Views
No views yet
1max_steps = 240
2seq_len = 8192
3
4[model]
5name = "/home/ubuntu/models/hex-toolcall-base-v2"
6
7[wandb]
8project = "xml-tool-thinking"
9name = "findfile-v2-from-hex140"
10
11[orchestrator]
12batch_size = 512
13rollouts_per_example = 16
14
15[orchestrator.sampling]
16max_tokens = 512
17temperature = 1.0
18
19[[orchestrator.env]]
20id = "findfile_xml"
21args = { max_turns = 8, hint_chars = 40, noise_dirs = 3, max_read_chars = 1500, data_path = "/home/ubuntu/dataproc/consensus_pass_v2" }
22
23[trainer.model]
24ac = { freq = 1 }
25
26[trainer.optim]
27lr = 1e-6
28max_norm = 0.001
29
30[trainer.scheduler]
31type = "linear"
32warmup_steps = 5
33decay_steps = 40
34min_lr = 0
35
36[inference.parallel]
37tp = 4
38
39[ckpt]