cutelemonlili/random_0FXEXE5zwkhtZEzH

logic_speculative_DeepScaleR_data

This model is a fine-tuned version of agentica-org/DeepScaleR-1.5B-Preview on the logic_speculative_DeepScaleR_data dataset. It achieves the following results on the evaluation set:

• Loss: 0.2231

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 4
• total_eval_batch_size: 4
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
0.3297	0.1541	100	0.2195
0.2183	0.3082	200	0.2199
0.2286	0.4622	300	0.2197
0.2119	0.6163	400	0.2199
0.2454	0.7704	500	0.2196
0.2651	0.9245	600	0.2185
0.1332	1.0786	700	0.2226
0.2727	1.2327	800	0.2235
0.2075	1.3867	900	0.2236
0.2613	1.5408	1000	0.2233
0.0864	1.6949	1100	0.2232
0.2049	1.8490	1200	0.2230

Framework versions

• Transformers 4.46.1
• Pytorch 2.6.0+cu124
• Datasets 3.1.0
• Tokenizers 0.20.3