x86-to-llvm-o2_epoch3-AMD

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

x86-to-llvm-o2_epoch3-AMD

This model is a fine-tuned version of adpretko/x86-to-llvm-o2_epoch2 on the x86-to-llvm-o2_part_00, the x86-to-llvm-o2_part_01, the x86-to-llvm-o2_part_02, the x86-to-llvm-o2_part_03, the x86-to-llvm-o2_part_04, the x86-to-llvm-o2_part_05, the x86-to-llvm-o2_part_06, the x86-to-llvm-o2_part_07, the x86-to-llvm-o2_part_08, the x86-to-llvm-o2_part_09, the x86-to-llvm-o2_part_10, the x86-to-llvm-o2_part_11, the x86-to-llvm-o2_part_12, the x86-to-llvm-o2_part_13, the x86-to-llvm-o2_part_14, the x86-to-llvm-o2_part_15, the x86-to-llvm-o2_part_16, the x86-to-llvm-o2_part_17, the x86-to-llvm-o2_part_18, the x86-to-llvm-o2_part_19, the x86-to-llvm-o2_part_20, the x86-to-llvm-o2_part_21, the x86-to-llvm-o2_part_22, the x86-to-llvm-o2_part_23, the x86-to-llvm-o2_part_24, the x86-to-llvm-o2_part_25, the x86-to-llvm-o2_part_26, the x86-to-llvm-o2_part_27, the x86-to-llvm-o2_part_28, the x86-to-llvm-o2_part_29, the x86-to-llvm-o2_part_30, the x86-to-llvm-o2_part_31, the x86-to-llvm-o2_part_32, the x86-to-llvm-o2_part_33, the x86-to-llvm-o2_part_34, the x86-to-llvm-o2_part_35, the x86-to-llvm-o2_part_36, the x86-to-llvm-o2_part_37, the x86-to-llvm-o2_part_38, the x86-to-llvm-o2_part_39, the x86-to-llvm-o2_part_40, the x86-to-llvm-o2_part_41, the x86-to-llvm-o2_part_42, the x86-to-llvm-o2_part_43, the x86-to-llvm-o2_part_44, the x86-to-llvm-o2_part_45, the x86-to-llvm-o2_part_46, the x86-to-llvm-o2_part_47 and the x86-to-llvm-o2_part_48 datasets.

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 8
• total_train_batch_size: 512
• total_eval_batch_size: 64
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 1.0

Training results

Framework versions

• Transformers 4.55.0
• Pytorch 2.8.0+rocm6.3
• Datasets 3.6.0
• Tokenizers 0.21.1