open-thought/reasoning-gym
1
0
Fork 0
mirror of https://github.com/open-thought/reasoning-gym.git synced 2026-04-29 17:35:16 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

cfg

Browse source
This commit is contained in:
Oliver 2025-04-22 20:33:04 +01:00
parent e372224ee1
commit 1ccd62bc1a
1 changed files with 4 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

8
training/configs/external_generalisation/math_curriculum_qwen_7b.yaml
View file

@ -136,7 +136,7 @@ actor_rollout_ref:
enforce_eager: True
free_cache_engine: True
load_format: dummy_dtensor
tensor_model_parallel_size: 4
tensor_model_parallel_size: 2
max_num_batched_tokens: 12288
max_num_seqs: 1024
log_prob_micro_batch_size: null # will be deprecated, use log_prob_micro_batch_size_per_gpu
@ -167,12 +167,12 @@ trainer:
balance_batch: True
total_epochs: 1
total_training_steps: 1500
project_name: inter-domain-generalisation
experiment_name: inter_reasoning_algebra_qwen_3b_composite
project_name: external-generalisation
experiment_name: math_curriculum_qwen_7b
logger: [ 'console', 'wandb' ]
val_generations_to_log_to_wandb: 0
nnodes: 1
n_gpus_per_node: 4
n_gpus_per_node: 2
save_freq: 100
# auto: find the last ckpt to resume. If can't find, start from scratch
resume_mode: auto # or auto or resume_path if