Shanmugamr's workspace
Runs
9
Name
2 visualized
State
Notes
User
Tags
Created
Runtime
Sweep
checkpointing.checkpoint_dir
checkpointing.enabled
checkpointing.higher_is_better
checkpointing.keep_top_k
checkpointing.metric_name
checkpointing.model_save_format
checkpointing.save_consolidated
checkpointing.save_period
cluster.gpus_per_node
cluster.num_nodes
data.dataset_name
data.max_input_seq_length
data.num_workers
data.prompt_file
data.shuffle
env.dapo.math_verify_impl
env.dapo.num_workers
env.math.math_verify_impl
env.math.num_workers
grpo.async_grpo.enabled
grpo.async_grpo.in_flight_weight_updates
grpo.async_grpo.max_trajectory_age_steps
grpo.async_grpo.recompute_kv_cache_after_weight_updates
grpo.batch_multiplier
grpo.dynamic_sampling_max_gen_batches
grpo.max_num_epochs
grpo.max_num_steps
grpo.max_rollout_turns
grpo.max_val_samples
grpo.normalize_rewards
grpo.num_generations_per_prompt
grpo.num_prompts_per_step
grpo.overlong_filtering
grpo.reward_scaling.enabled
grpo.reward_scaling.source_max
grpo.reward_scaling.source_min
grpo.reward_scaling.target_max
grpo.reward_scaling.target_min
grpo.reward_shaping.enabled
grpo.reward_shaping.max_response_length
grpo.reward_shaping.overlong_buffer_length
grpo.reward_shaping.overlong_buffer_penalty
grpo.seed
grpo.use_dynamic_sampling
Failed
shanmugamr
3h 36m 16s
-
results/grpo_megatron
false
true
3
val:accuracy
safetensors
false
10
1
1
OpenMathInstruct-2
512
1
examples/prompts/cot.txt
true
dapo_math_verify
8
hf_math_verify
8
false
false
1
false
1
10
1
500
1
256
true
16
32
false
false
1
0
1
0
false
512
128
1
42
false
Crashed
shanmugamr
3h 26m 16s
-
results/grpo_megatron
false
true
3
val:accuracy
safetensors
false
10
1
1
OpenMathInstruct-2
512
1
examples/prompts/cot.txt
true
dapo_math_verify
8
hf_math_verify
8
false
false
1
false
1
10
1
500
1
256
true
16
32
false
false
1
0
1
0
false
512
128
1
42
false
Crashed
shanmugamr
2h 49m 30s
-
results/grpo_megatron
false
true
3
val:accuracy
safetensors
false
10
1
1
OpenMathInstruct-2
512
1
examples/prompts/cot.txt
true
dapo_math_verify
8
hf_math_verify
8
false
false
1
false
1
10
1
500
1
256
true
16
32
false
false
1
0
1
0
false
512
128
1
42
false
Failed
shanmugamr
2h 36m 51s
-
results/grpo_megatron
false
true
3
val:accuracy
safetensors
false
10
1
1
OpenMathInstruct-2
512
1
examples/prompts/cot.txt
true
dapo_math_verify
8
hf_math_verify
8
false
false
1
false
1
10
1
500
1
256
true
16
32
false
false
1
0
1
0
false
512
128
1
42
false
Crashed
shanmugamr
1h 24m 1s
-
results/grpo_megatron
false
true
3
val:accuracy
safetensors
false
10
1
1
OpenMathInstruct-2
512
1
examples/prompts/cot.txt
true
dapo_math_verify
8
hf_math_verify
8
false
false
1
false
1
10
1
500
1
256
true
16
32
false
false
1
0
1
0
false
512
128
1
42
false
Failed
shanmugamr
2h 35m 57s
-
results/grpo_megatron
false
true
3
val:accuracy
safetensors
false
10
1
1
OpenMathInstruct-2
512
1
examples/prompts/cot.txt
true
dapo_math_verify
8
hf_math_verify
8
false
false
1
false
1
10
1
500
1
256
true
16
32
false
false
1
0
1
0
false
512
128
1
42
false
Crashed
shanmugamr
3h 58m 46s
-
results/grpo_megatron
false
true
3
val:accuracy
safetensors
false
10
1
1
OpenMathInstruct-2
512
1
examples/prompts/cot.txt
true
dapo_math_verify
8
hf_math_verify
8
false
false
1
false
1
10
1
500
1
256
true
16
32
false
false
1
0
1
0
false
512
128
1
42
false
Crashed
shanmugamr
3h 56m 46s
-
results/grpo_megatron
false
true
3
val:accuracy
safetensors
false
10
1
1
OpenMathInstruct-2
512
1
examples/prompts/cot.txt
true
dapo_math_verify
8
hf_math_verify
8
false
false
1
false
1
10
1
500
1
256
true
16
32
false
false
1
0
1
0
false
512
128
1
42
false
Crashed
shanmugamr
2h 21m 15s
-
results/grpo_megatron
false
true
3
val:accuracy
safetensors
false
10
1
1
OpenMathInstruct-2
512
1
examples/prompts/cot.txt
true
dapo_math_verify
8
hf_math_verify
8
false
false
1
false
1
10
1
500
1
256
true
16
32
false
false
1
0
1
0
false
512
128
1
42
false
1-9
of 9