Skip to main content
ucalyptus
Projects
gsm_hard_code_syntax
Log in
Sign up
Overview
Workspace
Runs
Automat.
Sweeps
Reports
Artifacts
Ucalyptus's workspace
Personal workspace
Automated workspace
Changes are only visible to you.
Runs (18)
Name
(18 visualized)
/data/grpo-folder/weights/qwen-2.5-1.5b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-1.5b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-3b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-3b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-3b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-3b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-7b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-7b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-3b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-3b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/phi-4-14b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/phi-4-14b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-1b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-1b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-1b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-1b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-1b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-1b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-7b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-7b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-1.5b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-1.5b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/phi-4-14b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/phi-4-14b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.1-8b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.1-8b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-1.5b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-1.5b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-3b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-3b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-1b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.2-1b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-1.5b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/qwen-2.5-1.5b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.1-8b/train_gsm_hard.jsonl/
/data/grpo-folder/weights/llama-3.1-8b/train_gsm_hard.jsonl/
1-18
of 18
Add panels
profiling
4
1-4 of 4
train
12
1-6 of 12
train/rewards/xml_format_reward
train/rewards/xml_format_reward
Showing first 10 runs
50
100
150
200
250
train/global_step
0
0.2
0.4
0.6
0.8
1
train/rewards/code_syntax_reward
train/rewards/code_syntax_reward
Showing first 10 runs
50
100
150
200
250
train/global_step
0
0.1
0.2
0.3
0.4
0.5
0.6
train/reward_std
train/reward_std
Showing first 10 runs
50
100
150
200
250
train/global_step
0
0.2
0.4
0.6
0.8
train/reward
train/reward
Showing first 10 runs
50
100
150
200
250
train/global_step
0
0.5
1
1.5
train/num_tokens
train/num_tokens
Showing first 10 runs
50
100
150
200
250
train/global_step
200000
400000
600000
800000
train/loss
train/loss
Showing first 10 runs
50
100
150
200
250
train/global_step
0
0.005
0.01
0.015
System
21
1-6 of 21
Add section