor.fsdp_config.use_torch_compile
actor_rollout_ref.actor.fsdp_config.wrap_policy.min_num_params
actor_rollout_ref.actor.grad_clip
actor_rollout_ref.actor.kl_loss_coef
actor_rollout_ref.actor.kl_loss_type
actor_rollout_ref.actor.loss_agg_mode
actor_rollout_ref.actor.optim._target_
actor_rollout_ref.actor.optim.betas
actor_rollout_ref.actor.optim.clip_grad
actor_rollout_ref.actor.optim.lr
actor_rollout_ref.actor.optim.lr_scheduler_type
actor_rollout_ref.actor.optim.lr_warmup_steps
actor_rollout_ref.actor.optim.lr_warmup_steps_ratio
actor_rollout_ref.actor.optim.min_lr_ratio
actor_rollout_ref.actor.optim.num_cycles
actor_rollout_ref.actor.optim.total_training_steps
actor_rollout_ref.actor.optim.weight_decay
Finished
listar2000
8m 45s
-
verl.workers.config.FSDPActorConfig
verl.trainer.config.CheckpointConfig
false
["model","optimizer","extra"]
["model","optimizer","extra"]
0.2
3
0.28
0.2
false
0
false
false
verl.workers.config.FSDPEngineConfig
false
false
false
false
-1
bfloat16
false
true
true
true
fsdp
1
false
true
0
1
0.001
low_var_kl
seq-mean-token-mean
verl.workers.config.FSDPOptimizerConfig
[0.9,0.999]
1
0.0001
constant
-1
0
0
0.5
3058300
0.01
Crashed
listar2000
4h 27m 1s
-
verl.workers.config.FSDPActorConfig
verl.trainer.config.CheckpointConfig
false
["model","optimizer","extra"]
["model","optimizer","extra"]
0.2
3
0.28
0.2
false
0
false
false
verl.workers.config.FSDPEngineConfig
false
false
false
false
-1
fp32
false
true
true
true
fsdp
1
false
true
0
1
0.001
low_var_kl
seq-mean-token-mean
verl.workers.config.FSDPOptimizerConfig
[0.9,0.999]
1
0.0001
constant
-1
0
0
0.5
3058300
0.01
Crashed
listar2000
3h 26m 45s
-
verl.workers.config.FSDPActorConfig
verl.trainer.config.CheckpointConfig
false
["model","optimizer","extra"]
["model","optimizer","extra"]
0.2
3
0.28
0.2
false
0
false
false
verl.workers.config.FSDPEngineConfig
false
false
false
false
-1
fp32
false
true
true
true
fsdp
1
false
true
0
1
0.001
low_var_kl
seq-mean-token-mean
verl.workers.config.FSDPOptimizerConfig
[0.9,0.999]
1
0.00005
constant
-1
0
0
0.5
3058300
0.01
Crashed
listar2000
5h 1m 31s
-
verl.workers.config.FSDPActorConfig
verl.trainer.config.CheckpointConfig
false
["model","optimizer","extra"]
["model","optimizer","extra"]
0.2
3
0.28
0.2
false
0
false
false
verl.workers.config.FSDPEngineConfig
false
false
false
false
-1
fp32
false
true
true
true
fsdp
1
false
true
0
1
0.001
low_var_kl
seq-mean-token-mean
verl.workers.config.FSDPOptimizerConfig
[0.9,0.999]
1
0.000001
constant
-1
0
0
0.5
3058300
0.01
1-4
of 4