Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,39 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
+
datasets:
|
4 |
+
- vicgalle/alpaca-gpt4
|
5 |
+
language:
|
6 |
+
- en
|
7 |
---
|
8 |
+
This is a finetuned version microsoft phi - 2. This is finetuned using the alpaca dataset.
|
9 |
+
|
10 |
+
|
11 |
+
## Training arguments
|
12 |
+
|
13 |
+
num_train_epochs=1
|
14 |
+
|
15 |
+
per_device_train_batch_size=1
|
16 |
+
|
17 |
+
gradient_accumulation_steps=8
|
18 |
+
|
19 |
+
optim="paged_adamw_32bit"
|
20 |
+
|
21 |
+
logging_steps=25
|
22 |
+
|
23 |
+
learning_rate=2e-4
|
24 |
+
|
25 |
+
weight_decay=0.001
|
26 |
+
|
27 |
+
max_grad_norm=0.3
|
28 |
+
|
29 |
+
warmup_ratio=0.03
|
30 |
+
|
31 |
+
lr_scheduler_type="cosine"
|
32 |
+
|
33 |
+
## Prompt format:
|
34 |
+
|
35 |
+
Use the below format
|
36 |
+
|
37 |
+
```
|
38 |
+
Below is an instruction that describes a task. Write a response that appropriately completes the request. ### Instruction:{prompt} ### {output}
|
39 |
+
```
|