Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-4.0
|
| 3 |
+
datasets:
|
| 4 |
+
- vicgalle/alpaca-gpt4
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
## Model Overview
|
| 10 |
+
Model license: cc-by-nc-4.0<br>
|
| 11 |
+
This model is trained based on [EleutherAI/pythia-1.4b-deduped](https://huggingface.co/EleutherAI/pythia-1.4b-deduped) model that is LoRA finetuned on [vicgalle/alpaca-gpt4](https://huggingface.co/datasets/vicgalle/alpaca-gpt4) dataset.<br>
|
| 12 |
+
|
| 13 |
+
## Prompt Template: `Alpaca`
|
| 14 |
+
```
|
| 15 |
+
<system_prompt>
|
| 16 |
+
|
| 17 |
+
### Instruction:
|
| 18 |
+
<user_message>
|
| 19 |
+
|
| 20 |
+
### Response:
|
| 21 |
+
<assistant_response>
|
| 22 |
+
```
|
| 23 |
+
|
| 24 |
+
## Intended Use
|
| 25 |
+
THIS IS A TEST MODEL, IT IS NOT INTENDED FOR REAL APPLICATIONS BY ANY MEANS. HOWEVER, A NEW MODEL IS COMING IN THE SAME TOPIC.<br>
|
| 26 |
+
This model series will be used for small but intense applications.
|
| 27 |
+
|
| 28 |
+
## Training Details
|
| 29 |
+
This model took `2:31:23` to train in QLoRA on a single `A100 40gb` GPU.<br>
|
| 30 |
+
- *epochs*: `1`
|
| 31 |
+
- *train batch size*: `12`
|
| 32 |
+
- *eval batch size*: `12`
|
| 33 |
+
- *gradient accumulation steps*: `1`
|
| 34 |
+
- *maximum gradient normal*: `0.3`
|
| 35 |
+
- *learning rate*: `2e-4`
|
| 36 |
+
- *weight decay*: `0.001`
|
| 37 |
+
- *optimizer*: `paged_adamw_32bit`
|
| 38 |
+
- *learning rate schedule*: `cosine`
|
| 39 |
+
- *warmup ratio (linear)*: `0.03`
|