Release 3x-0.5B-Chat (standard HF interface, no trust_remote_code needed)
Browse files- config.json +1 -1
- generation_config.json +10 -0
config.json
CHANGED
|
@@ -37,7 +37,7 @@
|
|
| 37 |
"attention"
|
| 38 |
],
|
| 39 |
"max_position_embeddings": 32768,
|
| 40 |
-
"model_type": "
|
| 41 |
"num_attention_heads": 16,
|
| 42 |
"num_hidden_layers": 24,
|
| 43 |
"num_key_value_heads": 16,
|
|
|
|
| 37 |
"attention"
|
| 38 |
],
|
| 39 |
"max_position_embeddings": 32768,
|
| 40 |
+
"model_type": "qwen2",
|
| 41 |
"num_attention_heads": 16,
|
| 42 |
"num_hidden_layers": 24,
|
| 43 |
"num_key_value_heads": 16,
|
generation_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151643,
|
| 4 |
+
"eos_token_id": 151645,
|
| 5 |
+
"output_attentions": false,
|
| 6 |
+
"output_hidden_states": false,
|
| 7 |
+
"pad_token_id": 151645,
|
| 8 |
+
"transformers_version": "5.2.0",
|
| 9 |
+
"use_cache": true
|
| 10 |
+
}
|