Update README.md
Browse files
README.md
CHANGED
|
@@ -23,6 +23,10 @@ The model was finetuned on ~128,000 curated transcripts across different domanin
|
|
| 23 |
|
| 24 |
It’s built to handle real-world Hindi, English, Indic noisy transcripts.
|
| 25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
Finetuning Parameters:
|
| 27 |
```
|
| 28 |
rank = 64 # kept small to know change inherent model intelligence but to make sure structured ectraction is followed
|
|
@@ -63,10 +67,6 @@ trainer = SFTTrainer(
|
|
| 63 |
),
|
| 64 |
)
|
| 65 |
```
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
`PS: VERY FEW EVALs WERE TAKEN FOR THE 0.8b MODEL`
|
| 69 |
-
|
| 70 |
|
| 71 |
Provide the below schema for best output:
|
| 72 |
```
|
|
|
|
| 23 |
|
| 24 |
It’s built to handle real-world Hindi, English, Indic noisy transcripts.
|
| 25 |
|
| 26 |
+
|
| 27 |
+
|
| 28 |
+
`PS: VERY FEW EVALs WERE TAKEN FOR THE 0.8b MODEL`
|
| 29 |
+
|
| 30 |
Finetuning Parameters:
|
| 31 |
```
|
| 32 |
rank = 64 # kept small to know change inherent model intelligence but to make sure structured ectraction is followed
|
|
|
|
| 67 |
),
|
| 68 |
)
|
| 69 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
| 70 |
|
| 71 |
Provide the below schema for best output:
|
| 72 |
```
|