metadata

license: other
base_model: Qwen/Qwen1.5-4B
tags:
  - generated_from_trainer
datasets:
  - tyzhu/lmind_hotpot_train8000_eval7405_v1_docidx
metrics:
  - accuracy
model-index:
  - name: lmind_hotpot_train8000_eval7405_v1_docidx_Qwen_Qwen1.5-4B_lora2
    results:
      - task:
          name: Causal Language Modeling
          type: text-generation
        dataset:
          name: tyzhu/lmind_hotpot_train8000_eval7405_v1_docidx
          type: tyzhu/lmind_hotpot_train8000_eval7405_v1_docidx
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.7691922246220302
library_name: peft

lmind_hotpot_train8000_eval7405_v1_docidx_Qwen_Qwen1.5-4B_lora2

This model is a fine-tuned version of Qwen/Qwen1.5-4B on the tyzhu/lmind_hotpot_train8000_eval7405_v1_docidx dataset. It achieves the following results on the evaluation set:

Loss: 1.0623
Accuracy: 0.7692

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: 0.0001
train_batch_size: 1
eval_batch_size: 2
seed: 42
distributed_type: multi-GPU
num_devices: 4
gradient_accumulation_steps: 8
total_train_batch_size: 32
total_eval_batch_size: 8
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: constant
lr_scheduler_warmup_ratio: 0.05
num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.6067	0.9997	839	1.8440	0.7197
1.5433	1.9994	1678	1.7661	0.7247
1.4167	2.9991	2517	1.6455	0.7310
1.2948	4.0	3357	1.5394	0.7366
1.1715	4.9997	4196	1.4463	0.7422
1.0458	5.9994	5035	1.3537	0.7484
0.9357	6.9991	5874	1.2456	0.7546
0.8269	8.0	6714	1.1735	0.7598
0.7262	8.9997	7553	1.0966	0.7649
0.6381	9.9970	8390	1.0623	0.7692

Framework versions

PEFT 0.5.0
Transformers 4.40.2
Pytorch 2.3.0
Datasets 2.19.1
Tokenizers 0.19.1