Upload 6 files
Browse files- config.json +7 -8
- tokenizer_config.json +1 -1
config.json
CHANGED
@@ -1,9 +1,8 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"architectures": [
|
4 |
"Phi3ForCausalLM"
|
5 |
],
|
6 |
-
"attention_bias": false,
|
7 |
"attention_dropout": 0.0,
|
8 |
"auto_map": {
|
9 |
"AutoConfig": "configuration_phi3.Phi3Config",
|
@@ -13,14 +12,14 @@
|
|
13 |
"embd_pdrop": 0.0,
|
14 |
"eos_token_id": 32000,
|
15 |
"hidden_act": "silu",
|
16 |
-
"hidden_size":
|
17 |
"initializer_range": 0.02,
|
18 |
-
"intermediate_size":
|
19 |
"max_position_embeddings": 4096,
|
20 |
"model_type": "phi3",
|
21 |
-
"num_attention_heads":
|
22 |
-
"num_hidden_layers":
|
23 |
-
"num_key_value_heads":
|
24 |
"original_max_position_embeddings": 4096,
|
25 |
"pad_token_id": 32000,
|
26 |
"resid_pdrop": 0.0,
|
@@ -29,7 +28,7 @@
|
|
29 |
"rope_theta": 10000.0,
|
30 |
"sliding_window": 2047,
|
31 |
"tie_word_embeddings": false,
|
32 |
-
"transformers_version": "4.
|
33 |
"use_cache": true,
|
34 |
"vocab_size": 32064
|
35 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "microsoft/Phi-3-mini-4k-instruct",
|
3 |
"architectures": [
|
4 |
"Phi3ForCausalLM"
|
5 |
],
|
|
|
6 |
"attention_dropout": 0.0,
|
7 |
"auto_map": {
|
8 |
"AutoConfig": "configuration_phi3.Phi3Config",
|
|
|
12 |
"embd_pdrop": 0.0,
|
13 |
"eos_token_id": 32000,
|
14 |
"hidden_act": "silu",
|
15 |
+
"hidden_size": 3072,
|
16 |
"initializer_range": 0.02,
|
17 |
+
"intermediate_size": 8192,
|
18 |
"max_position_embeddings": 4096,
|
19 |
"model_type": "phi3",
|
20 |
+
"num_attention_heads": 32,
|
21 |
+
"num_hidden_layers": 32,
|
22 |
+
"num_key_value_heads": 32,
|
23 |
"original_max_position_embeddings": 4096,
|
24 |
"pad_token_id": 32000,
|
25 |
"resid_pdrop": 0.0,
|
|
|
28 |
"rope_theta": 10000.0,
|
29 |
"sliding_window": 2047,
|
30 |
"tie_word_embeddings": false,
|
31 |
+
"transformers_version": "4.41.2",
|
32 |
"use_cache": true,
|
33 |
"vocab_size": 32064
|
34 |
}
|
tokenizer_config.json
CHANGED
@@ -121,7 +121,7 @@
|
|
121 |
"clean_up_tokenization_spaces": false,
|
122 |
"eos_token": "<|endoftext|>",
|
123 |
"legacy": false,
|
124 |
-
"model_max_length":
|
125 |
"pad_token": "<|endoftext|>",
|
126 |
"padding_side": "left",
|
127 |
"sp_model_kwargs": {},
|
|
|
121 |
"clean_up_tokenization_spaces": false,
|
122 |
"eos_token": "<|endoftext|>",
|
123 |
"legacy": false,
|
124 |
+
"model_max_length": 4096,
|
125 |
"pad_token": "<|endoftext|>",
|
126 |
"padding_side": "left",
|
127 |
"sp_model_kwargs": {},
|