Hand_off_DS_Llama8B_350steps_1e6rate_SFT

This model is a fine-tuned version of deepseek-ai/DeepSeek-R1-Distill-Llama-8B on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5504

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-06
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Use adafactor and the args are: No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 100
• training_steps: 350

Training results

Training Loss	Epoch	Step	Validation Loss
2.6766	0.0464	50	2.4440
0.7304	0.0929	100	0.7164
0.5979	0.1393	150	0.6025
0.5512	0.1857	200	0.5658
0.5287	0.2321	250	0.5530
0.5223	0.2786	300	0.5504
0.5666	0.3250	350	0.5504

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.3.1
• Tokenizers 0.21.0