Update README.md
Browse files
README.md
CHANGED
@@ -20,15 +20,20 @@
|
|
20 |
μ΄ ν©μ°λ λ°μ΄ν°λ μ½ 11GB μ
λλ€.
|
21 |
|
22 |
## νμ΅ μμΈ
|
23 |
-
- **BATCH_SIZE**:
|
24 |
-
- **ACCUMULATE**:
|
25 |
- **MAX_STEPS**: 12,500
|
26 |
-
- **Train Steps*Batch Szie**: 100M
|
27 |
- **WARMUP_STEPS**: 2,400
|
28 |
- **μ΅μ ν**: AdamW, LR 1e-3, BETA (0.9, 0.98), eps 1e-6
|
29 |
- **νμ΅λ₯ κ°μ **: linear
|
30 |
- **μ¬μ©λ νλμ¨μ΄**: 2x RTX 8000 GPU
|
31 |
|
|
|
|
|
|
|
|
|
|
|
32 |
## μ¬μ© λ°©λ²
|
33 |
```python
|
34 |
from transformers import AutoModel, AutoTokenizer
|
|
|
20 |
μ΄ ν©μ°λ λ°μ΄ν°λ μ½ 11GB μ
λλ€.
|
21 |
|
22 |
## νμ΅ μμΈ
|
23 |
+
- **BATCH_SIZE**: 112 (GPUλΉ)
|
24 |
+
- **ACCUMULATE**: 36
|
25 |
- **MAX_STEPS**: 12,500
|
26 |
+
- **Train Steps*Batch Szie**: **100M**
|
27 |
- **WARMUP_STEPS**: 2,400
|
28 |
- **μ΅μ ν**: AdamW, LR 1e-3, BETA (0.9, 0.98), eps 1e-6
|
29 |
- **νμ΅λ₯ κ°μ **: linear
|
30 |
- **μ¬μ©λ νλμ¨μ΄**: 2x RTX 8000 GPU
|
31 |
|
32 |
+
|
33 |
+
![Evaluation Loss Graph](https://cdn-uploads.huggingface.co/production/uploads/64a0fd6fd3149e05bc5260dd/-64jKdcJAavwgUREwaywe.png)
|
34 |
+
|
35 |
+
![Evaluation Accuracy Graph](https://cdn-uploads.huggingface.co/production/uploads/64a0fd6fd3149e05bc5260dd/LPq5M6S8LTwkFSCepD33S.png)
|
36 |
+
|
37 |
## μ¬μ© λ°©λ²
|
38 |
```python
|
39 |
from transformers import AutoModel, AutoTokenizer
|