Isaak Carter Augustus
commited on
Commit
•
099ea83
1
Parent(s):
05d4b81
Update README.md
Browse files
README.md
CHANGED
@@ -8,12 +8,34 @@ language:
|
|
8 |
|
9 |
# Model Card
|
10 |
This is the first modedl on my ongoing goal to cerate a smart autonmius Smart Assistant that manages your smart home. This is J.O.S.I.E.v4o, she is the first rendition of the first training stage.
|
11 |
-
She currently only accepts Text as the input but will be able to see and hear in real time like GPT4o.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
12 |
|
13 |
# Prompt Fortmats
|
14 |
Currently not trained to support tool calling.
|
15 |
|
16 |
-
|
17 |
## Main User role
|
18 |
|
19 |
```txt
|
|
|
8 |
|
9 |
# Model Card
|
10 |
This is the first modedl on my ongoing goal to cerate a smart autonmius Smart Assistant that manages your smart home. This is J.O.S.I.E.v4o, she is the first rendition of the first training stage.
|
11 |
+
She currently only accepts Text as the input but will be able to see and hear in real time like GPT4o. This moddel will be later trained with two full epochs.
|
12 |
+
|
13 |
+
|
14 |
+
# Training Arguments
|
15 |
+
|
16 |
+
```json
|
17 |
+
{
|
18 |
+
"dataset": "Isaak-Carter/j.o.s.i.e.v4.0.2o",
|
19 |
+
"base model": "Qwen/Qwen2-7B",
|
20 |
+
"framework": "unsloth",
|
21 |
+
"LoRA used": true,
|
22 |
+
"max_seq_length": 16384,
|
23 |
+
"dataset_num_proc": 2,
|
24 |
+
"packing": false,
|
25 |
+
"per_device_train_batch_size": 2,
|
26 |
+
"gradient_accumulation_steps": 4,
|
27 |
+
"warmup_steps": 5,
|
28 |
+
"max_steps": 1000,
|
29 |
+
"learning_rate": 2e-4,
|
30 |
+
"optim": "adamw_8bit",
|
31 |
+
"weight_decay": 0.01,
|
32 |
+
"lr_scheduler_type": "linear"
|
33 |
+
}
|
34 |
+
```
|
35 |
|
36 |
# Prompt Fortmats
|
37 |
Currently not trained to support tool calling.
|
38 |
|
|
|
39 |
## Main User role
|
40 |
|
41 |
```txt
|