sft-Dolphin-v0.2-on-memory_dialoges
This model is a fine-tuned version of cognitivecomputations/dolphin-2.8-mistral-7b-v02 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5359
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-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 10
- total_train_batch_size: 10
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 50
- training_steps: 500
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
2.0787 | 0.03 | 20 | 2.0334 |
1.8996 | 0.06 | 40 | 1.7453 |
1.4612 | 0.09 | 60 | 1.0596 |
0.8132 | 0.12 | 80 | 0.6634 |
0.6515 | 0.15 | 100 | 0.6216 |
0.61 | 0.18 | 120 | 0.6008 |
0.6141 | 0.21 | 140 | 0.5862 |
0.5982 | 0.24 | 160 | 0.5722 |
0.5911 | 0.27 | 180 | 0.5586 |
0.5379 | 0.3 | 200 | 0.5518 |
0.5556 | 0.33 | 220 | 0.5485 |
0.5657 | 0.36 | 240 | 0.5460 |
0.5553 | 0.39 | 260 | 0.5439 |
0.5659 | 0.42 | 280 | 0.5418 |
0.5466 | 0.45 | 300 | 0.5404 |
0.5415 | 0.48 | 320 | 0.5397 |
0.5605 | 0.51 | 340 | 0.5386 |
0.5501 | 0.55 | 360 | 0.5376 |
0.5504 | 0.58 | 380 | 0.5370 |
0.5592 | 0.61 | 400 | 0.5367 |
0.5315 | 0.64 | 420 | 0.5363 |
0.5458 | 0.67 | 440 | 0.5361 |
0.5568 | 0.7 | 460 | 0.5360 |
0.5649 | 0.73 | 480 | 0.5359 |
0.5464 | 0.76 | 500 | 0.5359 |
Framework versions
- PEFT 0.7.1
- Transformers 4.36.2
- Pytorch 2.1.2
- Datasets 2.15.0
- Tokenizers 0.15.1
- Downloads last month
- 0
Model tree for ghost613/sft-Dolphin-v0.2-on-memory_dialoges
Base model
mistral-community/Mistral-7B-v0.2
Adapter
this model