Training in progress, step 5000

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

1_Pooling/config.json

@@ -1,7 +1,7 @@
 {
-    "word_embedding_dimension": 384,
-    "pooling_mode_cls_token": false,
-    "pooling_mode_mean_tokens": true,
+    "word_embedding_dimension": 512,
+    "pooling_mode_cls_token": true,
+    "pooling_mode_mean_tokens": false,
     "pooling_mode_max_tokens": false,
     "pooling_mode_mean_sqrt_len_tokens": false,
     "pooling_mode_weightedmean_tokens": false,

Information-Retrieval_evaluation_val_results.csv

@@ -9,3 +9,4 @@ epoch,steps,cosine-Accuracy@1,cosine-Accuracy@3,cosine-Accuracy@5,cosine-Precisi
 -1,-1,0.83665,0.91045,0.9361,0.83665,0.83665,0.3034833333333333,0.91045,0.18722000000000003,0.9361,0.83665,0.8753945833333286,0.8793089583333286,0.9000254411118587,0.8812821493075779
 -1,-1,0.827675,0.903,0.928425,0.827675,0.827675,0.30099999999999993,0.903,0.18568500000000004,0.928425,0.827675,0.8671804166666619,0.8711970039682481,0.8922532953454642,0.87334664003711
 -1,-1,0.83295,0.9071,0.9329,0.83295,0.83295,0.3023666666666666,0.9071,0.18658000000000005,0.9329,0.83295,0.872013749999996,0.8760916468253912,0.8970951855878305,0.8781372459990227
+-1,-1,0.83545,0.911175,0.9366,0.83545,0.83545,0.303725,0.911175,0.18732000000000001,0.9366,0.83545,0.8751591666666616,0.8790415476190412,0.8999318372974409,0.8810239994800558

README.md

@@ -5,123 +5,51 @@ tags:
 - feature-extraction
 - dense
 - generated_from_trainer
-- dataset_size:713743
+- dataset_size:100000
 - loss:MultipleNegativesRankingLoss
-base_model: thenlper/gte-small
+base_model: prajjwal1/bert-small
 widget:
-- source_sentence: 'Abraham Lincoln: Why is the Gettysburg Address so memorable?'
+- source_sentence: How do I calculate IQ?
   sentences:
-  - 'Abraham Lincoln: Why is the Gettysburg Address so memorable?'
-  - What does the Gettysburg Address really mean?
-  - What is eatalo.com?
-- source_sentence: Has the influence of Ancient Carthage in science, math, and society
-    been underestimated?
+  - What is the easiest way to know my IQ?
+  - How do I calculate not IQ ?
+  - What are some creative and innovative business ideas with less investment in India?
+- source_sentence: How can I learn martial arts in my home?
   sentences:
-  - How does one earn money online without an investment from home?
-  - Has the influence of Ancient Carthage in science, math, and society been underestimated?
-  - Has the influence of the Ancient Etruscans in science and math been underestimated?
-- source_sentence: Is there any app that shares charging to others like share it how
-    we transfer files?
+  - How can I learn martial arts by myself?
+  - What are the advantages and disadvantages of investing in gold?
+  - Can people see that I have looked at their pictures on instagram if I am not following
+    them?
+- source_sentence: When Enterprise picks you up do you have to take them back?
   sentences:
-  - How do you think of Chinese claims that the present Private Arbitration is illegal,
-    its verdict violates the UNCLOS and is illegal?
-  - Is there any app that shares charging to others like share it how we transfer
-    files?
-  - Are there any platforms that provides end-to-end encryption for file transfer/
-    sharing?
-- source_sentence: Why AAP’s MLA Dinesh Mohaniya has been arrested?
+  - Are there any software Training institute in Tuticorin?
+  - When Enterprise picks you up do you have to take them back?
+  - When Enterprise picks you up do them have to take youback?
+- source_sentence: What are some non-capital goods?
   sentences:
-  - What are your views on the latest sex scandal by AAP MLA Sandeep Kumar?
-  - What is a dc current? What are some examples?
-  - Why AAP’s MLA Dinesh Mohaniya has been arrested?
-- source_sentence: What is the difference between economic growth and economic development?
+  - What are capital goods?
+  - How is the value of [math]\pi[/math] calculated?
+  - What are some non-capital goods?
+- source_sentence: What is the QuickBooks technical support phone number in New York?
   sentences:
-  - How cold can the Gobi Desert get, and how do its average temperatures compare
-    to the ones in the Simpson Desert?
-  - the difference between economic growth and economic development is What?
-  - What is the difference between economic growth and economic development?
+  - What caused the Great Depression?
+  - Can I apply for PR in Canada?
+  - Which is the best QuickBooks Hosting Support Number in New York?
 pipeline_tag: sentence-similarity
 library_name: sentence-transformers
-metrics:
-- cosine_accuracy@1
-- cosine_accuracy@3
-- cosine_accuracy@5
-- cosine_precision@1
-- cosine_precision@3
-- cosine_precision@5
-- cosine_recall@1
-- cosine_recall@3
-- cosine_recall@5
-- cosine_ndcg@10
-- cosine_mrr@1
-- cosine_mrr@5
-- cosine_mrr@10
-- cosine_map@100
-model-index:
-- name: SentenceTransformer based on thenlper/gte-small
-  results:
-  - task:
-      type: information-retrieval
-      name: Information Retrieval
-    dataset:
-      name: val
-      type: val
-    metrics:
-    - type: cosine_accuracy@1
-      value: 0.83545
-      name: Cosine Accuracy@1
-    - type: cosine_accuracy@3
-      value: 0.911175
-      name: Cosine Accuracy@3
-    - type: cosine_accuracy@5
-      value: 0.9366
-      name: Cosine Accuracy@5
-    - type: cosine_precision@1
-      value: 0.83545
-      name: Cosine Precision@1
-    - type: cosine_precision@3
-      value: 0.303725
-      name: Cosine Precision@3
-    - type: cosine_precision@5
-      value: 0.18732000000000001
-      name: Cosine Precision@5
-    - type: cosine_recall@1
-      value: 0.83545
-      name: Cosine Recall@1
-    - type: cosine_recall@3
-      value: 0.911175
-      name: Cosine Recall@3
-    - type: cosine_recall@5
-      value: 0.9366
-      name: Cosine Recall@5
-    - type: cosine_ndcg@10
-      value: 0.8999318372974409
-      name: Cosine Ndcg@10
-    - type: cosine_mrr@1
-      value: 0.83545
-      name: Cosine Mrr@1
-    - type: cosine_mrr@5
-      value: 0.8751591666666616
-      name: Cosine Mrr@5
-    - type: cosine_mrr@10
-      value: 0.8790415476190412
-      name: Cosine Mrr@10
-    - type: cosine_map@100
-      value: 0.8810239994800558
-      name: Cosine Map@100
 ---
 
-# SentenceTransformer based on thenlper/gte-small
+# SentenceTransformer based on prajjwal1/bert-small
 
-This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [thenlper/gte-small](https://huggingface.co/thenlper/gte-small). It maps sentences & paragraphs to a 384-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
+This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [prajjwal1/bert-small](https://huggingface.co/prajjwal1/bert-small). It maps sentences & paragraphs to a 512-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
 
 ## Model Details
 
 ### Model Description
 - **Model Type:** Sentence Transformer
-- **Base model:** [thenlper/gte-small](https://huggingface.co/thenlper/gte-small) <!-- at revision 17e1f347d17fe144873b1201da91788898c639cd -->
+- **Base model:** [prajjwal1/bert-small](https://huggingface.co/prajjwal1/bert-small) <!-- at revision 0ec5f86f27c1a77d704439db5e01c307ea11b9d4 -->
 - **Maximum Sequence Length:** 128 tokens
-- **Output Dimensionality:** 384 dimensions
+- **Output Dimensionality:** 512 dimensions
 - **Similarity Function:** Cosine Similarity
 <!-- - **Training Dataset:** Unknown -->
 <!-- - **Language:** Unknown -->
@@ -138,8 +66,7 @@ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [t
 ```
 SentenceTransformer(
   (0): Transformer({'max_seq_length': 128, 'do_lower_case': False, 'architecture': 'BertModel'})
-  (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
-  (2): Normalize()
+  (1): Pooling({'word_embedding_dimension': 512, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
 )
 ```
 
@@ -158,23 +85,23 @@ Then you can load this model and run inference.
 from sentence_transformers import SentenceTransformer
 
 # Download from the 🤗 Hub
-model = SentenceTransformer("redis/model-b-structured")
+model = SentenceTransformer("sentence_transformers_model_id")
 # Run inference
 sentences = [
-    'What is the difference between economic growth and economic development?',
-    'What is the difference between economic growth and economic development?',
-    'the difference between economic growth and economic development is What?',
+    'What is the QuickBooks technical support phone number in New York?',
+    'Which is the best QuickBooks Hosting Support Number in New York?',
+    'Can I apply for PR in Canada?',
 ]
 embeddings = model.encode(sentences)
 print(embeddings.shape)
-# [3, 384]
+# [3, 512]
 
 # Get the similarity scores for the embeddings
 similarities = model.similarity(embeddings, embeddings)
 print(similarities)
-# tensor([[ 1.0000,  1.0000, -0.0794],
-#         [ 1.0000,  1.0000, -0.0794],
-#         [-0.0794, -0.0794,  1.0000]])
+# tensor([[1.0000, 0.8563, 0.0594],
+#         [0.8563, 1.0000, 0.1245],
+#         [0.0594, 0.1245, 1.0000]])
 ```
 
 <!--
@@ -201,32 +128,6 @@ You can finetune this model on your own dataset.
 *List how the model may foreseeably be misused and address what users ought not to do with the model.*
 -->
 
-## Evaluation
-
-### Metrics
-
-#### Information Retrieval
-
-* Dataset: `val`
-* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
-
-| Metric             | Value      |
-|:-------------------|:-----------|
-| cosine_accuracy@1  | 0.8355     |
-| cosine_accuracy@3  | 0.9112     |
-| cosine_accuracy@5  | 0.9366     |
-| cosine_precision@1 | 0.8355     |
-| cosine_precision@3 | 0.3037     |
-| cosine_precision@5 | 0.1873     |
-| cosine_recall@1    | 0.8355     |
-| cosine_recall@3    | 0.9112     |
-| cosine_recall@5    | 0.9366     |
-| **cosine_ndcg@10** | **0.8999** |
-| cosine_mrr@1       | 0.8355     |
-| cosine_mrr@5       | 0.8752     |
-| cosine_mrr@10      | 0.879      |
-| cosine_map@100     | 0.881      |
-
 <!--
 ## Bias, Risks and Limitations
 
@@ -245,49 +146,23 @@ You can finetune this model on your own dataset.
 
 #### Unnamed Dataset
 
-* Size: 713,743 training samples
-* Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
+* Size: 100,000 training samples
+* Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>sentence_2</code>
 * Approximate statistics based on the first 1000 samples:
-  |         | anchor                                                                            | positive                                                                          | negative                                                                          |
+  |         | sentence_0                                                                        | sentence_1                                                                        | sentence_2                                                                        |
   |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
   | type    | string                                                                            | string                                                                            | string                                                                            |
-  | details | <ul><li>min: 6 tokens</li><li>mean: 16.07 tokens</li><li>max: 53 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 16.03 tokens</li><li>max: 53 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 16.81 tokens</li><li>max: 58 tokens</li></ul> |
+  | details | <ul><li>min: 6 tokens</li><li>mean: 15.79 tokens</li><li>max: 66 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 15.68 tokens</li><li>max: 66 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 16.37 tokens</li><li>max: 67 tokens</li></ul> |
 * Samples:
-  | anchor                                                                         | positive                                                                              | negative                                                                                                          |
-  |:-------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------|
-  | <code>Which one is better Linux OS? Ubuntu or Mint?</code>                     | <code>Why do you use Linux Mint?</code>                                               | <code>Which one is not better Linux OS ? Ubuntu or Mint ?</code>                                                  |
-  | <code>What is flow?</code>                                                     | <code>What is flow?</code>                                                            | <code>What are flow lines?</code>                                                                                 |
-  | <code>How is Trump planning to get Mexico to pay for his supposed wall?</code> | <code>How is it possible for Donald Trump to force Mexico to pay for the wall?</code> | <code>Why do we connect the positive terminal before the negative terminal to ground in a vehicle battery?</code> |
+  | sentence_0                                                       | sentence_1                                                       | sentence_2                                                                        |
+  |:-----------------------------------------------------------------|:-----------------------------------------------------------------|:----------------------------------------------------------------------------------|
+  | <code>Is masturbating bad for boys?</code>                       | <code>Is masturbating bad for boys?</code>                       | <code>How harmful or unhealthy is masturbation?</code>                            |
+  | <code>Does a train engine move in reverse?</code>                | <code>Does a train engine move in reverse?</code>                | <code>Time moves forward, not in reverse. Doesn't that make time a vector?</code> |
+  | <code>What is the most badass thing anyone has ever done?</code> | <code>What is the most badass thing anyone has ever done?</code> | <code>anyone is the most badass thing Whathas ever done?</code>                   |
 * Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
   ```json
   {
-      "scale": 7.0,
-      "similarity_fct": "cos_sim",
-      "gather_across_devices": false
-  }
-  ```
-
-### Evaluation Dataset
-
-#### Unnamed Dataset
-
-* Size: 40,000 evaluation samples
-* Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
-* Approximate statistics based on the first 1000 samples:
-  |         | anchor                                                                            | positive                                                                          | negative                                                                          |
-  |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
-  | type    | string                                                                            | string                                                                            | string                                                                            |
-  | details | <ul><li>min: 6 tokens</li><li>mean: 15.52 tokens</li><li>max: 74 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 15.51 tokens</li><li>max: 74 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 16.79 tokens</li><li>max: 69 tokens</li></ul> |
-* Samples:
-  | anchor                                                                                                                                                       | positive                                                                                                                                                     | negative                                                                                                                                                            |
-  |:-------------------------------------------------------------------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------|
-  | <code>Why are all my questions on Quora marked needing improvement?</code>                                                                                   | <code>Why are all my questions immediately being marked as needing improvement?</code>                                                                       | <code>For a post-graduate student in IIT, is it allowed to take an external scholarship as a top-up to his/her MHRD assistantship?</code>                           |
-  | <code>Can blue butter fly needle with vaccum tube be reused? Is it HIV risk? . Heard the needle is too small to be reused . Had blood draw at clinic?</code> | <code>Can blue butter fly needle with vaccum tube be reused? Is it HIV risk? . Heard the needle is too small to be reused . Had blood draw at clinic?</code> | <code>Can blue butter fly needle with vaccum tube be reused not ? Is it HIV risk ? . Heard the needle is too small to be reused . Had blood draw at clinic ?</code> |
-  | <code>Why do people still believe the world is flat?</code>                                                                                                  | <code>Why are there still people who believe the world is flat?</code>                                                                                       | <code>I'm not able to buy Udemy course .it is not accepting mine and my friends debit card.my card can be used for Flipkart .how to purchase now?</code>            |
-* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
-  ```json
-  {
-      "scale": 7.0,
+      "scale": 20.0,
       "similarity_fct": "cos_sim",
       "gather_across_devices": false
   }
@@ -296,49 +171,36 @@ You can finetune this model on your own dataset.
 ### Training Hyperparameters
 #### Non-Default Hyperparameters
 
-- `eval_strategy`: steps
-- `per_device_train_batch_size`: 128
-- `per_device_eval_batch_size`: 128
-- `learning_rate`: 0.0002
-- `weight_decay`: 0.0001
-- `max_steps`: 10000
-- `warmup_ratio`: 0.1
+- `per_device_train_batch_size`: 64
+- `per_device_eval_batch_size`: 64
 - `fp16`: True
-- `dataloader_drop_last`: True
-- `dataloader_num_workers`: 1
-- `dataloader_prefetch_factor`: 1
-- `load_best_model_at_end`: True
-- `optim`: adamw_torch
-- `ddp_find_unused_parameters`: False
-- `push_to_hub`: True
-- `hub_model_id`: redis/model-b-structured
-- `eval_on_start`: True
+- `multi_dataset_batch_sampler`: round_robin
 
 #### All Hyperparameters
 <details><summary>Click to expand</summary>
 
 - `overwrite_output_dir`: False
 - `do_predict`: False
-- `eval_strategy`: steps
+- `eval_strategy`: no
 - `prediction_loss_only`: True
-- `per_device_train_batch_size`: 128
-- `per_device_eval_batch_size`: 128
+- `per_device_train_batch_size`: 64
+- `per_device_eval_batch_size`: 64
 - `per_gpu_train_batch_size`: None
 - `per_gpu_eval_batch_size`: None
 - `gradient_accumulation_steps`: 1
 - `eval_accumulation_steps`: None
 - `torch_empty_cache_steps`: None
-- `learning_rate`: 0.0002
-- `weight_decay`: 0.0001
+- `learning_rate`: 5e-05
+- `weight_decay`: 0.0
 - `adam_beta1`: 0.9
 - `adam_beta2`: 0.999
 - `adam_epsilon`: 1e-08
-- `max_grad_norm`: 1.0
-- `num_train_epochs`: 3.0
-- `max_steps`: 10000
+- `max_grad_norm`: 1
+- `num_train_epochs`: 3
+- `max_steps`: -1
 - `lr_scheduler_type`: linear
 - `lr_scheduler_kwargs`: {}
-- `warmup_ratio`: 0.1
+- `warmup_ratio`: 0.0
 - `warmup_steps`: 0
 - `log_level`: passive
 - `log_level_replica`: warning
@@ -366,14 +228,14 @@ You can finetune this model on your own dataset.
 - `tpu_num_cores`: None
 - `tpu_metrics_debug`: False
 - `debug`: []
-- `dataloader_drop_last`: True
-- `dataloader_num_workers`: 1
-- `dataloader_prefetch_factor`: 1
+- `dataloader_drop_last`: False
+- `dataloader_num_workers`: 0
+- `dataloader_prefetch_factor`: None
 - `past_index`: -1
 - `disable_tqdm`: False
 - `remove_unused_columns`: True
 - `label_names`: None
-- `load_best_model_at_end`: True
+- `load_best_model_at_end`: False
 - `ignore_data_skip`: False
 - `fsdp`: []
 - `fsdp_min_num_params`: 0
@@ -383,23 +245,23 @@ You can finetune this model on your own dataset.
 - `parallelism_config`: None
 - `deepspeed`: None
 - `label_smoothing_factor`: 0.0
-- `optim`: adamw_torch
+- `optim`: adamw_torch_fused
 - `optim_args`: None
 - `adafactor`: False
 - `group_by_length`: False
 - `length_column_name`: length
 - `project`: huggingface
 - `trackio_space_id`: trackio
-- `ddp_find_unused_parameters`: False
+- `ddp_find_unused_parameters`: None
 - `ddp_bucket_cap_mb`: None
 - `ddp_broadcast_buffers`: False
 - `dataloader_pin_memory`: True
 - `dataloader_persistent_workers`: False
 - `skip_memory_metrics`: True
 - `use_legacy_prediction_loop`: False
-- `push_to_hub`: True
+- `push_to_hub`: False
 - `resume_from_checkpoint`: None
-- `hub_model_id`: redis/model-b-structured
+- `hub_model_id`: None
 - `hub_strategy`: every_save
 - `hub_private_repo`: None
 - `hub_always_push`: False
@@ -426,65 +288,32 @@ You can finetune this model on your own dataset.
 - `neftune_noise_alpha`: None
 - `optim_target_modules`: None
 - `batch_eval_metrics`: False
-- `eval_on_start`: True
+- `eval_on_start`: False
 - `use_liger_kernel`: False
 - `liger_kernel_config`: None
 - `eval_use_gather_object`: False
 - `average_tokens_across_devices`: True
 - `prompts`: None
 - `batch_sampler`: batch_sampler
-- `multi_dataset_batch_sampler`: proportional
+- `multi_dataset_batch_sampler`: round_robin
 - `router_mapping`: {}
 - `learning_rate_mapping`: {}
 
 </details>
 
 ### Training Logs
-| Epoch      | Step      | Training Loss | Validation Loss | val_cosine_ndcg@10 |
-|:----------:|:---------:|:-------------:|:---------------:|:------------------:|
-| 0          | 0         | -             | 3.6810          | 0.8566             |
-| 0.0448     | 250       | 1.5797        | 0.4480          | 0.8864             |
-| 0.0897     | 500       | 0.5396        | 0.4082          | 0.8901             |
-| 0.1345     | 750       | 0.4931        | 0.3876          | 0.8887             |
-| 0.1793     | 1000      | 0.4761        | 0.3822          | 0.8888             |
-| 0.2242     | 1250      | 0.462         | 0.3777          | 0.8899             |
-| 0.2690     | 1500      | 0.4452        | 0.3683          | 0.8896             |
-| 0.3138     | 1750      | 0.4356        | 0.3579          | 0.8899             |
-| 0.3587     | 2000      | 0.4303        | 0.3553          | 0.8902             |
-| 0.4035     | 2250      | 0.4176        | 0.3492          | 0.8915             |
-| 0.4484     | 2500      | 0.4118        | 0.3459          | 0.8918             |
-| 0.4932     | 2750      | 0.4082        | 0.3437          | 0.8929             |
-| 0.5380     | 3000      | 0.4017        | 0.3413          | 0.8930             |
-| 0.5829     | 3250      | 0.3987        | 0.3380          | 0.8930             |
-| 0.6277     | 3500      | 0.3955        | 0.3355          | 0.8945             |
-| 0.6725     | 3750      | 0.3899        | 0.3324          | 0.8945             |
-| 0.7174     | 4000      | 0.3885        | 0.3307          | 0.8943             |
-| 0.7622     | 4250      | 0.3852        | 0.3272          | 0.8944             |
-| 0.8070     | 4500      | 0.3798        | 0.3276          | 0.8952             |
-| 0.8519     | 4750      | 0.3791        | 0.3240          | 0.8958             |
-| 0.8967     | 5000      | 0.3762        | 0.3230          | 0.8962             |
-| 0.9415     | 5250      | 0.3744        | 0.3209          | 0.8966             |
-| 0.9864     | 5500      | 0.3706        | 0.3193          | 0.8962             |
-| 1.0312     | 5750      | 0.3591        | 0.3164          | 0.8964             |
-| 1.0760     | 6000      | 0.3541        | 0.3158          | 0.8970             |
-| 1.1209     | 6250      | 0.3531        | 0.3132          | 0.8968             |
-| 1.1657     | 6500      | 0.3516        | 0.3129          | 0.8974             |
-| 1.2105     | 6750      | 0.3511        | 0.3108          | 0.8973             |
-| 1.2554     | 7000      | 0.3494        | 0.3098          | 0.8975             |
-| 1.3002     | 7250      | 0.35          | 0.3086          | 0.8976             |
-| 1.3451     | 7500      | 0.3458        | 0.3081          | 0.8983             |
-| 1.3899     | 7750      | 0.3453        | 0.3072          | 0.8980             |
-| 1.4347     | 8000      | 0.3426        | 0.3066          | 0.8984             |
-| 1.4796     | 8250      | 0.3427        | 0.3042          | 0.8987             |
-| 1.5244     | 8500      | 0.342         | 0.3046          | 0.8992             |
-| 1.5692     | 8750      | 0.3404        | 0.3037          | 0.8994             |
-| 1.6141     | 9000      | 0.339         | 0.3027          | 0.8996             |
-| 1.6589     | 9250      | 0.3392        | 0.3015          | 0.8996             |
-| 1.7037     | 9500      | 0.3377        | 0.3012          | 0.8999             |
-| 1.7486     | 9750      | 0.3391        | 0.3007          | 0.8999             |
-| **1.7934** | **10000** | **0.3365**    | **0.3004**      | **0.8999**         |
-
-* The bold row denotes the saved checkpoint.
+| Epoch  | Step | Training Loss |
+|:------:|:----:|:-------------:|
+| 0.3199 | 500  | 0.4294        |
+| 0.6398 | 1000 | 0.1268        |
+| 0.9597 | 1500 | 0.1           |
+| 1.2796 | 2000 | 0.0792        |
+| 1.5995 | 2500 | 0.0706        |
+| 1.9194 | 3000 | 0.0687        |
+| 2.2393 | 3500 | 0.0584        |
+| 2.5592 | 4000 | 0.057         |
+| 2.8791 | 4500 | 0.0581        |
+
 
 ### Framework Versions
 - Python: 3.10.18

config.json

@@ -1,24 +1,60 @@
 {
+  "_sliding_window_pattern": 6,
   "architectures": [
-    "BertModel"
+    "Gemma3TextModel"
   ],
-  "attention_probs_dropout_prob": 0.1,
-  "classifier_dropout": null,
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "attn_logit_softcapping": null,
+  "bos_token_id": 2,
   "dtype": "float32",
-  "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.1,
-  "hidden_size": 384,
+  "eos_token_id": 1,
+  "final_logit_softcapping": null,
+  "head_dim": 256,
+  "hidden_activation": "gelu_pytorch_tanh",
+  "hidden_size": 768,
   "initializer_range": 0.02,
-  "intermediate_size": 1536,
-  "layer_norm_eps": 1e-12,
-  "max_position_embeddings": 512,
-  "model_type": "bert",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
+  "intermediate_size": 1152,
+  "layer_types": [
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 2048,
+  "model_type": "gemma3_text",
+  "num_attention_heads": 3,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 1,
   "pad_token_id": 0,
-  "position_embedding_type": "absolute",
+  "query_pre_attn_scalar": 256,
+  "rms_norm_eps": 1e-06,
+  "rope_local_base_freq": 10000.0,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 257,
   "transformers_version": "4.57.3",
-  "type_vocab_size": 2,
+  "use_bidirectional_attention": true,
   "use_cache": true,
-  "vocab_size": 30522
+  "vocab_size": 262144
 }

eval/Information-Retrieval_evaluation_val_results.csv

@@ -637,3 +637,24 @@ epoch,steps,cosine-Accuracy@1,cosine-Accuracy@3,cosine-Accuracy@5,cosine-Precisi
 1.7037302725968435,9500,0.83535,0.911475,0.93655,0.83535,0.83535,0.3038249999999999,0.911475,0.18731,0.93655,0.83535,0.8751141666666615,0.878991577380946,0.899889550569265,0.8809631243794993
 1.7485652797704447,9750,0.835325,0.9112,0.936325,0.835325,0.835325,0.3037333333333333,0.9112,0.18726500000000001,0.936325,0.835325,0.8750524999999947,0.8789839384920574,0.8998852971689221,0.8809643466594432
 1.793400286944046,10000,0.83545,0.911175,0.9366,0.83545,0.83545,0.303725,0.911175,0.18732000000000001,0.9366,0.83545,0.8751591666666616,0.8790415476190412,0.8999318372974409,0.8810239994800558
+0,0,0.639675,0.87685,0.9083,0.639675,0.639675,0.2922833333333333,0.87685,0.18166000000000004,0.9083,0.639675,0.7599624999999923,0.7643178273809481,0.8083956818551257,0.7670522600703508
+0.022417503586800575,250,0.82065,0.89945,0.9242,0.82065,0.82065,0.2998166666666666,0.89945,0.18483999999999998,0.9242,0.82065,0.8617137499999968,0.8657987599206315,0.8872724349019027,0.8680466843836286
+0.04483500717360115,500,0.81655,0.892425,0.91875,0.81655,0.81655,0.29747499999999993,0.892425,0.18375000000000002,0.91875,0.81655,0.8566166666666617,0.8607229662698371,0.882122156254855,0.863151001734206
+0.06725251076040172,750,0.8135,0.892075,0.917625,0.8135,0.8135,0.2973583333333333,0.892075,0.18352500000000005,0.917625,0.8135,0.8543795833333286,0.8584948015872975,0.8802034869577888,0.8609313205301325
+0.0896700143472023,1000,0.8244,0.898125,0.9234,0.8244,0.8244,0.299375,0.898125,0.18468000000000004,0.9234,0.8244,0.8632404166666602,0.8673072817460247,0.8881903001430266,0.8695682867122491
+0.11208751793400287,1250,0.825125,0.8975,0.92345,0.825125,0.825125,0.2991666666666666,0.8975,0.18469000000000002,0.92345,0.825125,0.8635195833333286,0.8675001488095195,0.8881479223773198,0.8698036797223343
+0.13450502152080343,1500,0.824825,0.89745,0.924675,0.824825,0.824825,0.29914999999999997,0.89745,0.18493500000000002,0.924675,0.824825,0.8635674999999957,0.867597232142853,0.8885981237608775,0.8699162638314769
+0.15692252510760402,1750,0.828125,0.9003,0.926725,0.828125,0.828125,0.3001,0.9003,0.18534500000000004,0.926725,0.828125,0.8663654166666617,0.8703199404761844,0.8910206909869869,0.8725571487400507
+0.1793400286944046,2000,0.001325,0.0084,0.022275,0.001325,0.001325,0.0028,0.0084,0.004455,0.022275,0.001325,0.007867083333333311,0.010877668650793688,0.0186950894860071,0.02490441055844739
+0.20175753228120516,2250,0.830025,0.902425,0.92865,0.830025,0.830025,0.3008083333333333,0.902425,0.18573,0.92865,0.830025,0.868462916666662,0.8723177777777715,0.8928818077159327,0.8744904419955204
+0.22417503586800575,2500,0.83385,0.905775,0.931825,0.83385,0.83385,0.30192499999999994,0.905775,0.18636500000000006,0.931825,0.83385,0.871977083333331,0.8756928670634883,0.8959559117645085,0.8778776795620055
+0.2465925394548063,2750,0.0001,0.0752,0.174675,0.0001,0.0001,0.025066666666666664,0.0752,0.034935,0.174675,0.0001,0.05248083333333305,0.055306488095238034,0.08872108035834274,0.06968419084794351
+0.26901004304160686,3000,2.5e-05,0.072325,0.142725,2.5e-05,2.5e-05,0.024108333333333332,0.072325,0.028544999999999997,0.142725,2.5e-05,0.04569916666666664,0.05269701388888883,0.08544234767952956,0.06723497471913427
+0.2914275466284075,3250,2.5e-05,0.072325,0.142725,2.5e-05,2.5e-05,0.024108333333333332,0.072325,0.028544999999999997,0.142725,2.5e-05,0.04569916666666664,0.05269701388888883,0.08544234767952956,0.06723497471913427
+0.31384505021520803,3500,2.5e-05,0.072325,0.142725,2.5e-05,2.5e-05,0.024108333333333332,0.072325,0.028544999999999997,0.142725,2.5e-05,0.04569916666666664,0.05269701388888883,0.08544234767952956,0.06723497471913427
+0.3362625538020086,3750,2.5e-05,0.072325,0.142725,2.5e-05,2.5e-05,0.024108333333333332,0.072325,0.028544999999999997,0.142725,2.5e-05,0.04569916666666664,0.05269701388888883,0.08544234767952956,0.06723497471913427
+0.3586800573888092,4000,2.5e-05,0.072325,0.142725,2.5e-05,2.5e-05,0.024108333333333332,0.072325,0.028544999999999997,0.142725,2.5e-05,0.04569916666666664,0.05269701388888883,0.08544234767952956,0.06723497471913427
+0.38109756097560976,4250,0.0,2.5e-05,7.5e-05,0.0,0.0,8.333333333333332e-06,2.5e-05,1.5000000000000002e-05,7.5e-05,0.0,2.3749999999999998e-05,2.6875e-05,4.409809989162802e-05,5.524355841383588e-05
+0.4035150645624103,4500,0.0,2.5e-05,7.5e-05,0.0,0.0,8.333333333333332e-06,2.5e-05,1.5000000000000002e-05,7.5e-05,0.0,2.3749999999999998e-05,2.6875e-05,4.409809989162802e-05,5.524355841383588e-05
+0.4259325681492109,4750,0.0,2.5e-05,7.5e-05,0.0,0.0,8.333333333333332e-06,2.5e-05,1.5000000000000002e-05,7.5e-05,0.0,2.3749999999999998e-05,2.6875e-05,4.409809989162802e-05,5.524355841383588e-05
+0.4483500717360115,5000,0.0,2.5e-05,7.5e-05,0.0,0.0,8.333333333333332e-06,2.5e-05,1.5000000000000002e-05,7.5e-05,0.0,2.3749999999999998e-05,2.6875e-05,4.409809989162802e-05,5.524355841383588e-05

final_metrics.json

@@ -1,16 +1,16 @@
 {
-  "val_cosine_accuracy@1": 0.83295,
-  "val_cosine_accuracy@3": 0.9071,
-  "val_cosine_accuracy@5": 0.9329,
-  "val_cosine_precision@1": 0.83295,
-  "val_cosine_precision@3": 0.3023666666666666,
-  "val_cosine_precision@5": 0.18658000000000005,
-  "val_cosine_recall@1": 0.83295,
-  "val_cosine_recall@3": 0.9071,
-  "val_cosine_recall@5": 0.9329,
-  "val_cosine_ndcg@10": 0.8970951855878305,
-  "val_cosine_mrr@1": 0.83295,
-  "val_cosine_mrr@5": 0.872013749999996,
-  "val_cosine_mrr@10": 0.8760916468253912,
-  "val_cosine_map@100": 0.8781372459990227
+  "val_cosine_accuracy@1": 0.83545,
+  "val_cosine_accuracy@3": 0.911175,
+  "val_cosine_accuracy@5": 0.9366,
+  "val_cosine_precision@1": 0.83545,
+  "val_cosine_precision@3": 0.303725,
+  "val_cosine_precision@5": 0.18732000000000001,
+  "val_cosine_recall@1": 0.83545,
+  "val_cosine_recall@3": 0.911175,
+  "val_cosine_recall@5": 0.9366,
+  "val_cosine_ndcg@10": 0.8999318372974409,
+  "val_cosine_mrr@1": 0.83545,
+  "val_cosine_mrr@5": 0.8751591666666616,
+  "val_cosine_mrr@10": 0.8790415476190412,
+  "val_cosine_map@100": 0.8810239994800558
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bb34c96d597d832aa5edd5352a2499b5d5c6df76a439f4cead08dce5b23e26ed
-size 133462128
+oid sha256:545b247747446f3974b25dce1c7156ba3f4d45cf47a42bc853d38de5dc798cfe
+size 1211486072

modules.json

@@ -10,11 +10,5 @@
     "name": "1",
     "path": "1_Pooling",
     "type": "sentence_transformers.models.Pooling"
-  },
-  {
-    "idx": 2,
-    "name": "2",
-    "path": "2_Normalize",
-    "type": "sentence_transformers.models.Normalize"
   }
 ]

special_tokens_map.json

@@ -1,34 +1,30 @@
 {
-  "cls_token": {
-    "content": "[CLS]",
+  "boi_token": "<start_of_image>",
+  "bos_token": {
+    "content": "<bos>",
     "lstrip": false,
     "normalized": false,
     "rstrip": false,
     "single_word": false
   },
-  "mask_token": {
-    "content": "[MASK]",
+  "eoi_token": "<end_of_image>",
+  "eos_token": {
+    "content": "<eos>",
     "lstrip": false,
     "normalized": false,
     "rstrip": false,
     "single_word": false
   },
+  "image_token": "<image_soft_token>",
   "pad_token": {
-    "content": "[PAD]",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
-  "sep_token": {
-    "content": "[SEP]",
+    "content": "<pad>",
     "lstrip": false,
     "normalized": false,
     "rstrip": false,
     "single_word": false
   },
   "unk_token": {
-    "content": "[UNK]",
+    "content": "<unk>",
     "lstrip": false,
     "normalized": false,
     "rstrip": false,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aa697a5587cabb6e578299c46e1745cc69a0c459a2f99c151e144c2158d43a4e
+oid sha256:0550728e68dcb4297a5ed369ce43e645f35ae6cb79deed45333999adba8e4c4e
 size 6161