henokyemam/safegate-granite-3.3-2b-instruct_dm_mix_1_trip

---

                               library_name: accelerate
                               license: apache-2.0
                               tags: [granite-3.3-2b-instruct, text-generation, chat]
                               datasets: [dm_mix_1_trip]
                               model-index:
                               - name: granite-3.3-2b-instruct (dm_mix_1_trip)
                               results: []
                               ---

                               # granite-3.3-2b-instruct — dm_mix_1_trip

                               ## Training hyperparameters
                               ```yaml
                               learning_rate: 5e-05
                               learning_scheduler: <accelerate.scheduler.AcceleratedScheduler object at 0x7fcfa2bd6300>
                               num_training_examples: 1922
                               batch_size_per_device: 1
                               gradient_accumulation_steps: 4
                               effective_batch_size: 16
                               max_steps: 600
                               num_epochs: 5
                               optimizer:  AdamW
                               num_warmup_steps: 30
                               lr_scheduler_type: linear
                               max_seq_length: 2880
                               precision: bf16
                               gradient_checkpointing: true
                               seed:42
                               ```