update model

README.md

@@ -21,6 +21,12 @@ This model was adapted from [ytu-ce-cosmos/turkish-small-bert-uncased](https://h
 - [nli_tr](https://huggingface.co/datasets/nli_tr)
 - [emrecan/stsb-mt-turkish](https://huggingface.co/datasets/emrecan/stsb-mt-turkish)
 
+:warning: **All texts were manually lowercased,** [as stated](https://huggingface.co/ytu-ce-cosmos/turkish-small-bert-uncased#%E2%9A%A0-uncased-use-requires-manual-lowercase-conversion) by the model's authors:
+
+ ```python
+text.replace("I", "ı").lower()
+```
+
 ## Usage (Sentence-Transformers)
 
 Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
@@ -85,10 +91,10 @@ print(sentence_embeddings)
 Achieved results on the [STS-b](https://huggingface.co/datasets/emrecan/stsb-mt-turkish) test split are given below:
 
 ```txt
-Cosine-Similarity :       Pearson: 0.7387 Spearman: 0.7244
-Manhattan-Distance:       Pearson: 0.7118 Spearman: 0.7156
-Euclidean-Distance:       Pearson: 0.7119 Spearman: 0.7155
-Dot-Product-Similarity:   Pearson: 0.7164 Spearman: 0.7081
+Cosine-Similarity :	    Pearson: 0.8227	Spearman: 0.8192
+Manhattan-Distance:	    Pearson: 0.8105	Spearman: 0.8079
+Euclidean-Distance:	    Pearson: 0.8110	Spearman: 0.8087
+Dot-Product-Similarity:	Pearson: 0.7908	Spearman: 0.7827
 ```
 
 ## Training
@@ -108,8 +114,8 @@ The model was trained with the parameters:
 Parameters of the fit()-Method:
 ```
 {
-    "epochs": 5,
-    "evaluation_steps": 45,
+    "epochs": 4,
+    "evaluation_steps": 9,
     "evaluator": "sentence_transformers.evaluation.EmbeddingSimilarityEvaluator.EmbeddingSimilarityEvaluator",
     "max_grad_norm": 1,
     "optimizer_class": "<class 'torch.optim.adamw.AdamW'>",
@@ -118,7 +124,7 @@ Parameters of the fit()-Method:
     },
     "scheduler": "WarmupLinear",
     "steps_per_epoch": null,
-    "warmup_steps": 45,
+    "warmup_steps": 36,
     "weight_decay": 0.01
 }
 ```

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "e5_b64_turkish_small_bert_uncased-mean-nli/",
+  "_name_or_path": "output/ytu_ce_cosmos-turkish_small_bert_uncased-b64-e4-nli/",
   "architectures": [
     "BertModel"
   ],

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f74195b8a9669f7049a5ebf54d48133741be60ce00ceb478e8a08728bb735530
+oid sha256:e9aae58e89afd0e0e65e353ad8c7c8e03b96b09ed7cb77e974fbd87b4414774c
 size 118109958