Views
No views yet
| Model | 1 Epoch Pct | Link |
|---|---|---|
| ML1-34b | 15% | Directory |
| ML1-34b | 50% | ~ |
| ML1-34b | 100% | ~ |
| ML1-mistral-7b | 50% | ~ |
| ML1-mistral-7b | 100% | ~ |
| ML1-70b | 15% | ~ |
| ML1-70b | 50% | ~ |
| ML1-70b | 100% | ~ |
1Below is an instruction that describes a task. Write a response that appropriately completes the request.
2
3### Instruction:
4{instruction}
5
6### Response:nmitchko/ML1-34b-previews is a large language model repository of LoRA checkpoints specifically fine-tuned to add text-book synthesized data in the style of Phi 1/1.5.
It is based on codellama-34b-hf at 34 billion parameters.bitsandbytes>=0.41.0
peft@main
transformers@mainemrgnt-cmplxty/sciphi-textbooks-are-all-you-need contains textbook synthesized data.| Item | Amount | Units |
|---|---|---|
| LoRA Rank | 64 | ~ |
| LoRA Alpha | 16 | ~ |
| Learning Rate | 1e-4 | SI |
| Dropout | 5 | % |
bitsandbytes quantization config was used during training: