Push to hub.
Browse files- config.json +7 -7
- model.safetensors +2 -2
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "hugosousa/smol-
|
3 |
"architectures": [
|
4 |
"ContextClassifier"
|
5 |
],
|
@@ -10,14 +10,14 @@
|
|
10 |
"finetuning_task": "text-classification",
|
11 |
"head_dim": 64,
|
12 |
"hidden_act": "silu",
|
13 |
-
"hidden_size":
|
14 |
"id2label": {
|
15 |
"0": ">",
|
16 |
"1": "<",
|
17 |
"2": "="
|
18 |
},
|
19 |
-
"initializer_range": 0.
|
20 |
-
"intermediate_size":
|
21 |
"is_llama_config": true,
|
22 |
"label2id": {
|
23 |
"<": 1,
|
@@ -28,9 +28,9 @@
|
|
28 |
"mlp_bias": false,
|
29 |
"model_type": "llama",
|
30 |
"n_score_hidden_layers": 2,
|
31 |
-
"num_attention_heads":
|
32 |
-
"num_hidden_layers":
|
33 |
-
"num_key_value_heads":
|
34 |
"pad_token_id": 0,
|
35 |
"pretraining_tp": 1,
|
36 |
"rms_norm_eps": 1e-05,
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "hugosousa/smol-135-c-3ed00d05",
|
3 |
"architectures": [
|
4 |
"ContextClassifier"
|
5 |
],
|
|
|
10 |
"finetuning_task": "text-classification",
|
11 |
"head_dim": 64,
|
12 |
"hidden_act": "silu",
|
13 |
+
"hidden_size": 576,
|
14 |
"id2label": {
|
15 |
"0": ">",
|
16 |
"1": "<",
|
17 |
"2": "="
|
18 |
},
|
19 |
+
"initializer_range": 0.041666666666666664,
|
20 |
+
"intermediate_size": 1536,
|
21 |
"is_llama_config": true,
|
22 |
"label2id": {
|
23 |
"<": 1,
|
|
|
28 |
"mlp_bias": false,
|
29 |
"model_type": "llama",
|
30 |
"n_score_hidden_layers": 2,
|
31 |
+
"num_attention_heads": 9,
|
32 |
+
"num_hidden_layers": 30,
|
33 |
+
"num_key_value_heads": 3,
|
34 |
"pad_token_id": 0,
|
35 |
"pretraining_tp": 1,
|
36 |
"rms_norm_eps": 1e-05,
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:71ef15999186260b50cf93f09b66ca46da33a8e656b5fe6a90508f3243ebd33c
|
3 |
+
size 322859302
|