Model save

README.md

@@ -242,7 +242,7 @@ model-index:
       value: 0.96
       name: Cosine Precision@1
     - type: cosine_precision@3
-      value: 0.33000000000000007
+      value: 0.3333333333333334
       name: Cosine Precision@3
     - type: cosine_precision@5
       value: 0.19999999999999996
@@ -254,7 +254,7 @@ model-index:
       value: 0.96
       name: Cosine Recall@1
     - type: cosine_recall@3
-      value: 0.99
+      value: 1.0
       name: Cosine Recall@3
     - type: cosine_recall@5
       value: 1.0
@@ -266,22 +266,22 @@ model-index:
       value: 0.96
       name: Cosine Ndcg@1
     - type: cosine_ndcg@5
-      value: 0.9832346581878777
+      value: 0.9839278926071438
       name: Cosine Ndcg@5
     - type: cosine_ndcg@10
-      value: 0.9832346581878777
+      value: 0.9839278926071438
       name: Cosine Ndcg@10
     - type: cosine_mrr@1
       value: 0.96
       name: Cosine Mrr@1
     - type: cosine_mrr@5
-      value: 0.9775
+      value: 0.9783333333333333
       name: Cosine Mrr@5
     - type: cosine_mrr@10
-      value: 0.9775
+      value: 0.9783333333333333
       name: Cosine Mrr@10
     - type: cosine_map@100
-      value: 0.9775
+      value: 0.9783333333333333
       name: Cosine Map@100
 ---
 
@@ -350,7 +350,7 @@ print(query_embeddings.shape, document_embeddings.shape)
 # Get the similarity scores for the embeddings
 similarities = model.similarity(query_embeddings, document_embeddings)
 print(similarities)
-# tensor([[ 0.8822, -0.1093,  0.1044]])
+# tensor([[ 0.8839, -0.1092,  0.1013]])
 ```
 
 <!--
@@ -391,20 +391,20 @@ You can finetune this model on your own dataset.
 | cosine_accuracy@5   | 1.0        |
 | cosine_accuracy@10  | 1.0        |
 | cosine_precision@1  | 0.96       |
-| cosine_precision@3  | 0.33       |
+| cosine_precision@3  | 0.3333     |
 | cosine_precision@5  | 0.2        |
 | cosine_precision@10 | 0.1        |
 | cosine_recall@1     | 0.96       |
-| cosine_recall@3     | 0.99       |
+| cosine_recall@3     | 1.0        |
 | cosine_recall@5     | 1.0        |
 | cosine_recall@10    | 1.0        |
 | cosine_ndcg@1       | 0.96       |
-| cosine_ndcg@5       | 0.9832     |
-| **cosine_ndcg@10**  | **0.9832** |
+| cosine_ndcg@5       | 0.9839     |
+| **cosine_ndcg@10**  | **0.9839** |
 | cosine_mrr@1        | 0.96       |
-| cosine_mrr@5        | 0.9775     |
-| cosine_mrr@10       | 0.9775     |
-| cosine_map@100      | 0.9775     |
+| cosine_mrr@5        | 0.9783     |
+| cosine_mrr@10       | 0.9783     |
+| cosine_map@100      | 0.9783     |
 
 <!--
 ## Bias, Risks and Limitations
@@ -488,6 +488,7 @@ You can finetune this model on your own dataset.
 - `push_to_hub`: True
 - `hub_model_id`: JacobLinCool/Qwen3-Embedding-0.6B-GIR-1
 - `hub_private_repo`: False
+- `gradient_checkpointing`: True
 - `eval_on_start`: True
 - `batch_sampler`: no_duplicates
 
@@ -580,7 +581,7 @@ You can finetune this model on your own dataset.
 - `hub_private_repo`: False
 - `hub_always_push`: False
 - `hub_revision`: None
-- `gradient_checkpointing`: False
+- `gradient_checkpointing`: True
 - `gradient_checkpointing_kwargs`: None
 - `include_inputs_for_metrics`: False
 - `include_for_metrics`: []
@@ -619,8 +620,7 @@ You can finetune this model on your own dataset.
 | Epoch   | Step   | Validation Loss | cosine_ndcg@10 |
 |:-------:|:------:|:---------------:|:--------------:|
 | 0       | 0      | 0.0042          | 0.9926         |
-| **1.0** | **25** | **0.0013**      | **0.9832**     |
-| -1      | -1     | -               | 0.9832         |
+| **1.0** | **25** | **0.0014**      | **0.9839**     |
 
 * The bold row denotes the saved checkpoint.