Commit
•
f4161a7
1
Parent(s):
a39955a
Update README.md
Browse files
README.md
CHANGED
@@ -70,7 +70,7 @@ Teeny-tiny-llama has been trained by leveraging scaling laws to determine the op
|
|
70 |
- **Size:** 162 million parameters
|
71 |
- **Dataset:** [Portuguese-Corpus-v3](https://huggingface.co/datasets/nicholasKluge/portuguese-corpus-v3)
|
72 |
- **Language:** Portuguese
|
73 |
-
- **Number of steps:**
|
74 |
- **Batch size:** 4
|
75 |
- **Optimizer:** `torch.optim.AdamW` (warmup_ratio = 0.01, learning_rate = 6e-4, epsilon = 1e-8)
|
76 |
- **GPU:** 1 NVIDIA A100-SXM4-40GB
|
|
|
70 |
- **Size:** 162 million parameters
|
71 |
- **Dataset:** [Portuguese-Corpus-v3](https://huggingface.co/datasets/nicholasKluge/portuguese-corpus-v3)
|
72 |
- **Language:** Portuguese
|
73 |
+
- **Number of steps:** 457,969
|
74 |
- **Batch size:** 4
|
75 |
- **Optimizer:** `torch.optim.AdamW` (warmup_ratio = 0.01, learning_rate = 6e-4, epsilon = 1e-8)
|
76 |
- **GPU:** 1 NVIDIA A100-SXM4-40GB
|