learning_rate: 0.0001
train_batch_size: 2
eval_batch_size: 1
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 4
total_train_batch_size: 64
total_eval_batch_size: 8
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.05
num_epochs: 3.0

Training Loss	Epoch	Step	Validation Loss	Token Acc
0.9035	0.1235	50	0.9303	0.7264
0.9639	0.2469	100	0.9595	0.7263
0.9271	0.3704	150	0.9436	0.7292
0.9139	0.4938	200	0.9303	0.7340
0.8831	0.6173	250	0.9085	0.7398
0.8678	0.7407	300	0.8973	0.7410
0.8481	0.8642	350	0.8721	0.7465
0.8405	0.9877	400	0.8632	0.7482
0.7477	1.1111	450	0.8376	0.7551
0.7283	1.2346	500	0.8215	0.7583
0.7169	1.3580	550	0.8048	0.7616
0.7011	1.4815	600	0.7884	0.7666
0.6865	1.6049	650	0.7673	0.7700
0.6773	1.7284	700	0.7466	0.7744
0.6495	1.8519	750	0.7298	0.7775
0.6359	1.9753	800	0.7114	0.7819
0.5476	2.0988	850	0.7051	0.7854
0.5545	2.2222	900	0.6890	0.7883
0.5256	2.3457	950	0.6809	0.7903
0.5741	2.4691	1000	0.6723	0.7922
0.5286	2.5926	1050	0.6659	0.7937
0.5133	2.7160	1100	0.6612	0.7949
0.4898	2.8395	1150	0.6593	0.7951
0.5075	2.9630	1200	0.6582	0.7952
0.4914	3.0	1215	0.6582	0.7953

boyuzhuGPT/qwen2_5_omni_all_1015_wo_audio

qwen2_5_omni_all_1015_wo_audio