Model save
Browse files
README.md
CHANGED
@@ -20,7 +20,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
20 |
|
21 |
This model is a fine-tuned version of [lightblue/suzume-llama-3-8B-japanese](https://huggingface.co/lightblue/suzume-llama-3-8B-japanese) on the generator dataset.
|
22 |
It achieves the following results on the evaluation set:
|
23 |
-
- Loss:
|
24 |
|
25 |
## Model description
|
26 |
|
@@ -53,13 +53,28 @@ The following hyperparameters were used during training:
|
|
53 |
|
54 |
### Training results
|
55 |
|
56 |
-
| Training Loss | Epoch
|
57 |
-
|
58 |
-
|
|
59 |
-
|
|
60 |
-
|
|
61 |
-
|
|
62 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
63 |
|
64 |
|
65 |
### Framework versions
|
|
|
20 |
|
21 |
This model is a fine-tuned version of [lightblue/suzume-llama-3-8B-japanese](https://huggingface.co/lightblue/suzume-llama-3-8B-japanese) on the generator dataset.
|
22 |
It achieves the following results on the evaluation set:
|
23 |
+
- Loss: 0.9951
|
24 |
|
25 |
## Model description
|
26 |
|
|
|
53 |
|
54 |
### Training results
|
55 |
|
56 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
57 |
+
|:-------------:|:------:|:----:|:---------------:|
|
58 |
+
| 2.4249 | 0.2222 | 1 | 1.8524 |
|
59 |
+
| 1.9685 | 0.4444 | 2 | 1.5389 |
|
60 |
+
| 1.72 | 0.6667 | 3 | 1.4369 |
|
61 |
+
| 1.5908 | 0.8889 | 4 | 1.3468 |
|
62 |
+
| 1.4306 | 1.1111 | 5 | 1.2865 |
|
63 |
+
| 1.368 | 1.3333 | 6 | 1.2355 |
|
64 |
+
| 1.2995 | 1.5556 | 7 | 1.1926 |
|
65 |
+
| 1.2591 | 1.7778 | 8 | 1.1429 |
|
66 |
+
| 1.1882 | 2.0 | 9 | 1.1208 |
|
67 |
+
| 1.0992 | 2.2222 | 10 | 1.0954 |
|
68 |
+
| 1.0199 | 2.4444 | 11 | 1.0754 |
|
69 |
+
| 1.0231 | 2.6667 | 12 | 1.0629 |
|
70 |
+
| 0.9314 | 2.8889 | 13 | 1.0317 |
|
71 |
+
| 0.8405 | 3.1111 | 14 | 1.0161 |
|
72 |
+
| 0.7706 | 3.3333 | 15 | 1.0216 |
|
73 |
+
| 0.782 | 3.5556 | 16 | 1.0055 |
|
74 |
+
| 0.8002 | 3.7778 | 17 | 0.9985 |
|
75 |
+
| 0.7849 | 4.0 | 18 | 0.9908 |
|
76 |
+
| 0.6481 | 4.2222 | 19 | 1.0070 |
|
77 |
+
| 0.5567 | 4.4444 | 20 | 0.9951 |
|
78 |
|
79 |
|
80 |
### Framework versions
|