Upload config.json
Browse files- llm/config.json +3 -3
llm/config.json
CHANGED
@@ -10,9 +10,9 @@
|
|
10 |
"hidden_size": 3584,
|
11 |
"initializer_range": 0.02,
|
12 |
"intermediate_size": 18944,
|
13 |
-
"max_position_embeddings":
|
14 |
"max_window_layers": 28,
|
15 |
-
"model_max_length":
|
16 |
"model_type": "qwen2",
|
17 |
"num_attention_heads": 28,
|
18 |
"num_hidden_layers": 28,
|
@@ -25,7 +25,7 @@
|
|
25 |
"rope_theta": 1062356830,
|
26 |
"sliding_window": null,
|
27 |
"tie_word_embeddings": false,
|
28 |
-
"tokenizer_model_max_length":
|
29 |
"tokenizer_padding_side": "right",
|
30 |
"torch_dtype": "bfloat16",
|
31 |
"transformers_version": "4.44.2",
|
|
|
10 |
"hidden_size": 3584,
|
11 |
"initializer_range": 0.02,
|
12 |
"intermediate_size": 18944,
|
13 |
+
"max_position_embeddings": 524288,
|
14 |
"max_window_layers": 28,
|
15 |
+
"model_max_length": 524288,
|
16 |
"model_type": "qwen2",
|
17 |
"num_attention_heads": 28,
|
18 |
"num_hidden_layers": 28,
|
|
|
25 |
"rope_theta": 1062356830,
|
26 |
"sliding_window": null,
|
27 |
"tie_word_embeddings": false,
|
28 |
+
"tokenizer_model_max_length": 524288,
|
29 |
"tokenizer_padding_side": "right",
|
30 |
"torch_dtype": "bfloat16",
|
31 |
"transformers_version": "4.44.2",
|