Hugging Face's logo

yenstdi
/
embbedding_text_1111

Sentence Similarity
sentence-transformers
Safetensors
bert
feature-extraction
dense
Generated from Trainer
dataset_size:50
loss:MultipleNegativesRankingLoss
text-embeddings-inference
Model card Files
xet
Community
yenstdi commited on
Commit
b91f9b9
·
verified ·
1 Parent(s): 4c0ff50

Upload folder using huggingface_hub

Browse files
Files changed (34) hide show
  1. README.md +24 -21
  2. checkpoint-12/1_Pooling/config.json +10 -0
  3. checkpoint-12/README.md +408 -0
  4. checkpoint-12/config.json +39 -0
  5. checkpoint-12/config_sentence_transformers.json +14 -0
  6. checkpoint-12/model.safetensors +3 -0
  7. checkpoint-12/modules.json +20 -0
  8. checkpoint-12/optimizer.pt +3 -0
  9. checkpoint-12/rng_state.pth +3 -0
  10. checkpoint-12/scheduler.pt +3 -0
  11. checkpoint-12/sentence_bert_config.json +4 -0
  12. checkpoint-12/special_tokens_map.json +37 -0
  13. checkpoint-12/tokenizer.json +0 -0
  14. checkpoint-12/tokenizer_config.json +58 -0
  15. checkpoint-12/trainer_state.json +142 -0
  16. checkpoint-12/training_args.bin +3 -0
  17. checkpoint-12/vocab.txt +0 -0
  18. checkpoint-8/1_Pooling/config.json +10 -0
  19. checkpoint-8/README.md +404 -0
  20. checkpoint-8/config.json +39 -0
  21. checkpoint-8/config_sentence_transformers.json +14 -0
  22. checkpoint-8/model.safetensors +3 -0
  23. checkpoint-8/modules.json +20 -0
  24. checkpoint-8/optimizer.pt +3 -0
  25. checkpoint-8/rng_state.pth +3 -0
  26. checkpoint-8/scheduler.pt +3 -0
  27. checkpoint-8/sentence_bert_config.json +4 -0
  28. checkpoint-8/special_tokens_map.json +37 -0
  29. checkpoint-8/tokenizer.json +0 -0
  30. checkpoint-8/tokenizer_config.json +58 -0
  31. checkpoint-8/trainer_state.json +106 -0
  32. checkpoint-8/training_args.bin +3 -0
  33. checkpoint-8/vocab.txt +0 -0
  34. model.safetensors +1 -1
README.md CHANGED
@@ -102,9 +102,9 @@ print(embeddings.shape)
102
  # Get the similarity scores for the embeddings
103
  similarities = model.similarity(embeddings, embeddings)
104
  print(similarities)
105
- # tensor([[1.0000, 0.6986, 0.1182],
106
- # [0.6986, 1.0000, 0.1618],
107
- # [0.1182, 0.1618, 1.0000]])
108
  ```
109
 
110
  <!--
@@ -203,9 +203,10 @@ You can finetune this model on your own dataset.
203
  #### Non-Default Hyperparameters
204
 
205
  - `eval_strategy`: epoch
206
- - `per_device_train_batch_size`: 64
207
- - `per_device_eval_batch_size`: 64
208
- - `num_train_epochs`: 10
 
209
  - `warmup_steps`: 100
210
 
211
  #### All Hyperparameters
@@ -215,22 +216,22 @@ You can finetune this model on your own dataset.
215
  - `do_predict`: False
216
  - `eval_strategy`: epoch
217
  - `prediction_loss_only`: True
218
- - `per_device_train_batch_size`: 64
219
- - `per_device_eval_batch_size`: 64
220
  - `per_gpu_train_batch_size`: None
221
  - `per_gpu_eval_batch_size`: None
222
  - `gradient_accumulation_steps`: 1
223
  - `eval_accumulation_steps`: None
224
  - `torch_empty_cache_steps`: None
225
- - `learning_rate`: 5e-05
226
  - `weight_decay`: 0.0
227
  - `adam_beta1`: 0.9
228
  - `adam_beta2`: 0.999
229
  - `adam_epsilon`: 1e-08
230
  - `max_grad_norm`: 1.0
231
- - `num_train_epochs`: 10
232
  - `max_steps`: -1
233
- - `lr_scheduler_type`: linear
234
  - `lr_scheduler_kwargs`: None
235
  - `warmup_ratio`: 0.0
236
  - `warmup_steps`: 100
@@ -336,16 +337,18 @@ You can finetune this model on your own dataset.
336
  ### Training Logs
337
  | Epoch | Step | Training Loss | Validation Loss |
338
  |:-----:|:----:|:-------------:|:---------------:|
339
- | 1.0 | 1 | 2.4771 | 0.4011 |
340
- | 2.0 | 2 | 2.5696 | 0.3978 |
341
- | 3.0 | 3 | 2.4096 | 0.3917 |
342
- | 4.0 | 4 | 2.4025 | 0.3832 |
343
- | 5.0 | 5 | 2.2429 | 0.3730 |
344
- | 6.0 | 6 | 2.1532 | 0.3615 |
345
- | 7.0 | 7 | 2.0347 | 0.3499 |
346
- | 8.0 | 8 | 1.8817 | 0.3384 |
347
- | 9.0 | 9 | 1.7143 | 0.3277 |
348
- | 10.0 | 10 | 1.4908 | 0.3180 |
 
 
349
 
350
 
351
  ### Framework Versions
 
102
  # Get the similarity scores for the embeddings
103
  similarities = model.similarity(embeddings, embeddings)
104
  print(similarities)
105
+ # tensor([[1.0000, 0.6704, 0.1423],
106
+ # [0.6704, 1.0000, 0.1834],
107
+ # [0.1423, 0.1834, 1.0000]])
108
  ```
109
 
110
  <!--
 
203
  #### Non-Default Hyperparameters
204
 
205
  - `eval_strategy`: epoch
206
+ - `per_device_train_batch_size`: 16
207
+ - `per_device_eval_batch_size`: 16
208
+ - `learning_rate`: 2e-05
209
+ - `lr_scheduler_type`: cosine
210
  - `warmup_steps`: 100
211
 
212
  #### All Hyperparameters
 
216
  - `do_predict`: False
217
  - `eval_strategy`: epoch
218
  - `prediction_loss_only`: True
219
+ - `per_device_train_batch_size`: 16
220
+ - `per_device_eval_batch_size`: 16
221
  - `per_gpu_train_batch_size`: None
222
  - `per_gpu_eval_batch_size`: None
223
  - `gradient_accumulation_steps`: 1
224
  - `eval_accumulation_steps`: None
225
  - `torch_empty_cache_steps`: None
226
+ - `learning_rate`: 2e-05
227
  - `weight_decay`: 0.0
228
  - `adam_beta1`: 0.9
229
  - `adam_beta2`: 0.999
230
  - `adam_epsilon`: 1e-08
231
  - `max_grad_norm`: 1.0
232
+ - `num_train_epochs`: 3
233
  - `max_steps`: -1
234
+ - `lr_scheduler_type`: cosine
235
  - `lr_scheduler_kwargs`: None
236
  - `warmup_ratio`: 0.0
237
  - `warmup_steps`: 100
 
337
  ### Training Logs
338
  | Epoch | Step | Training Loss | Validation Loss |
339
  |:-----:|:----:|:-------------:|:---------------:|
340
+ | 0.25 | 1 | 1.3902 | - |
341
+ | 0.5 | 2 | 1.6712 | - |
342
+ | 0.75 | 3 | 1.2991 | - |
343
+ | 1.0 | 4 | 1.3125 | 0.1941 |
344
+ | 1.25 | 5 | 1.6758 | - |
345
+ | 1.5 | 6 | 1.5893 | - |
346
+ | 1.75 | 7 | 1.2746 | - |
347
+ | 2.0 | 8 | 0.0071 | 0.1854 |
348
+ | 2.25 | 9 | 1.236 | - |
349
+ | 2.5 | 10 | 1.0984 | - |
350
+ | 2.75 | 11 | 1.208 | - |
351
+ | 3.0 | 12 | 0.3278 | 0.1744 |
352
 
353
 
354
  ### Framework Versions
checkpoint-12/1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 1024,
3
+ "pooling_mode_cls_token": true,
4
+ "pooling_mode_mean_tokens": false,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
checkpoint-12/README.md ADDED
@@ -0,0 +1,408 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - sentence-transformers
4
+ - sentence-similarity
5
+ - feature-extraction
6
+ - dense
7
+ - generated_from_trainer
8
+ - dataset_size:50
9
+ - loss:MultipleNegativesRankingLoss
10
+ base_model: BAAI/bge-large-zh-v1.5
11
+ widget:
12
+ - source_sentence: 定期定額投資的優缺點
13
+ sentences:
14
+ - 近年來大型語言模型與擴散模型在圖像與文本生成領域取得突破性進展。
15
+ - 國際間的生產與物流體系正在發生重大的組織變革與調整。
16
+ - 透過固定金額長期投入，投資者能有效攤平市場波動帶來的成本風險，但可能在強勁牛市中錯失更高的單筆申購報酬。
17
+ - source_sentence: 京都最適合賞楓的季節是什麼時候？
18
+ sentences:
19
+ - 秋季前往關西地區，十一月中旬到十二月初通常是觀賞紅葉的最佳時機。
20
+ - 使用 asyncio 庫可以實現非阻塞的 I/O 操作，顯著提升網路爬蟲或 API 請求的並發性能。
21
+ - 在快速變遷的職場環境中，持續獲取新知識與技能是維持個人競爭力與適應力的關鍵。
22
+ - source_sentence: 長期失眠該如何改善？
23
+ sentences:
24
+ - 建立規律的作息時間、減少睡前使用電子產品，並營造舒適的睡眠環境有助於緩解睡眠障礙。
25
+ - 植物透過葉綠體吸收太陽能，將二氧化碳與水轉化為葡萄糖並釋放氧氣，這是地球能量循環的基礎。
26
+ - 辦理信用貸款通常要求穩定的收入證明與良好的信用評分。
27
+ - source_sentence: 如何減少日常生活中的碳足跡
28
+ sentences:
29
+ - 在推動組織數位化過程中，往往會面臨技術債、員工抗拒改變以及缺乏清晰策略等難題。
30
+ - 該行動裝置的電力持久度表現優異，能滿足長時間使用的需求。
31
+ - 透過節能家電、搭乘大眾運輸及實踐蔬食生活，能有效降低個人的環境影響。
32
+ - source_sentence: 京都最值得造訪的歷史古蹟
33
+ sentences:
34
+ - 這座日本古都擁有眾多世界文化遺產，如清水寺、金閣寺與伏見稻荷大社，是體驗傳統文化的必經之地。
35
+ - 患者通常會感到胸口灼熱（俗稱火燒心）、胃酸逆流，有時還會伴隨慢性咳嗽或喉嚨發炎。
36
+ - 這種以植物性食物、橄欖油和適量深海魚為主的飲食模式，被證實能有效預防心血管疾病。
37
+ datasets:
38
+ - yenstdi/embbedding_text_1111
39
+ pipeline_tag: sentence-similarity
40
+ library_name: sentence-transformers
41
+ ---
42
+
43
+ # SentenceTransformer based on BAAI/bge-large-zh-v1.5
44
+
45
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [BAAI/bge-large-zh-v1.5](https://huggingface.co/BAAI/bge-large-zh-v1.5) on the [embbedding_text_1111](https://huggingface.co/datasets/yenstdi/embbedding_text_1111) dataset. It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
46
+
47
+ ## Model Details
48
+
49
+ ### Model Description
50
+ - **Model Type:** Sentence Transformer
51
+ - **Base model:** [BAAI/bge-large-zh-v1.5](https://huggingface.co/BAAI/bge-large-zh-v1.5) <!-- at revision 79e7739b6ab944e86d6171e44d24c997fc1e0116 -->
52
+ - **Maximum Sequence Length:** 512 tokens
53
+ - **Output Dimensionality:** 1024 dimensions
54
+ - **Similarity Function:** Cosine Similarity
55
+ - **Training Dataset:**
56
+ - [embbedding_text_1111](https://huggingface.co/datasets/yenstdi/embbedding_text_1111)
57
+ <!-- - **Language:** Unknown -->
58
+ <!-- - **License:** Unknown -->
59
+
60
+ ### Model Sources
61
+
62
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
63
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers)
64
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
65
+
66
+ ### Full Model Architecture
67
+
68
+ ```
69
+ SentenceTransformer(
70
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': True, 'architecture': 'BertModel'})
71
+ (1): Pooling({'word_embedding_dimension': 1024, '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})
72
+ (2): Normalize()
73
+ )
74
+ ```
75
+
76
+ ## Usage
77
+
78
+ ### Direct Usage (Sentence Transformers)
79
+
80
+ First install the Sentence Transformers library:
81
+
82
+ ```bash
83
+ pip install -U sentence-transformers
84
+ ```
85
+
86
+ Then you can load this model and run inference.
87
+ ```python
88
+ from sentence_transformers import SentenceTransformer
89
+
90
+ # Download from the 🤗 Hub
91
+ model = SentenceTransformer("sentence_transformers_model_id")
92
+ # Run inference
93
+ sentences = [
94
+ '京都最值得造訪的歷史古蹟',
95
+ '這座日本古都擁有眾多世界文化遺產，如清水寺、金閣寺與伏見稻荷大社，是體驗傳統文化的必經之地。',
96
+ '患者通常會感到胸口灼熱（俗稱火燒心）、胃酸逆流，有時還會伴隨慢性咳嗽或喉嚨發炎。',
97
+ ]
98
+ embeddings = model.encode(sentences)
99
+ print(embeddings.shape)
100
+ # [3, 1024]
101
+
102
+ # Get the similarity scores for the embeddings
103
+ similarities = model.similarity(embeddings, embeddings)
104
+ print(similarities)
105
+ # tensor([[1.0000, 0.6704, 0.1423],
106
+ # [0.6704, 1.0000, 0.1834],
107
+ # [0.1423, 0.1834, 1.0000]])
108
+ ```
109
+
110
+ <!--
111
+ ### Direct Usage (Transformers)
112
+
113
+ <details><summary>Click to see the direct usage in Transformers</summary>
114
+
115
+ </details>
116
+ -->
117
+
118
+ <!--
119
+ ### Downstream Usage (Sentence Transformers)
120
+
121
+ You can finetune this model on your own dataset.
122
+
123
+ <details><summary>Click to expand</summary>
124
+
125
+ </details>
126
+ -->
127
+
128
+ <!--
129
+ ### Out-of-Scope Use
130
+
131
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
132
+ -->
133
+
134
+ <!--
135
+ ## Bias, Risks and Limitations
136
+
137
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
138
+ -->
139
+
140
+ <!--
141
+ ### Recommendations
142
+
143
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
144
+ -->
145
+
146
+ ## Training Details
147
+
148
+ ### Training Dataset
149
+
150
+ #### embbedding_text_1111
151
+
152
+ * Dataset: [embbedding_text_1111](https://huggingface.co/datasets/yenstdi/embbedding_text_1111) at [610ac14](https://huggingface.co/datasets/yenstdi/embbedding_text_1111/tree/610ac1456cc501416303e62f7813f2ee87ee95e3)
153
+ * Size: 50 training samples
154
+ * Columns: <code>anchor</code> and <code>positive</code>
155
+ * Approximate statistics based on the first 50 samples:
156
+ | | anchor | positive |
157
+ |:--------|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
158
+ | type | string | string |
159
+ | details | <ul><li>min: 11 tokens</li><li>mean: 16.02 tokens</li><li>max: 26 tokens</li></ul> | <ul><li>min: 26 tokens</li><li>mean: 39.0 tokens</li><li>max: 55 tokens</li></ul> |
160
+ * Samples:
161
+ | anchor | positive |
162
+ |:-----------------------------------|:---------------------------------------------------------------------------|
163
+ | <code>尋找熟悉 React 生態系統的前端開發者</code> | <code>應徵者需具備 React, Redux 及 Next.js 的實作經驗，並能運用 TypeScript 撰寫高品質程式碼。</code> |
164
+ | <code>後端 Python 工程師職缺要求</code> | <code>精通 Django 或 FastAPI 框架，並有使用 Celery 處理非同步任務與分散式隊列的經驗。</code> |
165
+ | <code>雲端架構師 (AWS 專長) 招募中</code> | <code>負責維運 EC2、S3 與 Lambda 等雲端資源，並能有效配置 RDS 資料庫以確保系統效能。</code> |
166
+ * Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
167
+ ```json
168
+ {
169
+ "scale": 20.0,
170
+ "similarity_fct": "cos_sim",
171
+ "gather_across_devices": false
172
+ }
173
+ ```
174
+
175
+ ### Evaluation Dataset
176
+
177
+ #### embbedding_text_1111
178
+
179
+ * Dataset: [embbedding_text_1111](https://huggingface.co/datasets/yenstdi/embbedding_text_1111) at [610ac14](https://huggingface.co/datasets/yenstdi/embbedding_text_1111/tree/610ac1456cc501416303e62f7813f2ee87ee95e3)
180
+ * Size: 25 evaluation samples
181
+ * Columns: <code>anchor</code> and <code>positive</code>
182
+ * Approximate statistics based on the first 25 samples:
183
+ | | anchor | positive |
184
+ |:--------|:-----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
185
+ | type | string | string |
186
+ | details | <ul><li>min: 11 tokens</li><li>mean: 15.16 tokens</li><li>max: 20 tokens</li></ul> | <ul><li>min: 28 tokens</li><li>mean: 39.36 tokens</li><li>max: 54 tokens</li></ul> |
187
+ * Samples:
188
+ | anchor | positive |
189
+ |:-------------------------------|:----------------------------------------------------------------------|
190
+ | <code>這款手機的電池續航力令人印象深刻。</code> | <code>該行動裝置的電力持久度表現優異，能滿足長時間使用的需求。</code> |
191
+ | <code>什麼是機器學習中的過擬合現象？</code> | <code>當模型在訓練數據上表現極佳，但在未見過的測試數據上預測準確率大幅下降時，通常就是發生了 Overfitting。</code> |
192
+ | <code>2024年全球永續能源趨勢報告</code> | <code>隨著各國減碳政策的推進，太陽能與離岸風電在未來幾年將成為再生能源成長的核心動力。</code> |
193
+ * Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
194
+ ```json
195
+ {
196
+ "scale": 20.0,
197
+ "similarity_fct": "cos_sim",
198
+ "gather_across_devices": false
199
+ }
200
+ ```
201
+
202
+ ### Training Hyperparameters
203
+ #### Non-Default Hyperparameters
204
+
205
+ - `eval_strategy`: epoch
206
+ - `per_device_train_batch_size`: 16
207
+ - `per_device_eval_batch_size`: 16
208
+ - `learning_rate`: 2e-05
209
+ - `lr_scheduler_type`: cosine
210
+ - `warmup_steps`: 100
211
+
212
+ #### All Hyperparameters
213
+ <details><summary>Click to expand</summary>
214
+
215
+ - `overwrite_output_dir`: False
216
+ - `do_predict`: False
217
+ - `eval_strategy`: epoch
218
+ - `prediction_loss_only`: True
219
+ - `per_device_train_batch_size`: 16
220
+ - `per_device_eval_batch_size`: 16
221
+ - `per_gpu_train_batch_size`: None
222
+ - `per_gpu_eval_batch_size`: None
223
+ - `gradient_accumulation_steps`: 1
224
+ - `eval_accumulation_steps`: None
225
+ - `torch_empty_cache_steps`: None
226
+ - `learning_rate`: 2e-05
227
+ - `weight_decay`: 0.0
228
+ - `adam_beta1`: 0.9
229
+ - `adam_beta2`: 0.999
230
+ - `adam_epsilon`: 1e-08
231
+ - `max_grad_norm`: 1.0
232
+ - `num_train_epochs`: 3
233
+ - `max_steps`: -1
234
+ - `lr_scheduler_type`: cosine
235
+ - `lr_scheduler_kwargs`: None
236
+ - `warmup_ratio`: 0.0
237
+ - `warmup_steps`: 100
238
+ - `log_level`: passive
239
+ - `log_level_replica`: warning
240
+ - `log_on_each_node`: True
241
+ - `logging_nan_inf_filter`: True
242
+ - `save_safetensors`: True
243
+ - `save_on_each_node`: False
244
+ - `save_only_model`: False
245
+ - `restore_callback_states_from_checkpoint`: False
246
+ - `no_cuda`: False
247
+ - `use_cpu`: False
248
+ - `use_mps_device`: False
249
+ - `seed`: 42
250
+ - `data_seed`: None
251
+ - `jit_mode_eval`: False
252
+ - `bf16`: False
253
+ - `fp16`: False
254
+ - `fp16_opt_level`: O1
255
+ - `half_precision_backend`: auto
256
+ - `bf16_full_eval`: False
257
+ - `fp16_full_eval`: False
258
+ - `tf32`: None
259
+ - `local_rank`: 0
260
+ - `ddp_backend`: None
261
+ - `tpu_num_cores`: None
262
+ - `tpu_metrics_debug`: False
263
+ - `debug`: []
264
+ - `dataloader_drop_last`: False
265
+ - `dataloader_num_workers`: 0
266
+ - `dataloader_prefetch_factor`: None
267
+ - `past_index`: -1
268
+ - `disable_tqdm`: False
269
+ - `remove_unused_columns`: True
270
+ - `label_names`: None
271
+ - `load_best_model_at_end`: False
272
+ - `ignore_data_skip`: False
273
+ - `fsdp`: []
274
+ - `fsdp_min_num_params`: 0
275
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
276
+ - `fsdp_transformer_layer_cls_to_wrap`: None
277
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
278
+ - `parallelism_config`: None
279
+ - `deepspeed`: None
280
+ - `label_smoothing_factor`: 0.0
281
+ - `optim`: adamw_torch_fused
282
+ - `optim_args`: None
283
+ - `adafactor`: False
284
+ - `group_by_length`: False
285
+ - `length_column_name`: length
286
+ - `project`: huggingface
287
+ - `trackio_space_id`: trackio
288
+ - `ddp_find_unused_parameters`: None
289
+ - `ddp_bucket_cap_mb`: None
290
+ - `ddp_broadcast_buffers`: False
291
+ - `dataloader_pin_memory`: True
292
+ - `dataloader_persistent_workers`: False
293
+ - `skip_memory_metrics`: True
294
+ - `use_legacy_prediction_loop`: False
295
+ - `push_to_hub`: False
296
+ - `resume_from_checkpoint`: None
297
+ - `hub_model_id`: None
298
+ - `hub_strategy`: every_save
299
+ - `hub_private_repo`: None
300
+ - `hub_always_push`: False
301
+ - `hub_revision`: None
302
+ - `gradient_checkpointing`: False
303
+ - `gradient_checkpointing_kwargs`: None
304
+ - `include_inputs_for_metrics`: False
305
+ - `include_for_metrics`: []
306
+ - `eval_do_concat_batches`: True
307
+ - `fp16_backend`: auto
308
+ - `push_to_hub_model_id`: None
309
+ - `push_to_hub_organization`: None
310
+ - `mp_parameters`:
311
+ - `auto_find_batch_size`: False
312
+ - `full_determinism`: False
313
+ - `torchdynamo`: None
314
+ - `ray_scope`: last
315
+ - `ddp_timeout`: 1800
316
+ - `torch_compile`: False
317
+ - `torch_compile_backend`: None
318
+ - `torch_compile_mode`: None
319
+ - `include_tokens_per_second`: False
320
+ - `include_num_input_tokens_seen`: no
321
+ - `neftune_noise_alpha`: None
322
+ - `optim_target_modules`: None
323
+ - `batch_eval_metrics`: False
324
+ - `eval_on_start`: False
325
+ - `use_liger_kernel`: False
326
+ - `liger_kernel_config`: None
327
+ - `eval_use_gather_object`: False
328
+ - `average_tokens_across_devices`: True
329
+ - `prompts`: None
330
+ - `batch_sampler`: batch_sampler
331
+ - `multi_dataset_batch_sampler`: proportional
332
+ - `router_mapping`: {}
333
+ - `learning_rate_mapping`: {}
334
+
335
+ </details>
336
+
337
+ ### Training Logs
338
+ | Epoch | Step | Training Loss | Validation Loss |
339
+ |:-----:|:----:|:-------------:|:---------------:|
340
+ | 0.25 | 1 | 1.3902 | - |
341
+ | 0.5 | 2 | 1.6712 | - |
342
+ | 0.75 | 3 | 1.2991 | - |
343
+ | 1.0 | 4 | 1.3125 | 0.1941 |
344
+ | 1.25 | 5 | 1.6758 | - |
345
+ | 1.5 | 6 | 1.5893 | - |
346
+ | 1.75 | 7 | 1.2746 | - |
347
+ | 2.0 | 8 | 0.0071 | 0.1854 |
348
+ | 2.25 | 9 | 1.236 | - |
349
+ | 2.5 | 10 | 1.0984 | - |
350
+ | 2.75 | 11 | 1.208 | - |
351
+ | 3.0 | 12 | 0.3278 | 0.1744 |
352
+
353
+
354
+ ### Framework Versions
355
+ - Python: 3.12.12
356
+ - Sentence Transformers: 5.2.0
357
+ - Transformers: 4.57.6
358
+ - PyTorch: 2.9.0+cu126
359
+ - Accelerate: 1.12.0
360
+ - Datasets: 4.0.0
361
+ - Tokenizers: 0.22.2
362
+
363
+ ## Citation
364
+
365
+ ### BibTeX
366
+
367
+ #### Sentence Transformers
368
+ ```bibtex
369
+ @inproceedings{reimers-2019-sentence-bert,
370
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
371
+ author = "Reimers, Nils and Gurevych, Iryna",
372
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
373
+ month = "11",
374
+ year = "2019",
375
+ publisher = "Association for Computational Linguistics",
376
+ url = "https://arxiv.org/abs/1908.10084",
377
+ }
378
+ ```
379
+
380
+ #### MultipleNegativesRankingLoss
381
+ ```bibtex
382
+ @misc{henderson2017efficient,
383
+ title={Efficient Natural Language Response Suggestion for Smart Reply},
384
+ author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
385
+ year={2017},
386
+ eprint={1705.00652},
387
+ archivePrefix={arXiv},
388
+ primaryClass={cs.CL}
389
+ }
390
+ ```
391
+
392
+ <!--
393
+ ## Glossary
394
+
395
+ *Clearly define terms in order to be accessible across audiences.*
396
+ -->
397
+
398
+ <!--
399
+ ## Model Card Authors
400
+
401
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
402
+ -->
403
+
404
+ <!--
405
+ ## Model Card Contact
406
+
407
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
408
+ -->
checkpoint-12/config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertModel"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "dtype": "float32",
10
+ "eos_token_id": 2,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.1,
13
+ "hidden_size": 1024,
14
+ "id2label": {
15
+ "0": "LABEL_0"
16
+ },
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 4096,
19
+ "label2id": {
20
+ "LABEL_0": 0
21
+ },
22
+ "layer_norm_eps": 1e-12,
23
+ "max_position_embeddings": 512,
24
+ "model_type": "bert",
25
+ "num_attention_heads": 16,
26
+ "num_hidden_layers": 24,
27
+ "output_past": true,
28
+ "pad_token_id": 0,
29
+ "pooler_fc_size": 768,
30
+ "pooler_num_attention_heads": 12,
31
+ "pooler_num_fc_layers": 3,
32
+ "pooler_size_per_head": 128,
33
+ "pooler_type": "first_token_transform",
34
+ "position_embedding_type": "absolute",
35
+ "transformers_version": "4.57.6",
36
+ "type_vocab_size": 2,
37
+ "use_cache": true,
38
+ "vocab_size": 21128
39
+ }
checkpoint-12/config_sentence_transformers.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "5.2.0",
4
+ "transformers": "4.57.6",
5
+ "pytorch": "2.9.0+cu126"
6
+ },
7
+ "model_type": "SentenceTransformer",
8
+ "prompts": {
9
+ "query": "",
10
+ "document": ""
11
+ },
12
+ "default_prompt_name": null,
13
+ "similarity_fn_name": "cosine"
14
+ }
checkpoint-12/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aaaf117f7a606f380f476d15425eab0047f36fb55fb45fb7b92c4546a682d35f
3
+ size 1302134568
checkpoint-12/modules.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ },
14
+ {
15
+ "idx": 2,
16
+ "name": "2",
17
+ "path": "2_Normalize",
18
+ "type": "sentence_transformers.models.Normalize"
19
+ }
20
+ ]
checkpoint-12/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:879ba8df7f7dd89ec0fcdaf6dfbaaa450897f9d56a2ed1931ce55af82ce310da
3
+ size 2596114098
checkpoint-12/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1ea0bf1f1e5c8edf3e14bafad53f8103d77ffe7eaa1b70a6958eae020400152d
3
+ size 14645
checkpoint-12/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2814f0065eab971c741e00f027a2b389c3ff3fca3e57467e5539e719246067e
3
+ size 1465
checkpoint-12/sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": true
4
+ }
checkpoint-12/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
checkpoint-12/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-12/tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": true,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 512,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "BertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
checkpoint-12/trainer_state.json ADDED
@@ -0,0 +1,142 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 3.0,
6
+ "eval_steps": 500,
7
+ "global_step": 12,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.25,
14
+ "grad_norm": 22.533252716064453,
15
+ "learning_rate": 0.0,
16
+ "loss": 1.3902,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.5,
21
+ "grad_norm": 20.85812759399414,
22
+ "learning_rate": 2.0000000000000002e-07,
23
+ "loss": 1.6712,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.75,
28
+ "grad_norm": 20.013919830322266,
29
+ "learning_rate": 4.0000000000000003e-07,
30
+ "loss": 1.2991,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 1.0,
35
+ "grad_norm": 72.03453063964844,
36
+ "learning_rate": 6.000000000000001e-07,
37
+ "loss": 1.3125,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 1.0,
42
+ "eval_loss": 0.19411715865135193,
43
+ "eval_runtime": 0.3249,
44
+ "eval_samples_per_second": 76.938,
45
+ "eval_steps_per_second": 6.155,
46
+ "step": 4
47
+ },
48
+ {
49
+ "epoch": 1.25,
50
+ "grad_norm": 26.436626434326172,
51
+ "learning_rate": 8.000000000000001e-07,
52
+ "loss": 1.6758,
53
+ "step": 5
54
+ },
55
+ {
56
+ "epoch": 1.5,
57
+ "grad_norm": 19.47249412536621,
58
+ "learning_rate": 1.0000000000000002e-06,
59
+ "loss": 1.5893,
60
+ "step": 6
61
+ },
62
+ {
63
+ "epoch": 1.75,
64
+ "grad_norm": 17.667808532714844,
65
+ "learning_rate": 1.2000000000000002e-06,
66
+ "loss": 1.2746,
67
+ "step": 7
68
+ },
69
+ {
70
+ "epoch": 2.0,
71
+ "grad_norm": 0.7934295535087585,
72
+ "learning_rate": 1.4000000000000001e-06,
73
+ "loss": 0.0071,
74
+ "step": 8
75
+ },
76
+ {
77
+ "epoch": 2.0,
78
+ "eval_loss": 0.18537452816963196,
79
+ "eval_runtime": 0.336,
80
+ "eval_samples_per_second": 74.401,
81
+ "eval_steps_per_second": 5.952,
82
+ "step": 8
83
+ },
84
+ {
85
+ "epoch": 2.25,
86
+ "grad_norm": 17.14023780822754,
87
+ "learning_rate": 1.6000000000000001e-06,
88
+ "loss": 1.236,
89
+ "step": 9
90
+ },
91
+ {
92
+ "epoch": 2.5,
93
+ "grad_norm": 17.04801368713379,
94
+ "learning_rate": 1.8000000000000001e-06,
95
+ "loss": 1.0984,
96
+ "step": 10
97
+ },
98
+ {
99
+ "epoch": 2.75,
100
+ "grad_norm": 19.135517120361328,
101
+ "learning_rate": 2.0000000000000003e-06,
102
+ "loss": 1.208,
103
+ "step": 11
104
+ },
105
+ {
106
+ "epoch": 3.0,
107
+ "grad_norm": 40.10493850708008,
108
+ "learning_rate": 2.2e-06,
109
+ "loss": 0.3278,
110
+ "step": 12
111
+ },
112
+ {
113
+ "epoch": 3.0,
114
+ "eval_loss": 0.17443114519119263,
115
+ "eval_runtime": 0.3399,
116
+ "eval_samples_per_second": 73.543,
117
+ "eval_steps_per_second": 5.883,
118
+ "step": 12
119
+ }
120
+ ],
121
+ "logging_steps": 1,
122
+ "max_steps": 12,
123
+ "num_input_tokens_seen": 0,
124
+ "num_train_epochs": 3,
125
+ "save_steps": 500,
126
+ "stateful_callbacks": {
127
+ "TrainerControl": {
128
+ "args": {
129
+ "should_epoch_stop": false,
130
+ "should_evaluate": false,
131
+ "should_log": false,
132
+ "should_save": true,
133
+ "should_training_stop": true
134
+ },
135
+ "attributes": {}
136
+ }
137
+ },
138
+ "total_flos": 0.0,
139
+ "train_batch_size": 16,
140
+ "trial_name": null,
141
+ "trial_params": null
142
+ }
checkpoint-12/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:636997470068939e6673c062e721b262075587603009f7ce943d9eb64598a080
3
+ size 6097
checkpoint-12/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-8/1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 1024,
3
+ "pooling_mode_cls_token": true,
4
+ "pooling_mode_mean_tokens": false,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
checkpoint-8/README.md ADDED
@@ -0,0 +1,404 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - sentence-transformers
4
+ - sentence-similarity
5
+ - feature-extraction
6
+ - dense
7
+ - generated_from_trainer
8
+ - dataset_size:50
9
+ - loss:MultipleNegativesRankingLoss
10
+ base_model: BAAI/bge-large-zh-v1.5
11
+ widget:
12
+ - source_sentence: 定期定額投資的優缺點
13
+ sentences:
14
+ - 近年來大型語言模型與擴散模型在圖像與文本生成領域取得突破性進展。
15
+ - 國際間的生產與物流體系正在發生重大的組織變革與調整。
16
+ - 透過固定金額長期投入，投資者能有效攤平市場波動帶來的成本風險，但可能在強勁牛市中錯失更高的單筆申購報酬。
17
+ - source_sentence: 京都最適合賞楓的季節是什麼時候？
18
+ sentences:
19
+ - 秋季前往關西地區，十一月中旬到十二月初通常是觀賞紅葉的最佳時機。
20
+ - 使用 asyncio 庫可以實現非阻塞的 I/O 操作，顯著提升網路爬蟲或 API 請求的並發性能。
21
+ - 在快速變遷的職場環境中，持續獲取新知識與技能是維持個人競爭力與適應力的關鍵。
22
+ - source_sentence: 長期失眠該如何改善？
23
+ sentences:
24
+ - 建立規律的作息時間、減少睡前使用電子產品，並營造舒適的睡眠環境有助於緩解睡眠障礙。
25
+ - 植物透過葉綠體吸收太陽能，將二氧化碳與水轉化為葡萄糖並釋放氧氣，這是地球能量循環的基礎。
26
+ - 辦理信用貸款通常要求穩定的收入證明與良好的信用評分。
27
+ - source_sentence: 如何減少日常生活中的碳足跡
28
+ sentences:
29
+ - 在推動組織數位化過程中，往往會面臨技術債、員工抗拒改變以及缺乏清晰策略等難題。
30
+ - 該行動裝置的電力持久度表現優異，能滿足長時間使用的需求。
31
+ - 透過節能家電、搭乘大眾運輸及實踐蔬食生活，能有效降低個人的環境影響。
32
+ - source_sentence: 京都最值得造訪的歷史古蹟
33
+ sentences:
34
+ - 這座日本古都擁有眾多世界文化遺產，如清水寺、金閣寺與伏見稻荷大社，是體驗傳統文化的必經之地。
35
+ - 患者通常會感到胸口灼熱（俗稱火燒心）、胃酸逆流，有時還會伴隨慢性咳嗽或喉嚨發炎。
36
+ - 這種以植物性食物、橄欖油和適量深海魚為主的飲食模式，被證實能有效預防心血管疾病。
37
+ datasets:
38
+ - yenstdi/embbedding_text_1111
39
+ pipeline_tag: sentence-similarity
40
+ library_name: sentence-transformers
41
+ ---
42
+
43
+ # SentenceTransformer based on BAAI/bge-large-zh-v1.5
44
+
45
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [BAAI/bge-large-zh-v1.5](https://huggingface.co/BAAI/bge-large-zh-v1.5) on the [embbedding_text_1111](https://huggingface.co/datasets/yenstdi/embbedding_text_1111) dataset. It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
46
+
47
+ ## Model Details
48
+
49
+ ### Model Description
50
+ - **Model Type:** Sentence Transformer
51
+ - **Base model:** [BAAI/bge-large-zh-v1.5](https://huggingface.co/BAAI/bge-large-zh-v1.5) <!-- at revision 79e7739b6ab944e86d6171e44d24c997fc1e0116 -->
52
+ - **Maximum Sequence Length:** 512 tokens
53
+ - **Output Dimensionality:** 1024 dimensions
54
+ - **Similarity Function:** Cosine Similarity
55
+ - **Training Dataset:**
56
+ - [embbedding_text_1111](https://huggingface.co/datasets/yenstdi/embbedding_text_1111)
57
+ <!-- - **Language:** Unknown -->
58
+ <!-- - **License:** Unknown -->
59
+
60
+ ### Model Sources
61
+
62
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
63
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers)
64
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
65
+
66
+ ### Full Model Architecture
67
+
68
+ ```
69
+ SentenceTransformer(
70
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': True, 'architecture': 'BertModel'})
71
+ (1): Pooling({'word_embedding_dimension': 1024, '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})
72
+ (2): Normalize()
73
+ )
74
+ ```
75
+
76
+ ## Usage
77
+
78
+ ### Direct Usage (Sentence Transformers)
79
+
80
+ First install the Sentence Transformers library:
81
+
82
+ ```bash
83
+ pip install -U sentence-transformers
84
+ ```
85
+
86
+ Then you can load this model and run inference.
87
+ ```python
88
+ from sentence_transformers import SentenceTransformer
89
+
90
+ # Download from the 🤗 Hub
91
+ model = SentenceTransformer("sentence_transformers_model_id")
92
+ # Run inference
93
+ sentences = [
94
+ '京都最值得造訪的歷史古蹟',
95
+ '這座日本古都擁有眾多世界文化遺產，如清水寺、金閣寺與伏見稻荷大社，是體驗傳統文化的必經之地。',
96
+ '患者通常會感到胸口灼熱（俗稱火燒心）、胃酸逆流，有時還會伴隨慢性咳嗽或喉嚨發炎。',
97
+ ]
98
+ embeddings = model.encode(sentences)
99
+ print(embeddings.shape)
100
+ # [3, 1024]
101
+
102
+ # Get the similarity scores for the embeddings
103
+ similarities = model.similarity(embeddings, embeddings)
104
+ print(similarities)
105
+ # tensor([[1.0000, 0.6583, 0.1509],
106
+ # [0.6583, 1.0000, 0.1927],
107
+ # [0.1509, 0.1927, 1.0000]])
108
+ ```
109
+
110
+ <!--
111
+ ### Direct Usage (Transformers)
112
+
113
+ <details><summary>Click to see the direct usage in Transformers</summary>
114
+
115
+ </details>
116
+ -->
117
+
118
+ <!--
119
+ ### Downstream Usage (Sentence Transformers)
120
+
121
+ You can finetune this model on your own dataset.
122
+
123
+ <details><summary>Click to expand</summary>
124
+
125
+ </details>
126
+ -->
127
+
128
+ <!--
129
+ ### Out-of-Scope Use
130
+
131
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
132
+ -->
133
+
134
+ <!--
135
+ ## Bias, Risks and Limitations
136
+
137
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
138
+ -->
139
+
140
+ <!--
141
+ ### Recommendations
142
+
143
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
144
+ -->
145
+
146
+ ## Training Details
147
+
148
+ ### Training Dataset
149
+
150
+ #### embbedding_text_1111
151
+
152
+ * Dataset: [embbedding_text_1111](https://huggingface.co/datasets/yenstdi/embbedding_text_1111) at [610ac14](https://huggingface.co/datasets/yenstdi/embbedding_text_1111/tree/610ac1456cc501416303e62f7813f2ee87ee95e3)
153
+ * Size: 50 training samples
154
+ * Columns: <code>anchor</code> and <code>positive</code>
155
+ * Approximate statistics based on the first 50 samples:
156
+ | | anchor | positive |
157
+ |:--------|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
158
+ | type | string | string |
159
+ | details | <ul><li>min: 11 tokens</li><li>mean: 16.02 tokens</li><li>max: 26 tokens</li></ul> | <ul><li>min: 26 tokens</li><li>mean: 39.0 tokens</li><li>max: 55 tokens</li></ul> |
160
+ * Samples:
161
+ | anchor | positive |
162
+ |:-----------------------------------|:---------------------------------------------------------------------------|
163
+ | <code>尋找熟悉 React 生態系統的前端開發者</code> | <code>應徵者需具備 React, Redux 及 Next.js 的實作經驗，並能運用 TypeScript 撰寫高品質程式碼。</code> |
164
+ | <code>後端 Python 工程師職缺要求</code> | <code>精通 Django 或 FastAPI 框架，並有使用 Celery 處理非同步任務與分散式隊列的經驗。</code> |
165
+ | <code>雲端架構師 (AWS 專長) 招募中</code> | <code>負責維運 EC2、S3 與 Lambda 等雲端資源，並能有效配置 RDS 資料庫以確保系統效能。</code> |
166
+ * Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
167
+ ```json
168
+ {
169
+ "scale": 20.0,
170
+ "similarity_fct": "cos_sim",
171
+ "gather_across_devices": false
172
+ }
173
+ ```
174
+
175
+ ### Evaluation Dataset
176
+
177
+ #### embbedding_text_1111
178
+
179
+ * Dataset: [embbedding_text_1111](https://huggingface.co/datasets/yenstdi/embbedding_text_1111) at [610ac14](https://huggingface.co/datasets/yenstdi/embbedding_text_1111/tree/610ac1456cc501416303e62f7813f2ee87ee95e3)
180
+ * Size: 25 evaluation samples
181
+ * Columns: <code>anchor</code> and <code>positive</code>
182
+ * Approximate statistics based on the first 25 samples:
183
+ | | anchor | positive |
184
+ |:--------|:-----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
185
+ | type | string | string |
186
+ | details | <ul><li>min: 11 tokens</li><li>mean: 15.16 tokens</li><li>max: 20 tokens</li></ul> | <ul><li>min: 28 tokens</li><li>mean: 39.36 tokens</li><li>max: 54 tokens</li></ul> |
187
+ * Samples:
188
+ | anchor | positive |
189
+ |:-------------------------------|:----------------------------------------------------------------------|
190
+ | <code>這款手機的電池續航力令人印象深刻。</code> | <code>該行動裝置的電力持久度表現優異，能滿足長時間使用的需求。</code> |
191
+ | <code>什麼是機器學習中的過擬合現象？</code> | <code>當模型在訓練數據上表現極佳，但在未見過的測試數據上預測準確率大幅下降時，通常就是發生了 Overfitting。</code> |
192
+ | <code>2024年全球永續能源趨勢報告</code> | <code>隨著各國減碳政策的推進，太陽能與離岸風電在未來幾年將成為再生能源成長的核心動力。</code> |
193
+ * Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
194
+ ```json
195
+ {
196
+ "scale": 20.0,
197
+ "similarity_fct": "cos_sim",
198
+ "gather_across_devices": false
199
+ }
200
+ ```
201
+
202
+ ### Training Hyperparameters
203
+ #### Non-Default Hyperparameters
204
+
205
+ - `eval_strategy`: epoch
206
+ - `per_device_train_batch_size`: 16
207
+ - `per_device_eval_batch_size`: 16
208
+ - `learning_rate`: 2e-05
209
+ - `lr_scheduler_type`: cosine
210
+ - `warmup_steps`: 100
211
+
212
+ #### All Hyperparameters
213
+ <details><summary>Click to expand</summary>
214
+
215
+ - `overwrite_output_dir`: False
216
+ - `do_predict`: False
217
+ - `eval_strategy`: epoch
218
+ - `prediction_loss_only`: True
219
+ - `per_device_train_batch_size`: 16
220
+ - `per_device_eval_batch_size`: 16
221
+ - `per_gpu_train_batch_size`: None
222
+ - `per_gpu_eval_batch_size`: None
223
+ - `gradient_accumulation_steps`: 1
224
+ - `eval_accumulation_steps`: None
225
+ - `torch_empty_cache_steps`: None
226
+ - `learning_rate`: 2e-05
227
+ - `weight_decay`: 0.0
228
+ - `adam_beta1`: 0.9
229
+ - `adam_beta2`: 0.999
230
+ - `adam_epsilon`: 1e-08
231
+ - `max_grad_norm`: 1.0
232
+ - `num_train_epochs`: 3
233
+ - `max_steps`: -1
234
+ - `lr_scheduler_type`: cosine
235
+ - `lr_scheduler_kwargs`: None
236
+ - `warmup_ratio`: 0.0
237
+ - `warmup_steps`: 100
238
+ - `log_level`: passive
239
+ - `log_level_replica`: warning
240
+ - `log_on_each_node`: True
241
+ - `logging_nan_inf_filter`: True
242
+ - `save_safetensors`: True
243
+ - `save_on_each_node`: False
244
+ - `save_only_model`: False
245
+ - `restore_callback_states_from_checkpoint`: False
246
+ - `no_cuda`: False
247
+ - `use_cpu`: False
248
+ - `use_mps_device`: False
249
+ - `seed`: 42
250
+ - `data_seed`: None
251
+ - `jit_mode_eval`: False
252
+ - `bf16`: False
253
+ - `fp16`: False
254
+ - `fp16_opt_level`: O1
255
+ - `half_precision_backend`: auto
256
+ - `bf16_full_eval`: False
257
+ - `fp16_full_eval`: False
258
+ - `tf32`: None
259
+ - `local_rank`: 0
260
+ - `ddp_backend`: None
261
+ - `tpu_num_cores`: None
262
+ - `tpu_metrics_debug`: False
263
+ - `debug`: []
264
+ - `dataloader_drop_last`: False
265
+ - `dataloader_num_workers`: 0
266
+ - `dataloader_prefetch_factor`: None
267
+ - `past_index`: -1
268
+ - `disable_tqdm`: False
269
+ - `remove_unused_columns`: True
270
+ - `label_names`: None
271
+ - `load_best_model_at_end`: False
272
+ - `ignore_data_skip`: False
273
+ - `fsdp`: []
274
+ - `fsdp_min_num_params`: 0
275
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
276
+ - `fsdp_transformer_layer_cls_to_wrap`: None
277
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
278
+ - `parallelism_config`: None
279
+ - `deepspeed`: None
280
+ - `label_smoothing_factor`: 0.0
281
+ - `optim`: adamw_torch_fused
282
+ - `optim_args`: None
283
+ - `adafactor`: False
284
+ - `group_by_length`: False
285
+ - `length_column_name`: length
286
+ - `project`: huggingface
287
+ - `trackio_space_id`: trackio
288
+ - `ddp_find_unused_parameters`: None
289
+ - `ddp_bucket_cap_mb`: None
290
+ - `ddp_broadcast_buffers`: False
291
+ - `dataloader_pin_memory`: True
292
+ - `dataloader_persistent_workers`: False
293
+ - `skip_memory_metrics`: True
294
+ - `use_legacy_prediction_loop`: False
295
+ - `push_to_hub`: False
296
+ - `resume_from_checkpoint`: None
297
+ - `hub_model_id`: None
298
+ - `hub_strategy`: every_save
299
+ - `hub_private_repo`: None
300
+ - `hub_always_push`: False
301
+ - `hub_revision`: None
302
+ - `gradient_checkpointing`: False
303
+ - `gradient_checkpointing_kwargs`: None
304
+ - `include_inputs_for_metrics`: False
305
+ - `include_for_metrics`: []
306
+ - `eval_do_concat_batches`: True
307
+ - `fp16_backend`: auto
308
+ - `push_to_hub_model_id`: None
309
+ - `push_to_hub_organization`: None
310
+ - `mp_parameters`:
311
+ - `auto_find_batch_size`: False
312
+ - `full_determinism`: False
313
+ - `torchdynamo`: None
314
+ - `ray_scope`: last
315
+ - `ddp_timeout`: 1800
316
+ - `torch_compile`: False
317
+ - `torch_compile_backend`: None
318
+ - `torch_compile_mode`: None
319
+ - `include_tokens_per_second`: False
320
+ - `include_num_input_tokens_seen`: no
321
+ - `neftune_noise_alpha`: None
322
+ - `optim_target_modules`: None
323
+ - `batch_eval_metrics`: False
324
+ - `eval_on_start`: False
325
+ - `use_liger_kernel`: False
326
+ - `liger_kernel_config`: None
327
+ - `eval_use_gather_object`: False
328
+ - `average_tokens_across_devices`: True
329
+ - `prompts`: None
330
+ - `batch_sampler`: batch_sampler
331
+ - `multi_dataset_batch_sampler`: proportional
332
+ - `router_mapping`: {}
333
+ - `learning_rate_mapping`: {}
334
+
335
+ </details>
336
+
337
+ ### Training Logs
338
+ | Epoch | Step | Training Loss | Validation Loss |
339
+ |:-----:|:----:|:-------------:|:---------------:|
340
+ | 0.25 | 1 | 1.3902 | - |
341
+ | 0.5 | 2 | 1.6712 | - |
342
+ | 0.75 | 3 | 1.2991 | - |
343
+ | 1.0 | 4 | 1.3125 | 0.1941 |
344
+ | 1.25 | 5 | 1.6758 | - |
345
+ | 1.5 | 6 | 1.5893 | - |
346
+ | 1.75 | 7 | 1.2746 | - |
347
+ | 2.0 | 8 | 0.0071 | 0.1854 |
348
+
349
+
350
+ ### Framework Versions
351
+ - Python: 3.12.12
352
+ - Sentence Transformers: 5.2.0
353
+ - Transformers: 4.57.6
354
+ - PyTorch: 2.9.0+cu126
355
+ - Accelerate: 1.12.0
356
+ - Datasets: 4.0.0
357
+ - Tokenizers: 0.22.2
358
+
359
+ ## Citation
360
+
361
+ ### BibTeX
362
+
363
+ #### Sentence Transformers
364
+ ```bibtex
365
+ @inproceedings{reimers-2019-sentence-bert,
366
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
367
+ author = "Reimers, Nils and Gurevych, Iryna",
368
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
369
+ month = "11",
370
+ year = "2019",
371
+ publisher = "Association for Computational Linguistics",
372
+ url = "https://arxiv.org/abs/1908.10084",
373
+ }
374
+ ```
375
+
376
+ #### MultipleNegativesRankingLoss
377
+ ```bibtex
378
+ @misc{henderson2017efficient,
379
+ title={Efficient Natural Language Response Suggestion for Smart Reply},
380
+ author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
381
+ year={2017},
382
+ eprint={1705.00652},
383
+ archivePrefix={arXiv},
384
+ primaryClass={cs.CL}
385
+ }
386
+ ```
387
+
388
+ <!--
389
+ ## Glossary
390
+
391
+ *Clearly define terms in order to be accessible across audiences.*
392
+ -->
393
+
394
+ <!--
395
+ ## Model Card Authors
396
+
397
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
398
+ -->
399
+
400
+ <!--
401
+ ## Model Card Contact
402
+
403
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
404
+ -->
checkpoint-8/config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertModel"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "dtype": "float32",
10
+ "eos_token_id": 2,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.1,
13
+ "hidden_size": 1024,
14
+ "id2label": {
15
+ "0": "LABEL_0"
16
+ },
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 4096,
19
+ "label2id": {
20
+ "LABEL_0": 0
21
+ },
22
+ "layer_norm_eps": 1e-12,
23
+ "max_position_embeddings": 512,
24
+ "model_type": "bert",
25
+ "num_attention_heads": 16,
26
+ "num_hidden_layers": 24,
27
+ "output_past": true,
28
+ "pad_token_id": 0,
29
+ "pooler_fc_size": 768,
30
+ "pooler_num_attention_heads": 12,
31
+ "pooler_num_fc_layers": 3,
32
+ "pooler_size_per_head": 128,
33
+ "pooler_type": "first_token_transform",
34
+ "position_embedding_type": "absolute",
35
+ "transformers_version": "4.57.6",
36
+ "type_vocab_size": 2,
37
+ "use_cache": true,
38
+ "vocab_size": 21128
39
+ }
checkpoint-8/config_sentence_transformers.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "5.2.0",
4
+ "transformers": "4.57.6",
5
+ "pytorch": "2.9.0+cu126"
6
+ },
7
+ "model_type": "SentenceTransformer",
8
+ "prompts": {
9
+ "query": "",
10
+ "document": ""
11
+ },
12
+ "default_prompt_name": null,
13
+ "similarity_fn_name": "cosine"
14
+ }
checkpoint-8/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a851b841148810169ddb97c329e324d8be9adf2dd1d636dbb1c1b23963b50aed
3
+ size 1302134568
checkpoint-8/modules.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ },
14
+ {
15
+ "idx": 2,
16
+ "name": "2",
17
+ "path": "2_Normalize",
18
+ "type": "sentence_transformers.models.Normalize"
19
+ }
20
+ ]
checkpoint-8/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a44f3b733d028136ce283dfc10e6bfb7a86cb477d828a5c4085e6fcfc2be310
3
+ size 2596114098
checkpoint-8/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3df96807d359a545e2e36d2c65b9aaa03b57acf0d99bb0153a827928956540bb
3
+ size 14645
checkpoint-8/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cad61be6495c80766576a3e44e770fd294d8133989ff7efdf554772ee260b662
3
+ size 1465
checkpoint-8/sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": true
4
+ }
checkpoint-8/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
checkpoint-8/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-8/tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": true,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 512,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "BertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
checkpoint-8/trainer_state.json ADDED
@@ -0,0 +1,106 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 2.0,
6
+ "eval_steps": 500,
7
+ "global_step": 8,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.25,
14
+ "grad_norm": 22.533252716064453,
15
+ "learning_rate": 0.0,
16
+ "loss": 1.3902,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.5,
21
+ "grad_norm": 20.85812759399414,
22
+ "learning_rate": 2.0000000000000002e-07,
23
+ "loss": 1.6712,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.75,
28
+ "grad_norm": 20.013919830322266,
29
+ "learning_rate": 4.0000000000000003e-07,
30
+ "loss": 1.2991,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 1.0,
35
+ "grad_norm": 72.03453063964844,
36
+ "learning_rate": 6.000000000000001e-07,
37
+ "loss": 1.3125,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 1.0,
42
+ "eval_loss": 0.19411715865135193,
43
+ "eval_runtime": 0.3249,
44
+ "eval_samples_per_second": 76.938,
45
+ "eval_steps_per_second": 6.155,
46
+ "step": 4
47
+ },
48
+ {
49
+ "epoch": 1.25,
50
+ "grad_norm": 26.436626434326172,
51
+ "learning_rate": 8.000000000000001e-07,
52
+ "loss": 1.6758,
53
+ "step": 5
54
+ },
55
+ {
56
+ "epoch": 1.5,
57
+ "grad_norm": 19.47249412536621,
58
+ "learning_rate": 1.0000000000000002e-06,
59
+ "loss": 1.5893,
60
+ "step": 6
61
+ },
62
+ {
63
+ "epoch": 1.75,
64
+ "grad_norm": 17.667808532714844,
65
+ "learning_rate": 1.2000000000000002e-06,
66
+ "loss": 1.2746,
67
+ "step": 7
68
+ },
69
+ {
70
+ "epoch": 2.0,
71
+ "grad_norm": 0.7934295535087585,
72
+ "learning_rate": 1.4000000000000001e-06,
73
+ "loss": 0.0071,
74
+ "step": 8
75
+ },
76
+ {
77
+ "epoch": 2.0,
78
+ "eval_loss": 0.18537452816963196,
79
+ "eval_runtime": 0.336,
80
+ "eval_samples_per_second": 74.401,
81
+ "eval_steps_per_second": 5.952,
82
+ "step": 8
83
+ }
84
+ ],
85
+ "logging_steps": 1,
86
+ "max_steps": 12,
87
+ "num_input_tokens_seen": 0,
88
+ "num_train_epochs": 3,
89
+ "save_steps": 500,
90
+ "stateful_callbacks": {
91
+ "TrainerControl": {
92
+ "args": {
93
+ "should_epoch_stop": false,
94
+ "should_evaluate": false,
95
+ "should_log": false,
96
+ "should_save": true,
97
+ "should_training_stop": false
98
+ },
99
+ "attributes": {}
100
+ }
101
+ },
102
+ "total_flos": 0.0,
103
+ "train_batch_size": 16,
104
+ "trial_name": null,
105
+ "trial_params": null
106
+ }
checkpoint-8/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:636997470068939e6673c062e721b262075587603009f7ce943d9eb64598a080
3
+ size 6097
checkpoint-8/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:64f8dd72cc5492b2190a42c09c2e63751a388035624b4b32e90e400b504ec6ae
3
  size 1302134568
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aaaf117f7a606f380f476d15425eab0047f36fb55fb45fb7b92c4546a682d35f
3
  size 1302134568