Add yarn scaling
Browse filesdummy pr for yarn
- config.json +6 -1
config.json
CHANGED
|
@@ -20,7 +20,12 @@
|
|
| 20 |
"num_key_value_heads": 8,
|
| 21 |
"pretraining_tp": 1,
|
| 22 |
"rms_norm_eps": 1e-05,
|
| 23 |
-
"rope_scaling":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
"rope_theta": 10000.0,
|
| 25 |
"tie_word_embeddings": false,
|
| 26 |
"torch_dtype": "float32",
|
|
|
|
| 20 |
"num_key_value_heads": 8,
|
| 21 |
"pretraining_tp": 1,
|
| 22 |
"rms_norm_eps": 1e-05,
|
| 23 |
+
"rope_scaling": {
|
| 24 |
+
"factor": 4.0,
|
| 25 |
+
"original_max_position_embeddings": 1024,
|
| 26 |
+
"rope_type": "yarn",
|
| 27 |
+
"type": "yarn"
|
| 28 |
+
},
|
| 29 |
"rope_theta": 10000.0,
|
| 30 |
"tie_word_embeddings": false,
|
| 31 |
"torch_dtype": "float32",
|