metadata
license: llama2
base_model: meta-llama/Llama-2-7b-hf
tags:
- generated_from_trainer
datasets:
- tyzhu/lmind_hotpot_train8000_eval7405_v1_qa
metrics:
- accuracy
model-index:
- name: lmind_hotpot_train8000_eval7405_v1_qa_3e-4_lora2
results:
- task:
name: Causal Language Modeling
type: text-generation
dataset:
name: tyzhu/lmind_hotpot_train8000_eval7405_v1_qa
type: tyzhu/lmind_hotpot_train8000_eval7405_v1_qa
metrics:
- name: Accuracy
type: accuracy
value: 0.5883291139240506
lmind_hotpot_train8000_eval7405_v1_qa_3e-4_lora2
This model is a fine-tuned version of meta-llama/Llama-2-7b-hf on the tyzhu/lmind_hotpot_train8000_eval7405_v1_qa dataset. It achieves the following results on the evaluation set:
- Loss: 2.9650
- Accuracy: 0.5883
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: 0.0003
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 4
- 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: 20.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.7554 | 1.0 | 250 | 1.7940 | 0.6093 |
| 1.5248 | 2.0 | 500 | 1.8274 | 0.6085 |
| 1.2054 | 3.0 | 750 | 1.9718 | 0.6027 |
| 0.8989 | 4.0 | 1000 | 2.1519 | 0.5987 |
| 0.6306 | 5.0 | 1250 | 2.3293 | 0.5961 |
| 0.4712 | 6.0 | 1500 | 2.5599 | 0.5936 |
| 0.3797 | 7.0 | 1750 | 2.7329 | 0.5936 |
| 0.3527 | 8.0 | 2000 | 2.8185 | 0.5913 |
| 0.3314 | 9.0 | 2250 | 2.8250 | 0.592 |
| 0.3265 | 10.0 | 2500 | 2.9242 | 0.5911 |
| 0.3148 | 11.0 | 2750 | 3.0013 | 0.5912 |
| 0.3184 | 12.0 | 3000 | 2.9315 | 0.5906 |
| 0.3101 | 13.0 | 3250 | 2.9116 | 0.5897 |
| 0.3164 | 14.0 | 3500 | 2.9208 | 0.5902 |
| 0.3074 | 15.0 | 3750 | 2.9385 | 0.5909 |
| 0.3107 | 16.0 | 4000 | 2.9519 | 0.5892 |
| 0.3054 | 17.0 | 4250 | 3.0108 | 0.5898 |
| 0.309 | 18.0 | 4500 | 3.0037 | 0.5904 |
| 0.3005 | 19.0 | 4750 | 3.0279 | 0.5898 |
| 0.3127 | 20.0 | 5000 | 2.9650 | 0.5883 |
Framework versions
- Transformers 4.34.0
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.14.1