Commit
•
e4fc86d
1
Parent(s):
1697874
Upload Qwen2ForCausalLM
Browse files- config.json +3 -1
- generation_config.json +2 -0
config.json
CHANGED
|
@@ -10,12 +10,13 @@
|
|
| 10 |
"hidden_size": 1536,
|
| 11 |
"initializer_range": 0.02,
|
| 12 |
"intermediate_size": 8960,
|
| 13 |
-
"max_position_embeddings":
|
| 14 |
"max_window_layers": 28,
|
| 15 |
"model_type": "qwen2",
|
| 16 |
"num_attention_heads": 12,
|
| 17 |
"num_hidden_layers": 28,
|
| 18 |
"num_key_value_heads": 2,
|
|
|
|
| 19 |
"quantization_config": {
|
| 20 |
"_load_in_4bit": true,
|
| 21 |
"_load_in_8bit": false,
|
|
@@ -37,6 +38,7 @@
|
|
| 37 |
"tie_word_embeddings": true,
|
| 38 |
"torch_dtype": "bfloat16",
|
| 39 |
"transformers_version": "4.44.2",
|
|
|
|
| 40 |
"use_cache": true,
|
| 41 |
"use_mrope": false,
|
| 42 |
"use_sliding_window": false,
|
|
|
|
| 10 |
"hidden_size": 1536,
|
| 11 |
"initializer_range": 0.02,
|
| 12 |
"intermediate_size": 8960,
|
| 13 |
+
"max_position_embeddings": 131072,
|
| 14 |
"max_window_layers": 28,
|
| 15 |
"model_type": "qwen2",
|
| 16 |
"num_attention_heads": 12,
|
| 17 |
"num_hidden_layers": 28,
|
| 18 |
"num_key_value_heads": 2,
|
| 19 |
+
"pad_token_id": 151665,
|
| 20 |
"quantization_config": {
|
| 21 |
"_load_in_4bit": true,
|
| 22 |
"_load_in_8bit": false,
|
|
|
|
| 38 |
"tie_word_embeddings": true,
|
| 39 |
"torch_dtype": "bfloat16",
|
| 40 |
"transformers_version": "4.44.2",
|
| 41 |
+
"unsloth_fixed": true,
|
| 42 |
"use_cache": true,
|
| 43 |
"use_mrope": false,
|
| 44 |
"use_sliding_window": false,
|
generation_config.json
CHANGED
|
@@ -1,6 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"bos_token_id": 151643,
|
| 3 |
"eos_token_id": 151643,
|
|
|
|
| 4 |
"max_new_tokens": 2048,
|
|
|
|
| 5 |
"transformers_version": "4.44.2"
|
| 6 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"bos_token_id": 151643,
|
| 3 |
"eos_token_id": 151643,
|
| 4 |
+
"max_length": 131072,
|
| 5 |
"max_new_tokens": 2048,
|
| 6 |
+
"pad_token_id": 151665,
|
| 7 |
"transformers_version": "4.44.2"
|
| 8 |
}
|