xlam_8B_all_zhtw_lr5e-7_ep3_16_64

This model is a fine-tuned version of Salesforce/Llama-xLAM-2-8b-fc-r on the multi_turn_miss_func_zh_tw_function_mix_sharegpt, the multi_turn_miss_param_zh_tw_function_mix_sharegpt, the multi_turn_zh_tw_function_mix_sharegpt, the irrelevance_zh_tw_sharegpt3000 and the apigen_zhtwV3_remove_sys_sharegpt datasets. It achieves the following results on the evaluation set:

• Loss: 0.3667

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: 5e-07
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Use OptimizerNames.ADAMW_TORCH 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: 0.03
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
0.6001	0.3238	1550	0.5461
0.4547	0.6476	3100	0.4493
0.4005	0.9714	4650	0.4115
0.2563	1.2952	6200	0.3927
0.4162	1.6190	7750	0.3787
0.3712	1.9428	9300	0.3722
0.4585	2.2666	10850	0.3687
0.3812	2.5904	12400	0.3673
0.2111	2.9142	13950	0.3671

Framework versions

• PEFT 0.17.0
• Transformers 4.55.2
• Pytorch 2.4.1+cu121
• Datasets 4.0.0
• Tokenizers 0.21.4